INTERNATIONAL SOCIETY FOR

SOIL MECHANICS AND

GEOTECHNICAL ENGINEERING

This paper was downloaded from the Online Library of

the International Society for Soil Mechanics and
Geotechnical Engineering (ISSMGE). The library is
available here:

https://www.issmge.org/publications/online-library

This is an open-access database that archives thousands

of papers published under the Auspices of the ISSMGE and
maintained by the Innovation and Development
Committee of ISSMGE.
 Behaviour of transdanubian clay under unloading and reloading
Comportement des transdanubiennes sols argileuses pendant chargement et déchargement

E. Koch
Széchenyi István University, GyĘr, Hungary

ABSTRACT
In recent years, in the field of geotechnical design, software based on FEM has come to the front. Advanced computer programs make
it possible to use more developed soil models besides the most current elastic-plastic Mohr-Coulomb model. By using these computer
programs, nonlinear behaviour of the soil can be followed better, even in the case of more complicated load events. Among the
constitutive models incorporated in the commercial software the Hardening Soil Model (HS) and Soft Soil Model (SS) are the most
promising ones. Observations and experience gained in tunnel construction, deep open excavation or preloaded embankment prove
that with these soil models reality can be followed more accurately, especially in cases where unloading and reloading are present.
In order to produce the software input parameters, in case of more complicated soil model, more demanding laboratory tests are
needed. In case of HS model one of the basic demands is to determine the power for stress-level dependency of stiffness (m), tangent
stiffness for primary oedometer loading ( Eoedref ) and unloading/reloading stiffness ( Eurref ).
The paper focuses on the behaviour of transdanubian clay common in Hungary. Sampling, laboratory investigations and evaluation
aimed to determine the input parameters for the HS are described. Results based on a number of oedometric tests accomplished with
unloading and reloading proved to be adequate for the computational purposes.
RÉSUMÉ
Dans ces derniéres années, des logiciels monté sur la methode des elements finis se mettre en avant dans le domain du projet de
géotechnique. Ces nouvelles logiciels permettent d’utiliser diverses modéles avancées á coté de la modéle plastique-elastique (Mohr-
Coulomb) qui est la plus courante. En utilisent ces nouvelles modéles, on peut suivre mieux la comportement de la sol en cas des
chargement compliqués. Celles la modéle durcissemente (Hardening Soil Model-HS) et la modéle sol spongieux (Soft Soil Model-SS)
qui sont les plus prometteuses. En juger par l’experience des mesures, il est possible de suivre la réalité plus précisement,
particulierement dans les cas de déchargement et/ou réchargement, par example des tunnels, des fouilles de travaux ou des remblais
prechargés.
Pour identifier les paramétres d’inputs d’une modéle de la sol plus compliquée, on a besoin des test de laboratoire plus détaillé.
Definir du changement d’indice d’une puissance de la coefficient de compression tipique de la premiere chargement et la definition
des coefficient relatifs pendant déchargement et rechargement - ces sont des exigences essentielles (m, Eoedref , Eurref)
On a fait des recherches sur le comportement des sols argileuses hongroises, utilisent les échantillons effectué sur le Sud de la
Transdanubia. L’objectif de cette investigation scientifique a été la determination des paramétres d’ inputs de la modéle HS dans le
cas de la sol argileuse transdanubienne. Les resultats se basent sur les nombreuses test oedometrique, accomplissent
réchargement/déchargement.

Keywords : nonlinear constitutive model, Plaxis, Hardening Soil Model, clay

1 INTRODUCTION Geotechnicans using these computational facilities often face

the problem of having inadequate or inaccurate data for
In civil engineering lately we have several projects where preliminary geotechnical estimations or suggestions for
unloading and reloading are present. The typical case is creating structural design. To overcome this difficulty the laboratory at
a deep excavation in an urban area or applying preload to an Széchenyi István University has performed approximately two
embankment. Movements induced by consecutive excavations hundred oedometer tests using unloading and reloading during
demand more accurate calculation and analysis of soil-structure the last few years. We have analyzed the test results and
interaction especially in urban areas. In such cases the strains established correlations with the computational model
and displacements cannot be descried accurately by the parameters. The aim of this research was to determine reliable
conventional constitutive model: the elasto-plastic model with relationships to substitute or estimate missing data and serve as
Mohr-Coulomb failure criteria. Excavation and reloading can be a design aid during the preliminary phase of a project.
analyzed only by a constitutive model capable of describing
nonlinear behavior and hardening process. Advanced programs 2 INVESTIGATED SOIL PROPERTIES
like PLAXIS now make it possible to use more complex soil
models. The two most promising new models are: the Table 1. shows the main parameters of the tests. The plasticity
Hardening Soil Model (HS); and the Soft Soil Model (SS). index ranges between 9 % and 60 % with an average value of
Observations and experience gained in tunnel construction, deep 21 %. The consistency index is between 0,5 and 1,6, the
excavations, and preloaded embankments show that these soil average value is 1.0. The average value of the oedometer
models more accurately describe field behavior. This is modulus due to primary compression is 12 MPa, and the
especially true where there are complex loading and unloading average elastic unloading/reloading modulus is 47 MPa. Note
sequences in construction. that the unloading/reloading is about four times the primary

Proceedings of the 17th International Conference on Soil Mechanics and Geotechnical Engineering 463
M. Hamza et al. (Eds.)
© 2009 IOS Press.
doi:10.3233/978-1-60750-031-5-463
464 E. Koch / Behaviour of Transdanubian Clay

compression. This is in line with recommendations in the Plaxis Application of the HS model requires the user to have known
manual which states that the quotient should be between three m, Eoedref modulus and Eurref modulus.
and five. The collected data have been evaluated statistically
and correlations have been evaluated, and these results are 60
presented below.
The samples originated from the southern part of Hungary.
The points are above the A-line and diverge from the A-line 40

IP (%)
with increasing liquid limit values (Figure 1). The equation of A line
the best-fit line is Ip=0,85·(wL-18,2) with a correlation
20
coefficient r = 0,95. The tested samples plot are above the A-
line, so they are mainly medium and high plasticity clay but
with some silt too. 0
Figure 2. shows the relationship between water content and 20,0 30,0 40,0 50,0 60,0 70,0 80,0
consistency index. The linear correlation is descried by the
w L (%)
equation Ic=1,9-0,04·w, with a correlation coefficient r=0,7.

3 PARAMETERS OF HARDENING SOIL MODEL Figure 1. Correlation of Ip - wL

The Hardening Soil model (HS) is an advanced model for

simulating the behavior of different types of soil, both soft and 1,8
hard. A basic feature of the present HS model is the stress 1,6
dependency of soil stiffness. The HS model uses moduli both
1,4
from oedometer tests and from triaxial tests. Oedometer test
results demonstrate the dependence of stiffness on confining 1,2
stress. The increase of the oedometric modulus depends on the 1,0

Ic (-)
mean hardening stress. This is described by the equation
0,8
m
§ · 0,6
ref ⋅ ¨ c ⋅ ctgϕ − σ′1
E oed = E oed ¸ (1)
¨ ref ¸ 0,4
© c ⋅ ctgϕ + p ¹ 0,2
where the Eoedref is the reference oedometer modulus, and the 0,0
strength and stress parameters are raised to the power m. The 5,0 10,0 15,0 20,0 25,0 30,0 35,0
reference pressure, pref is typically 100 kPa. Under triaxial w (%)
conditions, the parameter E50 is the confining stress dependent
stiffness modulus for primary loading and is given by the equation Figure 2. Best-fit linear regression between consistency index, Ic and
water content, w
m
§ ·
ref ⋅ ¨ c ⋅ ctgϕ − σ′3
E50 = E50 ¸ (2)
¨ ref ¸
© c ⋅ ctgϕ + p ¹

where E50ref is a reference stiffness modulus corresponding to

the reference confining pressure pref. For unloading and
reloading stress paths, another stress-dependent stiffness
modulus is used (figure 3) :

m
§ c ⋅ ctgϕ − σ′ ·
E ur = E ref ¨ 3 ¸ (3)
ur ⋅ ¨ ref ¸
© c ⋅ ctgϕ + p ¹

where Eurref is the reference Young’s modulus for unloading

and reloading, corresponding to the reference pressure pref. Figure 3. Deviatoric stress vs. axial strain showing E50 and Eur

Table 1. Properties of the investigated soil

In order to determine the modulus input parameter, we
minimum maximum model the stress-strain behavior as a power curve (as suggested
parameter average value
value value by Janbu) :
w % 10,0 35,0 22,5
σ
ε z = A( z ) B (4)
Ip % 9,0 59,9 20,9 p
Ic - 0,5 1,6 1,0
where εz – the vertical specific strain, σz - the vertical stress, A
Eoed MPa 3,7 40,4 12,3 – scale factor relating stress and strain, B – power factor relating
stress and strain, p=100 kPa. With this formula we get the same
Eur MPa 8,6 125,6 46,6 shape as it is in the HS model. By taking the derivative of the
A - 0,007 0,118 0,032
p σ
B - 0,22 0,92 0,57
Es = ⋅ ( z )1−B (5)
A⋅B p
m - 0,08 0,78 0,43
 E. Koch / Behaviour of Transdanubian Clay 465

equation, we can formulate the oedometric tangent modulus as a needed. The basic demands are to determine the power for
function of σz: stress-level dependency of stiffness (m), tangent stiffness for
Figure 4. shows the relation between A and B. The primary oedometer loading ( Eoedref ) and unloading/reloading
correlation can be descried by the equation: stiffness (Eurref) when using the Hardening Model.

0,125
B= (6) 160,0
A 0, 41 140,0
120,0
The correlation coefficient is r = 0,86, implying good
correlation The compression curve can now be descried exactly 100,0

Eur (MPa)
by these A and B parameters. 80,0
60,0
40,0
1,2 20,0
1 0,0
0,0 10,0 20,0 30,0 40,0 50,0
0,8
Eoed (MPa)
B (-)

0,6
Figure 6. Relation between Eoed and Eu
0,4

0,2
1
0
0 0,02 0,04 0,06 0,08 0,1 0,12 0,14 0,8
A (-)
0,6
m (-)
Figure 4. A and B parameters from oedometer test results 0,4

0,2
Further investigations were performed to establish
correlations for other parameters. The relationship between 0
parameter A and water content gives the best result (figure 5). 0,4 0,6 0,8 1 1,2 1,4 1,6
The correlation coefficient is only r=0,64, which means I c (-)
moderately close relationship. The equation of the line is
Figure 7. Relation between m and consitency index, Ic
3 ⋅ w − 35
A= The correlations determined from over 200 oedometer and
1000 (7) index tests are: A = (3·w-35)/1000, B = 0,125/A0,41, Eur =
Eoed·3,6, m = 0,45.
The relation of the oedometer modulus due to primary These parameters are in agreement with the recommended
compression and the modulus due to unloading/reloading is value in the Plaxis manual. They can be applied by geotechnical
shown on figure 6. The second modulus is 3,6 times bigger than engineers in cases where only index properties are available or
the first modulus with a correlation coefficient r = 0,7. when they are performing preliminary evaluations. These
Figure 7. shows the correlation between the stress correlations also add confidence to the analyst when performing
dependent stiffness according to a power law (m = 1-B) and the computations using Plaxis or other sophisticated programs.
consistency index. It can be seen that the average m parameter
hardly changes with the consistency index. The average value is
m = 0,45 in case of plastic and hard clay. ACKNOWLEDGEMENTS
0,1
This project has been undertaken by the Department of
0,08 Structural Engineering, Széchenyi István University, Hungary.

0,06
A (-)

REFERENCES
0,04
Atkinson, J. : The mechanics of soils and foundation, Taylor & Francis
0,02 Group, London, 2007, pp. 161–201.
Brinkgreve, R. B. J., Vermeer, P. A.: Plaxis, Balkema (1998)
0
Hans-Georg Kempfert, Berhane Gebreselassie : Excavations and
14 18 22 26 30
Foundations In Soft Soils. Springer, Berlin, 2006, pp. 57-111.
w (%) Szepesházi Róbert.: Dispalcements around deep excavations
Figure 5. Relation between parameter A and water content (manuscript, in Hungarian), Miskolc, Hungary, 2007
Szymansky, A. : The use of constitutive soil models in consolidation
analysis of organic subsoil under embankment. Proceeding of the
Workshop on Advances in Understanding and modelling the
4 SUMMARY mechanical behaviour of peat, Balkema, Rotterdam:231-240, 1994

In order to produce the software input parameters, for more

complicated soil models, more demanding laboratory tests are