COMPDYN 2015

5th ECCOMAS Thematic Conference on

Computational Methods in Structural Dynamics and Earthquake Engineering
M. Papadrakakis, V. Papadopoulos, V. Plevris (eds.)
Crete Island, Greece, 25–27 May 2015

CLASS A PREDICTION FOR SEISMIC CENTRIFUGE MODELING OF

MULTI-BLOCK QUAY-WALL: UNDRAINED (PLAXIS) VERSUS
COUPLED (FLAC) EFFECTIVE STRESS ANALYSIS

Panagiota Tasiopoulou1, Nikos Gerolymos2, and George Gazetas3

1
National Technical University of Athens, Greece
9, Iroon Polytechniou Str., Polytechnic Campus. Zografos P.C 15780
e-mail: ptasiopoulou@gmail.com
2
National Technical University of Athens, Greece
9, Iroon Polytechniou Str., Polytechnic Campus. Zografos P.C 15780
gerolymos@gmail.com
3
National Technical University of Athens, Greece
9, Iroon Polytechniou Str., Polytechnic Campus. Zografos P.C 15780
gazetas@ath.forthnet.gr

Keywords: quay-wall, seismic effective stress analysis, coupled flow dynamic response,
FLAC, PLAXIS.

Abstract. Effective stress analyses are conducted aiming to provide Class A predictions for a
seismic centrifuge test of a multi-block gravity quay-wall. Two different codes are used: i) the
finite difference code FLAC 2D (coupled effective- stress analysis) and ii) the finite element
code, PLAXIS 2D (undrained effective-stress analysis). Two versions of the original
UBCSAND constitutive model, implemented in each code, are used after meticulous calibra-
tion in order to reproduce equivalent liquefaction resistance curves. Apart from predicting
the experimental response, the study aims at performing a comparison between the two codes
widely-used in practice.

2787
 Panagiota Tasiopoulou, Nikos Gerolymos and George Gazetas

1 INTRODUCTION
The latest advances in port and maritime industry have redefined the role of harbor facili-
ties in the economy. Ports are nowadays multipurpose lifeline facilities that function as em-
barkation, storage and maintenance facilities for the transportation of cargo and passengers.
Waterfront structures form the core of nearly all harbor facilities (be it commercial, industrial,
or simply passenger terminals). Therefore maintaining their integrity (as well as a minimum
serviceability threshold) during seismic events, constitutes a necessity for the sustainability pf
the facility as a whole, since most of the components comprising any waterfront plant (e.g.
pipelines, cranes, storage tanks etc.) are either founded, or directly dependent, upon them.
Gravity quay wall structures have repeatedly suffered substantial outward displacement
and rotation even when subjected to moderate earthquake shaking. (e.g. [1]-[7]). The most
astonishing case study comes from the port of Kobe, where during the 1995 earthquake, wall
displacements reached as much as 5 m [8]. Yet, even during the mild Lefkada (Greece)
earthquake of 2003 (Ms = 6.4) most of the coastal structures sustained relatively large
displacements up to 25 cm. An even more recent example is the damage of the Lixouri har-
bour quay wall in the two 2014 Cephalonia (Greece) earthquakes, despite the relatively small
magnitude (Ms  6) of the events.
The dynamic response of gravity quay walls is strongly affected by non-linear soil behav-
iour. Development of excess pore pressures and accumulation of shear and volumetric strains
both at the retained and the foundation soil, produces shear strength degradation which may
be accompanied by liquefaction. The above phenomena are further complicated when ac-
counting for soil-structure interaction. The strong rocking of quay walls (due only to their in-
ertial forces), when founded on a compliant foundation soil in combination with the one-sided
action of earth pressures leads to the accumulation of horizontal displacement and rotation
towards the seaside. Evidently, the deformation modes that synthesize the response of the
quay wall at large displacements and near failure conditions cannot be realistically assessed
by conventional design procedures. The use of suitable constitutive soil models [22] that bal-
ance simplicity and effectiveness in conjunction with powerful numerical techniques is a key-
step for a successful prediction.
In this paper, effective stress numerical analyses are conducted aiming to provide Class A
predictions for a seismic centrifuge test of a multi-block gravity quay-wall, replica of a typical
wall at Piraeus port in Greece. The centrifuge testing was conducted at University of Dundee
and the experimental measurements will be published by [9]. Two different codes are used: i)
finite difference code, FLAC 2D [10] and ii) finite element code, PLAXIS 2D. Two different
versions of the original elasto-plastic effective stress constitutive model [11]-[12],
UBCSAND, have been implemented in each code: i) UBCSAND-904aR for FLAC [13] and
ii) UBC3D-PLM for PLAXIS [14]. These versions are calibrated appropriately in an attempt
to achieve equivalence in their predictions of liquefaction resistance stress ratios for the same
number of loading cycles under undrained conditions. The goal of this study is twofold: a) to
predict of the experimental response and b) to compare the computed response by two codes
widely-used in practice, FLAC (coupled effective- stress analysis) and PLAXIS (undrained
effective-stress analysis).

2 SEISMIC CENTRIFUGE MODELING

A dynamic centrifuge model test was conducted at the University of Dundee centrifuge fa-
cility is examined. The model used a fine quartz based silica sand (HST95) and simulated the
response of a multi-block gravity quay wall made of aluminium alloy, as a replica of a typical

2788
 Panagiota Tasiopoulou, Nikos Gerolymos and George Gazetas

wall at Piraeus port in Greece. The sand bed was formed into an ESB (equivalent shear beam)
model container by carrying out air pluviation at a relative density of D r = 80%. It was then
saturated with water and subjected to an acceleration field of 60g. A sketch of the experi-
mental setup (including the instrumentation layout) is illustrated in figure 1. While in flight, a
sequence of actual ground-acceleration records were applied at the base of the model as input
motion. However, only numerical results for the first acceleration time history (a record from
the ML = 5.9 L’ Aquila 2009 earthquake, shown in Figure 2) are presented as class A predic-
tions. Details on the experimental procedure and measurements will be published by [9].

Figure 1. Centrifuge model setup and instrument locations for Class A prediction.

Figure 2. Base input motion: Acceleration time history and corresponding spectrum

2789
 Panagiota Tasiopoulou, Nikos Gerolymos and George Gazetas

3 NUMERICAL MODELING
Numerical effective stress analysis of a section of the centrifuge model is performed in pro-
totype scale with both finite difference code FLAC and finite element code PLAXIS. The
analysis is conducted taking into account for material (in the soil) and geometric (interface)
nonlinearities. The aluminium alloy frames and rubber spacing layers of the ESB model con-
tainer were also modelled in detail, assuming elastic behaviour. Prescribed displacements
were imposed on the horizontal boundaries of each frame prohibiting their movement in the
vertical direction, and kinematic constraints were assigned to the external and internal vertical
edges of the model allowing it to move as a laminar box ([14]- [17]).
The contact conditions between the blocks of the quay wall as well as between the quay
wall and the adjacent soil were modelled with special interface elements allowing for slippage
and gapping via a Coulomb frictional law. Special interface elements were also placed along
the inner edges of the ESB model container. The friction interface angles were assumed equal
to 18o between the blocks of the quay wall, 10o and 14o at the back and at the base of the wall,
respectively, and 10o for the inner vertical edges of the container. The waterfront was simulat-
ed through hydrostatic pressures applied to the front side of the wall, as well as the seabed. To
avoid spurious oscillations at very small deformations and for high frequency components of
motion, Rayleigh damping was also introduced into the model, accounting for equivalent hys-
teretic damping values between 1.5% and 3% in the range of 0.2 Hz and 2 Hz. The initial hor-
izontal effective stresses were set to 0.5 times the vertical effective stresses. The input motion
in prototype scale, in Figure 2, was applied to the base of the numerical models.

3.1 COUPLED ANALYSIS - FLAC

The finite difference mesh of the model, portrayed in Figure 3, involves a grid spacing of
0.5 m x 0.5 m. FLAC allows for coupled flow dynamic analysis, accounting for interaction of
the poro mechanical soil properties. The coefficient of hydraulic permeability was estimated
to k = 3 x 10-4 m/s (in prototype scale) and assumed to be constant throughout the analysis.

Figure 3. Numerical model in FLAC.

2790
 Panagiota Tasiopoulou, Nikos Gerolymos and George Gazetas

3.2 UNDRAINED ANALYSIS - PLAXIS

Undrained effective stress analysis was conducted with finite element code PLAXIS 2D
AE. Both the quay wall and the soil are modelled with 15-node triangular plane strain ele-
ments, elastic for the former and nonlinear for the latter (3130 elements in total). The finite
element mesh is shown in Figure 4.

Figure 4. The finite element model (PLAXIS)

4 CONSTITUTIVE MODEL
Cyclic soil behaviour is described through an elasto-plastic effective stress constitutive
model, UBCSAND, originally developed at University of British Columbia by [11] and [12].
It involves two yield surfaces (a primary and a secondary one) of the Mohr-Coulomb type.
The primary surface evolves according to an isotropic hardening law while a simplified kine-
matic hardening rule is used for the second yield surface. The elastic response is described by
the elastic shear and bulk moduli given by:

 pa   p / pa 
ne
Ge  kG
e
(1)

Be  kBe  pa   p / pa 
me
(2)

in which kGe and k Be are the elastic shear and bulk numbers in respect, pa is the reference
stress, p is the mean effective stress and ne , me are exponents for stress dependency. The
plastic shear modulus is given by:
2
 η 
G p  Gip  1  Rf  (3)
 ηf 
 

2791
 Panagiota Tasiopoulou, Nikos Gerolymos and George Gazetas

in which Gip component variates for primary, secondary and post-dilation loading, η is the
 
current stress ratio, η f is the stress ratio at failure, equal to sin φ p , φ p being the peak fric-
tion angle and R f is a failure ratio that truncates hyperbolic curve. The plastic flow rule is
non-associated and is based on the Drucker-Prager’s law and Rowe’s stress dilatancy hypoth-
esis:

dεvp   sin  φcv   η  dγ p (4)

where φcv is the phase transformation friction angle.

Different versions and extensions of the model, based on the above-described framework,
have been developed by various researchers. The most widely-used ones, are those imple-
mented in finite difference code, FLAC and finite element code, PLAXIS, such as
UBCSAND-904aR [13] and UBC3D-PLM [14], respectively. One of the most significant di-
vergences of the two versions, when performing seismic effective stress analysis, lies on their
approach on the stiffness degradation of the secondary plastic shear modulus, which practical-
ly controls the number of loading cycles to cause liquefaction.

4.1 UBCSAND 904aR in FLAC

In this version of the model, Gip component has the following form:
np
 p 
Gip
p
 kGp  a   
 f hfac1 ,hfac2 ,ncyc ...  (5)
p  pa 

where kGp is the plastic shear modulus number, np is a constant, hfac1 is a factor controlling
the number of cycles to trigger liquefaction and hfac2 is a factor refining the shape of pore
water pressure rise with cycles, ncyc is the number of loading cycles etc. Obviously, the deci-
sive parameter to define the liquefaction resistance versus number of loading cycles is hfac1 .
The smaller the value of hfac1 , the greater the excess pore water pressure development and the
lesser the liquefaction resistance.

4.2 UBC3D-PLM in PLAXIS

In this version, Gip component is described as:

np
 p 
Gip
p
 kGp  a   
 f fac hard ,ncyc,...  (6)
p  pa 
in whih fachard affects the number of cycles for liquefaction occurrence. This parameter has
the same trend as hfac1 , in the previous version; thus, lower values lead to lesser number of
cycles required to cause liquefaction.

4.3 Calibration Methodology

[13] proposed a set of equations for the calibration of the UBCSAND model parameters,
with the corrected SPT value (N1)60 being the sole variable. The calibration procedure aimed

2792
 Panagiota Tasiopoulou, Nikos Gerolymos and George Gazetas

at matching the cyclic resistance ratio indicated by the NCEER/NSF curve for a given cor-
rected SPT blow count to induce liquefaction at 15 uniform loading cycles. In order to apply
this methodology to our study, the empirical correlation between (N1)60 and relative density,
Dr , proposed by [18] was used:

 N1 60 46  Dr 
2
(7)

Thus, applying the methodology by Beaty and Byrne (2011) for both versions of UBCSAND ,
while assuming that φ p = 42o, as indicated by experiments of HST95 Silica sand for
Dr  80% , the values of the common model parameters were derived, as shown in Table 1.
Then, the different parameters of the two versions, hfac1 for UBCSAND-904aR and fachard
for UBC3D-PLM, were calibrated based on experimental liquefaction resistance curves for
various sands with Dr  80% ([19]-[21]), as shown in Figure 5. The goal of this calibration
process was to achieve equivalence of the two versions in predicting the number of loading
cycles to cause liquefaction under specific cyclic stress ratios. Finally, in order to incorporate
stress dependency, or the so-called Kσ effects, the model parameters, hfac1 and fachard, are
given as functions of the vertical effective stress, as illustrated in Figures 6 and 7. Representa-
tive results of computed response for both versions are shown in figure 8, for CSR = 0.3, ini-
tial effective stress  v0 = 100 kPa and lateral earth pressure coefficient at rest K0 = 0.5.

Figure 5. Comparison between predicted and experimental liquefaction resistance curves in undrained cyclic
simple shear testing for Dr = 80%.

2793
 Panagiota Tasiopoulou, Nikos Gerolymos and George Gazetas

UBC3D- UBCSAND-
Parameters Unit Description PLM 904aR
(PLAXIS) (FLAC)
φp (deg) Peak friction angle 42 42
Phase transformation friction
φcv (deg) 36.1 36.1
angle
k eB - Elastic bulk modulus number 937 937

Ke G - Elastic shear modulus number 1338.6 1338.6

kpG - Plastic shear modulus number 3580.5 3580.5

Exponent for stress dependen-
me - 0.5 0.5
cy of elastic bulk modulus
Exponent for stress dependen-
ne - 0.5 0.5
cy of elastic shear modulus
Power for stress dependency
np - 0.4 0.4
of plastic shear modulus
Rf - Failure ratio 0.662 0.662

pa (kPa) Reference stress 100 100

Fitting parameter to adjust
fachard - number of cycles to liquefac- see Fig. 7 N/A
tion
Fitting parameter to adjust
facpost - 0.01 N/A
post- dilation behaviour
Fitting parameter to adjust
hfac1 - number of cycles to liquefac- N/A see Fig. 6
tion
Fitting parameter to refine
hfac2 - shape of pore pressure rise N/A 1
with cycles
(N1)60 - Corrected SPT blow counts 29.4 29.4

Table 1. Constitutive model parameters for UBC3D-PML (PLAXIS) and UBCSAND-904aR (FLAC).

2794
 Panagiota Tasiopoulou, Nikos Gerolymos and George Gazetas

Figure 6. (a) Cyclic stress ratios versus number of equivalent uniform loading cycles in undrained direct simple
shear loading to cause ru=98% for Dr = 80% and varying vertical effective stress. (b) Corresponding values of
hfac1 parameter of UBCSAND-904aR (in FLAC).

Figure 7. (a) Cyclic stress ratios versus number of equivalent uniform loading cycles in undrained direct simple
shear loading to cause ru=98% for Dr = 80% and varying vertical effective stress. (b) Corresponding values of
fachard parameter of UBC3D-PLM (in PLAXIS).

Figure 8. Simulated cyclic direct simple shear tests with (a) UBCSAND-904aR (in FLAC) and (b) UBC3D-PLM
(in PLAXIS) for Dr = 80%.

2795
 Panagiota Tasiopoulou, Nikos Gerolymos and George Gazetas

5 CLASS A PREDICTION: FLAC AND PLAXIS

Numerical predictions obtained from both FLAC and PLAXIS are shown in terms of time
histories at the measurement locations shown in Figure 1. Initially, comparison of the devel-
oped deformation mechanism of the quay-wall-soil-container system is held in order to ensure
that both models can successfully reproduce the outward quay-wall displacement and rotation,
as has been observed in various case histories. Figure 9 confirms that both models produce
similar deformed meshes after the end of shaking. The contours of horizontal displacements
obtained from both models, illustrated in Figure 10, indicate that sliding at the base of the
quay-wall prevailed against bearing capacity failure.
The numerical analyses provide similar results in terms of the quay-wall outward dis-
placement and rotation, as well as the settlement at the backfill (see Figures 11 and 12). Both
models predict a residual horizontal displacement of the quay-wall in the order of 22 cm and a
residual rotation of 0.24 deg.
Acceleration time histories, shown in Figure 13, also compare well, apart from some high-
frequency spikes in case of FLAC analysis. The inward accelerations are systematically larger
than their outward (seaward) counterparts which appear to have been curtailed due to exces-
sive sliding at the base of the wall. The absence of long period pulses in the accelerogram ob-
tained at the backfill is a sign of either no or limited soil liquefaction occurrence.
However, both analyses predict positive excess pore pressure development at the backfill
away from the quay-wall, close to the base of the model (see Figure 14). In particular, in case
of PLAXIS, where totally undrained conditions apply and no flow (dissipation) is accounted
for, there is a pore pressure build up close to liquefaction. On the other hand, analysis in
FLAC allows for concurrent generation and dissipation of pore water pressure leading thus, to
less positive excess pore water pressure development. In contrast to what happens in the back-
fill away from the quay-wall, negative excess pore pressures develop close to the wall-soil
interface (point 11 in Figure 1) due to the outward displacement of the quay-wall, in agree-
ment with previous studies ([6], [7]).

6 CONCLUSIONS
Class A predictions were presented for seismic centrifuge modeling of a multi-block gravi-
ty quay-wall supporting a dense backfill of Dr=80%. Numerical analyses were performed
with two different codes: i) finite difference code FLAC (coupled flow effective stress analy-
sis) and ii) finite element code PLAXIS (undrained effective stress analysis). The constitutive
models used for soil behavior are extensions of the original elasto-plastic model, UBCSAND.
The two versions were extensively calibrated in order to render them equivalent in terms of
cyclic liquefaction resistance. The analyses provided results in remarkable agreement, espe-
cially for the quay-wall performance. In detail, both analyses predicted that the quay-wall
moved seawards by 22 cm at the top and rotated 0.24 deg. Sliding at the base of the quay-wall
was the predominant mechanism to induce outward quay-wall displacement.

2796
 Panagiota Tasiopoulou, Nikos Gerolymos and George Gazetas

Figure 9. Deformed meshes after the end of shaking magnified 7 times: (a) FLAC and (b) PLAXIS.

Figure 10. Contours of horizontal displacements after the end of shaking: (a) FLAC and (b) PLAXIS.

2797
 Panagiota Tasiopoulou, Nikos Gerolymos and George Gazetas

Figure 11. Predicted quay-wall rotation (top) and horizontal displacement at top (bottom).

Figure 12. Predicted settlement of the backfill at location LVDT3 in Figure 1.

2798
 Panagiota Tasiopoulou, Nikos Gerolymos and George Gazetas

Figure 13. Predicted acceleration time histories at the top block of the quay-wall (top) and at the backfill 2m be-
low to surface (bottom).

Figure 14. Predicted pore water pressures at locations shown in Figure 1.

2799
 Panagiota Tasiopoulou, Nikos Gerolymos and George Gazetas

ACKNOWLEDGEMENT
This research has been co-financed by the European Union (European Social Fund – ESF)
and Greek national funds through the Operational Program "Education and Lifelong
Learning" of the National Strategic Reference Framework (NSRF) - Research Funding
Program: Thales. Investing in knowledge society through the European Social Fund, Project
ID "UPGRADE".

REFERENCES
[1] Pitilakis K. and Moutsakis A., Seismic analysis and behaviour of gravity retaining walls
– the case of Kalamata harbour quaywall. Soil and Foundations, 29(1), 1-17, 1989.
[2] Egan J.A., Hayden R.F., Scheibel. L.L., Otus M., Serventi G.M., Seismic repair at Sev-
enth Street Marine Terminal. Grouting Soil Improvement and Geosynthetics, Geotech-
nical Special Publication, ASCE, 30, 867-878, 1992.
[3] Iai S., Matsunaga Y., Morita T., Miyata M., Sakurai H., Oishi H., Ogura H., Ando Y.,
Tanaka Y., Kato M., Effects of remedial measures against liquefaction at 1993 Kushiro
– Oki earthquake. Proc. Fifth U.S. – Japan Workshop on Earthquake resistant design of
Lifeline Facilities and Countermeasures against Soil Liquefaction, National Center for
Earthquake Engineering Research, NCEER-94-0026, 135-152, 1994.
[4] Sugano T., and Iai S., Damage to port facilities. The 1999 Kocaeli Earthquake Turkey –
Investigation into Damage to Civil Engineering Structures, Earthquake Engineering
Committee, Japan Society of Civil Engineers, 6-1 – 6-14, 1999.
[5] Zarzouras O., Gerolymos N., Gazetas G., Seismic Response of Caisson Quay-Wall in a
Liquefied Environment. Analysis of a Case History. Proc. of the 6th Hellenic Confer-
ence on Geotechnical and Geoenvironmental Engineering, Vol. 1, 549–556, Volos,
Greece, September 29–October 1, 2010.
[6] Tasiopoulou P., Gerolymos N., Tazoh T. and Gazetas G., Pile-Group Response to Large
Soil Displacements and Liquefaction: Centrifuge Experiments Versus a Physically Sim-
plified Analysis. Journal of Geotechnical and Geoenvironmental Engineering, ASCE ,
139(2), 223-233, 2013.
[7] Tasiopoulou P., Gerolymos N, Gazetas G., Seismic Effective Stress Analysis of Gravity
Block-Type Quay Walls: Application to Piraeus Port. 2nd European Conference on
Earthquake Engineering and Seismology, Istanbul, Turkey, 24-29 August, 2014.
[8] Inagaki H., Iai S., Sugano T., Yamazaki H. and Inatomi T., Performance of caisson type
quay walls at Kobe port. Soils and Foundations, Special Issue on Geotechnical Aspects
of the January 17 1995 Hyogoken-Nambu Earthquake, 119-136, 1996.
[9] Anastasopoulos I., Loli M., Antoniou M., Knappet J., Brennan A., Centrifuge testing of
multi-block quay walls. SECED 2015 Conference: Earthquake Risk and Engineering
towards a Resilient World, Cambridge UK, 9-10 July 2015.
[10] Itasca, Fast Lagrangian Analysis of Continua, Itasca Consulting Group Inc., Minneap-
olis, Minnesota, 2005.

2800
 Panagiota Tasiopoulou, Nikos Gerolymos and George Gazetas

[11] Puebla H., Byrne P.M. and Phillips R. Analysis of CANLEX Liquefaction Embank-
ments: Prototype and Centrifuge Models. Canadian Geotechnical Journal, 34(5), 641-
657, 1997.
[12] Beaty M. and Byrne P., An effective stress model for predicting liquefaction behaviour
of sand. Geotechnical Earthquake Engineering and Soil Dynamics III, ASCE Geotech-
nical Special Publication, 75, 766-777, 1998
[13] Beaty M. and Byrne P., UBCSAND constitutive model, Version 904aR. Documentation
Report: UBCSAND constitutive model on Itasca UDM Web Site, 2011.
[14] Galavi V., Petalas A. and Brinkgreve R.B.J, Finite element modeling of seismic lique-
faction in soils. Geotechnical Engineering Journal of the SEAGS & AGSSEA, 44(3), 55-
64, 2013.
[15] Zienkiewicz, O.C., Bicanic N. and Shen F.Q., Earthquake input definition and the
transmitting boundary conditions. Proceedings of Advances in Computational Nonline-
ar Mechanics I, Doltsinis (eds), Springer-Verlag, 109-138, 1988.
[16] Gerolymos N., Giannakou A., Anastasopoulos I., Gazetas G., Evidence of Beneficial
Role of Inclined Piles: Observations and Summary of Numerical Analyses. Bulletin of
Earthquake Engineering, 6(4), 705-722, 2008.
[17] Zafeirakos A. and Gerolymos N., On the Seismic Response of Under-Designed Caisson
Foundations. Bulletin of Earthquake Engineering, 11(5), 1337-1372, 2013.
[18] Idriss, I. M. and Boulanger, R. W., Soil liquefaction during earthquakes, Monograph
MNO-12, Earthquake Engineering Research Institute, Oakland, CA, 2008.
[19] Kammerer J W., Riemer M., Pestana J. and Seed R., A new mult-directional direct sim-
ple shear testing database. 13th World Conference on Earthquake Engineering, Vancou-
ver, B.C., Canada, Paper No. 2083, August 1-6, 2004.
[20] Sriskandakumar, S., Cyclic loading response of fraser sand for validation of numerical
models simulating centrifuge tests. Master's thesis. The University of British Columbia,
Department of Civil Engineering, 2004.
[21] Tatsuoka F., Ochi K., Fujii, S. and Okamoto, M., Cyclic triaxial and torsional strength
of sands for different preparation methods. Soils and Foundations, 26( 3), 23-41, 1986.
[22] Tasiopoulou, P. and Gerolymos, N., Development of a modified elastoplasticity model
for sand. Proceedings of the Second International Conference on Performance–Based
Design in Earthquake Geotechnical Engineering, Taormina (Italy), 28-30 May, 2012.

2801