Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
PII:
S0266-
George A. Athanasopoulos
Department of Civil Engineering, School of Engineering, University of Patras, GR-26500,
Patras, Greece
ABSTRACT
The results of direct shear tests on near-saturated silty clay samples reinforced
with woven and nonwoven geotextiles are presented and analyzed in terms of
the strength increase, shear and volumetric deformation of reinforced soil, and
soil/reinforcement interface bond development. The analyses--which are
based on total stresses--indicate that the inclusion of nonwoven geotextiles
resulted in a significant strength increase of the wet cohesive soil. The inclusion of woven geotextiles, however, did not offer any strength increase. This
differentiation of behaviour was attributed to the nil water transmissivity of
the woven geotextile. By further analyzing the test results it was found that the
magnitude of the interface bond increased with the transmissivity of the
geotextile and varied with the normal interfacial stress. By utilizing a simple
soil-reinforcement interaction model, values of interface friction angles were
computed for different assumed values of adhesion efficiency between cohesive
soil and geotextile. 1997 Elsevier Science Ltd.
NOMENCLATURE
A t
As
C
Cot
e
Gs
Ko
620
LL
PL
W
Wopt
z
7t
~'dmax
6
At
(7
(7v
T
G. A. Athanasopoulos
INTRODUCTION
In geosynthetic-reinforced soil applications, such as retaining walls, slopes
and embankments, the mobilisation of interface bond between backfill
material and reinforcement is of paramount importance. During the early
development of soil-reinforcement techniques the proposed backfill materials were granular, especially in the case of metallic reinforcements. In the
case of geosynthetics, however, several successfull applications have demonstrated the great potential of using cohesive soils as backfill materials
(Delmas et al., 1988, 1992; Wayne and Wilcosky, 1995; Tatsuoka et al., 1996;
Kojima et al., 1996). As a result of the growing interest in utilizing on-site
cohesive soils in reinforced soil structures (associated with significant cost
reduction) the research on the subject of the mechanical behaviour of
geosynthetic-reinforced cohesive soils as well as on the subject of the
geosynthetic/cohesive soil interface behaviour has intensified during the last
decade (Ling and Tatsuoka, 1994; Bergado et al., 1993; Porbaha and Goodings, 1996; Zornberg and Mitchell, 1994; Mitchell and Zornberg, 1995). This
interface behaviour is described by a friction angle 6 and an adhesion c~
which are usually expressed as fractions of the corresponding shear strength
parameters of the soil material, q5 and c. The corresponding ratios are usually
termed efficiencies, i.e. the friction efficiency = ~/4~ (or tan 6/tan q~) and the
adhesion efficiency = c~/c (Ingold, 1994; Koerner, 1994).
In this paper the results are presented of a preliminary experimental
investigation of the mechanical behaviour of geosynthetic-reinforced cohesive soil and of the corresponding interface behaviour. The tests were carried
out using a laboratory direct shear apparatus and following the unit-cell
approach (Ling and Tatsuoka, 1994). The effects of several parameters such
621
as the type of geotextile, its water transmissivity and the normal interfacial
stress are discussed and conclusions are drawn regarding the practical applications of reinforced cohesive-soil structures.
REVIEW OF PREVIOUS W O R K
The behaviour of clays reinforced with geotextiles was investigated at the
beginning of the 1980s by triaxial tests, direct shear tests, pull-out tests and
physical models (Ingold, 1981, 1983; Ingold and Miller, 1983). Values of
adhesion efficiency were very low in the pull-out mode (0-18) and relatively
high (0-89) in the direct shear mode. The triaxial tests indicated that under
undrained conditions the use of geotextiles with no capability of in-plane
flow actually reduced the shear strength of cohesive soils especially when the
degree of saturation was greater than 70%. Under drained conditions,
however, the shear strength of reinforced clay was found to increase significantly.
Subsequent investigations by direct shear and pull-out testing showed that
the adhesion efficiency may range from 0-52 to 1-65 for nonwoven and 0.37
to 1.08 for woven geotextiles (Fourie and Fabian, 1987). Bouazza and
Djafer-Khodja (1994) reported values of efficiencies equal to 1.23 for friction
and 0-68 for adhesion between a peat with water content equal to 250% and
a nonwoven geotextile. They also reported that the values of efficiencies were
reduced with increasing values of normal interfacial stress. Values of adhesion efficiencies greater than 0.50 were also reported by Nataraj et al. (1995)
for clay/geotextile interfaces. The significant effect of the moisture content of
fiber reinforced cohesive soils on their shear strength was reported by
Alwahab and Al-ourna (1995) who found that for water contents higher by
5% or more than the optimum water content of the cohesive soil, there was
no increase in shear strength.
The interaction between geogrids and cohesive soils has also been investigated by direct shear and pull-out tests. Fourie and Fabian (1987) have
reported values of adhesion efficiencies ranging from 0.46 to 1.02, whereas
Bergado et al. (1993) conclude that the apertures of a geogrid help in developing shear resistance equal to or greater than the cohesive soil resistance.
Farrag and Griffin (1993) have also found that an increase of water content
above the optimum value derived from compaction tests, results in a
decrease in the pull-out resistance.
The above-mentioned experimental evidence for the dependence of friction
and adhesion efficiencies on the interfacial normal stress implies that the
shearing resistance along the soil-geosynthetic interface cannot be described
adequately by the linear relationship
622
G. A. Athanasopoulos
z = c, + a t a n 6
(1)
"d
~
-
hyperbolic relationship
623
EXPERIMENTAL PROCEDURE
As was mentioned in a previous section, the unit-cell testing approach was
used in this study, i.e. it was assumed that a unit-cell represented an element
of the geosynthetic-reinforced soil. In our case, the unit-cell coincided with
the reinforced soil specimen in the direct shear box. Fakher and Jones (1996)
have recently summarised the advantages and disadvantages of unit-cell tests
and have concluded that they represent a quick, inexpensive and accurate
way of modelling the reinforced soil mechanisms.
The direct shear tests described in this investigation were conducted in
a square shear box measuring 6 3 m m x 6 3 m m . The geotextile was placed
perpendicular to the plane of shear as shown in Fig. 2. A perpendicular
direction of the reinforcement sheet was chosen in order to facilitate its
installation in the shear box, although it is known that it does not
represent the optimum orientation for obtaining the maximum reinforcing effect [a most recent publication on this subject is the one by
Fakher and Jones (1996)]. The findings of the present investigation are
not affected, however, by this deviation from the optimum orientation of
reinforcements since they have mainly a relative character, i.e. they
compare the behaviour of different types of geotextiles. This type of
unit-cell testing has been used in the past by the author for testing
tY
6.3 c m
Fig. 2. D i r e c t s h e a r t e s t o n g e o t e x t i l e - r e i n f o r c e d c o h e s i v e soil.
624
G. A. Athanasopoulos
625
19.0
Wop t = 13%
Ydmax = 18.55 k N / m 3
Z
.~
18.0
l
A S T M D 6 9 8 - 7 8 ( M e t h o d A)
17.0
~
5.0
i
10.0
Water
15.0
content,
20.0
Fig. 3. Results of compaction tests on the cohesive soil material used in this investigation.
of geosynthetics and should not be taken as the values pertaining to t h e
pieces of reinforcement used in the tests.
The rate of shear used in the tests was 0.4mm/min. This rate is in the
range recommended in the literature (Smith and Criley, 1995; Ingold, 1994).
For the cohesive soil used in the tests, this rate of shearing is taken t o
represent undrained loading conditions. All tests were run following immediately the placement and compaction of soil in the shear box; therefore the
tests are considered unconsolidated-undrained and represent mainly the
short-term conditions developed in the corresponding field applications. It is
recognised, however, that the test results also reflect, to a smaller degree, the
behaviour of reinforced soil under anisotropic consolidation conditions. The!
values of normal stress in the direct shear tests ranged from 25 kPa to
600 kPa.
G. A. Athanasopoulos
626
6 6 6 ~
*"
6 6 6 ~
0
~ < ~
627
G. A. Athanasopoulos
628
300
Unreinforced
tp = 14.8 , c = 43.2 kPa
200
7
o
100
0
0
100
200
300
400
500
600
700
~, k N / m 2
I+50kPa
--I---100kPa ~ 1 5 0 k P a
0 200kPa ~ 3 0 0 k P a ]
- X - 4 0 0 kPa --~-400 (c~ctv)-0---500 kPa @ 600 kPa
'
"
50kPa
--II--100kPa
~k 150kPa
--O---200kPa
--~-- 300kPa
600 kPa
0.5
O~
,-&
-0.5
-1
-1.5
Shear displacement, mm
F i g . 4. R e s u l t s o f d i r e c t s h e a r t e s t s o n u n r e i n f o r c e d
silty c l a y
)k !
629
200
j J
100
J
100
I
200
l
300
I
400
L
500
600
700
~, kN/m 2
50 kPa
I ----41---400 kPa
---II-----75 kPa
t
500 kPa
1.
100 kPa
[]
600 kPa
200 kPa
300 kPa I 1
I
,g
E
E
0.5
50 kPa
--IIt-- 75 kPa
400 kPa
- - A - 500kPa
100 kPa
200 kPa
~(
300 kPal
e~
-0.5
-1.5
1
Shear displacement,
Fig. 5. Results
of direct shear
mm
with Terram
500.
G. A. Athanasopoulos
630
200
Z
~d, ~
lO0
I
100
I
200
I
300
I
400
I
500
1
600
t
700
O, k N / m 2
I
50kPa - - i ~ l O O k P a
500 kPa
I 600 kPa
200kPa - - 3 0 0 k P a
~ 400kPa]
[
g.
2
r~
+
0.5
50 kPa
[ --l--
400 kPa
100 kPa
-..-0.-- 500kPa
200 kPa
300 kPa [
600 kPa
i
i
-1.5
Shear d i s p l a c e m e n t , m m
631
200
J
~
~ ~ ~ ~ ~ ~L~e~,~ ~ced
I
100
I
100
I
200
I
300
o, kN/m
@
.
25kPa
@50kPa
400 kPa
O ' 500 kPa
.
I
400
I
500
t
600
i
700
150kPa
I " 600 kPa
--{~-200kPa
.
@,300kPa
]
[
g.
2:
01
1,
I
0.5
25 kPa
--WI--- 50 kPa
400 kPa
----0----500kPa
150 kPa
200 kPa
- - ~ - - 300 kPa 1
600 kPa
01
t~
e~
u~ -0.5
e~
L~
0~
-t
-1.5
~
0
t
1
~
2
~
3
-4
Shear displacement, mm
Fig. 7. Results of direct shear tests o n silty clay reinforced with T e r r a m 4000.
G. A. A thanasopoulos
632
Silty clay (CL)
300
A m o c o 4545
200
2:
t-;
100
100
200
300
400
500
600
7OO
~, kN/m 2
50 kPa
I00 kPa
,Ik
200 kPa
400 kPa
500 kPa
[]
600 kPa
50kPa
lOOkPa
-----Am 200kPa
400 kPa
500kPa
e-,
r~
I ~
0.5
- - , ~ - - 300kPa
t~
~, -0.5
"~
0)
-1
Shear displacement, mm
Fig, 8. Results o f direct shear tests on silty clay reinforced with A m o c o 4545.
633
200
Z
100
100
200
300
400
500
600
700
~, k N / m z
--'-50
kPa
+ 4 0 0
kPa
-41-100
kPa
-*-600
kPa
-~r-200
kPa
-o-300
kPa
+ 4 0 0
kPa
t~
t-
+
--
0.5
--
50 kPa
400 kPa
----~...~-500 kPa
100 kPa
200 kPa
300 kPa
600 kPa
.3
o=
or
*-
--
A A
A.
-0.5
"5
-1.5
0
Shear displacement, mm
Fig. 9. Results of direct shear tests on silty clay reinforced with Nicoton 66448.
634
G. A. Athanasopoulos
strong curvature and eventually intersects the failure envelope of unreinforced soil (Fig. 5). This type of behaviour indicates a negligible (or zero)
increase of shearing strength at high values of normal stress, a fact that could
be explained by the reasoning presented later in this section.
For the Nicolon woven geotextile the failure envelope of reinforced soil
almost coincides with the failure envelope of unreinforced soil. This means
that this geotextile did not offer any strength increase to soil samples
(Fig. 9). This differentiation of behaviour can be attributed to the different
water transmissivities of nonwoven and woven geotextiles. The pore water
pressures developed during shearing along the soil-geotextile interface can be
rapidly dissipated in the case of nonwoven geotextiles due to the high transmissivity of these fabrics. Thus the interface shearing conditions in this case
can be characterized as being drained resulting in a substantial amount of
interface friction in addition to a possible interface adhesion. In the case of
the woven geotextile, however, its practically nil in-plane flow capability
precludes any rapid dissipation of pore water pressures and thus the shearing
conditions remain undrained. Under such conditions, and when the pore
water pressures are high enough, it becomes very difficult or impossible for
any friction resistance to be developed along the soil/reinforcement interface.
Regarding the adhesion component it is expected that it will have a rather
low value which could not provide an overall strength increment high
enough to be detected by the small size direct shear tests of this investigation.
It should also be pointed out that the water transmissivity of nonwoven
geotextiles depends on the value of normal stress acting on the interface
(Koerner, 1994). Higher values of normal interfacial stress result in a
decrease in the thickness and consequently of the transimissivity of the
geotextile thus favouring the development of an undrained state of stress and
of higher pore water pressures, similar to the case of woven geotextiles. This
behaviour could explain the curved or the flattened shape of the failure
envelopes of reinforced soil, shown in Figs 5-7. It may be argued, then, that
in the case of moist (or near-saturated) cohesive soils reinforced with
geosynthetics, a bilinear (or curved in general) form of failure envelope,
expressed in terms of total stresses, does not necessarily mean a transition
from the slip-type to the stretching-type of failure. It may, instead, indicate a
continuous decrease of water transmissivity of geosynthetic under increasing
values of normal interfacial stress which results in the generation of pore
water pressures which in turn inhibit any further increase of effective normal
interfacial stresses and corresponding strength increase.
The plots of Figs 5-9 indicate that the cohesion intercept of the reinforced
soil is affected--though not significantly--by the geotextile reinforcement. It
should be recognised, however, that in order to be able to draw conclusions
on this aspect, direct shear tests under very low values of normal stress
635
The diagrams of Figs 5-9 include plots of the shear stress versus shear
displacement and of vertical deformation of specimen versus shear displacement for all values of normal stresses applied in the direct shear tests.
Regarding the shear stress versus shear displacement behaviour of the reinforced cohesive soil specimens, an inspection of the pertinent plots indicates
that the peak shearing restistance of reinforced soil is developed at a higher
shear displacement compared to the unreinforced soil. This behaviour is similar
to the behaviour observed in reinforced granular soils (Athanasopoulos, 1993).
To facilitate a comparison between reinforced and unreinforced soil, regarding
the development of shearing stiffness, the plots of Fig. 10 were prepared. In
these plots the shear stress versus shear displacement curves of unreinforced and
reinforced cohesive soil are shown for the three geotextiles used in the tests and
for a low (100 kPa) and high (500 kPa) value of vertical stress, av. It is observed
that the Terram 4000 geotextile offers a shear stiffness increase for all values of
shear displacement. The Amoco geotextile, however, resulted in an initial
decrease of shear stiffness for low values of shear displacement. This type of
behaviour has been also observed in tests on granular soils and is usually
attributed to the particular test conditions (Athanasopoulos, 1993). As it might
be expected, the Nicolon reinforcement did not offer any substantial change in
the shear stiffness of reinforced specimens.
A comparison between reinforced and unreinforced soil with respect to the
development of vertical deformation versus shearing displacement can be
made with the help of the plots of Fig. 11. According to these plots the effect
of geotextile reinforcement on the volumetric deformation of cohesive soil
depends on the value of normal stress av. For low values of normal stress
(e.g. av = 1 0 0 k P a ) the reinforcement results in a decrease of vertical
compression of soil specimen for all types of geotextiles including the woven
Nicolon fabric. For higher values of normal stress (e.g. av = 500 kPa) the
behaviour is reversed, i.e. the vertical compression of the reinforced specimen
is increased. It is interesting to note that this increase is rather substantial for
the woven Nicolon geotextile.
636
G. A. Athanasopoulos
300
250
g~ 200
~g
150
250
v = 500 kPa.
av i 500 kPa
"
.............................................
100
50
......... Unreinforced
Reinforced w / T e r ~
............................ ~
~
...-
100 kP~
a v = 100 kPa"
.........................................................
......... Unreinforced
Amoco 4545
% = 500 kPa
~, 200
150
u~
1oo
r~
a v = 100 kPa
50
......---i
250
......... Unreinforced
Nicolon 66448
200
o v = 500 kP;
150
~ lOO
er~
o v = 100 kPa
50
L
1
I
7
Shear displacement, mm
Fig. 10. Development of shearing resistance with shear displacement for unreinforced and
reinforced silty clay.
637
......... U n r e i n f o r c e d
Terram 4000
..3
e-
0.0
E
100kP
-0.4
100 kP~
-0.8
-1.2
.........
kPa
500
Unreinforced
Amoco 4545
0.0
..Y
c v
= 100 k P a
a)
-0.4
a v = 100 kPa
~3
-0.8
a v = 500 k P a
>
-1.2
.........
Unreinforced
Nicolon 66448
o
o v = 100 kPa
-0.4
~3
.2
a v = 100 kP~
fly = 600 k P a
-0.8
~ v = 6 0 0 kP~
-1.2
Shear displacement,
Fig.
11.
Development
of vertical
mm
for unreinforced
and
silty clay.
638
G. A . A t h a n a s o p o u l o s
[A'A~- Asc~]
6 = tan -1 [
A-~-d~-, J
(2)
where:
A t =
A s =
A~=
o- V
Xo =
Ko.av
Terram 4000 ~
639
Amoco 4545J
Ca/C = 0.0
9-
'
Terrain 500
---II.---Terram 1500
Terram 4000 ~
Amoco 4545
&
Terram 4000 +
Amoco 4545
Ca/C = 0.5
oO
Terram 500
Terram 1500
Ca/C = 1.0
2
9-
"
0
i
50
i
100
~
150
200
a
250
300
640
G. A. Athanasopoulos
stress. It is believed that this behaviour does not reflect the actual field
behaviour of the interface bond development, being an artifact of the testing
procedure under low confining stresses.
According to the plots of Fig. 12 the Terram 4000 geotextile developed
the highest efficiency ratio ranging from 2.90 to 2.07 for values of c~/c
ranging from 0.0 to 1-0. These high efficiency values should not be
considered as unreasonable since they are referenced to the undrained
angle of the cohesive soil, ~b= 14.8 . The corresponding friction angles are
6 =43 30.6 . The interface friction angles for the other two Terram and
the Amoco geotextiles ranged from 1.5 to 22 . It is interesting to note that
the values of friction angle 6 for the three Terram geotextiles increased
with the thickness of geotextile, a fact that could be explained by the
expected increasing transmissivity of the geotextiles. It is also interesting to
note that the Amoco geotextile, despite its low tensile strength, developed
relatively high friction angles, a fact that could be attributed to its high
water transmissivity.
CONCLUSIONS
Based on the results of direct shear tests on a geotextile-reinforced nearsaturated cohesive soil the following conclusions can be drawn:
(1) Nonwoven geotextiles possessing the capability of in-plane flow of
water (transmissivity) can be used for reinforcing near-saturated
cohesive soils provided these fabrics also possess a reasonable amount
of tensile strength. The failure envelope of the composite material
(soil+ reinforcement) takes, in general, a curved shape indicating a
continuous transition from drained to undrained behaviour as the
value of normal interfacial stress is increased. Although a critical
value of normal interfacial total stress can be derived from the failure
envelope of composite material its value should not be taken to
represent the transition from slippage failure to stretching failure of
the reinforcement.
(2) Woven geotextiles with practically nil water transmissivity, despite
their high tensile strength, cannot provide substantial strength
increase to cohesive soils with a water content much higher than the
optimum value obtained from compaction tests. When the structure
of these geotextiles contains openings that can be filled with the
particles of cohesive soil, then some adhesion-type strength increase
can be obtained with values of adhesion efficiencies ranging from 0.3
to 1.0 (as is also the case with geogrid reinforcement).
641
ACKNOWLEDGEMENTS
The author wishes to express his thanks to former civil engineering students
of the University of Patras, D. Papandropoulos, A. Papageorgiou and Chr.
Tsarouhis for the careful performance of the laboratory tests described in
this paper. Thanks are also expressed to the graduate students of geotechnical engineering at the Department of Civil Engineering of the University of
Patras, P. Pelekis and E. Leonidou for assisting in the preparation of some of
the figures of this paper.
642
G. A. Athanasopoulos
REFERENCES
643
644
G. A. Athanasopoulos
Shin, E. C., Das, B. M., Puvi, V. K., Yen, S. C. & Cook, E. E. (1993). Effect of
particle size on the mechanical behaviour of sand-geotextile composites.
Geotechnical Testing Journal, ASTM, 16, 534-541.
Smith, M. E.& Criley, K. (1995). Effect of particle size on the mechanical behaviour
of sand-geotextile composites. Geoteehnical Fabrics Report, April, 28-31.
Tatsuoka, F., Uchimura, T., Tateyama, M. & Koseki, J. (1996). Geosynthetic-reinforced soil retaining walls as important permanent structures. Geosynthetics:
Applications, Design and Construction, Proceedings of the First European
Geosynthetics Conference, Maastricht, pp. 3-24.
Wayne, M. H. & Wilcosky, E. (1995). Effect of particle size on the mechanical
behaviour of sand-geotextile composites. Geotechnical Fabrics Report, March,
10.
Zornberg, J. G. & Mitchell, J. K. (1994). Effect of particle size on the mechanical
behaviour of sand-geotextile composites. Geosynthetics International, 1(2), 103148.