You are on page 1of 27

Geotextiles and Geomembranes 20 (2002) 395421

Overall stability of geosynthetic-reinforced


embankments on soft soils
Jose! Leit*ao Borges*, Antonio
! Silva Cardoso
Department of Civil Engineering, Faculty of Engineering, University of Porto, Rua Dr. Roberto Frias,
4200-465 Porto, Portugal
Received 12 September 2001; received in revised form 3 March 2002; accepted 11 June 2002

Abstract

Overall stability of geosynthetic-reinforced embankments on soft soils is analysed using two


different methodologies: application of a numerical model based on the nite element method;
use of a limit equilibrium method. These two methodologies are described and also applied on
three geosynthetic-reinforced embankments on soft soils. One of the cases is a case history
constructed up to failure. Considering the analysis of the results, some conclusions are
formulated on the limit equilibrium method accuracy, namely regarding the critical slip
surface, overall safety factor and overturning and resisting moments.
r 2002 Elsevier Science Ltd. All rights reserved.

Keywords: Reinforced embankment; Geosynthetic; Overall stability; Limit equilibrium; Finite element
analysis

1. Introduction

To design embankments on soft soils it is essential to take into account the


multiple constructive techniques that allow to solve the problems usually associated
with this kind of construction: overall stability deciency and large settlements that
develop slowly.
The constructive solutionsusually based on both foundation soil properties
improvement and construction procedures or ll properties alterationprovide one
or more of the following effects: increase of overall stability, consolidation
acceleration and decrease of long term settlements.

*Corresponding author. Tel.: +351-22508-1928; fax: +351-22508-1440.


E-mail address: leitao@fe.up.pt (J.L. Borges).

0266-1144/02/$ - see front matter r 2002 Elsevier Science Ltd. All rights reserved.
PII: S 0 2 6 6 - 1 1 4 4 ( 0 2 ) 0 0 0 1 4 - 6
396 J.L. Borges, A.S. Cardoso / Geotextiles and Geomembranes 20 (2002) 395421

With textile industry development in the 20th century, especially following the
1960s, geosynthetic reinforcement has been added to the list of possible solutions
when embankments must be constructed on very soft foundations. In many cases,
the use of one geotextile or geogrid can signicantly increase the safety factor,
improve performance in terms of displacements and reduce costs in comparison with
more conventional solutions.
Especially due to their simplicity, overall stability of reinforced embankments on
soft soils is usually computed by limit equilibrium methods along potential slip
circles. The reinforcement effect is considered by a resisting force due to the
geosynthetic.
Theoretically, however, because rigidplastic behaviour is tacitly assumed for the
materials (soil and reinforcement), the use of these methods may raise some
reticences, as strains are not taken into account before overall failure, as well as
stress redistribution caused by the geosynthetic. This often determines the use of the
nite element method in the study of geosynthetic-reinforced embankments (Rowe,
1984; Humphrey, 1986; Soderman, 1986; Rowe and Soderman, 1987; Kwok, 1987;
Rowe and Mylleville, 1990; Mylleville and Rowe, 1991; Russell, 1992; Borges, 1995).
In the paper, overall stability of geosynthetic-reinforced embankments on soft
soils is analysed adopting two different methodologies: (i) using the results obtained
from a numerical model based on the nite element method (Borges, 1995); along
each analysed slip circle, acting and resisting tangential forces are obtained from
numerical results and strength characteristics of the materials; (ii) applying a limit
equilibrium method, based on the formulation proposed by Kaniraj and Abdullah
(1993) but with some improvements (see Section 3) performed by Borges (1995).
The two methodologies are also applied on three geosynthetic-reinforced
embankments on soft soils. The third embankment is a case history that was
constructed up to failure (Quaresma, 1992).

2. Overall stability analysis using results from nite element method application

2.1. General characteristics of the finite element model

Numerical analysis is simulated by a model, developed by Borges (1995), based on


the nite element method, with the following theoretical hypotheses: (a) plane strain
conditions; (b) coupled formulation of the ow and equilibrium equations,
considering soil constitutive relations formulated in effective stresses (Biots
consolidation theory) (Lewis and Schreer, 1987; Britto and Gunn, 1987); this
formulation is applied to all phases of the problem, both during the embankment
construction and in the post-construction period; (c) utilisation of the p2q2y critical
state model (Lewis and Schreer, 1987; Britto and Gunn, 1987; Borges, 1995; Borges
and Cardoso, 1998) to simulate constitutive behaviour of the foundation and
embankment soils (see below); (d) utilisation of a hardening elasto-plastic model to
simulate constitutive behaviour of the reinforcement; (e) simulation of constitutive
behaviour of the soilgeosynthetic interfaces using a hardening elasto-plastic model.
J.L. Borges, A.S. Cardoso / Geotextiles and Geomembranes 20 (2002) 395421 397

Geosynthetic and soilgeosynthetic interfaces are modelled by bar and interface


elements, respectively. For 2D elements, triangular elements are used, namely six-
noded elements (at the vertices and middle of the edges) with 12 displacement
degrees of freedom (for ll elements) or with three more excess pore pressure degrees
of freedom at the vertices of the triangle (for foundation elements where
consolidation analysis is considered).
In order to verify accuracy of the numerical model used in this paper, Borges
(1995) compared numerical and eld results of two reinforced embankments on soft
soils, one constructed up to failure (Quaresma, 1992) and the other observed until
the end of the consolidation (Yeo, 1986). The accuracy was considered adequate in
both cases, as numerical and eld results are similar, namely in terms of settlements,
pore pressures and strains or tensile forces in the reinforcement. Only some
quantitative differences were observed in the horizontal displacements, despite an
overall qualitative similarity too.

2.2. Evaluation of safety factors

Using effective stresses obtained from numerical model application, one can
compute overall safety factor and also partial safety factors of the different materials
(reinforcement and foundation and ll soils) as described below (Borges, 1995;
Borges and Cardoso, 1997).
Firstly, for each potential slip circle, the intersection points of the circle with the
edges of the nite elements of the mesh are determined. Therefore, the slip circle is
divided into small line segments, each of them located inside of only one of the nite
elements of the mesh. Afterwards, the average values of effective stresses (s0x ; s0y and
txy ; normal and shear stresses in the directions of x- and y-axes) at each of those
segments are computed extrapolating from the stresses at the Gauss points of the
corresponding nite element. Mathematical procedures for this extrapolation
depend on the type of the 2D nite element, i.e. position and number of nodes
and Gauss points. For the 2D element used in this studysix-noded triangular
element with seven Gauss points (see Fig. 1), whose strains vary linearly within the
elementthe following scheme was assessed adequate: (i) rstly, stresses at each
node of the element are computed by linear extrapolation from two pointsthe

Node
Gauss point

Line segment of the


slip surface

Fig. 1. 2D nite element.


398 J.L. Borges, A.S. Cardoso / Geotextiles and Geomembranes 20 (2002) 395421

Gauss point at middle of the element and the Gauss point that is nearest the node;
(ii) secondly, stresses at each extremity of the segment line of the slip circle are
computed by linear interpolation from stresses at the two nearest nodes of the
element; (iii) nally, stresses at each segment are the average values obtained from
stresses at the two extremities of the segment.
Considering the slip circle divided into line segments, partial safety factor of the
soil along the slip surface (foundation and ll) is dened by
PN
qfi li
Fs Pi1
N
; 1
i1 ti li

where li is the i-segment length; N is the number of 2D elements of the mesh


intersected by the circle; ti is the average value of shear stress at i-segment
(determined from effective stresses s0x ; s0y and txy ; known the angle that denes
i-segment inclination); qfi is the average value of soil shear strength at i-segment.
Limiting the sum only to the segments in the foundation or in the ll, the partial
safety factors for corresponding soils along the slip circle can also be obtained with
Eq. (1).
Once the nite element method model uses a constitutive model, p2q2y critical
state model, that formulates the soil strength or failure (critical state) by the Mohr
Coulomb criteria (see Fig. 2), soil shear strength at i-segment, qfi ; can be estimated
by the equation
qfi c0i cos f0i p0i sin f0i ; 2
where p0i s0xi s0yi =2 s01i s03i =2 is effective volumetric stress and c0i and f0i are
the cohesion and friction angle of soil at i-segment dened in effective terms (s01i ; s03i
are the principal effective stresses; s0xi ; s0yi are the normal effective stresses in x- and
y-directions).

Fig. 2. Yield and critical state surfaces of p2q2y critical state model in principal effective stress space.
J.L. Borges, A.S. Cardoso / Geotextiles and Geomembranes 20 (2002) 395421 399

As shown by Borges and Cardoso (2001) and Borges (1995), if consolidation is not
signicant, the typical effective stress path in the foundation corresponds to low
variations of p0 s01 s03 =2 with signicant increases of q s01  s03 =2: As an
approximation, p0 can be considered with the same value at acting stress state
(obtained from nite element analysis) and at failure stress state, i.e. the
corresponding Mohr circles can be considered concentric. Soil shear strength, qf ;
is the radius of the failure Mohr circle, given by Eq. (2), as deduced in Fig. 3.
The partial safety factor of the geosynthetic is determined by the equation
Tr
Fg ; 3
Ta
where Ta is the geosynthetic acting tensile force at the cut point (point I in Fig. 4),
intersection of the slip circle with the geosynthetic (or geosynthetics, if there are
several reinforcement layers) and Tr is the corresponding resisting tensile force. Ta is
obtained from the numerical results interpolating from tensile stresses at the two
nearest Gauss points of the bar element that contains point I: Tr is given by
Tr minTrg ; Trp 4


=c'+'n tan'

qf = c + p sin =
('n , ) tan
qf
c' ' = c cos + p sin
'3f '3 p' '1 '1f 3f p 1f 'n
n

p'= ('1+'3)/2 failure Mohr total stress


c'/tan' circle
acting Mohr Mohr circle
circle

Fig. 3. Relation between soil shear strength and effective cohesion and friction angle.

A geosynthetic I B

Fig. 4. Overall stability analysis.


400 J.L. Borges, A.S. Cardoso / Geotextiles and Geomembranes 20 (2002) 395421

being Trg and Trp the maximum forces that can be mobilised at point I taking into
account the geosynthetic strength and pull-out force of the soilgeosynthetic
interfaces, respectively.
So, Trg is a property of the geosynthetic and, because equilibrium has to be
veried, Trp is calculated by the equation
Trp minTrp2AI ; Trp2IB ; 5
where Trp2AI and Trp2IB correspond, respectively, to the integrals of the maximum
adhesive and frictional stresses that can be mobilised in the soilgeosynthetic
interfaces on the left, segment AI, and on the right, segment IB, from the cut point I
(see Fig. 4):
Z
Trp2AI al au s0n tan dl tan du  dl; 6a
AI
Z
Trp2IB al au s0n tan dl tan du  dl; 6b
IB

where al ; dl are the adhesion and frictional angle of the lower interface; au ; du are
the adhesion and frictional angle of the upper interface; and s0n is the normal stress
acting on the interface.
It should be noted that the values of s0n are known at the Gauss points of the
elements that model the interface, so Eqs. (6a) and (6b) can be calculated
approximately considering linear the variation of s0n between two contiguous Gauss
points. This is an adequate approximation because the nite element meshes usually
used in this kind of problems must have sufciently small elements.
Finally, overall safety factor is dened as follows:
PN PN g
i1 qfi li j1 Trj coscr yj 
F PN PN g ; 7
i1 ti li j1 Taj coscr yj 

where Ng is the number of reinforcements (in most cases Ng 1), yj the angle
between the reinforcement direction and the tangent to the slip circle at the cut point,
cr a reduction coefcient (it varies between 0 and 1) related to the direction
considered for the forces Trj and Taj (resisting and acting tensile forces of
j-geosynthetic, as dened above), and li ; N; ti and qfi have the same meaning that
in Eq. (1). It should be noted that the reinforcement force has been considered to act
in its original orientation (usually horizontally) by some investigators (Broms, 1977;
Tensar, 1982; Jewell, 1982; Ingold, 1982, 1983; Brakel et al., 1982; Duncan and
Wong, 1984; Milligan and La Rochelle, 1984), tangentially to the circle by some
others (Haliburton, 1981; Quast, 1983) assuming that the local deformations
associated with the formation of failure surface result in a local reorientation of the
geosynthetic, and between the above two directions by Huisman (1987). However, as
indicated by Soderman (1986), the assumption that provides better agreement with
observed embankment performance is, likely, dependent on the particular case.
Because the assumption that assumes the geosynthetic remains in its original
J.L. Borges, A.S. Cardoso / Geotextiles and Geomembranes 20 (2002) 395421 401

C' C

COMPLEMENTARY SURFACE

INITIAL SURFACE

Fig. 5. Symmetric problem regarding one slip surface that does not include whole continuum.

orientation (cr 1) is the most conservative assumption, in doubt, one should


consider it in design.
Sometimes, because the problem is symmetric, the nite element mesh does not
include the whole continuum. In this case, symmetry allows solving this question
considering a complementary surface, as indicated in Fig. 5. In this second surface
one can have all data and results regarding the part of initial surface that is out of the
mesh limits.

3. Application of a limit equilibrium method on the overall stability design

Most of the design methods used to assess the overall stability of reinforced
embankments on soft soils are limit equilibrium methods which use circular failure
surfaces and consider a resisting force due to the reinforcement (Rowe, 1984; Jewell,
1982; Ingold, 1982; Haliburton, 1981; Brakel et al., 1982; Fowler, 1982; Milligan and
LaRochelle, 1984).
Overall safety factor, F ; can be computed by
MR
F ; 8
MO
where MR is the resisting moment and MO the overturning moment, which can be
given by (Fig. 6):
MO Wx; 9
X 
MR tri Dli R Tg R cosy  a; 10
402 J.L. Borges, A.S. Cardoso / Geotextiles and Geomembranes 20 (2002) 395421

R y

W geosynthetic
Tg

l i
ri

Fig. 6. Potential failure surface: rotational stability analysis.

where W is the weight of the ll mass inside the slip circle, tri the resisting shear
stress along Dli long arc of the slip circle, and Tg the maximum force that
reinforcement can mobilise at the intersection point with the circle (as seen in Section
2.2, this force depends on geosynthetic strength and the pull-out force that can be
developed along the soilgeosynthetic interfaces).
Alternatively to Eq. (8), in agreement with more recent design methodologies,
partial safety factors can be used in overall failure analysis. These methodologies
consider: (i) factors that multiply actions, Zg for unit weight of the ll soil and Zn for
eventual variable actions; (ii) factors that divide materials properties, Gc and Gf for
cohesion and friction angle of the ll soil, Gsu for undrained shear strength of the
foundation soil, Gr for reinforcement strength, and Ga and Gd for adhesion and
frictional angle of the soilreinforcement interfaces.
As known, using partial factors, one has to verify the inequation:
MOd pMRd ; 11
where MOd and MRd are, respectively, design overturning and resisting moments, i.e.
determined with the above partial factors. Of course, if all partial factors of actions
and materials are imposed equal to 1, an overall safety factor can be obtained as
dened by Eq. (8).
As said in Section 1, a limit equilibrium method is used in this paper, based on the
assumptions proposed by Kaniraj and Abdullah (1993) but with some improvements
implemented by Borges (1995). These improvements are: (a) introduction of partial
safety factors; (b) correct calculation of the moment due to resisting forces in the
foundation soil along the slip surface (which is done approximately by Kaniraj and
Abdullah (1993), depending on the number of layers that the foundation is divided);
(c) calculation of the pull-out force of the geosynthetic taking into account that its
J.L. Borges, A.S. Cardoso / Geotextiles and Geomembranes 20 (2002) 395421 403

Fig. 7. Slip circle in a reinforced embankment on soft soil (Kaniraj and Abdullah, 1993).

value, due to equilibrium reasons, is the minimum of the two values computed on the
left and on the right of the cut point (intersection of the reinforcement with the slip
circle).
The equations used to calculate MOd and MRd are presented in appendix assuming
the geometry as shown in Fig. 7, which shows an arbitrary slip circle in a reinforced
embankment. The origin of (X ;Y )-axes is assumed to be at the toe.

4. Overall stability analysis of geosynthetic-reinforced embankments on soft soils

4.1. Description of the illustrative cases

In this section overall stability of three geosynthetic-reinforced embankments on


soft soils is analysed using the two methodologies described in Sections 2 and 3.
Case 1 is a 28-day continuous construction of a 2 m height symmetric
embankment, with a 10.6 m crest width and 23 (V/H) inclined slopes. The foundation
is a 5 m thick saturated clay layer lying on a rigid and impermeable soil, which
constitutes the lower boundary. The clay is lightly overconsolidated to 1.8 m depth
and normally consolidated from 1.8 to 5 m. One geosynthetic-reinforcement level is
considered between the foundation and embankment soils. Taking into account the
high permeability of the geosynthetic and of the embankment soil, the upper
foundation surface is a drainage boundary.
Fig. 8 shows the nite element mesh used in the numerical analysis; only 2D
elements are represented; y-axis is the symmetry line.
404 J.L. Borges, A.S. Cardoso / Geotextiles and Geomembranes 20 (2002) 395421

Fig. 8. Finite element mesh for Case 1.

Table 1
Geotechnical properties of the foundation and of the embankment for Case 1

g (kN/m3) n0 c0 (kPa) f0 (1) kx (m/s) ky (m/s) p2q2y critical state model

l k G N
9 9
Foundation 17 0.25 0 30 10 10 0.22 0.02 3.26 3.40
Embankment 20 0.30 0 35 0.03 0.005 1.80 1.817

Table 2
At rest earth pressure coefcient, k0 ; and over-consolidation ratio, OCR, in the foundation for Case 1

Depth (m) k0 OCR

01 0.7 2.43
11.8 0.70.5 2.431
1.85 0.5 1

The constitutive relations of both the embankment and foundation soils were
simulated using the p2q2y critical state model (Lewis and Schreer, 1987; Britto
and Gunn, 1987; Borges, 1995; Borges and Cardoso, 1998) with the parameters
indicated in Table 1 (l; slope of normal consolidation line and critical state line; k;
slope of swelling and recompression line; G; specic volume of soil on the critical
state line at mean normal stress equal to 1 kPa; N; specic volume of normally
consolidated soil at mean normal stress equal to 1 kPa). Table 1 also shows other
geotechnical properties: g; unit weight; n0 ; Poissons ratio for drained loading; c0 and
f0 ; cohesion and angle of friction dened in effective terms; kx and ky ; coefcients of
permeability in x- and y-directions. Table 2 indicates the variation with depth of the
at rest earth pressure coefcient, k0 ; and over-consolidation ratio (OCR), in the
foundation. All these parameters were dened taking into account typical
experimental values for this kind of soils (Borges, 1995; Lambe and Whitman, 1969).
J.L. Borges, A.S. Cardoso / Geotextiles and Geomembranes 20 (2002) 395421 405

200

FORCE (kN/m) 150

100

50

0
0 5 10 15 20
(a) STRAIN (%)


n = 15kPa
10
SHEAR STRESS (kPa)

0
0 1 2 3 4 5 6 7 8 s
(b) RELATIVE DISPLACEMENT (mm)

Fig. 9. Constitutive curve for Case 1 of the: (a) geosynthetic; (b) soilgeosynthetic interface for normal
stress of 15 kPa.

Fig. 9 shows the constitutive curves of the geosynthetic and soilgeosynthetic


interfaces. The geosynthetic thickness is 2 mm and its elastic modulus is
1.5 106 kPa. Normal and tangential stiffnesses of soilreinforcement interfaces
are 2.0 107 and 1.6 104 kPa/m, respectively. Geosynthetic strength is 200 kN/m.
Adhesion, a, and frictional angle, d; for both upper and lower soilgeosynthetic
interfaces are 0 kPa and 33.71, respectively.
Mechanical behaviour of the reinforcement and soilreinforcement interfaces were
simulated in the numerical analysis by elasto-plastic models that incorporate the
following hardening law (Prevost and Hoeg, 1975; Owen and Hinton, 1980; Thomas,
1984):
c 2 h c 3 h2
Y h c1 ; 12
1 c 4 h c 5 h2
where h is the hardening parameter and c1 ; c2 ; c3 ; c4 and c5 are parameters that
characterise the material. For the geosynthetic, the yielding function is expressed by
f s  Y h; 13
where s is the tensile stress, Y h sY ep is the hardening law, h ep the hardening
parameter, sY the yielding stress and ep the plastic tensile strain.
406 J.L. Borges, A.S. Cardoso / Geotextiles and Geomembranes 20 (2002) 395421

Table 3
Values of the hardening law parameters for Case 1

c1 c2 c3 c4 c5
4 6
Geosynthetic 7.5 10 (kPa) 8.824 10 (kPa) 0 (kPa) 35.29 0
Soilgeosynthetic interface 0.333 417.094 (m1) 0 (m2) 1251.408 (m1) 0 (m2)

Table 4
Shear strength in the foundation for Case 1

Depth from ground surface (m) Shear strength, qf (kPa)

0 8.49
1.0 8.49
1.8 4.725
5.0 13.125

Shear strength varies linearly between two contiguous shear strength values.

For the soilgeosynthetic interfaces, the yielding function is:


f jtj  sn Y h; 14

P the tangential and normal stresses, Y h tan dsp


where t and sn are, respectively,
is the hardening law, h jsp j the hardening parameter, d the interface friction
angle and sp the plastic tangential relative displacement.
Table 3 shows parameters c1 ; c2 ; c3 ; c4 and c5 adopted.
For the foundation clay, the correspondence between friction angle f0 (used in the
numerical analysis) and shear strength qf at the end of the embankment construction
(used in the limit equilibrium method) was determined by the application of Eq. (2).
As said in Section 2 (Borges and Cardoso, 2001; Borges, 1995), if consolidation is not
signicant, one can consider approximately p0 p00 (where p00 is the initial effective
volumetric stress), because the typical effective stress path during construction
corresponds to low variations of effective volumetric stress with signicant increases
of shear stress. Being c0 0; Table 4 shows the values of shear strength qf in the
foundation for Case 1, obtained by the application of the above considerations,
except from 0 to 1.0 m depth, where signicant increases of effective volumetric
stress, by consolidation, were observed during construction (about 14 kPa in average
terms, as shown by the element nite analysis). To also consider the consolidation
effect in the limit equilibrium method application, shear strength in the foundation,
from 0 to 1.0 m, was determined applying Eq. (2) at the middle of the layer (0.5 m
depth) with p0 p00 14 kPa. A transition layer is considered from 1 to 1.8 m.
Case 2 is a 28-day continuous construction of a 5 m height symmetric
1
embankment, with a 30.26 m crest width and 1:23 (V/H) inclined slopes. The
foundation is a 21 m thick saturated clay layer lying on a rigid and impermeable soil.
The clay is lightly overconsolidated to 7 m depth and normally consolidated from 7
to 21 m. The variation with depth of the at rest earth pressure coefcient, k0 ; and
OCR in the foundation are indicated in Table 5. One geosynthetic-reinforcement
J.L. Borges, A.S. Cardoso / Geotextiles and Geomembranes 20 (2002) 395421 407

Table 5
At rest earth pressure coefcient, k0 ; and over-consolidation ratio, OCR, in the foundation for Case 2

Depth (m) k0 OCR

04.5 0.7 2.43


4.57.0 0.70.5 2.431
7.021.0 0.5 1

Symmetry line

0 5m

Fig. 10. Finite element mesh for Case 2.

level is considered in the embankment soil at 0.9 m height. Geotechnical properties of


the embankment and foundation soils are considered with the same values as in Case
1 (indicated in Table 1), as well as properties of geosynthetic and soilgeosynthetic
interfaces. Taking into account the high permeability of the embankment soil, the
ground surface is a drainage boundary. Fig. 10 shows the nite element mesh used in
the numerical analysis; only 2D elements are represented.
Table 6 shows the values of shear strength qf in the foundation for Case 2,
obtained by the same assumptions as in Case 1, i.e. application of Eq. (2) with
p0 p00 from 7 to 21 m depth (where the numerical analysis showed that
consolidation is not signicant). Considering the increases of effective volumetric
stress, by consolidation, during construction in the foundation from 0 to 4.5 m depth
(about 63 kPa from 0 to 1 m and 16 kPa from 1 to 4.5 m, in average terms, as shown
by the element nite analysis), p0 p00 63 and p00 16 kPa were the values adopted
at middle of the corresponding layers. A transition layer is considered from 4.5 to
7 m.
Case 3 is a case history of one asymmetrical embankment on soft soils, constructed
1
up to failure (8.75 m height), with a 19 m crest width and 1:23 (V/H) inclined slopes,
which is documented in detail by Quaresma (1992). From the in situ and laboratorial
tests, Quaresma (1992) identied the following layers in the foundation: an
overconsolidated clay to 1 m depth, which becomes less overconsolidated from 1
to 3m; an organic soft soil from 3 to 7 m; and a lightly organic clay from 7 to 24 m
lying on a gravel layer. The element nite mesh used in the numerical analysis is
shown in Fig. 11; only 2D elements are represented. The lower boundary is at 24 m
depth and the ground surface is a drainage boundary. Fig. 12 illustrates the
408 J.L. Borges, A.S. Cardoso / Geotextiles and Geomembranes 20 (2002) 395421

Table 6
Shear strength in the foundation for Case 2

Depth from ground surface (m) Shear strength, qf (kPa)

01.0 33.0
1.04.5 16.0
4.57.0 16.018.375
7.021.0 18.37555.125

Fig. 11. Finite element mesh for Case 3.


EMBANKMENT HEIGHT (m)

9
8
7
6
5
4
3 Real sequence
2 Simulated sequence
1
0
5 15 25 35 135 145 155 165 175 185 195 205 215 225 235 245 255 265 275 285 295 305 315 325 335 345 355 TIME (hour)

Fig. 12. Embankment construction sequence for Case 3.

construction sequence: rstly, the construction of a 4 m height global platform was


undertaken and, secondly, embankment itself, up to 8.75 m height. One geosyn-
thetic-reinforcement level was considered in the embankment soil at 0.9 m height.
Adopting the average values dened by Quaresma (1992), Tables 7 and 8 summarize
the geotechnical properties of the foundation and embankment soils. Geosynthetic
elastic modulus is 1818 kN/m and normal and tangential stiffnesses of the soil
reinforcement interfaces are 2.0 107 and 1.28 104 kPa/m, respectively. Geosyn-
thetic strength is 200 kN/m. Adhesion and frictional angle for both upper and lower
soilgeosynthetic interfaces are 0 kPa and 30.961, respectively. Table 9 shows
parameters c1 ; c2 ; c3 ; c4 and c5 adopted in the numerical analysis for elasto-plastic
models that simulate the geosynthetic and soilgeosynthetic interfaces.

4.2. Analysis of the results

Fig. 13 shows the critical slip circles for Case 1and corresponding overall safety
factors (F )determined by the methodologies described in Section 2 using nite
J.L. Borges, A.S. Cardoso / Geotextiles and Geomembranes 20 (2002) 395421 409

Table 7
Geotechnical properties of the foundation for Case 3

Depth (m) l k G N f (1) n0 g (kN/m3) kx ky (m/s) qf su (kPa) k0 OCR


10
01 0.217 0.011 3.046 3.189 30 0.25 18 6.5 10 43 1.740 22.5
13 0.261 0.013 3.374 3.546 33 0.25 16.6 6.5 1010 31 0.712 3.067
37 0.651 0.065 7.052 7.459 32 0.25 13.5 7 1010 30 0.616 1.964
718 0.304 0.019 3.440 3.638 31 0.25 17 2 1010 32 0.534 1.270
1824 0.304 0.019 3.440 3.638 31 0.25 17.5 2 1010 32 0.534 1.270

Table 8
Geotechnical properties of the embankment for Case 3

Layer Height (m) l k G N f (1) n0 g (kN/m3) k0 OCR

1 01 0.021 0.002 1.8000 1.8132 33 0.20 17.2 0.455 1


2 18.75 0.010 0.001 1.7689 1.7752 35 0.30 21.9 0.426 1

Table 9
Values of the hardening law parameters for Case 3

c1 c2 c3 c4 c5

Geosynthetic 60 150 (kPa) 1 804 432.003 (kPa) 0 (kPa) 119.9915 0


Soilgeosynthetic interface 0.267 413.1437 (m1) 0 (m2) 1240.6716 (m1) 0 (m2)

element analysis (slip circle A) and in Section 3 by the limit equilibrium method (slip
circle B). For slip circles A and B, Table 10 shows the values of the overturning
moment (MO ), total resisting moment (MR ) and partial resisting moments due to
the reinforcement (MRR ) and due to the forces in the embankment (MRE ) and in the
foundation (MRF ). Fig. 14 and Table 11 illustrate the corresponding results for Case
2. In order to compare results obtained by the two methodologies, Tables 10 and 11
also show the results obtained by the limit equilibrium method along slip circle A:
The results of Case 3 are shown for two embankment heights: (i) 7 m, which still
corresponds to an equilibrium situation, before overall failure; and (ii) 8.75 m, the
embankment failure height, as reported by Quaresma (1992). According to all eld
measurements, the embankment began to slip approximately at 7.5 m height when
strength of the soil along the slip surface was reached. The slipping and the increase
of the embankment height caused the increase of the geosynthetic strains and
determined its failureand overall failureat 8.75 m height.
For 7 m height, the critical slip circles obtained by the two methodologies and
corresponding values of the moments are illustrated in Fig. 15 and Table 12. For
8.75 m height, as expected, because the embankment is no longer in an equilibrium
situation, an adequate numerical solution could not be reached for this height.
Fig. 16 shows the slip circle obtained by the limit equilibrium method (slip circle B)
410 J.L. Borges, A.S. Cardoso / Geotextiles and Geomembranes 20 (2002) 395421

2m

8.49
geosynthetic
8.49

4.725
Slip circle B 5m
(F=1.74)
Slip circle A
(F=1.88)

13.125
Variation of shear HARD
strength with depth (kPa) STRATRUM

Fig. 13. Critical slip circles for Case 1 and corresponding overall safety factors (F ) obtained from nite
element analysis (slip circle A) and determined by the limit equilibrium method (slip circle B).

Table 10
Slip circles for Case 1: overall safety factors (F ) and overturning and resisting moments

X0 Y0 R Overturning Resisting moment F


(m) (m) (m) moment, MO
(kN m/m) Total Foundation Embankment Geosynthetic
MR MRF MRE MRR
(kN m/m) (kN m/m) (kN m/m) (kN m/m)

Slip circle Aa 1.0 1.0 5.74 553.71 1039.75 754.11 85.64 200.00 1.88
Slip circle Bb 1.0 1.0 5.24 521.66 907.37 647.12 60.25 200.00 1.74
Slip circle Ab 1.0 1.0 5.74 631.38 1114.70 849.25 65.45 200.00 1.77
a
Determined by the element nite method.
b
Determined by the limit equilibrium method.

and the slip surface estimated from eld measurements (Quaresma, 1992). Table 13
illustrates the corresponding values of the moments for slip circle B.
For the three embankments, all partial safety factors, Zg ; Gc ; Gf ; Gsu ; Gr ; Ga and Gd ;
were imposed equal to 1, and the geosynthetic force was considered acting
horizontally (the most conservative assumption).
The analysis of the results allows to point out that the slip circles obtained by
the two methodologies are similar (except for Case 3 at 7m height, where a
larger discrepancy can be observed), as well as the values of the overall safety
factor.
In spite of this overall similaritywhich is important in practical terms and can be
considered as an adequate accuracy of the limit equilibrium method to assess the
overall safety factor in this kind of problems, the values shown in Tables 1012 point
out that the most important differences between the two methods are related to
the values of the overturning and resisting moments. In fact, these moments are
substantially smaller in the methodology based on the numerical analysis (due to the
J.L. Borges, A.S. Cardoso / Geotextiles and Geomembranes 20 (2002) 395421 411

EMBANKMENT
5m

33.00 geosynthetic

16.00
Slip circle B
18.375
(F=1.15)
Slip circle A 21 m
(F=1.22)
SOFT SOILS

55.125
Variation of shear HARD STRATUM
strength with depth (kPa)

Fig. 14. Critical slip circles for Case 2 and corresponding overall safety factors (F ) obtained from nite
element analysis (slip circle A) and determined by the limit equilibrium method (slip circle B).

Table 11
Slip circles for Case 2: overall safety factors (F ) and overturning and resisting moments

X0 Y0 R Overturning Resisting moment F


(m) (m) (m) moment, MO
(kN m/m) Total Found. Embank. Geosyn.
MR MRF MRE MRR
(kN m/m) (kN m/m) (kN m/m) (kN m/m)

Slip circle Aa 2.0 5.0 12.94 5694.88 6966.52 5894.40 252.12 820.00 1.22
Slip circle Bb 3.0 8.0 15.04 9540.50 10972.20 8368.50 1183.70 1420.00 1.15
Slip circle Ab 2.0 5.0 12.94 7667.20 9133.30 7360.60 952.73 820.00 1.19
a
Determined by the element nite method.
b
Determined by the limit equilibrium method.

forces in the embankment and in the foundation), as can be seen from the results of
the two methodologies along slip circle A. This indicates that inside of the soil mass
some stress redistribution occurs causing an effect that decreases overturning and
resisting moments. In fact, as shown by Borges (1995) and Borges and Cardoso
(2001), during the construction period, the external forces (weight of ll soil), which
are equilibrated by the internal stresses in the materials (soil and geosynthetic),
determine that critical state (strength of the soil) is reached in some areas of the
foundation earlier than in others. For those areas, additional external forces due to
the increase of the embankment height cannot be equilibrated by stress increments.
This determines that stresses have to migrate to areas where critical state has not
412 J.L. Borges, A.S. Cardoso / Geotextiles and Geomembranes 20 (2002) 395421

Embankment 7m

43
31

30 geotextile

24 m

32 Soft soils
Slip circle A
(F=1.15) Slip circle B
(F=1.19)
Variation of shear Hard stratum
strength with depth (kPa)
Fig. 15. Case 3 Critical slip circles for 7 m height and corresponding overall safety factors (F ) obtained
from nite element analysis (slip circle A) and determined by the limit equilibrium method (slip circle B).

Table 12
Case 3Slip circles for 7 m height: overall safety factors (F ) and overturning and resisting moments

X0 Y0 R Overturning Resisting moment F


(m) (m) (m) moment, MO
(kN m/m) Total Found. Embank. Geosyn.
MR MRF MRE MRR
(kN m/m) (kN m/m) (kN m/m) (kN m/m)

Slip circle Aa 4.80 8.0 21.83 20638.96 23793.62 22255.68 117.94 1420.00 1.15
Slip circle Bb 3.80 15.0 31.47 63870.00 75754.50 68087.00 4847.50 2820.00 1.19
Slip circle Ab 4.80 8.0 21.83 34166.03 42695.51 36451.01 4824.50 1420.00 1.25
a
Determined by the element nite method.
b
Determined by the limit equilibrium method.

been reached yet. This effect is caught by the nite element analysis but it is not
obviously considered in the simplied hypotheses that support the limit equilibrium
method.
The results also indicate that the differences between the values of the moments in
the two methods are larger for embankments with smaller values of the overall safety
factor, which is the case of Cases 2 and 3 compared with Case 1. This clearly
corroborates that smaller values of the overall safety factor are obviously related to
larger stress redistribution inside of the soil mass, which implies larger differences
between the moments obtained by the two methodologies.
J.L. Borges, A.S. Cardoso / Geotextiles and Geomembranes 20 (2002) 395421 413

Embankment

43
31

30
geotextile

Slip surface 24 m
estimated
from field Soft soils
measurements Slip circle B
32
(Quaresma,1992) determined by
the limit
equilibrium method (F=0.99)
Variation of shear Hard stratum
strength with depth (kPa)

Fig. 16. Case 3 Critical slip circle for 8.75 m height and corresponding overall safety factor (F ) determined
by the limit equilibrium method (slip circle B) and slip surface estimated from eld measurements.

Table 13
Case 3 Slip circle for 8.75 m height determined by the limit equilibrium method: overall safety factor (F )
and overturning and resisting moments

X0 Y0 R Overturning Resisting moment F


(m) (m) (m) moment, MO
(kN m/m) Total Found. Embank. Geosyn.
MR MRF MRE MRR
(kN m/m) (kN m/m) (kN m/m) (kN m/m)

Slip circle B 4.80 14.0 28.80 65116.60 64784.50 56419.10 5745.30 2620.10 0.99

5. Conclusions

Overall stability of geosynthetic-reinforced embankments on soft soils was


analysed with two different methodologies described in the paper: (i) application
of a numerical model based on the nite element method (Borges, 1995); and (ii) use
of a limit equilibrium method, based on the assumptions proposed by Kaniraj and
Abdullah (1993) and Borges (1995).
These two methodologies were applied on three geosynthetic-reinforced embank-
ments on soft soils; one of them is a case history that was constructed up to failure.
Considering the analysis of the results, the following conclusions can be
formulated:
The critical slip circles obtained by the two methodologies are similar for the three
cases, as well as the values of the safety factor, which determines an adequate
accuracy of the limit equilibrium method to assess overall safety factor in reinforced
embankments on soft soils.
414 J.L. Borges, A.S. Cardoso / Geotextiles and Geomembranes 20 (2002) 395421

The most important differences between the two methodologies are related to the
values of the overturning and resisting moments, which are smaller in the
methodology based on the numerical analysis.
These differences which are larger in embankments with smaller values of the
safety factor are related to some stress redistribution that occurs inside of the soil
mass, simulated by the nite element method but not considered in the simplied
hypotheses of the limit equilibrium method.

Appendix A. Equations of the limit equilibrium method

A.1. Design overturning moment

The design overturning moment MOd is contributed by the soil mass ABCEI
(Fig. 7). As deduced by Kaniraj and Abdullah (1993), and adding partial safety
factors, it is given by
He
MOd Zg K1 for Xe pn1 H b; A:1
2
 
gHe gH  He
MOd Zg K1 K2 for Xe > n1 H b; A:2
2 2
where
He2
K1  n21 1 n1 X0 Y0 He
3
XI  2XI X0  2YI Y0 YI2 ;
2
A:3

K2 b n1 n2 Hb n1 H  n2 He  2X0
n22  n21 2
n1 n2 X0 H He H He2 He H; A:4
3
where XI and YI are the coordinates of point I (in this case YI 0); instead of this
point, the coordinates of any other point of the circle can be taken. Xe and He are the
coordinates of point E; extremity of the slip circle as shown in Fig. 7 (0pHe pH).

A.2. Design resisting moment

The design resisting moment MRd depends on the materials properties and can be
expressed as
MRd MREd MRFd MRRd ; A:5
where MREd is the moment due to resisting forces in the embankment along slip arc
EI; MRFd the moment due to resisting forces in the foundation along slip arc MJI ;
MRRd the moment due to the reinforcement resisting force that can be mobilised at
point G:
J.L. Borges, A.S. Cardoso / Geotextiles and Geomembranes 20 (2002) 395421 415

A.2.1. Design resisting moment MREd

MREd can be given by (Low, 1989):


  
c tan f
MREd R lgHe ye R; A:6
Gc Gf
where R is the radius of the slip circle, l is an averaging coefcient for normal stress
along arc EI; He is the Y -coordinate of point E (intersection of slip circle with slope
or crest) and ye (angle IOE in Fig. 7) is computed by

Y0 Y0  He
ye sin1  sin1 : A:7
R R
As proposed by Low (1989), limiting its value approximately to 0.5, l can be
expressed by

0:02n1
l 0:19 A:8
R  Y0 =He
where n1 denes slope inclination.

A.2.2. Design resisting moment MRFd

The design resisting moment MRFd is computed considering the foundation


divided into a number of layers with the variation of undrained shear strength su
being linear in each layer along vertical direction (see Fig. 17).

0 ( X0 , Y0 )

R
Ai

Bi

(0,0) X

su -A
i
A 'i Ai
layer i
B i' Bi su - B
i

Fig. 17. Linear variation of undrained shear strength along vertical direction in each foundation layer.
416 J.L. Borges, A.S. Cardoso / Geotextiles and Geomembranes 20 (2002) 395421

Therefore, the partial resisting moment MRFdi due to the forces in the ith layer
(acting along arc Ai Bi and its symmetric A0i B0i ) is given by
Z yAi
su
MRFdi 2R R dy: A:9
yBi Gsu

Integrating this equation (note that su does not vary linearly along arc Ai Bi ) and
considering contribution of all layers, MRFd is obtained by (Borges, 1995):
X
N
MRFd 2R2 mi Rsin yAi  sin yBi
i1
1
mi Y0 bi yAi  yBi  ; A:10
Gsu
where
Y0  YAi
yAi cos1 ; A:11
R
Y0  YBi
yBi cos1 ; A:12
R
suAi  suBi
mi  ; A:13
YAi  YBi
bi suAi mi YAi A:14
with YAi ; YBi being the Y -coordinates at top and bottom of ith layer, suAi ; suBi the
undrained shear strength at top and bottom of ith layer, N the number of layers in
the foundation intersected by the slip circle.

A.2.3. Design resisting moment MRRd

Assuming that the resisting force T due to the geosynthetic acts at an angle a to
the horizontal (as said above, a 0 is the most conservative assumption), MRRd is
given by (Fig. 7)
MRRd TR cosyr  a; A:15
where
Y0  Yc
yr cos1 A:16
R
and Xc and Yc are the coordinates of the cut point G:
T is dened by the equation
 
Tg
T min ; Tp ; A:17
Gr
where Tg is the geosynthetic strength and Tp the maximum force that can be
mobilised at point G taking into account the strength of soilgeosynthetic interfaces
reduced by their partial factors.
J.L. Borges, A.S. Cardoso / Geotextiles and Geomembranes 20 (2002) 395421 417

b
W2

n1 W1 n2
1 1 H
M N
G( X c , Y c )

l1 l2 Z

geosynthetic

Fig. 18. Lengths of soilgeosynthetic interfaces conditioning maximum force mobilised at point G:

Because equilibrium has to be veried, Tp is calculated by the equation


Tp minTp1 ; Tp2 ; A:18
where Tp1 and Tp2 are computed from shear stress of soilgeosynthetic interfaces
along segments MG and GN (Fig. 18). Therefore,

al au tan dl tan du
Tp1 l 1 W1 ; A:19
Ga Gd

al au tan dl tan du
Tp2 l 2 W2 ; A:20
Ga Gd
where al ; dl are the adhesion and frictional angle of the lower soilgeosynthetic
interface, au ; du the adhesion and frictional angle of the upper soilgeosynthetic
interface, l1 the length of segment MG (Fig. 18), l2 the length of segment GN
(Fig. 18), W1 the weight of ll soil on the segment MG; W2 the weight of ll soil on
the segment GN:
If g is the unit weight of the embankment material, then W1 and W2 can be given
by
Xc  n1 Yc 2
W1 0:5 g for Xc pn1 H; A:21
n1

W1 0:52Xc  n1 H Yc H  Yc g for n1 HoXc on1 H b; A:22

l n1 Yc  Xc 2
W1 W  0:5 g for Xc Xn1 H b; A:23
n2

W 2 W  W1 ; A:24
418 J.L. Borges, A.S. Cardoso / Geotextiles and Geomembranes 20 (2002) 395421

where W is the total weight of ll soil on the geosynthetic and l is the geosynthetic
length (length of segment MN in Fig. 18) and can be expressed by
W 0:5l bH  Yc g; A:25

l n1 n2 H  Yc b: A:26
If there are N number of reinforcement levels, Eqs. (A.17)(A.26) are applied to
each reinforcement, and MRRd is the sum of all moments MRRdi due to those
reinforcements
X
N X
N
MRRd MRRdi Ti R cosyri  ai : A:27
i1 i1

A.3. Sequence of computation

Taking into account the above equations, the rotational stability analysis of a
geosynthetic-reinforced embankment on soft soils can be computed as the following
sequence:
1. Reading of data (see Fig. 19):
(a) Geometry of the problem: n1 ; n2 ; b; H and D;
(b) Data on the potential slip circles: X0 min ; Y0 min ; X0 max ; Y0 max ; Dxy and Dr (for
each point O of the mesh of circle centres, several concentric circles are analysed
with radius varying from a minimum value Rmin to a maximum value Rmax ; Rmin
can be considered as the longer value of the distances from point O to the two
extremities of the segment AB, and Rmax as the smaller value of the distances

Dxy Y

O (X0max , Y 0max )
Dxy

(X0min , Y 0min ) B C
n1 n2
1 geosynthetic 1 H
Z
A X

Hard stratum
Fig. 19. Geometry of reinforced embankment on soft soils.
J.L. Borges, A.S. Cardoso / Geotextiles and Geomembranes 20 (2002) 395421 419

from point O to point Z and to the hard stratum under the soft soils; Dr is radius
length increment for concentric circles);
(c) Properties of the foundation and embankment materials:
* fill soil: unit weight, cohesion and friction angle;

* foundation soils: number of levels where s is indicated (with linear variation


u
between two contiguous levels along vertical direction); value of su and Y -
coordinate for each level;
* geosynthetic: number of reinforcement levels and, for each level,
Y -coordinate, geosynthetic strength, adhesion and frictional angle of the
lower and upper soilgeosynthetic interfaces;
(d) Partial safety factors: Zg , Gc ; Gf ; Gsu ; Gr ; Ga andGd (if all these factors are
imposed equal to 1, an overall safety factor is obtained);
(e) Coefficient ai =yri : denition of geosynthetic force inclination;

2. For each potential slip circle, calculate:


(a) Design overturning moment MOd (Eqs. (A.1)(A.4));
(b) Design resisting moment MREd (Eqs. (A.7), (A.8) and (A.6));
(c) Design resisting moment MRFd (Eqs. (A.11)(A.14) and (A.10));
(d) Design resisting moment MRRd (Eqs. (A.21)(A.26), (A.19), (A.20), (A.18),
(A.17), (A.16) and (A.15) (or A.27));
(e) Total design resisting moment MRd (Eq. (A.5));
(f) Value of E given by
E MRd =MOd X1 A:28
which is the excess of safety in relation to that assumed by the use of partial
safety factors (if these factors are all equal to 1, E is the overall safety factor);

3. Selection of the critical slip circle corresponding to the smallest value of E


obtained.

References

Borges, J.L., 1995. Geosynthetic-reinforced embankments on soft soils. Analysis and design. Ph.D. Thesis
in Civil Engineering, Faculty of Engineering, University of Porto, Portugal, 1995 (in Portuguese).
Borges, J.L., Cardoso, A.S., 1997. Stability of geosynthetic-reinforced embankments on soft soils. Revista
Geotecnia, 80, 4357 (in Portuguese).
Borges, J.L., Cardoso, A.S., 1998. Numerical simulation of the pqy critical state model in embankments
on soft soils. Revista Geotecnia, 84, 3963 (in Portuguese).
Borges, J.L., Cardoso, A.S., 2001. Structural behaviour and parametric study of reinforced embankments
on soft clays. Computers and Geotechnics 28/3, 209233.
Brakel, J., Coppens, M., Magdenberg, A.C., Risseuw, P., 1982. Stability of slopes constructed with
polyester reinforcing fabric, test section at AlmereHolland, 79. Proceedings of the Second
International Conference on Geotextiles, Las Vegas, Vol. 3, 727732.
Britto, A.M., Gunn, M.J., 1987. Critical Soil Mechanics Via Finite Elements. Ellis Horwood Limited,
England.
420 J.L. Borges, A.S. Cardoso / Geotextiles and Geomembranes 20 (2002) 395421

Broms, B.B., 1977. Polyester fabric as reinforcement in soil. Proceedings of the International Conference
on the Use of Fabrics in Geotechnics, Paris, Vol. 1, pp. 129135.
Duncan, J.M., Wong, K.S., 1984. STABGM: a computer solution for slope stability analysis with circular
slip surfaces and geogrid reinforcement. Microcomputer version. A Tensar Corporation Users
Manual.
Fowler, J., 1982. Theoretical design considerations for fabric reinforced embankments. Proceedings of the
Second International Conference on Geotextiles, Las Vegas, Vol. 2, pp. 665670.
Haliburton, T.A., 1981. Use of engineering fabric in road and embankment construction. Seminar on the
Use of Synthetic Fabrics in Civil Engineering, November 19, Toronto, pp. 6694.
Huisman, M.J.H., 1987. Design guideline for reinforced embankments on soft soils using Stabilenka
reinforcing mats. Enka Technical Report, Arnhem.
Humphrey, D.N., 1986. Design of reinforced embankments. Report FHWA/IN/JHRP-86/17 Joint
Highway Research Project, School of Civil Engineering, Purdue University, West Lafayette, Indiana,
423pp.
Ingold, T.S., 1982. An analytical study of geotextile reinforced embankments. Proceedings of the Second
International Conference on Geotextiles, Las Vegas, Vol. 2, pp. 683688.
Ingold, T.S., 1983. Some factors in the design of geotextile reinforced embankments. In Improvement of
Ground. Proceedings of the Eighth European Conference on Soil Mechanics and Foundation
Engineering, Helsinki, Vol. 2, pp. 503508.
Jewell, R.A., 1982. A limit equilibrium design method for reinforced embankments on soft foundations.
Proceedings of the Second International Conference on Geotextiles, Las Vegas 3, 671676.
Kaniraj, S.R., Abdullah, H., 1993. Rotational stability of narrow-crested reinforced embankments on soft
soils. Geotextiles and Geomembranes 12, 599614.
Kwok, C.M., 1987. Finite elements studies of reinforced embankments on soft ground. Ph.D. Thesis,
University of Shefeld, Shefeld.
Lambe, T.W., Whitman, R.V., 1969. Soil Mechanics. Wiley, New York.
Lewis, R.W., Schreer, B.A., 1987. The Finite Element Method in the Deformation and Consolidation of
Porous Media. Wiley, New York.
Low, B.K., 1989. Stability analysis of embankments on soft ground. Journal of Geotechnical Engineering,
Division, ASCE 115 (2), 211227.
Milligan, V., La Rochelle, P., 1984. Design methods for embankments over weak soils. Symposium on
Polymer Grid Reinforcement in Civil Engineering, Institute of Civil Engineers, London, UK, Paper
No. 3.4.
Mylleville, B.L.J., Rowe, R.K., 1991. On the design of reinforced embankments on soft brittle clays.
Geosynthetics 91 Conference, Atlanta, USA, pp. 395408.
Owen, D.R.J., Hinton, E., 1980. Finite Elements in Plasticity: Theory and Practice. Pineridge Press
Limited, Swansea.
Prevost, J.H., Hoeg, K., 1975. Effective stressstrain-strength model for soils. Journal of Geotechnical
Engineering, Division, ASCE 101 (GT3), 259278.
Quaresma, M.G., 1992. Behaviour and modelling of an embankment over soft soils reinforced by
geotextile. Ph.D. Thesis, Universite Joseph Fourier, Grenoble I (in French).
Quast, P., 1983. Polyester reinforcement fabric mats for the improvement of the embankment stability. In
improvement of ground. Proceedings of the Eighth European Conference on Soil Mechanics and
Foundation Engineering, Helsinki, Vol. 2, pp. 531534.
Rowe, R.K., 1984. Reinforced embankments: analysis and design. Journal of Geotechnical Engineering,
ASCE 110 (2), 231246.
Rowe, R.K., Mylleville, B.L.J., 1990. Implications of adopting an allowable geosynthetic strain in
estimating stability. Proceedings of the Fourth International Conference on Geotextiles, Geomem-
branes and Related Products, The Hauge, Netherlands, pp. 131136.
Rowe, R.K., Soderman, K.L., 1987. Stabilisation of very soft soils using high strength geosynthetics: the
role of nite elements analysis. Geotextiles and Geomembranes 6 (1), 5380.
Russell, D., 1992. Finite element analysis of embankments on soft ground incorporating reinforcement
and drains. Ph.D. Thesis, University of Shefeld, Shefeld.
J.L. Borges, A.S. Cardoso / Geotextiles and Geomembranes 20 (2002) 395421 421

Soderman, K.L., 1986. The behaviour of geotextile reinforced embankments. Ph.D. University of Western
Ontario, Ontario.
Tensar Corporation, 1982. Designing with Tensar. Morrow, Georgia and Toronto, Ont., 17pp.
Thomas, J.N., 1984. An improved acelerated initial stress procedure for elasto-plastic nite element
analysis. International Journal for Numerical and Analytical Methods in Geomechanics 8, 359379.
Yeo, K.C., 1986. Simplied foundation data to predictors. Proceedings of the Prediction Symposium on a
Reinforced Embankment on Soft Ground, Kings College, London.

You might also like