Dissertation Suthasinee
Dissertation Suthasinee
Dissertation Suthasinee
ds
= fraction of grid surface area providing the direct shear
resistance
20
If
ds
is equal to zero, it will be the case of soil shearing over soil and then f
ds
will be
equal to one; but if
ds
is equal to one, it will be the case of soil shearing over the
surface are of grid reinforcement and then f
ds
will be equal to tan/tan
ds
.
Being relatively simple to perform, direct shear tests are often conducted to determine
shear parameters not only between soil and soil, but soil and reinforcement as well. If
such parameters are obtained by performing separate tests, the efficiency values of
grid reinforcement either on cohesion or on friction could be determined by using the
following equations:
E
c
= 100 x
c
c
a
|
\
|
(2.3)
E
= 100 x
tan
tan
|
|
\
|
(2.4)
where: E
c
= efficiency of grid reinforcement on cohesion
E
n
= applied normal stress to the interface between soil and grid
reinforcement
= skin friction angle between soil and grid reinforcement surface
(or interface friction angle)
However, when reaching the shear strength, shear stress,, can be assumed as a
function of the corresponding displacement at the maximum frictional resistance, d
cr
,
and can be expressed in the simple form as follows:
cr s
d k = (2.8)
where k
s
is the shear stiffness.
The linear elastic-perfect plastic model can be simply expressed by Eqs. 2.7 and 2.8,
and finally the model are in terms of shear stiffness, k
s
, as follows:
k
s
=
cr
n a
d
tan c +
(2.9)
22
This model is simple and the shear parameters, c, , and k
s
can be determined from
direct shear tests. The model, however, cannot simulate neither hardening nor
softening behavior of the interface between soil and grid reinforcement. Apart from
the direct shear test, shear stiffness, k
s
, can be also calculated from pullout test results
of a single longitudinal member embedded in soil fill (Srikongsri, 1999). Usually, the
shear stiffness obtained from pullout tests is better than that from direct shear tests
because it integrates the shear resisting forces along the total length, and can be used
either for determining an accurate profile or for estimating an interface slippage limit
(Gurung et al., 1999).
2.10.6 Bearing resistance-displacement relationship
The mobilized pullout resistance on each reinforcement segment, especially
inextensible grid reinforcements (e.g. steel grids) used in laboratory tests can be
practically treated as equal. Thus, the relationship between pullout bearing resistance
and displacement of a single bearing member can be investigated by directly using the
pullout test results of a steel grid. The bearing resistance can be obtained by deducting
the frictional resistance from the pullout resistance. The pullout displacement is
normalized by the bearing or transverse member thickness for analyzing the results of
the pullout test. Then, the mobilization process of the bearing resistance for different
transverse member thicknesses can be compared by using the normalized
displacement. The laboratory test results revealed that the relationship of pullout
bearing stress,
b
, and normalized displacement, d
n
, of an individual bearing member
could be modeled by adopting a hyperbolic function (Chai, 1992), and such a model
can be expressed in the following equation:
b
=
bult
n
ip
n
d
E
1
d
+
(2.10)
where: E
ip
= initial slope of bearing resistance-normalized displacement
curve
bult
= ultimate value of bearing stress
This hyperbolic model was developed successively by Wongsawanon (1998),
Srikongsri (1999), and Kongkitkul (2001) in order to predict the bearing resistance
from the pullout tests on hexagonal wire meshes, which were used as inextensible
reinforcement in silty sand.
2.10.7 Displacement at the maximum pullout force
Chai (1992) conducted pullout tests on polymer geogrid reinforcements in weathered
Bangkok clay. The polymer geogrids used as reinforcing materials were Tensar
geogrids (SR 80) with spacing of transverse and longitudinal members of 160 and
22.5 mm respectively, the widths of the transverse and longitudinal members were 16
and 10 mm respectively. The thicknesses of the transverse and longitudinal members
were 3.6 and 1.4 mm respectively. The curves of displacements at the maximum
pullout force versus distance from the pullout face are shown in Fig. 2.45. In the
figure, it was observed that most of the displacements occurred at the pullout face
while at the tail of geogrid reinforcements the displacements rarely occurred. In
23
addition, the curves were non-linear due to the high extension of the geogrid
reinforcements which were highly extensible geogrids.
2.11 Finite Element Modeling of Reinforced Earth Structure
The behavior of reinforced earth structure on soft ground, and its influence factors
have been analyzed using finite element method (FEM) by several researchers (Rowe
and Ho, 1997, Long, 1996; Chai, 1992; Hird and Kwok; 1989, etc.). There are two
general approaches to the finite element analysis of reinforced soil system, namely:
discrete material and composite material. Although the discrete approach needs more
computer time, it is preferable because the properties and responses of
soil/reinforcement interaction properties are key factors that control its performance.
Hence, the discrete approach is used in this research, and is discussed in detail in the
following sections.
2.11.1 Soil model
The commonly used soil model to simulate stress- strain relationships of soil can be
divided into two groups: nonlinear elastic models such as hyperbolic model and
elastic plastic models such as Cam Clay Model and Soft Soil Model. The details of
each models are shown in following section.
2.11.2 Hyperbolic model
In this model, the stress/strain curve of the soil is approximated by a hyperbola
(Kondner and Zelasko, 1963) as expressed in the following equation:
( )
1
1
3 1
b a +
= (2.11)
where:
1,
3
= major and minor principal stresses
1
= axial strain
a,b = constant, whose value may be determined from
conventional triaxial test. (1/a is the initial tangent
modulus E
i
and 1/b is the ultimate principal stress
difference)
Duncan and Chang (1970) improved the hyperbolic model by: (1) obtaining the
relationship for the initial tangent modulus, E
i
,
as given by Janbu (1963) from the
following expression:
n
a
3
a i
P
kP E
|
|
\
|
=
(2.12)
where: P
a
= atmospheric pressure
k = modulus number
n = modulus number
and (2) utilizing the Mohr- Coulomb failure criterion to determine the value of b.
The resulting tangent modulus, E
i
, can be written as :
24
( )( )
n
a
3
a
2
3
2 1 f
t
P
kP
sin 2 cos c 2
sin 1 R
1 E
|
|
\
|
|
|
\
|
+
= (2.13)
where: c = cohesion of the soil
= friction angle of the soil
R
f
= failure ratio,
( )
( )
ult
fail
f
R
3 1
3 1
=
The unloading and reloading modulus of soil, E
ur
, is expressed as follows:
n
a
3
a ur ur
P
P k E
|
|
\
|
=
(2.14)
where: k
ur
is the unloading-reloading modulus number. In most cases the Poissons
ratio is taken as constant. This model needs six parameters: k, k
ur,
n, R
f
, C and as
previously defined.
The Poissons ratio can be taken as variant,
t
, which may be related to the tangent
Youngs modulus, E
t
, and bulk modulus, B, as follows:
|
\
|
=
B 6
E
5 . 0
t
t
(2.15)
The value of bulk modulus, B, can be obtained from triaxial test, and is defined as
follows:
( )
( )
v
3 2 1
3
B
+ +
= (2.16)
where
1
,
2
and
3
are principal stresses,
v
is the volumetric strain corresponding to
the stress condition.
Variation of bulk modulus, B, with confining pressure may be obtained by the
following empirical equation:
m
a
3
a b
P
P k B
|
|
\
|
=
(2.17)
where k
b
and m are bulk modulus number and bulk modulus exponent, respectively.
The model becomes an eight parameters model consist of k, k
ur
, n, R
f
, c, , k
b
and m.
2.11.3 Elasto-plastic model
The elasto- plastic model is characterized by: the yield function, the flow rule, and the
hardening law. The strain due to any increment of stress can be divided into two
components: elastic strain and plastic strain. For conditions where the stress state is in
yield locus, an increment of stress can only cause elastic strain. If the stress conditions
correspond to a point on the yield surface and if the material is stable, the increment
of stress produces elastic and plastic strain. There are several elasto-elastic models for
soils as discussed below.
25
2.11.4 Elastic-perfectly-plastic Mohr Coulomb model
The relationship between effective stress rate and strain rate for elastoplasticity
(Smith and Griffith, 1988) are given as follows:
(
(
=
e
T
'
e e '
D
g
f g
D
d
D (2.18)
'
e
'
T
g
D
f
d
= (2.19)
where: f = yield function
g = plastic potential function
= effective stress tensor
= strain tensor
D
e
= elastic constitutive matrix
The term is used as a switch if material behavior is elastic in which, the value of
is zero. For plastic behavior, is unity.
The yield function for Mohr Coulomb model is defined as three yield functions which
formulated in terms of principal stress (Smith and Griffith, 1988) as follows:
0 cos c sin
2
1
2
1
f
3 2 3 2 1
+ + = (2.20)
0 cos c sin
2
1
2
1
f
1 3 1 3 2
+ + = (2.21)
0 cos c sin
2
1
2
1
f
2 1 2 1 3
+ + = (2.22)
where: = friction angle of soil
c = cohesion intercept
The plastic potential functions are defined as follows:
+ + = sin
2
1
2
1
g
3 2 3 2 1
(2.23)
+ + = sin
2
1
2
1
g
1 3 1 3 2
(2.24)
+ + = sin
2
1
2
1
g
2 1 2 1 3
(2.25)
where: is the dilation angle
In order to model the influence of stress level on the material stiffness, a simple power
law for the shear modulus is introduced (Vermeer and Brinkgreve, 1995):
26
m
ref
*
ref
P
P
G G
|
|
\
|
= (2.26)
where: P
*
= ( ) cot
3
1
3 2 1
c + + +
G
ref
= reference shear modulus corresponding to
P
*
=P
ref
P
ref
= reference pressure
m = power number
Thus, the Mohr Coulomb model required a total of given parameters which are G
ref
,
P
ref
, m, , , c, , and are generally familiar to most geotechnical engineers.
2.11.5 Modified Cam clay model
This model was proposed by Roscoe and Burland (1968) and has been widely used
because it is one of the simplest elasto-plastic model that can simulate the stress-strain
behavior of lightly and normally consolidated clays. Cam Clay models were derived
from the critical state energy equation and the associated flow rule, using volumetric
strain as hardening parameter. A brief review of Modified Cam Clay models is
presented as follow.
The Modified Cam Clay model assumes an associated flow rule, so the yield function
and the plastic potential function are the same.
'
c
' 2 2 ' 2 2
p p M p M q = +
(2.27)
where: q =
1
-
3
p = (
1
+
2
+
3
)/3
M = slope of failure line in (p,q) axis
p
c
= the value of p when yield locus intersect the p
axis in (pq) plane.
The failure surface of the model is straight line in p, q space passing the origin
described by:
'
Mp q = (2.28)
where: q =
1
-
3
p
= (
1
+
2
+
3
)/3
M = slope of failure line in (p,q
) axis
The strain hardening law uses consolidation parameter, and , obtained by isotropic
loading and unloading normally consolidated soil. The parameter is the slope at the
void ratio versus ln (p) during loading while , is the initial value of the slope at
rebound. In elastic state, the elastic shear strain is neglected. The elastic volumetric
strain is given as follows:
27
( )
'
'
v
p e 1
dp
d
+
= (2.29)
where: e = void ratio
The increment of plastic volumetric strain,
v,
and plastic shear strain are given as
follows:
|
|
\
|
+
|
\
|
+
= d
M
2
p
dp
e 1
d
2 2
'
p
V
(2.30)
Using Equation (2.30), the total volumetric strain can be expressed by;
( )
'
'
2 2
'
v
p
dp
e 1
d
M
2
p
dp
e 1
d
+
+
|
|
\
|
+
|
\
|
+
= (2.31)
The shear strain is equal to:
|
|
\
|
+
|
\
|
+
= d
M
2
p
dp
e 1
M
2
d
2 2
'
2 2
S
(2.32)
where: = Cc/2.303; Cc is compression index
= Cr/2.303; Cr is recompression index
G = shear modulus
In finite element formulation, the Modified Cam Clay model only needs five
fundamental parameters such as: M, , , or G, and the specific volume on critical
state line at unit mean stress (p).
2.11.6 Soft soil model
This model was proposed by Vermeer and Brinkgreve (1995). It is derived from the
Cam Clay type model, which resembles the Modified Cam Clay model. In this model,
it is assumed that there is a logarithmic relation between the volumetric strain and the
mean effective pressure. This can be formulated as follows:
|
|
\
|
=
0
'
* 0
V V
p
p
ln (2.33)
|
|
\
|
=
0
'
* e 0
V
e
V
p
p
ln (2.34)
where:
v
= volumetric strain
v
0
= initial volumetric strain
v
e
= elastic volume strain
28
v
e0
= initial elastic volumetric strain
p = mean effective stress
p
0
= initial mean effective stress
*
= modified compression index
* = modified swelling index
The yield function of the Soft Soil model is defined as:
c
*
p f f =
(2.35)
where f* is a function of the stress state (p,q) and the preconsolidation stress p
c
is a
function of plastic strain:
( )
'
' 2
2
*
p
cot c p exp M
q
f +
+
= (2.36)
|
|
\
|
=
* *
p
V 0
c c
exp p p (2.37)
where: = angular of friction of soil due to Mohr Coulomb.
c = cohesion
*
= modified swelling index
* = modified compression index
v
p
= plastic volumetric strain
M = function of at rest earth pressure (k
0
nc
) and
Poisson ratio ()
( )
( )
( )( )
( )( ) ( )( ) + +
|
\
|
+
+
=
1 k 1 / 2 1 k 2 1
1
2 1 k 1
k 2 1
k 1
M
nc
0
* * nc
0
*
*
nc
0
nc
0
2
nc
0
(2.38)
where: k
0
nc
= earth pressure at rest of normal consolidate state
= Poisson ratio
In finite element modeling, soft soil model needs six parameters: , c,
*
, *, , M or
K
0.
2.11.7 Soil and reinforcement interface model
In PLAXIS Software, the stress-strain behavior at soil-interface is simulated by
elastic, perfectly-plastic interface model. The model parameters at soil-structure
interface can be generated from that soil using the interaction coefficient, R, defined
as the ratio of the shear strength of the interface to the corresponding shear strength of
the soil (Vermeer and Brinkgreve, 1995) as follows:
29
= tan R tan (2.39)
Rc c
i
= (2.40)
G R G
2
i
= (2.41)
where: G = shear modulus of soil that contact with reinforcement
G
i
= shear of interface element
= angular of friction of soil contact with reinforcement
= angular of friction of interface element
2.11.8 Finite element analysis of consolidation
Consolidation analysis corresponds to a coupling between the laws governing the
behavior of the skeleton of the soil and the flow of pore fluid. In a variation form, the
analysis of this problem leads to the search for a field of displacements and a
distribution of hydraulic head satisfying the simultaneous equation. Biots
consolidation theory (Biot, 1941) is a rigorous solution to this problem, when the soil
skeleton is linear elastic and the pore fluid is incompressible. The equation used in the
application for finite element analysis is shown below (Vermeer and Brinkgreve,
1995):
( ) 0
t
p
k
n
t
m p p
k
w
T
steady w
w
p
T
=
|
|
\
|
(2.42)
where k
p
is the permeability matrix
(
=
y
x
p
k 0
0 k
k (2.43)
where: k
x
= permeability in x-direction
k
y
= permeability in y-direction
p,p
steady state
= pore water pressure
n = porosity
= strain tensor
k
w
= bulk modulus of water
In general, when a nonlinear material model is used, iteration is needed to arrive at the
correct solution. So in finite element analysis, the consolidation for finite element
analysis can be obtained using standard isoparameter finite element formulation
procedure (Britto and Gunn, 1987).
2.12 Behavior of Embankment on Soft Ground
2.12.1 Total settlement
Soil, which is a product of natural process, is highly variable and displays very
complex behavior. The range in properties can be tremendous. Furthermore, soil
properties can change with time, moisture and stress. Because of this large variation
in soil properties, the opportunity of working with widely scattered data cannot be
avoided. In view of this aforementioned rationale, the use of semi-empirical
approaches based on the knowledge of fundamental soil behavior, select soil results,
and detailed soil in-situ observation are necessary to come up with the needed
parameters for settlement analysis by back analyses.
30
Generally, the settlement of foundation may be regarded as consisting of three
separate components of settlement as follows:
s c i
+ + = (2.44)
where: = total ultimate settlement,
i
= immediate settlement resulting from the constant volume
distortion of the loaded soil mass,
c
= consolidation settlement resulting from the time dependent flow of
water from the loaded area under the influence of the load that
generate excess pore pressure which in itself dissipated by the flow
s
= secondary settlement or creep which is also time dependent but
may occur at essentially constant effective stress.
2.12.2 Immediate settlement
The immediate settlement is a constant volume deformation contribution to the final
settlement and therefore it occurs simultaneously as the load is applied. DAppolonia
et al. (1971) developed a simplified method, which considers the local yielding
beneath the embankment as it approaches failure. This local yielding occurs when the
shear stresses at some points of the soil mass are relatively higher compared to the
shear strength during the undrained loading. The immediate settlement with the effect
of the local yielding can be estimated from the following equation:
) SR ( E
) 1 ( qBI
u
2
i
= (2.45)
where is the undrained Poissons ratio, and SR is the settlement ratio obtained from
Fig. 2.46 and a function of the initial shear stress (f), the applied stress ratio (q/q
ult
),
and the geometry of the problem. This method uses an initial shear stress ratio, f,
given as:
u
0 v 0
s 2
) k 1 (
f
= (2.46)
where k
0
is the coefficient of earth pressure at rest,
v0
is the effective
overburden
pressure,
and s
u
is the undrained shear strength.
A design chart for estimating the immediate settlement given by Janbu et al. (1956) is
shown in Fig. 2.47. The chart provides estimates of the average immediate settlement
of uniformly loaded, flexible areas of strip, either rectangular or circular in shape. The
immediate settlements are obtained from
u
2 1
i
E
qB
= (2.47)
2.12.3 Consolidation settlement
The observation made by Bjerrum (1972) on various instrumented test fills shows that
almost all clays demonstrate a maximum past pressure,
p
, the magnitude of which
depends on the plasticity and geological history of the clay deposit. For increment of
loading within the range of maximum past pressure,
p
, its settlement are small, the
excess pore pressure is small and it dissipates rapidly. For loads that exceed maximum
31
past pressure,
p
, the settlement is high, the excess pore pressure is high and dissipate
slowly. Hence, the embankment settlement can be analyzed in two scenarios. If the
stress in the clay is lower than
p
, the consolidation settlement can be predicted from
the equation, defined as:
z v
z
o c
d m =
(2.48)
In cases, however, in which the stresses in the clay exceed
p
, the consolidation
settlement is composed of two contributions. The first contribution settlement
occurring when the load is increased from
vo
to
p
, calculated as:
z vo p v
z
o 1 c
d ) ( m = (2.49)
The second contribution is the consolidation settlement which occurs when
p
is
exceeded is shown as:
p
z p
o
c z
o 2 c
d
log
e 1
C
+
+
= (2.50)
The total consolidation settlement, therefore, is the sum of the two settlement
contributions:
2 c 1 c c
+ = (2.51)
A method in which the final primary consolidation is calculated from three-
dimensional excess pore pressure obtained under undrained axis symmetric triaxial
stress conditions and can be written in terms of the consolidation settlements
computed from oedometer tests as:
oed c
= (2.52)
The correction factor is a function of the pore pressure coefficient, A, and the
geometry of the problem as shown in Fig. 2.48.
2.12.4 Secondary settlement
The secondary consolidation is the continued settlement under sustained loading at
the end of primary consolidation (that is, after complete dissipation of excess pore
pressure). The variation of the void ratio e with time t for a given load increment is
shown in Fig. 2.30. The secondary compression index can be defined as:
|
|
\
|
1
2
t
t
log
e
C (2.53)
Then, the magnitude of secondary consolidation can be calculated from:
|
|
\
|
=
1
2 '
s
t
t
log H C (2.54)
32
where: C
= C
/ (1+e
p
) and C
= (2.58)
33
where:
0
= the interpret of the fitted straight line
1
= the slope of the fitted line
f
= final total settlement
The point of intersection with the 45 line gives the final consolidation settlement,
f
as shown in Fig. 2.51. The coefficient of consolidation is obtained from the following
relationships:
dt
ln
H
6
1
C
1 2
v
= , for the case of double drainage (2.59)
dt
ln
H
12
5
C
1 2
v
= , for the case of single drainage (2.60)
2.12.7 Excess pore water pressures
The increase of pore water pressure in the soil due to various loading conditions
without drainage is important in both theoretical and applied soil mechanics. If the
load is applied very slowly on a soil such that sufficient time is allowed for pore
pressure to drain out, there will be practically no increase of pore pressure. However,
when a soil is subjected to rapid loading and if the permeability is small, there will be
insufficient time for drainage of pore pressure. This will lead to an increase of the
excess pore pressure. The excess pore pressure generated beneath the embankment
can be calculated by:
v
u = (2.61)
where: = the pore pressure correlation factor
v
= the increase in vertical stress due to embankment
loading
The value of in Eq. 2.61 is 1 for Terzaghi and Peck (1967) method. Skempton and
Bjerrum (1957) suggested that is a function of geometry of loading and degree of
consolidation. The value for Bangkok clay can be determined from charts shown in
Fig. 2.52.
2.12.8 Lateral displacements
It is the purpose of the present study to check the validity of the above description of
the foundation behavior in terms of lateral displacements to analyze further the
development of long-term lateral deformation. The magnitude and distribution with
depth of the lateral deformation depend on the location of the inclinometer with
respect to the embankment.
To investigate the variations of the magnitude of lateral deformation with both
embankment load and time it has been found appropriate to refer to the maximum
lateral deformation, y
m
, observed along a vertical soil profile and to compare its
variation to those of maximum settlement, s, generally observed under the centerline
of embankment. This study will consist of on analysis of y
m
as a function of s during
and after construction. The conclusion drawn from this analysis will be used to
investigate the characteristic of the distribution of the lateral deformations, y, with the
depth, z.
34
2.12.9 Strain measurements
Shivashankar (1991) monitored the strains in the longitudinal members of the grid
reinforcements obtained through the use of self-temperature compensating electrical
resistant strain gages. A Wheatstone bridge circuit and a differential voltage
measurement system was employed to measure the strain. The wire leads from the
gages were bundle. In each bundle, the lead wires were connected to a 12-pin quick-
connected plug which was plugged onto switching unit of AM416 multiplexer
connected to the 21X data logger when taking the readings.
The initial reading on strain gage was taken from zero tension. Subsequent readings
were then taken as the wall was constructed and past construction. From the strains
calculated thus, the tensions in the wire can be computed as:
= EA T (2.62)
where: T = axial tension in reinforcing wires
E = modulus of elasticity of steel
A = cross-sectional area of the reinforcing wire, and
= axial strain in the reinforcing wires
2.12.10 Tension in reinforcement
The tensile stress in the reinforcement can be obtained by the measured strain.
Shivashankar (1991) plotted the tension with the height of the embankment then
compared it with the K values during construction in Figs. 2.53a, b. Figures 2.53a, b
shows the tensile forces compared with the distance from the face of the wall during
construction. The line of maximum tension was found to vary between 0.61-1.83 m
from the face of the wall during construction. Figure 2.54 shows the mobilized
tensions in each layer of reinforcement. From the end of construction, it was seen that
some reinforcing mats displayed sudden increases in the stresses at small distances
behind the facing and decreased at distances far behind the facing. Then, the
monitoring was continued after 89, 203 and 286 days. In Fig. 2.54, the coherent
gravity failure plane and the tie-back wedge failure plane are also plotted for
comparison. Shivashankar (1991) commented that the settlement had influenced the
tensile forces in the reinforcement. The settlements at the center were not large
enough to cause substantial release of tensile stresses at points far behind the facing
and to cause any increase in tensile stresses at points closer to the facing.
2.13 Overview of Reinforced Full Scale Test Embankment with Geosynthetics
The USAID embankment (Shivashankar, 1991) was a study of the field behavior of a
fully instrumented mechanically stabilized embankment using welded wire
reinforcement. The welded wire wall was constructed at the AIT campus. The
embankment system was extensively instrumented both in the subsoil and within the
embankment in order to monitor the behavior of the wall. The instrumentation
program of this study consist of surface settlement plates, subsurface settlement
gages, temporary benchmarks, hydraulic piezometers, pneumatic piezometers, earth
pressure cells, and inclinometer. Figure 2.55 shows the schematic plan view layout of
the field instrumentation and the dimensions of embankment. The details of the
instrumentation program for welded wire wall/embankment system are shown in Figs.
2.56 to 2.57. The test was divided into three sections along its length, with three
35
different locally available backfill soils used for each section, namely: weathered
Bangkok clay, lateritic residual soils and clayey sand.
Alfaro (1996) investigated the soil reinforcement interaction on the soft ground
foundation through the performances of two field test facilities with different
reinforcement materials with the same backfill soil. One test facility used steel grid
(metallic) reinforcements and the other test facility used polymer grid (geosynthetic)
reinforcement as shown in Fig. 2.58. The test facility with steel grid reinforcement is
hereafter referred to as Test Facility I while that with polymer grid reinforcements is
referred to as Test Facility II. The plan layout and instrumentation program of the test
facility I is USAID embankment (Shivashankar, 1991) and test facility II is shown in
Fig. 2.59.
Long (1996) studied the field behavior of reinforced embankment by using geotextile
reinforcement on soft ground. In order to study the improvement of embankment
stability on soft ground, two test embankments were constructed to failure. One test
embankment was reinforced with high-strength, nonwoven geotextile as the base
reinforcement. The reinforcement consisted of one layer of high-strength geotextile
placed directly on the natural ground surface at the bottom of the embankment fill.
For comparison, another fullscale, unreinforced embankment using the same fill
material as that of the reinforced one was also constructed to failure at the adjacent
site. The height at failure of the reinforced embankment was 6 m. The study indicated
that the high-strength, nonwoven geotextile as base reinforcement, can considerably
increase the ultimate height of embankment on soft clay. It was also shown that the
rupture occur at large deformation of foundation subsoils. The layout of the test
embankment is given in Fig. 2.60. The instrumentation program consisting of
settlement gauges, piezometers, inclinometer, and earth pressure cell is shown in Fig.
2.61.
Voottipruex (2000) conducted a fully instrumented test on wall/embankment
reinforced with hexagonal wire constructed on soft Bangkok clay foundation in
Thailand, as shown in schematic view in Fig. 2.62. The reinforced wall/embankment
consisted of 10 degree inclined gabion facing with hexagonal wire mesh
reinforcements on one side and sloping unreinforced sandfill in the opposite side with
a total height of 6.0 m. Strain gages were affixed to the wire mesh to monitor tensile
stresses. High strength wires were also used to measure the displacement in the
hexagonal wire reinforcement. The lateral earth pressure coefficient, K, measured
during the wall construction varies from a value corresponding to the active condition,
K
a
, at the base of the wall, to a value nearly at at-rest condition, K
o
, at the top of the
wall with K/K
a
equal to 1.6 which is between geogrids and metal strips. The
maximum tension line interpreted from the observations of the strains induced in the
hexagonal wire mesh reinforcement appeared to be in between the tie-back wedge or
the Coulomb/Rankine failure plane and the coherent gravity failure plane. The
maximum displacement measured from high strength wire extensometer agreed well
with maximum lateral movement measured by the inclinometer. The maximum
ground surface settlement of 450 mm was observed at the front face of embankment
at 630 days from beginning of construction. The lateral movement in the subsoil was
about 35 mm at the weakest zone of about 2.5 m to 4.00 m depth below the general
ground surface. The excess pore pressure observed below the embankment at 3 m and
36
6 m depths indicate the pore pressures build up at first and gradually dissipated during
soft clay consolidation
Abdullah and Edil (2007) studied the test of embankment which was constructed over
soft ground to evaluate the performance of different types of load transfer platform
(LTP) supported on rammed aggregate piers (called geopiers). The length of the test
embankment was approximately 90 m, the width was14.5 m, and the height was 3.5
m. The side slopes of the embankment were 1V:1.5H. The test embankment was
divided into four major sections (Sections 14), and two control sections (C1 and C2)
at the two ends of the embankment. The major LTP sections and control sections are
shown in Fig. 2.63. Three types of LTP were constructed in accordance with the
recommended design for each: a geosynthetic reinforced LTP with two layers of
geogrid (catenary LTP), a geosynthetic reinforced LTP with three or more layers of
geogrid (beam LTP), and a reinforced concrete LTP. The results indicate that the
differential settlement between the geopiers and the matrix soil is relatively small in
all LTP sections, with the smallest in the beam LTP. The tensile strain in the geogrid
was approximately 60% of the allowable design strain of 5% in the beam LTP. In the
catenary LTP, the tensile strain in the geogrid was approximately 20% of the
allowable design strain of 6%, although the differential settlement was larger than the
beam LTP. The total and differential settlements observed indicate that the use of less
stiff geopier columns together with LTPs provides an attractive option for supporting
low embankments on soft ground with tolerable total and differential settlements. The
presence of LTPs and the supporting columns also tend to reduce the lateral
movement of the foundation soil.
Arai et al. (2007) studied full-scale model test for a new reinforced soil retaining wall
system, having a vertical layer which absorbs the deformation between facing
concrete brocks and reinforced backfill. The wall was constructed to have a vertical
height of 9m by piling up 0.9m high facing concrete blocks in ten rows. The
reinforced soil retaining wall was installed with a layer for absorbing deformation
between facing concrete blocks and reinforced soil to mitigate the earth pressure
acting on the facing concrete blocks by the deformation of the soil during and after
the construction of the wall. Structure of reinforced soil retaining wall and overall
view of the wall are shown in Fig. 2.64. The reinforcement geogrid in front view and
cross is shown in Fig. 2.65. The calculated and measured of horizontal earth
pressures and horizontal displacements acting on the facing concrete blocks were
measured. The results show that earth pressure acting on the facing concrete blocks
was very small and large strains were observed near the wall.
Tanchaisawat (2008) investigated the full scale field test embankment which was
constructed by using rubber tire chips-sand mixtures as the lightweight geomaterials
in the campus of Asian Institute of Technology (AIT). Section view of embankment
with instrument location and the schematic 3D view of full scale test embankment
with instrumentation are shown in Fig. 2.66 and 2.67. The geogrid reinforced
embankment system was extensively instrumented in the subsoil and within the
embankment itself in order to observe its behavior during construction and post
construction phases, and thereby evaluate its performance. The unit weight of rubber
tire chip-sand mixtures with 30:70 % by weight is 13.6 kN/m
3
compared to
conventional sand backfill of 18.0 kN/m
3
. The former is lighter by about 75 % than
the latter. Figure 2.68 shows comparison of settlement between conventional and
lightweight backfill. The total settlement magnitude of 122 mm at ground surface is
37
67.5% less when compared to the corresponding value of 400 mm for conventional
backfill without foundation treatments. The excess pore water starts to build up after
15 days since the end of construction and starts to dissipate after 50 days. The excess
pore water pressure becomes constant after 120 days since the end of construction as
shown in Fig 2.69. This lightweight geomaterials reinforced with geogrid can be used
for embankment construction on soft ground area to reduce total settlement of
structure.
38
CHAPTER 3
LABORATORY MATERIALS INVESTIGATION
3.1 General
New types of natural fiber reinforcing materials have been introduced successively in
geotechnical applications; for example, jute, coir, and geosynthetic reinforcements.
Thus, the knowledge of the mode of interaction on natural fiber materials is important,
and has to be investigated before constructing reinforced earth structures. This chapter
discusses the test materials, the experimental apparatus and procedures, testing
program of the study.
Natural fibers used for this study were water hyacinth (Eichhornia crassipes), and
roselle or Thai kenaf (Hibiscus sabdariffa var altissima) as shown in Fig. 3.1.
Though, these natural fibers are widely used in ropes, twines, rugs, mats, mattresses,
and handcrafted articles, a large quantity of these economic and renewable resources
is still under-utilized. In considering the aforementioned fibers for geotextiles, plain
weaving patterns of natural fiber were obtained from the handicraft communities in
Nakhon Pathom, Buriram, Khon Kaen and Petchburi Provinces, respectively. Roselle
or Kenaf and water hyacinth fibers were fabricated to dimension of 300 mm x 300
mm and 4 mm opening size. Hexagonal, plain and knot-plain pattern were selected for
this study to find the best suited pattern of natural fibers.
3.2 Investigation Properties of the Woven LLGs
Woven geotextiles from the selected plant fibers was fabricated to dimension of 300
mm x 300 mm. Plain pattern was selected for this study to find the best suited pattern
of natural fibers as shown in Figs. 3.2 and 3.3. The following properties of geotextiles
were investigated:
3.2.1 Mass per unit area (weight) test
Mass per unit area is the weight of geotextile was given in grams per square meter
(g/m
2
) according to ASTM D5261. Mass per unit area (weight) was measured in plain
pattern of geotextiles before testing tensile strength. The average mass per unit area of
Kenaf LLGs and water hyacinth LLGs for plain pattern are 1157 g/m
2
and 1469 g/m
2
,
respectively, as tabulated in Table 3.1.
3.2.2 Thickness
The thickness was measured as the distance between the upper and lower surfaces of
the fabric, measured at a specified pressure by using ASTM D5199. The thickness of
a geotextile was measured to an accuracy of at least 0.02 mm. The average
thicknesses of Kenaf LLGs and water hyacinth LLGs for plain pattern are 5.27 mm
and 6.96 mm, respectively, as tabulated in Table 3.2.
3.2.3 Tensile strength test
The tensile strength and elongation of the geotextiles were measured wherein the
entire width of a 200mm wide specimen is gripped between two clamps with a
39
separation of 100mm. Test Method ASTM D4595-86 Standard Test Method for
Tensile Properties of Geotextiles by the Wide-Width Strip Method was obtained by
using testing machine (Instron 5567) at a speed of 10 mm/min with a gauge length of
100 mm and 3 tons of load cell. These fibers were tested at temperature 22 C. The
200 mm wide specimen was gripped between two hydraulic clamps of the machine.
Then, the load was applied to the specimens until failure. The tensile testing machine
and sample between two clamps are shown in Figs. 3.4 and 3.5 respectively. The
tensile strength and elongation of Kenaf and water hyacinth LLGs in machine and
cross direction were obtained to compare their performance in each type.
3.2.3.1 Tensile strength test of water hyacinth LLGs
The average tensile strength of knot-plain patterns were 3.21 kN/m and 2.38 kN/m for
machine and cross direction, respectively. For hexagonal patterns, were 6.78 kN/m
and 9.48 kN/m for machine and cross direction, respectively. For plain pattern,
average tensile strengths were 10.98 kN/m and 9.92 kN/m for machine and cross
direction, respectively. The tensile strength was not much different for both
directions.
The comparison of three patterns is shown in Fig. 3.6. The plain pattern had the
highest tensile strength, followed by hexagonal and knot-plain, respectively. The
elongation or percentages of strain, knot-plain patterns have highest elongations
which were 80.52% and 87.83% for machine and cross direction, respectively. The
percentages of strains of hexagonal patterns were 69.45% and 64.92% for machine
and cross direction, respectively. The lowest elongations have been found in plain
pattern which were 19.71% and 32% for machine and cross direction, respectively.
3.2.3.2 Tensile strength test of Kenaf LLGs
The results of tensile strength of knot-plain pattern was 7.72 kN/m for machine
direction and 6.42 kN/m for cross direction. The tensile strength of hexagonal pattern
was 19.99 kN/m for machine direction and 16.43 kN/m for cross direction. For plain
pattern, the results were 22.79 kN/m and 21.26 kN/m for machine and cross direction,
respectively, which have similar results in both directions. The comparison of all
pattern is plotted in Fig. 3.7. The plain pattern had the highest tensile strength,
followed by hexagonal pattern and knot-plain, respectively. Elongation or percentages
of strain, knot-plain patterns has highest percent of strain which were 150.80% and
160.93% for machine and cross direction, respectively. The percentages of strain for
hexagonal pattern were 103.80% and 97.15% for machine and cross direction,
respectively, while the plain pattern had the lowest elongation which were 14.70% for
machine direction and 14.30% for cross direction.
The comparison of tensile strength of plain pattern of Kenaf LLGs and water hyacinth
LLGs are plotted in Fig. 3.8. It can be seen that tensile strength for plain patterns of
Kenaf were higher than water hyacinth. Consequently, from the results of tensile
strength, Kenaf LLGs can be applied for soil reinforcement and water hyacinth LLGs
can be applied for soil erosion control in geotechnical field.
40
3.3 Investigation Properties of Water Hyacinth LLGs for Erosion Control
Woven Limited Life Geotextile made from Water Hyacinth was used in this study
which has 2 different opening size dimensions of 8 mm and 12 mm in plain pattern.
The diameter of the LLGs yarn is 2 mm as shown in Fig. 3.9.
3.3.1 Mass per unit area (weight) test
Mass per unit area is the weight of geotextile was given in grams per square meter
(g/m
2
) according to ASTM D5261. Mass per unit area (weight) was measured in plain
pattern of geotextiles before testing tensile strength. The average mass per unit area of
water hyacinth LLGs for 8 mm and 12 mm opening size are 854 g/m
2
and 648 g/m
2
,
respectively, as tabulated in Table 3.3.
3.3.2 Thickness
The thickness was measured as the distance between the upper and lower surfaces of
the fabric, measured at a specified pressure according to ASTM D5199. The thickness
of a geotextile was measured to an accuracy of at least 0.02 mm. The average
thicknesses water hyacinth LLGs for 8 mm and 12 mm opening size are 6.96 mm and
6.94 mm, respectively, as tabulated in Table 3.3.
3.3.3 Tensile strength test
The wide-width strip tensile strength method was conducted to provide a more
reliable assessment of fabric on selected fibers in accordance with ASTM D4595
Standard Test Method for Tensile Properties of Geotextiles by the Wide-Width Strip
Method was obtained by using testing machine (Instron 5567) at a speed of 10
mm/min with a gauge length of 100 mm. These fibers were tested at temperature 22
C. The 200 mm wide specimen was gripped between two hydraulic clamps of the
machine. Then, the load was applied to the specimens until failure. The tensile
strength and elongation of water hyacinth LLGs in machine and cross direction were
obtained.
The average tensile strength of water hyacinth for plain pattern, the results were 6.6
kN/m for 8 mm and 4.2 kN/m for 12 mm opening size and elongation were 16 % and
13% for machine direction, respectively, as plotted in Fig. 3.10 and 3.11. The
comparison of tensile strength of 8 mm and 12 mm of water hyacinth LLGs are
plotted in Fig. 3.12. Consequently, from the results of tensile strength of 8 mm
opening size of water hyacinth LLGs has higher strength than 12 mm opening size.
3.3.4 Permittivity test
The major function that geotextile perform is filtration. The liquid flows
perpendicularly through the geotextile, it is important that geotextile allow this flow to
occur without being impeded. Permittivity test was performed in accordance with
ASTM D4491 (Standard Test Methods for Water Permeability of Geotextiles by
Permittivity) for determining the hydraulic conductivity (water permeability) of
geotextiles with constant head. The volumetric ow rate of water per unit cross
sectional area per unit head under laminar flow conditions, in the normal direction
41
through a geotextile was measured. A photo of device shows in Fig. 3.13. The device
consists of an upper and lower unit, which fasten together. The specimen was
positioned in the bottom of the upper unit. There was a standpipe for measuring the
constant head value. The rotating discharge pipe allows adjustment of the head of
water at the bottom of the specimen. In this study, a head of 50 and 100 mm of water
was obtained on the water hyacinth woven LLGs. The quantity of flow was measured
versus time.
Four specimens were selected equally spaced along a diagonal line extending from the
lower left hand corner to the upper right hand corner of the sample. Specimens were
cut to fit the testing apparatus which is 73 mm in diameter for the device as shown in
Fig. 3.14. Specimens was soaked in a closed container of de-aired water at room
conditions for a period of 2 hour. Then, de-aired water was flowed into the system
through the water inlet, discharge pipe was adjusted along with the rate of water
flowing into the apparatus. The values of time (t), quantity of flow (Q) was recorded.
The hydraulic heads of 50 and 100 mm of water were utilized on the water hyacinth
LLGs. The quantity of flow was measured versus time. The permeability can be
calculated as follow:
(3.2)
where = permeability, s
-1
= quantity of flow,mm
3
= head of water on the specimen, mm,
= cross-sectional area of test area of specimen, mm
2
,
= time for flow ( ). s, and
R
t
= temperature correction factor determined using Eq (3.2)
c
(3.3)
where
t
= water viscosity at test temperature, millipoises
20c
= water viscocity at 20 C, mP.
The volumetric ow rate of water per unit cross sectional area per unit head under
laminar flow conditions was measured. The average flow rate of 8 mm spacing size
with head of 50 and 100 mm were 234 l/m-s and 280 l/m-s, respectively. For
specimens spacing 12 mm with head of 50 and 100 mm, the average flow rate were,
249 l/m-s and 350 l/m-s. Higher results of flow rate have found in 12 mm spacing
specimen while average flow rate of 8 mm and 12 mm are not much different. It can
be seen that at head of 100 mm, the average flow rate has increased at higher head
loss. The comparisons of the average flow rate are plotted in Fig. 3.15.
42
3.3.5 Hydraulic transmittivity test
In plane permeability for the drainage function is defined from transmittivity test.
Transmittivity test was performed in accordance with ASTM D4716 (Standard Test
Method for Determining the (In-plane) Flow Rate per Unit Width and Hydraulic
Transmittivity of a Geosynthetic Using a Constant Head) to determining the flow rate
per unit width within the manufactured plane of sample under varying normal
compressive stresses and a constant head. A photo of device shows in Fig. 3.16.
Water hyacinth woven LLGs specimen opening sizes of 8 mm and 12 mm with
dimensions of 300 by 300 mm were used in this study. Three hydraulic gradient (i)
values; 0.10, 0.50 and 1.0 was applied to perform flow rate of testing with three
applied normal stresses following values; 100, 200 and 300 kPa. Specimen was placed
on the test device base. The sides of the specimen parallel to the direction of ow by
wrapping the test specimen in a thin sheet of rubber membrane as shown in Fig. 3.17.
Plate was seat at the top on the test assembly and water was filling slowly to the
reservoir ow through the test specimen. From this point forward, the specimen was
kept saturated at all times.
Hydraulic gradient was computed with the difference in water elevations between the
reservoir and dividing this value by the length of the specimen subjected to the normal
compressive stress. Elevation of the water level in the reservoir box was adjusted to
change the system gradient at 0.10, 0.50 and 1.0. Normal compressive stress was
increased and the three different hydraulic gradients were applied also.
In plane permeability of 8 mm and 12 mm spacing size were investigated with 0.10,
0.50 and 1.0 hydraulic gradient (i) values with normal stresses following values; 100,
200 and 300 kPa. At pressure 100 kPa, the averages of transmissivity were 4.5x10
-6
cm/s, 6.2x10
-6
cm/s and 8.2x10
-6
cm/s at 0.10, 0.50 and 1.0 hydraulic gradient,
respectively. For 200 kPa of pressure, the averages of transmissivity were 2.5x10
-6
cm/s, 4.3x10
-6
cm/s and 6.3x10
-6
cm/s at 0.10, 0.50 and 1.0 hydraulic gradient,
respectively. For 300 kPa of pressure, the averages of transmissivity were 0.9 x10
-6
cm/s, 1.2x10
-6
cm/s and 1.6x10
-6
cm/s at 0.10, 0.50 and 1.0 hydraulic gradient,
respectively. It can be seen that the averages of transmissivity has decreased when
applied more pressure. The comparisons of the averages of transmissivity of 8 mm
specimen spacing size are plotted in Fig. 3.18 and the averages of transmissivity of 12
mm specimen spacing size are plotted in Fig. 3.19.
3.4 Investigation of Backfill Properties in the Laboratory
3.4.1 Sieve analysis test of sand backfill
Sieve analysis tests with ASTM D422-63, (Standard Test Method for Particle-Size
Analysis of Soils) were conducted by shaking soil sample through a series of ten
screens consisting of the following number of sieve following number of sieve: 4, 10,
20, 40, 60, 100, and 200 with a pan at the bottom. After sieving, the material retained
on each sieve was weighed and the weight was recorded. Figure 3.20 show the grain
size distribution curve for Ayuttaya sand backfill samples as percent finer by weight
with the particle size decreasing from left to right on the horizontal axis. From the
distribution curve, there is 1% fines passing No. 200 sieve According to Unified Soil
Classification System (USCS), this sand can be classified as silty sand (SM).
43
3.4.2 Compaction test of sand backfill
Standard Proctor Compaction test was conducted in accordance with ASTM D698
(Standard Test Methods for Laboratory Compaction Characteristics of Soil Using
Standard Effort) to find optimum moisture content to achieve the maximum dry
density under a designated compactive effort for a specific soil. The result of
compaction test is shown in Fig. 3.21. The maximum unit weight (
dmax
) and optimum
moisture content (OMC) of the backfill sand was 18.1 KN/m
3
and 10% respectively.
According to the flat shape compaction curve, the maximum unit weight of Ayuttaya
sand is not much influenced by the water content of soil. A representative portion of
the mixed soil samples provided was taken, prior to the preparation of specimens such
as direct shear test and pullout test. The result of this test would be used to check the
degree of compaction effort applied in each fill material layer in the pullout test.
3.5 Investigation of Interaction Coefficients of Kenaf LLGs and Backfill soils
In this study with the size of 0.5 m x 0.9 m of woven Limited Life Geotextile made
from kenaf were prepared. The opening size of LLGs is 4 mm and the diameter of
the LLGs yarn is 2 mm as shown in Fig. 3.22.
3.5.1 Backfill materials
The local silty sand with a specific gravity of 2.65 was used for pullout and direct
shear tests. Based on standard Proctor compaction test, the maximun dry unit weight
was 18.1 kN/m3 and optimum moisture content was 10%. The water contents before
the test were maintained within 1 % of the desired optimum water content.
Consequently, the backfill samples for pullout and direct shear tests were kept in
plastic containers for 24 hours before tests to achieve uniform distributions of
moisture contents. The backfill soil for both large scale direct shear and pullout tests
were compacted to 95% of maximum dry density measured by sand cone method.
3.5.2 Pullout test
Pullout test was conducted to determine displacement and structure of LLGs
reinforcement layer needed to achieve active limit state in order to exploit
reinforcements load capacity properly. In this study, normal pressures 20, 40 and 60
kPa were applied to cover the range of possible reinforcement failures (i.e. slippage
and breakage).
3.5.2.1 Pullout apparatus
Pullout apparatus has been used by several previous researchers (e.g. Chai 1992, Mir
1996, Kabiling 1997, Teerawattanasuk 1997, Wongsawanon 1998, Srikongsri 1999,
Kongkitkul 2001, Supawiwat 2002, Youwai and Bergado 2003, Prempramote 2005,
Shezad 2008 and Tin 2009). There are 15.9 mm thick steel plates around the pullout
box and 12.7 mm thick steel plate at the bottom. These plates are welded together to
form the channel section. Moreover, there are additional 12.7 mm thick steel plates
were used as removable back and front walls and 102 mm steel channels were used
to support the pullout box. The inside dimension of pullout box were 1270 mm long,
760 mm wide and 508 mm high (Fig. 3.23). The front wall is composed of two parts,
the upper and the lower one; there is a 50 mm slot between them.
44
A 225 kN capacity electro-hydraulic controlled jack was used to transmit pullout
force through the steel reaction frame in front of the pullout box. The pullout clamp
with dimensions of 100mm x 600mm x 12.7mm which is composed of two steel
plates with grooves was used to grip the reinforcement. The detail of the clamp in
the schematic diagram is illustrated in Fig. 3.24. The longitudinal bars of the steel
grids were kept between the upper plate and the lower plate to prevent slippage of
reinforcement during pulling procedure. The load cell attached between the pullout
jack and reinforcement was used to measure pullout force.
The inflated air bag put between the flexible steel plate and the top cover of the
pullout box was used to apply the normal pressure. The top cover consist of a 12.7
mm thick steel plate reacted by the steel angles. The pressure regulator valve was
connected between the air compressor and the air bag was used to adjust the normal
pressure maintained constant during the pullout test. A pullout machine is shown in
Fig. 3.25.
Pullout force was measured by a load cell connected to the 21X data logger. The
displacement of the woven Kenaf LLGs was recorded by a No.3 LVDT and a dial
gage in front of the pullout box (Fig. 3.26). High strength wires were connected to the
longitudinal bars and the other ends were connected to the LVDTs to measure the
displacement of woven Kenaf LLGs at different positions on the longitudinal bars
(Fig. 3.27). The reading values would be transferred to the 21X data logger then to the
computer. Pulling rate was controlled by the electronic control components which
were interfaced with a potentiometer pot. This was used to regulate manually the
input signal to the servo valve to adjust the velocity and direction of the hydraulic
cylinder. Using the potentiometer along with the dial gage reading and stop watch,
required pulling rate could be controlled.
3.5.2.2 Setup for the pullout test
The backfill soils was put into the pullout box then compacted into three successful
layers with the height of 0.15 m for each. The required density for each layer is 95%
of the maximum dry density which was obtained from the result of Standard Proctor
Test. Sand cone test and oven-dried method was used to check the density and water
content of each layer. The woven Kenaf LLGs was gripped with the clamp at one
end. After that, the upper front cover plate was put and the next soil layers were
poured above the sample. The requirements for density and water content of the
latter layers were the same as that of the first layer. When compaction completed,
steel plate was put above the sand then the air bag was installed. Finally, the top
cover was put and three bolted beams were placed to act as reaction equipment.
The normal pressure was applied through the air bag. To ensure that the pressure
was transmitted uniformly on the filling material, this normal pressure was kept in 1
hour before applying the pullout force. The pullout rate of the test was controlled at
1 mm/min during the test. The data logger was used to record the pullout resistance
and pullout displacement then it was transmitted readings value to the computer. The
test stopped when the sample has been pulled 25 mm (Shivashankar, 1991). The
tests were performed at three different normal pressures 20, 40, and 60 kPa and for
the same samples.
45
3.5.2.3 Pullout resistance of Kenaf LLGs reinforcement
Pullout tests on woven Kenaf LLGs were conducted at normal pressures at 20 kPa, 40
kPa, and 60 kPa, receptively in order to evaluate the contribution of the interface
frictional resistance to the overall pullout- resistance. Pullout resistance was measured
by the load cell connected to the data logger and the pullout displacement was
measured by the dial gage in front of the pullout box. The relationship between the
pullout resistances and pullout displacements of Kenaf LLGs reinforcement during
the pullout test at different normal pressures are shown in Fig. 3.28.
In general, the pullout resistance increases when normal pressure increases resulting
from the increase of confinement on the woven Kenaf LLGs reinforcement.
Moreover, the pullout resistance has been observed to increase rapidly at small pullout
displacement. After the peak values of pullout resistance at small displacement, the
pullout resistance continues to increase very slowly to the maximum pullout
resistance. The test was terminated when a maximum displacement of 80 mm was
reached for three values of normal pressures. The maximum pullout resistances
observed at the end of the test were 12.57 kN/m, 22.04 kN/m, and 26.03 kN/m for
normal pressures of 20 kPa, 40 kpa, and 60 kPa, respectively. Furthermore, the
relationships between maximum pullout resistance and normal pressure for woven
Kenaf LLGs are plotted in Fig. 3.29.
The failure mode of woven Kenaf LLGs reinforcements can be divided into two
modes, namely: slippage failure and tension failure, depending on the magnitudes of
the applied normal stresses. At lower normal stresses of 20 and 40 kPa, the failure
modes of Kenaf LLGs reinforcements were slippage failure with no visible damages
in the reinforcement. The pullout force increased with displacement as shown in Fig.
3.28. In contrast, the Kenaf LLGs were damaged at normal stress of 60 kPa because
the failure mode was tension failure. Due to tension failure of the reinforcement, the
pullout force reduced with displacement.
3.5.3 Large scale direct shear test with LLGs
Direct shear testing apparatus was modied to be used as the interface shear
equipment in accordance with ASTM D 5321 (Standard Test Method for
Determining the Coefficient of Soil and Geosynthetic or Geosynthetic and
Geosynthetic Friction by the Direct Shear Method). The tests were carried out up to
high relative displacements to evaluate the residual shear resistance. The inner
dimensions of rectangular shear boxes were 300 mm long, 300 mm wide, 50 mm deep
for upper and lower box. This direct shear apparatus has movable lower shear box
with horizontally supported upper box and a rigid load plate to apply the normal
stress. The normal and shear stresses were computed after the area of the contact
surface between the tested materials has been corrected. Two dial gauges were used to
measure the horizontal displacement and a dial gauge installed at the top of the
loading plate measured the vertical displacement. The bottom shear box was moved
relative to the xed upper shear box under a constant normal compressive stress. A
photograph of device is shown in Fig. 3.30.
The large-scale direct test conducted for evaluating the friction between backfill soil
only and between kenaf LLGs and backfill soil. The backfill sand only being tested
was used initially as the substrate in the lower and upper box. Normal loads were
applied to the backfill sand specimen and the specimen sheared across the pre-
determined horizontal plane between the two halves of the shear box. The
displacement rate 1 mm/min was used for all the tests. The normal loads applied in
46
the direct shear tests for this study are 40 kPa, 80 kPa, and 120 kPa, respectively. The
shearing was carried out until a displacement of 50 mm was achieved. Measurements
of shear load, shear displacement and normal displacement were recorded.
Kenaf LLGs was prepared by dimension of 300 mm by 500 mm and it was folded at
one end and placed between compacted backfill soils which compacted to optimum
moisture content in the upper and lower direct shear box, as shown in Fig. 3.31.
During the shear tests, the soil is forced to slide along a Kenaf LLGs under a constant
rate of displacement, while a constant load is applied normal to the plane of relative
movement. The normal and shear forces applied to determine the point of failure. The
test was repeated under three different normal pressures at 40 kPa, 80 kPa, and 120
kPa, respectively.
In this study, large scale direct shear tests were carried out to determine the value of
friction angel () by shearing Kenaf LLGs with sand backfill material and only sand
backfill material. The tests were conducted under three different normal pressures at
40 kPa, 80 kPa, and 120 kPa, receptively. The tests were ended at maximum
displacement of 150 mm. After the tests, the relationship between shear stress and
horizontal displacement were plotted in Fig. 3.32 for sand backfill only and Kenaf
LLGs with sand backfill. From the result of only silty sand backfill specimen, the
peak friction angles was 35.63 degrees and cohesion was 11.33 kPa. In case of Kenaf
LLGs with silty sand backfill, the peak friction angle was found to be 27.66 degrees
and cohesion was 9.26 kPa. The relationship between shear stress and normal stress is
shown in Fig. 3.33.
3.5.4 Interaction between soil and Kenaf LLGs reinforcement
Shear parameters from soil to reinforcement were obtained from the direct shear and
pullout tests. The efciency values of Kenaf LLGs reinforcement by the direct shear
test on cohesion and on friction could be determined by using the following equations
(Bergado and Chai, 1994):
E
c
= 100
c
ci
(1)
E
= 100
tan
tan
(2)
where E
c
is the efciency of Kenaf LLGs reinforcement on cohesion; E
is the
efciency of Kenaf LLGs reinforcement on friction; c
i
is the adhesion between soil
and Kenaf LLGs reinforcement; c is the cohesion between soil and soil; is skin
friction angle between soil and Kenaf LLGs reinforcement from shearing resistance
test between Kenaf LLGs and sand backfill material and is friction angle between
soil and soil. The aforementioned parameters were obtained from large scale direct
shear test results as plotted in Fig. 3.33. From this study, the efciency of Kenaf LLGs
reinforcement on cohesion, E
c
can be calculated as 81.73% and the efciency of
Kenaf LLGs reinforcement on friction, E
n
is the applied normal stress or confining stress. The pullout interaction coefcients,
R
in
, of Kenaf LLGs are 1.11, 1.07 and 0.88 for applied normal load as 20 kPa, 40 kPa
and 60 kPa, respectively.
The interface parameters from direct shear and pullout tests are tabulated in Table 3.4.
Since the data from large scale direct shear test was lower, its interface coefficient 0.8
was used for analyses.
3.6 Summary and Concluding Remarks
Natural fiber woven LLGs were investigated to find the suitable pattern for geotextile
application. Based on the results of tensile test, the tensile strength for knot-plain,
hexagonal and plain patterns of Kenaf were higher than water hyacinth. Plain pattern
is the most appropriate pattern due to the highest tensile strengths. Thus, plain pattern
of Kenaf can be applied for soil reinforcement and plain pattern of water hyacinth can
be applied for soil erosion control in geotechnical field.
Based on grain size distribution curve, this backfill sand can be classified as silty
sand. Based on standard Proctor compaction test, the backfill sand has 18.1 kN/m
3
of
maximun dry unit weight and 10% of optimum moisture content. Large scale direct
shear tests were performed to find the interface strength of Kenaf woven LLGs with
sand backfill material as well as the sand backfill materials under different normal
confining pressures of 40, 80, and 120 kPa. Furthermore, pullout tests were performed
using normal confining pressures of 20, 40 and 60 kPa. Slippage failure occurred
below the normal confining pressure of 40 kPa and tensile failure occurred at higher
normal confining pressures. The interaction coefficients of Kenaf LLGs decreased as
increasing the normal stress. Since the data from large scale direct shear test was
lower, its interface coefficient 0.8 was used for simulate the behavior of Kenaf LLGs
reinforcement.
48
According to the plain pattern of woven Kenaf LLGs, tensile force can be provided to
the soil if sufficient bond between soil and reinforcement is mobilized. The silty sand
embankment materials can develop the interface strengths with Kenaf LLGs. Based
on the results of the interaction between the Kenaf LLGs reinforcement and backfill
soil, the direct shear and pullout resistances indicate that the Kenaf LLGs can be used
for temporary soil reinforcement applications.
Water hyacinth LLGs which has two different opening size dimensions of 8x8 mm
and 12x12 mm in plain pattern can be obtained for erosion control application. Based
on laboratory test results, the tensile strength of 8x8 mm opening size of water
hyacinth LLGs has higher strength than 12x12 mm opening size. Hydraulic
transmissivity test were performed, the flow rates per unit width within the plane of
geotextile decreased when the normal pressure were decreased for both 8 mm and 12
mm opening size of woven water hyacinth LLGs as expected. Permittivity test were
performed on woven water hyacinth LLGs, the volumetric flow rate of water per unit
cross-sectional area per unit head has increased at higher hydraulic gradient as
expected. Thus, the water hyacinth LLGs can be applied for erosion control.
49
CHAPTER 4
FULL SCALE REINFORCED EMBANKMENT TEST
4.1 General
The full-scale test embankment was constructed using silty sand backfill materials and
reinforced by 4x4 mm opening size of woven Kenaf LLGs which used silty sand for
backfill material on soft Bangkok clay. The embankment was instrumented with
settlement plates, piezometers, and wire extensometers. The instrumentations in the
subsoil that were installed prior to the construction of the woven Kenaf LLGs
reinforced embankment. Consequently, the embankment behavior was monitored
concerning its surface settlement, subsurface settlement, excess pore water pressure
and deformation of reinforcements.
4.2 Test Location and Soil Profile
The test embankment was constructed at the northern part of the campus of the Asian
Institute of Technology (AIT) in KlongLuang, Pathumthani, Thailand in February,
2011. The layout of the test embankment and the location of field vane tests and
boreholes are also shown in Fig. 4.1.
The general soil profile and soil properties of the subsoil in the uppermost three layers
at the AIT campus are presented in Fig. 4.2 (Jamsawang, 2009). The uppermost 10 m
can be divided into 3 layers. The uppermost 2m thick layer is the weathered crust
consisting of heavily overconsolidated reddish brown clay. The second layer from 2 m
depth down to 8 m depth is the soft clay layer. The third layer is the medium stiff clay
layer with silt seams and fine sand lenses exists at 8 to 10 m depths. Underneath the
medium stiff clay layer is stiff clay layer which extends to 30 m depth. The soft
Bangkok clay at the AIT campus is comprised of 1 to 8% sand, 20 to 40% silt and 50
to 70% clay (Kanjanophas, 1969; Dumn, 1977). The specific gravity of the soil
particles on an average is about 2.7. The mineralogical composition of the soft
Bangkok clay was found to contain: illite (35%), kaolinite (10%), montmorillonite
(30%), quartz (15%) and other unidentified mineral (10%) (Cox, 1968). These
composition varies with locations.
4.3 Laboratory and In-Situ Testing Program
4.3.1 Investigation soil index properties
The soil samples used in this study obtained from northern part of Asian Institute of
Technology where embankment was constructed. The soft clay samples extracted
from a depth of 3 m and 6 m of soft clay layer below weathered crust layer.
Undisturbed samples were obtained by digging the soils up to required depth and
collect the soil samples and then placed in the humidity room. The color of sample
was greenish gray. In this study, index properties of soft Bangkok clay sample
including vane shear test in the field, and consolidation test were conducted.
Properties of soil at 3m and 6m depth are tabulated in Table 4.1.
50
4.3.1.1 Vane shear tests
The in-situ strength of the subsoil was measured by performing the field vane shear
test in the Asian Institute of Technology (AIT) campus near the embankment area as
shown in Fig. 4.3. The vane shear test basically consists of placing a four-bladed vane
in the undisturbed soil at the desired depth and rotating to cause a cylindrical surface
to be sheared by the vane. Once the vane was in position in the undisturbed soil, a
torque was applied to the vane at a maximum rate of 0.1 degree/sec. The undisturbed
strength was ascertained from the maximum torque required to shear the cylindrical
surface. A Geonor vane apparatus with blade dimensions of 130 mm height and 65
mm radius was used. Measurements of the vane strength were made at 3 m and 6 m
depth. Corrected vane shear undrained strength, S
u
was 20 and 17 kPa at 3 m and 6 m
depth, respectively. The undrained strength increases linearly with the depth which
has similar results to Jamsawang (2009).
4.3.1.2 Consolidation test for undisturbed sample
Consolidation test was performed in accordance with ASTM D2435 (Standard Test
Method for One-Dimensional Consolidation Properties of Soils) to determine the
magnitude and rate of volume change that a laterally confined soil specimen
undergoes when subjected to different vertical pressures. Metal rings was pressed into
the sample and continued until the sample protruded a short distance through the
bottom of the ring. Then, the sample was trimmed and placed to consolidation cells.
Porous stones and the filter papers was placed on the top and bottom surfaces of the
test specimen. Water was added until the specimen is completely covered and
saturated. In this study, samples from 3 m and 6 m depths were performed different
vertical pressures of 12.5, 25, 50, 100, 200, 400, and 800 kPa and unloading pressures
of 200 kPa.
Coefficient of conslidation, Cv is a very important parameter for prediction of
settlement. The value of Cv decreases as the applied pressure increases, then increases
as applied pressure is about 100-200 kN/m
2
. From the measured data, the relationship
between applied pressure and void ratio at 3m and 6m depth are also plotted as well.
From the recorded data, maximum past pressure of the soil at 3 m depth is 73 kPa, the
overconsolidation ratio (OCR) is 1.55, compression ratio (CR) is 0.178 and
recompression ratio (RR) is 0.042 as found from Fig. 4.4 and Coefficient of
consolidation (C
v
) is 0.5 m
2
/yr as found from Fig. 4.5. For clay at 6m depth, the
maximum past pressure is 85 kPa, the overconsolidation ratio (OCR) is 1.0,
Compression Ratio (CR) is 0.216 and recompression ratio (RR) is 0.034 are recorded
form the Fig. 4.6 and Coefficient of consolidation (C
v
) is 0.95 m
2
/yr as found from
Fig. 4.7, respectively. The results from consolidation were used for prediction the
settlement in the field.
4.4 Instrumentation Program
The instrumentations in the subsoil were installed prior to the construction of the
reinforced embankment wall consisting of the surface settlement plates, subsurface
settlement gauges, temporary bench marks, open standpipe, groundwater table
observation wells, , dummy open standpipe. Figures 4.8 shows the schematic plan
view layout give instrumentation of the embankment reinforcement.
51
4.4.1 Temporary bench marks and dummy area
A permanent benchmark anchored to about 200 m depth, this permanent benchmark
located about 100 m from the temporary benchmark. Precise leveling survey was also
run from the permanent benchmark to the temporary benchmark, once every week to
monitor the effects of ground subsidence. Two dummy open standpipes were installed
at the area nearby the embankment. These open standpipe piezometers comprised of
perforated steel pipe containing porous sand mixed with epoxy and were pushed
manually into the required depth by means of a special adapter and a steel pipe.
4.4.2 Surface settlements plates
Two surface settlement plates placed beneath the embankment wall at 3 m depth.
Settlements measured by precise leveling with reference to a bench mark. The surface
settlement plates consisted of a 16 mm diameter steel rod connected to a 0.4 m x 0.4
m base plate and protected by a 19 mm diameter casing pipe as shown in Fig. 4.8. The
location of the settlement points are shown in Fig. 4.9.
4.4.3 Subsurface settlements gauges
The measurement of the subsurface settlements was similar to that of the surface
settlements, i.e. by precise leveling with reference to the temporary benchmark. Four
subsurface settlements gauges were installed at 6 m depth and the rest at 3 m depth
below the general ground surface at Kenaf LLGs coated and non-coated sections as
shown in Figs. 4.8 and 4.9. The subsurface settlements gauges consisted of 250 mm
diameter steel screw head connected to a 16 mm diameter steel rod. These were
screwed down into the ground by means of 19 mm inner diameter steel tube which
was attached to the screw head by a key way (Fig. 4.11). After installation to the
desired depth the outer tube was lifted from the keyway by about 200 mm. The outer
tube was left in the ground and filled with oil to eliminate the subsequent drag down
forces from the inner rod.
4.4.4 Open stand pipe
The conventional openstandpipe hydraulic piezometers (AIT Type) as shown in Fig.
4.12 were installed to monitor the pore pressure in the subsoil. Four of these
piezometers were installed in the soft clay subsoil at 3 m and 6 m depths at kenaf
LLGs coated and non-coated side. Moreover, two dummy open standpipe piezometers
were installed at the area nearby the embankment. These open standpipe piezometers
comprised of a porous tip and were pushed manually into the required depth by means
of a special adaptor and steel pipe.
4.4.5 Wire extensometer
Wire extensometers were used to measure the displacements of the kenaf LLGs
reinforcement and the surrounding soil as well. The wire extensometer consists of an
inner 2 mm diameter high-strength stainless cable and an outer flexible PVC tube.
The PVC tube with 4 mm inner and 6 mm outer diameters enabled the free movement
of the inner wire. One end of the wire was fixed at the measured point. The other end
was connected to a counterweight of about 0.8 kg through a pulley system at the
readout board located behind the embankment (Fig. 4.10).
52
4.5 Slope Stability Analysis
Limit equilibrium analysis using the method of slices has been widely used in solving
slope stability problems. The simplified Bishops Method is a method for analysis of
the stability of slopes. It is an extension of the method of slices. In this study, the
Slide 5.0 software was used to analyze embankment on soft ground by using Bishop
simplified method. The embankment plan view is shown in Fig. 4.8, while the section
views are shown in Figs. 4.9 and 4.10. Kenaf LLGs were applied for embankment
reinforcement. The properties of Kenaf LLGs were obtained from laboratory tests
such as tensile strength, pullout tests, and direct shear tests. The most important
properties of biodegradable geotextiles such as vegetable fibres for soil reinforcement
are their high initial tensile strength (Mwasha, 2009a). Tensile force can be mobilized
to the soil if sufficient bond between soil and reinforcement is warranted. The soil
and parameters used in Slide 5.0 software are tabulated in Table 4.2
The resulting factor of safety values including: slope stability analysis with and
without Kenaf LLGs, and slope stability analysis with and without Kenaf LLGs
including applied surcharge load 25 kN/m
2
and without Kenaf LLGs including
applied surcharge load of 25 kN/m
2
. On the side slope 1:1, the values of 2.2 and 2.6
safety factors were calculated using without and with Kenaf LLGs, respectively, as
shown in Fig. 4.13 and Fig. 4.14. The reinforced embankment with Kenaf LLGs
indicated higher values of the factor of safety. Kenaf LLGs reinforcement can
stabilize the embankment on soft ground when backfill materials can develop the
interface strength. Thus, the stability of the embankment increased resulting in higher
factor of safety about 17%. Moreover, with 25 kN/m
2
applied for surcharge load, the
results show that 1.83 and 1.87 safety factors in case of without and with Kenaf LLGs
as shown in Figs. 4.15 and 4.16, respectively.
4.6 Full Scale Reinforced Embankment Test
4.6.1 Materials
The materials used in this study were Kenaf fiber yarns fabricated in plain pattern
which has dimension of 4 mm opening size. The Kenaf LLGs dimensions were 1 m
wide and 5 m long. Polyurethane coatings provided the protection against
environmental inuences, including moisture, radiation, biological deterioration or
damage from mechanical or chemical origin. Polyurethane coatings have been utilized
on many different materials, to improve their appearance and lifespan (Chattopadhyay
and Raju, 2007). In this study, there were two types of Kenaf LLGs, namely: coated
and non-coated with polyurethane. For the first time in this study, the polyurethane
coating was used to reduce water absorption and mitigate the effect of weathering on
the Kenaf LLGs increasing its lifetime. Two coated with polyurethane samples were
installed at the side of embankment and other sides were installed two non-coated
samples as shown in Fig. 4.17.
4.6.2 Embankment construction
A full scale experimental embankment was constructed to investigate the behavior
and performance of natural fibers LLGs for reinforcement and erosion control. Kenaf
LLGs was used for reinforcement in the embankment. The embankment was
constructed by using silty sand backfill until 3 m height and covered by 1m thick
compacted weathered clay at the top, back and side slope. Moreover, the side and
53
back slopes consisted of 1 vertical to 1.5 horizontal and front slope consisted of 1
vertical to 1 horizontal (Figs. 4.8, 4.9, and 4.10). Six layers of woven Kenaf LLGs
with 4 mm opening size were utilized as reinforcement with vertical spacing of 0.5 m
(Fig. 4.17). Both coated with polyurethane and non-coated Kenaf LLGs were utilized.
The polyurethane coating decreased water absorbtion and, thereby prolong the life of
Kenaf LLGs. As shown in Fig. 4.17, the LLGs reinforcements were instrumented to
measure its deformations using high strength wire extensometers riveted to the LLGs
and encased with plastic tubes. The other ends of the high strength wire extensometers
were connected to roller and pulley system in the readout board (Figs. 10 and 21).
The instrumentation in the subsoil were installed prior to the construction of
reinforced embankment consisting of the surface settlement plates, subsurface
settlement plates at 3 and 6 m depths, and piezometers at 3 and 6 m depths (Figs.
4.8, 4.9, and 4.10). Two surface settlement plates were installed beneath the
embankment at 3 m depth. The settlements measured by precise leveling with
reference to a bench mark. The measurements of the subsurface settlements were
similar to that of the surface settlements. Four subsurface settlements gauges were
installed at 3 m depth and at 6 m depth below the ground surface at Kenaf LLGs
coated and non-coated sections as shown in Fig. 4.9. The silty sand backfill was then
spread out in 0.15 m compaction lifts to the total thickness of 0.5 m to the required
density of 95% of standard Proctor and compacted with roller compactor. The
degree of compaction and the moisture contents were checked regularly at several
points with sand cone tests. Thereafter, the construction was repeated, layer by layer,
until 3 m embankment height with the 6 layers of Kenaf LLGs reinforcements.
Then, a 1 m thick compacted weathered clay covered the silty sand backfill
embankment at the top as well as at the side and back slopes. The completed
embankment is shown in Fig. 4.18. Water hyacinth LLGs which has opening size of
8 mm and 12 mm were used for erosion control on the embankment. Two type of
LLGs including coated water hyacinth LLGs and non-coated water hyacinth LLGs
with polyurethane.
A full scale embankment on was constructed in AIT campus, Thailand. The project
site is located at AIT campus about 50 kilometers north of Bangkok, the capital city of
Thailand, and it is underlain by the known soft Bangkok clay. The embankment
construction was completed within 15 days in February 2011. The main objective for
the construction of the test embankment was to evaluate the performance of the
natural fiber geotextiles reinforcements.
4.7 Behavior of the Full Scale Test Embankment
The Kenaf LLGs reinforcement effect at critical time period for stability or very
shortly, after construction. The stability of the system will improve in time and so the
stabilizing force, which needs to be provided by the geotextile, will diminish. After a
certain time (typically between several months and a few years) the embankment will
be stable without any assistance from the geotextile due to the process dissipation of
pore water pressure the foundation gain shear strength (Sarsby, 2007). The monitoring
data of the Kenaf LLGs reinforced embankment behavior during and after
construction phases were obtained to evaluate its performance.
54
4.7.1 Observed and predicted surface and subsurface settlements
The observed surface and subsurface settlements of the test embankment are shown in
Fig. 4.19. This figure shows the relationship between the settlement distributions with
time. During construction, the rates of settlement slowly increased and immediate
elastic settlements occurred. After construction, the rates of settlement rapidly
increased up to 250 days from the end of construction. At 250 days after construction
period, the settlements on coated Kenaf LLGs side at the surface, 3 m and 6 m depths
were 295mm, 179mm and 79mm, respectively. The corresponding of settlements in
non-coated Kenaf LLGs side were 298mm, 183mm and 82mm, respectively. Thus,
similar magnitudes and settlement rates were obtained. Moreover, the surface
settlements were higher than the subsurface settlements at 3 m and 6 m depths as
expected.
It can be seen the settlement of coated and non-coated Kenaf LLGs reinforcements,
similar magnitudes of settlement rate were obtained. Moreover, the rate of surface
settlement is higher than the rates of subsurface settlement at 3m and 6m depths,
respectively.
The computed time settlement curves computed using Terzaghis one-dimensional
consolidation theory, Skempton and Bjerrum (1957) and Asaoka (1978) methods were
compared to the average total consolidation settlements. The comparison of observed
and predicted settlement at the surface, 3m and 6m depths are illustrated in Figs. 4.20,
4.21 and Fig. 4.22, respectively. From Asaoka (1978) method, the total surface
settlement was obtained as 300mm at 250 days. The total subsurface settlement at 3m
and 6m depths were 180mm and 98mm, respectively, at 250 days, from the end of
construction. According to Skempton and Bjerrum (1957) method, the consolidation
behavior is 30 % overestimated from measured values in the field. The correction
factor () decreases with increasing overconsolidation ratio (OCR). The correction
factor () at the surface, 3m and 6m depths was 0.7. Settlements computed using
Asaoka (1978) method agreed with the observed data while the settlements computed
using Skempton and Bjerrum (1957) method closely followed the observed data. The
Terzaghis one-dimensional consolidation method overpredicted the observed values
as expected.
4.7.2 Observed and predicted excess pore water pressures
The performance of full scale test embankment constructed on soft Bangkok
foundation was indicated by the building of excess pore water pressures. Afterwards
consolidation settlements occurred and pore water pressure started to dissipate with
time. The build up and dissipation of excess pore pressure were monitored by using
open standpipe piezometers. Four open standpipe piezometer were used to monitor
the excess pore water pressure beneath the reinforced embankment at 3 m and 6 m
depths. The excess pore water pressures dissipated very fast from 15 to 120 days after
construction and dissipated with slower rate thereafter. After 250 days, the excess
pore water pressures decreased to 10 kPa and were almost constant with time. The
excess pore water pressures below coated and non-coated Kenaf LLGs sides at 3m
were 37 kPa and 35 kPa, respectively. The corresponding values at 6m depths were 33
kPa and 32 kPa, respectively. The maximum excess pore water pressures occurred at
55
7 days just after the full height of embankment was completed. The excess pore
pressures below coated and non-coated LLGs sides were similar in magnitudes.
The predicted excess pore water pressures below the embankment using Terzaghis
one-dimensional method and by using Skempton and Bjerrum (1957) three-
dimensional method were compared with the average observed field data as plotted in
Figs. 4.23 and 4.24 for 3m and 6m depths, respectively. The excess pore water
pressures by Terzaghis one-dimensional method were 70 kPa and 60 kPa at 3m and
6m depth, respectively, while the corresponding values from the Skempton and
Bjerrum (1957) three-dimensional method were 42 kPa and 38 kPa at 3m and 6m
depths. The excess pore water pressures from one-dimensional method were
overpredicted while the prediction from the Skempton and Bjerrum (1957) method
agreed with the observed data at 3m and 6m depths.
4.7.3 Observed deformation and stress of reinforcement
To measure the deformation of Kenaf LLGs, high strength wire extensometers were
utilized. The measurement points were located at 1.25m, 2.50m, 3.75m and 5m from
the edge of Kenaf LLGs reinforcement in the silty sand backfill material zone. The
extensometers were installed at all the layers (6 layers) of Kenaf LLGs
reinforcements. The observed deformations were found to be continually changing in
response to the settlements of the test embankment. The observed deformations of the
Kenaf LLGs reinforcement were monitored once a week since the end of construction
and every month afterwards until 250 days.
The deformations along the length of Kenaf LLGs reinforcements were plotted with
the height of full scale test embankment in each layer at 14 and 250 days after
construction as shown in Fig. 4.25. The average deformations at the middle of Kenaf
LLGs reinforcements were higher than the edges in both coated and non-coated Kenaf
LLGs following the settlement pattern. Moreover, the deformations in each layer
increased with time. The maximum deformations of Kenaf LLGs reinforcement layers
were almost similar in magnitudes but far lower than the elongation at break.
The stresses in Kenaf LLGs reinforced from the observed field data were calculated at
14 days and 250 days after construction based on correlations with the laboratory
pullout test results as shown in Fig 4.326. The stresses of Kenaf LLGs reinforcement
at the bottom of full scale test embankment were almost similar at the top following
the deformation patterns (Figs. 4.25). In addition, the stresses of Kenaf LLGs
reinforcement at the middle have higher stresses than at the edges following its
deformations in response to the settlement pattern.
4.7.4 Effect of Kenaf LLGs coated and non-coated with polyurethane
From the results of this study, the polyurethane coating did not affect the embankment
behavior. The results of overall behavior were similar in both coated and non-coated
Kenaf LLGs section. The Kenaf LLGs improved stability of embankment at critical
period immediately after construction.
56
4.8 Summary and Concluding Remarks
For the first time, based on the results of the interaction between the Kenaf LLGs
reinforcement and backfill soil, the direct shear and pullout resistances indicate that
the Kenaf LLGs can be used for temporary soil reinforcement applications. The full
scale reinforced test embankment was constructed at the AIT campus and monitored
to evaluate the performance of the woven Kenaf LLGs natural fibers geotextiles
reinforcements. In order to monitor the behavior of the embankment, it was
extensively instrumented both in soft subsoil and within the embankment itself. The
embankment construction began on February 6, 2011 and completed after 10 days on
February 15, 2011. The monitoring period during construction was measured
everyday day and after construction was measured weekly for 2 months thereafter,
every months until 250 days after construction due to flooding crisis in Thailand in
the later part of 2011.
After construction, the rates of settlement rapidly increased until 250 days from the
end of construction. The settlements predicted by Asaokas method agreed well with
the measured values in the field while the consolidation behavior was overestimated
by the Terzaghi one-dimensional consolidation method. The calculated settlements
using the method of Skempton and Bjerrum 3D method closely followed the observed
values. Rapid dissipation of excess pore water pressures were observed at the end of
construction. The excess pore water pressures slowly dissipated and tend to be
constant after 250 days. The predicted pore pressures using the Skempton and
Bjerrum 3D method were in good agreement with the measured excess pore water
pressures. The time to reach 90% average degree of consolidation (t
90
) based on the
observation was 250 days. The Kenaf LLGs have estimated working life about 1 year
(365 days). Thus, the embankment still has reinforcement at 250 days corresponding
to 90% for degree of consolidation (t
90
) was less than the life of Kenaf LLGs.
In addition, at middle of the embankment, the Kenaf LLGs reinforcement had higher
deformations in both coated and non-coated Kenaf LLGs and the deformations
increased with time. The stresses of Kenaf LLGs reinforcement at the middle have
higher stresses than at the edges following its deformations in response to the
settlement pattern. Furthermore, the mobilized tensile force (T
mob
) was 20 kN/m at the
estimated 4% of strain from wide width tensile test and and 22 kN/m from pullout
test. Using the mobilized tensile force of 20 kN/m in the reinforcement, the factor of
safety increased by 17 %.
The polyurethane coating did not affect to the embankment behavior as its purpose is
to extend the life of Kenaf LLGs. The results of overall behavior were similar in both
coated and non-coated Kenaf LLGs section. The results presented are limited to the
short-term of properties of Kenaf LLGs reinforcement. Kenaf is biodegradable
material its tensile strength can be loss due to in situ deterioration. Based on the
performance of the test embankments, woven Kenaf LLGs can be applied for soil
reinforcement on soft clay for short term application.
57
CHAPTER 5
2D FINITE ELEMENT MODELING OF
FULL SCALE TEST EMBANKMENT
5.1 General
The finite element method (FEM) has been used in many fields of engineering
practice. Moreover, the FEM are using widely for analyzing geotechnical problems.
The full scale reinforced embankment on soft foundation has been analyzed using
numerical model based on the finite element method. Natural fibers are increasingly
applied in geotechnical application for reinforcement. Natural fibers used in this study
consist of woven Roselle or Kenaf (Hibiscus sabdariffa var altissima). This chapter
presents the modeling of the full scale test embankment reinforced with natural fiber
by PLAXIS software on two-dimensional (2D) under plane strain condition.
Moreover, the results from PLAXIS computer program are compared with the
observed field data from full scale reinforced embankment in term of settlement,
excess pore water pressure and deformation of reinforcement.
5.2 Material Parameters for Modeling
PLAXIS software was performed to support a number of material models to simulate
the behavior of soil. The models and parameters for analysis are described in detail in
the Material Models manual. The summary of some essential features of the program
was obtained.
5.2.1 Mohr-Coulomb model
In general, this model is used as a first approximation of soil behavior. It is
recommended to use for considering the problem at the first analysis. The five
parameters involve in this model, namely Young's modulus, (E), Poisson's ratio, (),
the cohesion, (c), the friction angle, (), and the dilatancy angle, (). The advantage of
the Mohr-Coulomb in the undrained condition with the effective parameters is that the
shear strength will automatically increase with consolidation.
The Mohr-Coulomb model also has limitation such as the increase of stiffness with
depth can be taken into account but this model dose neither includes stress-
dependency nor stress-path dependency of stiffness. Moreover, in the undrained
conditions, this model can be used with the friction angle, set to 0 and the cohesion,
c set to c
u
(s
u
) to enable control the undrianed shear strength but this model does not
automatically increase the shear strength with consolidation.
5.2.2 Soft soil model
This is Cam-clay type model, which can be used to simulate the behavior of soft soil
like normally consolidated clays and peat. This was proposed by Vermeer and
Brinkgreve (1995). The model performs best in situations of primary compression the
SSM requires the following material constants: * = the modified compression index
taken from (
v
ln p) plot. This constant is not the same with the intrinsic
compression index as defined by Burland (1990); * = modified swelling index (slope
58
of swelling line of above mentioned (
v
ln p) plot; c = cohesion; = friction angle
and = dilatancy angle.
5.2.3 Basic parameters of mohr-coulomb model and soft soil models.
The Mohr-Coulomb model requires a total of five parameters. Generally, the five
parameters can be obtained from basic tests on soil samples. These parameters and
their standard units are described below:
E : Young's modulus
[kN/m
2
]
: Poisson's ratio [-]
: Friction angle [
o
]
c : Cohesion
[kN/m
2
]
: Dilatancy angle [
o
]
The Soft Soil Creep model requires the similar parameters as the Mohr-Coulomb
model. In addition, it involves parameters more than the Mohr-Coulomb model. In
brief, the Soft Soil Creep model requires the following parameters
Failure parameters as in the Mohr-Coulomb model:
c : Cohesion
[kN/m2]
: Friction angle [
o
]
: Dilatancy angle [
o
]
Basic stiffness parameters:
* : Modified swelling index [-]
* : Modified compression index [-]
* : Modified creep index [-]
Advanced parameters (it is advised to use the default setting):
ur
: Poisson's ratio for unloading-reloading (default 0.15) [-]
0
NC
K : 'xx / 'yy stress ratio in a state of normal consolidation [-]
M :
0
NC
K -related parameters [-]
5.2.4 Youngs modulus (E)
The Youngs modulus, E is the basic stiffness modulus in the elastic model and the
Mohr-Coulomb model that has the dimension of stress. In soil mechanics, the initial
slope and is the secant modulus at 50% strength usually indicated as E
0
and E
50
,
respectively. In case of sand and near normally consolidated clays subjected to
loading, it can be more appropriate to use E
50
.
59
5.2.5 Poissons ratio ()
Generally, the value of Poissons ratio is in the range between 0.3 to 0.4 for most soil.
In case of undrained behavior, it is recommended to input the effective value of
Poissons ratio and to select the undrained in the type of behavior and the value of
Poissons ratio should be less than 0.35.
5.2.6 Cohesion (c)
The unit of cohesive strength has the dimension of stress. For cohesionless like sands
(c = 0), it is recommended to use the small value (use c > 0.2 kPa) to avoid the
complications.
Mohr-Coulomb model, the cohesion parameter may be use to model the effective
cohesion c and effective friction angle for both drained and undrained behaviors.
For the advantage of undrained behavior type, is that the strength will increase after
consolidation. On the other hand, undrained shear strength as obtained from the model
may incorrect from the reality because of different in the actual stress path being
followed.
5.2.7 Friction angle ()
The friction angle, (phi), is specified in degrees (
o
). The friction angle mostly
determines the shear strength by means of Mohr's stress circles.
5.2.8 Dilatancy angle ()
The dilatancy angle, (psi), is entered in degrees (
o
). For soft clay and silt tend to
show little dialatancy ( 0) but for sand depend on the density and friction angle
which about ( 30
o
)
5.3 Material Behavior Models
The modeling of full scale reinforced embankment that construct on soft foundation
used the PLAXIS software based on the finite element method for a realistic
simulation of the construction sequences and the inclusion of reinforcement and the
interface elements at any stage of the analysis without any significant changes in the
input data and finite element mesh. The finite element model of reinforced
embankment consisted of LLGs reinforcements, soil-to-reinforcement interaction
elements and connection elements. The fifteen-node triangular element was used in
model simulation The Soft Soil Model (SSM) was used to model the behavior of the
soft clay foundation which is divided into 3 layers. The elastic perfectly plastic with
MohrCoulomb failure criterion was used to model the behavior of the weathered
crust, medium stiff clay, and the backfill material. The linear elastic material model
was used to model Kenaf LLGs reinforcement. The model parameters are presented
following.
5.3.1 Backfill material
Silty sand was used for backfill soil in the embankment. The MohrCoulomb model
with drained behavior was used for the backfill material. The friction angle and
cohesion of silty sand were obtained from large-scale direct shear test (cohesion, c=
60
11.3 and friction angle, = 35.63). The material parameters that use for analysis is
elasticity, E = 7500 kPa, Poissons ratio,
= 0.3. The material properties of the
backfill for the FEM simulations are tabulated in Table 5.1.
Weather crust of soft Bangkok clay was used for backfill soil in the embankment. The
MohrCoulomb model with drained behavior was used for the backfill material. The
friction angle and cohesion of weather crust were obtained from large-scale direct
shear test (cohesion, c= 10 and friction angle, = 26). The material parameters that
use for analysis is elasticity, E = 3000 kPa, Poissons ratio,
= 0.25. The material
properties of the backfill for the FEM simulations are tabulated in Table 5.1.
5.3.2 Weathered crust
At the construction site, the upper layer 0 to 2m depth consists of the weathered crust,
which is heavily overconsolidated. The elastic perfectly plastic with MohrCoulomb
failure criterion model was used for this soil with a constant value for the Poissons
ratio,
= 0.25. The strength parameters were obtained from the existing test data on
Bangkok clay by using cohesion, c= 10 and friction angle, = 23, elasticity, E =
3000 kPa (Balasubramaniam et al, 1978). The modulus of elasticity and strength
parameters used for the weathered crust are shown in Table 5.1.
5.3.3 Soft clay
The Cam Clay model has been widely used for representing the stressstrain
relationship of the soft Bangkok clay, which is normally consolidated to lightly
overconsolidated. In the PLAXIS software used, the Modified Cam Clay model which
is a Cam Clay type model and the Soft Soil Model are available. For clay with OCR
<2 which is slightly overconsolidated, the Soft Soil model is more appropriate rather
than modified Cam Clay due to the mesh sensitivity of softening behavior in the latter
model and the improved behavior in one-dimension compression of the former
(Vermeer and Brinkgreve, 1995). The friction angle of the soft clay layer was
assumed to be 23 as suggested by Balasubramaniam et al. (1978) for soft Bangkok
clay.
The FEM input parameters for the soft clay layers beneath weather crust which is
between 2.00 to 8.00m. The soft soil layer can be divided into three layers for
analysis. The strength parameters were determined based on one-dimensional
consolidation tests data on Bangkok clay by using strength parameters cohesion, c= 3
and friction angle, = 23. The values for analysis were as follows: modified
compression index, * = 0.14, modified swelling index, * = 0.028. The modulus of
elasticity and strength parameters for the soft clay layers are presented in Table 5.1.
5.3.4 Medium stiff clay
The soil layer from 8 to 10 m depth is referred to as medium stiff clay layers. These
are modeled using the Mohr-Coulomb Model (MCM) with undrained behavior. The
strength parameters were obtained from the previous test data on Bangkok clay
(Balasubramaniam et al, 1978). The FEM input parameters that use for analysis is
cohesion, c= 10 and friction angle, = 25, elasticity, E = 5000 kPa, Poissons ratio,
= 0.25.
61
5.3.5 Stiff clay
The stiff clay layer is the final layer that is 10 to 30 m depth. The Mohr-Coulomb
Model (MCM) with undrained behavior is used for analysis in this soil layer. For
FEM analysis is used modulus of elasticity, E = 9000 kPa, Poissons ratio,
= 0.25
and strength parameters cohesion, c= 30 and friction angle, = 26.
5.3.6 Kenaf LLGs reinforcement
The Kenaf LLGs reinforcement was modeled as geogrid in the software. This element
needs only an axial stiffness, EA, where E is the modulus of elasticity of the material
and A is the cross-sectional area per unit width in out of plane. The axial stiffness
was obtained from wide width tensile test as shown in Fig. 3.7. The relation between
tensile strength and elongation, the tensile strength in machine side has higher than
the tensile strength in cross side. As a result, the average tensile strength was 20
kN/m, and the elongation was 15%.The properties of LLGs reinforcement used in this
analysis is tabulated in Table 5.2.
5.3.7 Soil/LLGs interface model
The elastic perfectly plastic model was used to simulate the constitutive relation of the
soil/geogrid interface. In the finite element software used, the compression modulus is
related the shear modulus assuming a fixed value of 0.25 for the effective Poissons
ratio. The shear modulus and strength parameter are automatically calculated from the
soil parameter using the interaction coefficient, R. The direct shear interaction mode
dominates the behavior at working condition with some zone having pullout
interaction mode (Alfaro et al., 1997).
In the Kenaf LLGs reinforce embankment, the interaction coefficients of silty sand
backfill and Kenaf LLGs can be determined into two types: a) direct shear resistance
and b) pullout resistance. Direct shear resistance can be investigated as backfill sliding
over the reinforcing material. On the other hand, pullout resistance is the pulling of
reinforcements out from the backfill material is shown in Fig. 3.28 and Fig. 3.29. Both
direct shear resistance and pullout resistance can be investigated by using various soil
backfill materials and a range of normal stresses, respectively. In this study, the
interaction coefficient was obtained from large-scale direct shear tests which
determined from the relationship between shear stress and horizontal displacement are
shown in Fig. 3.32 and the relationship between shear stress and normal stress are
shown in Fig. 3.33. The PLAXIS manual recommended using the interaction
coefficient from direct shear. The interaction coefficient of silty sand backfill and
Kenaf LLGs is 0.8. The mobilized tensile force (T
mob
) was 20 kN/m at the estimated
4% of strain from wide width tensile test.
5.4 Finite Element Simulation by Plaxis Software
The numerical modeling of the full-scale test embankment was performed using a
finite element software. The program allowed for plane strain and axisymmetric
idealizations including simulation of the construction sequences. The embankment
was modeled as a plane strain, two-dimensional problem by using PLAXIS 2D
version 2010.01 for the simulation finite element model.
62
Under the staged construction feature of the software, the program allowed for the
incremental fill placement to be simulated. The zone near the face and the top layer of
the subsoil, high stress gradient were considered and the mesh was refined in these
zones. The in-situ stresses in the foundation soil were generated by the K
o
procedure.
The backfill which was divided into 6 layers for Kenaf LLGs, as in the field, was
placed on the foundation soil layer-by-layer. After the placement of a compacted fill
layer the reinforcement was placed on top of the layer before the next layer was
added. The compacted fill in a given layer was assigned the material parameters
according to the stress state induced after the addition of the layer. During this
construction stage, undrained analysis was applied to simulate the layer by layer
construction. After the completion of the full height of the embankment, drained
analysis was applied to simulate the consolidation process.
5.4.1 Plaxis 2D numerical simulation of full scale test
For two-dimensional simulation, PLAXIS 2D version 2010.1 was used to model the
embankment as shown in Fig. 5.1. PLAXIS software allows for plane strain
idealization including simulation of construction sequence. The full scale test
embankment was modeled as a plane strain, two-dimensional problem for finite
element analysis consist of x and y directions. Defining boundary model, boundary
model was created at distance of eight times the width of reinforced embankment in
x-direction. In addition, distance of four times the width of reinforced embankment
was created in y-direction. For boundary condition, the nodes at the bottom boundary
was fixed, at the boundary sides, the node was fixed only in a horizontal direction.
Under the staged construction feature of the software, the program allowed for the
incremental fill placement to be simulated. The finite element model of reinforced
embankment consisted of geotextile reinforcements, soil-to-reinforcement interaction
elements and connection elements. The fifteen-node triangular element was assigned
in model simulation with very fine mesh. The soft soil model (SSM), which is similar
to the Cam Clay Model, was assigned to model the behavior of a soft clay foundation.
The linear elastic material model was used to model Kenaf LLGs reinforcement. The
elastic perfectly plastic with Mohr-Coulomb failure criterion was used to model the
behavior of weathered crust, medium stiff clay, stiff clay and embankment material.
The compacted fill in a given layer was assigned the material parameters according to
the stress state induced after the addition of the layer. During this construction stage,
undrained analysis was used to simulate the layer by layer construction. After the
completion of the full height of the embankment, drained analysis was used to
simulate the consolidation process. The material model is shown in Table. 5.1.
5.4.2 Stages of construction
The numerical simulation of Kenaf LLGs reinforced embankment was divided into 2
sections, the first section of construction which has 8 stages during 8 days, as well as
next section from phase 9 to phase 22 were investigated consolidation within a period
500 days. The numerical procedure for creating the reinforced embankment was done
by placing silty sand backfill material and inserting the Kenaf LLGs reinforcement at
interval of 0.5 m vertical spacing per stage until 3 m height with six layers of plain
pattern of Kenaf LLGs, Then, the clay backfill with thickness of 1 m which covered
the silty sand backfill for completion. In numerical model, the front side slope of full
scale test embankment consisted of 1 vertical to 1.5 horizontal and back slope
63
consisted of 1 vertical to 1 horizontal. The flow chart on the overall process was
involved in running PLAXIS software is shown in Fig. 5.2.
5.5 Finite Element Back-Analyses of Parameters
The PLAXIS finite element software was used for back-analysis of the full scale
embankment test with LLGs reinforcements. Parameters from previous researchers of
reinforced embankment on soft soil foundation at the Asian Institute of Technology
(AIT) were used for back-analysis. Appropriate parameters were selected for
reinforcement models, estimation of soil permeability of the soft clay foundation, and
the interface parameters between the backfill and Kenaf LLGs reinforcement
corresponding to the interaction mechanism. The predicted values included
settlements, excess pore water pressures, and the deformations of Kenaf LLGs
reinforcements. The predictions were subsequently compared with the corresponding
observed values. Flow chart on back analysis process involved in running PLAXIS
software is shown in Fig. 5.3.
5.6 Results and Discussions
The full-scale test embankment was constructed using silty sand backfill materials and
reinforced by woven Kenaf LLGs which used silty sand for backfill material on soft
Bangkok clay. The embankment was instrumented with settlement plates,
piezometers, and wire extensometers. Consequently, the embankment behavior was
monitored concerning its surface settlement, subsurface settlement, excess pore water
pressure and deformation of reinforcements. The finite element input soil parameters
were compared base on observation field data. In this study, a realistic simulation of
reinforced embankment was analyzed to investigate boundary condition found on
two-dimensional and three-dimensional by PLAXIS finite element software in order
to model and investigate the behavior and performance of the full scale test
embankment.
5.6.1 Settlement
Beneath the center point of coated and non-coated Kenaf LLGs reinforced
embankment, the observed field settlement at surface (0.2m depth below the ground
surface) was predicted to compare with observed the data. The simulated settlements
by FEM 2D analyses were also compared with observed field data at the surface, 3m
and 6m depths.
During the construction, the rates of settlement slowly increased, and immediate
elastic settlements occurred. After construction, the rates of settlement rapidly
increased until 250 days after the end of the construction. 250 days after the end of the
construction, the settlements on coated Kenaf LLGs side at the surface, 3m and 6m
depths were 295mm, 179mm and 79mm, respectively. The corresponding settlements
in the non-coated Kenaf LLGs sections at the surface, 3m and 6m depths were
298mm, 183mm and 82mm, respectively as show in Figs. 5.4 to 5.6. The settlement
of coated and non-coated Kenaf LLGs reinforcements, similar magnitudes of
settlement rate were obtained. Moreover, the rate of surface settlement is higher than
the rates of subsurface settlement at 3m and 6m depths, respectively.
64
In addition, the time settlement curves computed using Terzaghis one-dimensional
consolidation theory, Asaoka (1978) method and FEM 2D method were used to
compare the average total consolidation settlement. Settlements computed using
Asaoka (1978) method closely followed the observed data while the settlements
computed using the one-dimensional consolidation theory overpredicted the observed
values. The observed surface settlement at the beginning of construction is greater
than the predicted values due to the drained behavior effect of the soft clay foundation
at the early stages of the construction which can be related to the method of applying
the embankment loading during the construction.
Additionally, the rates of settlement from FEM 2D method were higher than the
observed settlement data at the surface, 3m and 6m depths as plotted together in Fig.
5.7. The FEM 2D results from output program are shown the settlement at after
construction, 14 days, and 250 days, are illustrated in Fig. 5.8.
5.6.2 Excess pore water pressure
The performance of the full scale test embankment constructed on soft Bangkok
foundation was indicated by the building of excess pore water pressures. Four open
standpipe piezometers were used to monitor excess pore water pressure beneath the
reinforced embankment at 3 m and 6 m depths. Starting 7 days after the end of
construction (at full height of embankment), the excess pore water pressures rapidly
increased to the maximum pore water pressure. The excess pore water pressures
dissipated very quickly with time after 15 days to 120 days and dissipated with a
slower rate after 120 days. After 250 days, the excess pore water pressures decreased
to 10 kPa and it remained almost constant with time. The maximum excess pore water
pressures occurred at 7 days just after the full height of embankment construction.
The maximum excess pore water pressures beneath the coated and non-coated Kenaf
LLGs at 3m depth were 37 kPa and 35 kPa, respectively, 7 days at the end of
construction. The corresponding maximum excess pore water pressures below the
coated and non-coated Kenaf LLGs sections at 6m depth were 33 kPa and 32 kPa,
respectively, 7 days from the end of construction. From those results, the maximum
excess pore water pressures occurred at 7 days just after full height of embankment
construction. The results obtained from the coated and non-coated LLGs sections
were similar in magnitudes of excess pore water pressures.
The predicted excess pore water pressures below the embankment were estimated by
using Terzaghis one-dimensional model, predicted data by using Skempton and
Bjerrums three-dimensional method and FEM 2D method were compared with
average observed field data. Results are also plotted in Fig. 5.9 for 3m depth and Fig.
5.10 for 6m depth. The excess pore water pressures predicted by Terzaghis one-
dimensional model were 70 kPa and 60 kPa at 3m and 6m depths, respectively. The
excess pore water pressures obtained by the Skempton and Bjerrums three-
dimensional method were 42 kPa and 38 kPa at 3m and 6m depths, respectively.
In the case of the numerical simulation by the FEM 2D method, the maximum pore
water pressures were 49 kPa and 46 kPa at 3m and 6m depths, respectively.
Consequently, the excess pore water pressures from the one-dimensional method
overpredicted the measured values while the prediction from the Skempton and
Bjerrum (1957) method were in good agreement with the observed data at 3m and 6m
depths. Moreover, it can be seen that the predicted maximum excess pore-water
65
pressure at 3m and 6m depths obtained from the FEM 2D analyses agreed reasonably
well with the observed field data. The FEM 2D results from output program are
shown the excess pore water pressure at after construction, 14 days, and 250 days, are
illustrated in Fig. 5.11.
5.6.3 Deformation of reinforcement
The reinforcement deformations were obtained from the observed field data. The wire
extensometer were installed in 6 layers of reinforcement at elevations 0.00, +0.50,
+1.00, +1.50, +2.00 and +2.50 m above original ground surface. The average
deformations at the middle of Kenaf LLGs reinforcement were higher in both coated
and non-coated Kenaf LLGs. The maximum deformations of Kenaf LLGs
reinforcement for each layer were located at the middle point of the embankment
reinforcement corresponding to the levels of embankment settlements. Moreover, the
deformations in each layer increased with time. The observed deformation of the
Kenaf LLGs reinforcement and predicted Kenaf LLGs deformations by FEM 2D were
compared as shown in Fig 5.12.
5.6.4 Stresses of reinforcement versus time curve
The deformations were obtained from the observed field data. The wire extensometer
were installed in 6 layers at elevation 0.00, +0.50, +1.00, +1.50, +2.00 and +2.50 m
above original ground surface. Figures 5.13 show stresses in Kenaf LLGs reinforced
compared between FEM 2D simulations with observed field data. The stresses in
Kenaf LLGs reinforced from the observed field data were calculated at 14 days and
250 days after construction based on correlations with the laboratory pullout test
results. The stresses of Kenaf LLGs reinforcement at the bottom of full scale test
embankment were almost similar at the top following the deformation patterns (Fig.
5.12). In addition, the stresses of Kenaf LLGs reinforcement at the middle have higher
stresses than at the edges following its deformations in response to the settlement
pattern.
5.6.5 Stability Analyses of the Embankment by Finite Element
PLAXIS software based on phi-c reduction was used to calculate factor of safety at
time after construction and at time more than one year after construction. FEM 2D
simulation was performed in order to compare the stability of the test embankment
reinforced with and without Kenaf LLGs. Figure 5.14 shows a comparison of the test
embankment in the stability analyses using numerical finite element for stability by
analyses using the factor of safety. The zones of failure in stability analyses using
FEM 2D is shown in Fig. 5.15 for analysis full scale embankment test without Kenaf
LLGs reinforced. Moreover, Fig. 5.16 is shown the zones of failure in stability
analyses using FEM 2D with Kenaf LLGs reinforced.
From FEM 2D analysis, the values factor of safety were decreased from 1.9 to 1.32
then increased to 1.8 for without Kenaf LLGs whereas the factor of safety were
decreased from 2 to 1.45 then increased to 1.85 for improved with Kenaf LLGs.
66
5.7 Summary and Concluding Remarks
For 2D FEM simulation using, PLAXIS 2D version 2010.1 based on plane strain
condition were carried out to study the behavior of a full scale test embankment
reinforced with Kenaf LLGs on soft Bangkok Clay. The two-dimensional (FEM 2D)
problem for finite element analysis consisting of x and y directions was modeled. The
surface settlements, subsurface settlements, excess pore water pressures, deformations
and stresses of Kenaf LLGs reinforcement were predicted. The predicted from FEM
2D analyses of surface and subsurface settlements at 3m and 6m depths was not
agreed with the observed data. The predicted excess pore water pressures by FEM 2D
analyses and the Skempton and Bjerrum (1957) method agreed with the observed data
at 3m and 6m depths while the one-dimensional method overpredicted the results as
expected. The predicted deformations and stresses of Kenaf LLGs from FEM 2D
overpredicted the observed field data due to the geometric effects and plan
dimensions of the test embankment. Stability Analyses using FEM 2D, the factor of
safety for test embankment reinforced with Kenaf LLGs was higher than without
Kenaf LLGs reinforcement.
67
CHAPTER 6
3D FINITE ELEMENT MODELING OF
FULL SCALE TEST EMBANKMENT
6.1 Introduction
The finite element methods are growing rapidly from two-dimensional (2D) to three-
dimensional (3D) analyses because the 3D simulation can simulate real behavior. The
3D finite element model in the PLAXIS 3D Program is created through a two-
dimensional (2D) in vertical cross-section model in x-y plane and later extended into
the third dimension (z-direction). PLAXIS 3D Program had been carried out to
investigate the soil behavior of full scale embankment test. The FEM 3D prediction
results showed that the predicted settlements, excess pore water pressure, deformation
were underestimated with observed settlement at 3m and 6m depth due to the three
three-dimensional loading and boundary conditions. Three-dimensional numerical
simulation has been carried out to investigate in many geotechnical problems. There
are many previous researches that investigate behavior of embankment on soft
foundation by used three-dimensional (3D) numerical analyses such as hexagonal
wire reinforced wall with sand backll and precast concrete facing panels (Lai et al.,
2006), lightweight rubber tire chips-sand backll embankment with geogrid
reinforcements (Tanchaisawat et al., 2009), and SDCM and DCM piles under axial
and lateral loads and under embankment (Suksawat, 2009).
6.2 Material Parameter for Modeling
In this research used the Mohr-Coulomb model and Soft Soil model for full scale
embankment analysis model. The Mohr-Coulomb model is used for simple analysis
such as weathered crust layer, medium stiff layer and stiff clay layer. On the other
hand, The Soft Soil model is used for very soft soil layer.
6.2.1 Mohr-Coulomb model
The Mohr-Coulomb model is a linear elastic perfectly-plastic. The numerical result
depend on five input parameters, i.e. young's modulus (E), poisson's ratio ( ), friction
angle (), cohesion (c), dilatancy angle (). Elasticity result depend on two influence
parameters that are E and. For plasticity, it depend on two parameters effect that are
and c. The Mohr-Coulomb model represents a first-order approximation of soil or
rock behavior. A first analysis recommends this model for soil behavior analysis. For
each layer one estimates a constant average stiffness or a stiffness that increases
linearly with depth. Due to this constant stiffness, computations tend to be relatively
fast and one obtains a first estimate of deformations.
Although the increase of stiffness with depth can be taken into account, the Mohr-
Coulomb model does neither include stress-dependency nor stress-path dependency of
stiffness or anisotropic stiffness. In general, effective stress states at failure are quite
well described using the Mohr-Coulomb failure criterion with effective strength
parameter and c. For undrained materials, the Mohr-Coulomb model may be used
with the friction angle set to 0 and the cohesion c set to c
u
(S
u
), to enable a direct
68
control of undrained shear strength. In that case note that the model does not
automatically include the increase of shear strength with consolidation.
6.2.2 Soft soil model
The Soft Soil model is a Cam-Clay type model especially meant for primary
compression of near normally-consolidated clay-type soils. This was proposed by
Vermeer and Brinkgreve (1995). The model performs best in situations of primary
compression the SSM requires the following material constants: * = the modified
compression index taken from (
v
ln p) plot. This constant is not the same with the
intrinsic compression index as defined by Burland (1990); * = modified swelling
index (slope of swelling line of above mentioned (
v
ln p) plot; c = cohesion; =
friction angle and = dilatancy angle. Although the modeling capabilities of this
model are superseded by the Hardening Soil model, the Soft Soil model is still
retained in the current version, because existing PLAXIS users might be comfortable
with this model and still like to use it in their applications.
6.2.3 Interfaces
Interface elements are generally modeled by means of the bilinear Mohr-Coulomb
model. When a more advanced model is used for the corresponding cluster material
data set, the interface element will only pick up the relevant data that are young's
modulus (E), poisson's ratio ( ), friction angle (), cohesion (c), dilatancy angle ()
from the Mohr-Coulomb model. In some cases, the interface stiffness is set equal to
the elastic soil stiffness. Therefore, E = E
ur
where E
ur
is stress level dependent,
following a power law with E
ur
proportional to
m
. For the Soft Soil model, Soft Soil
Creep model and Modified Cam-Clay model the power m is equal to 1 and E
ur
is
largely determined by the swelling constant
*
.
6.2.4 Undrained behavior
In general, care must be taken in undrained conditions, since the effective stress path
that is followed in any of the models may deviate significantly from reality. Although
PLAXIS has options to deal with undrained behavior in an effective stress analysis,
the use of undrained shear strength (c
u
or s
u
) may be preferred over the use of
effective strength properties ( and c) in such cases. Please note that direct input on
undrained shear strength does not automatically include the increase of shear strength
with consolidation If, for any reason, the user decides to use effective strength
properties in undrained condition, it is strongly recommended to check the resulting
mobilized shear strength using the corresponding option in the PLAXIS Output
program.
6.3 Finite Element Simulation by PLAXIS
PLAXIS finite element software was used to simulate the behavior of the test
embankment using PLAXIS 3D version 2010.01 at three-dimensional. The finite
element input parameters were analyzed based on the observed field data at coated
and non-coated Kenaf LLGs. The material properties of the backfill for the FEM
simulations are tabulated in Table 5.1. During this construction stage, undrained
analysis was applied to simulate the layer by layer construction. After the completion
of the full height of the embankment, drained analysis was applied to simulate the
69
consolidation process. The comparisons between observed field data and predicted
results on settlements at surface and subsurface, excess pore water pressures,
deformation of reinforcement are discussed in the following sections.
6.3.1 Plaxis 3D numerical simulation of full scale test
The 3D finite element analyses had been carried out to investigate the behavior of
embankment when it is subjected to loading. PLAXIS 3D version 2010.01 software
was used to model and investigate the behavior and performance of the full scale test
embankment as shown in Fig. 6.1. This is a three-dimensional finite element computer
program used to perform deformation and stability analyses. Determination of
accurate effective stresses is an essential task to accomplish this objective. Complex
geotechnical structure can be solved with special features built within the program.
The 3D finite element model in the PLAXIS 3D version 2010.01 Program is created
through a two-dimensional (2D) in vertical cross-section model in x-y plane and later
extended into the third dimension (z-direction). Moreover, the mobilized tensile force
(T
mob
) was 20 kN/m at the estimated 4% of strain from wide width tensile test. The
material model is shown in Table. 5.1.
6.3.2 3D condition
The modeling of full scale embankment test on soft ground was analyzed in three-
dimensional analysis by using 3D version 2010.01 software. For the modeling of full
scale embankment test with plain pattern of Kenaf LLGs, geometry dimensions was
assigned 140m, 140m, and 75m at x-direction, y-direction, and z-direction,
respectively.
The full scale embankment modeling was divided into seven layers. The embankment
was assigned silty sand backfill properties until 3 m height with six layers of plain
pattern of Kenaf LLGs. Moreover, weather crust properties of soft Bangkok clay was
assign the silty sand backfill embankment until 4m. In addition, the side slope
consisted of 1 vertical to 1.5 horizontal and back slope consisted of 1 vertical to 1
horizontal. The soil profile model was divided into 4 layers.
6.3.3 3D initial and boundary condition
The boundary condition for three-dimensional simulation, standard fixities use fixed
boundary in x, y, and z direction at the bottom of soil boundary. Lateral movement
was fixed while vertical movement was allowed at the vertical plane of the foundation
soil. The initial stresses in model were influenced by the weight of the material and
the history of its formation. Initial stress condition was generated using K
o
procedure
built in the program by the coefficient of lateral earth pressure. The stages of
construction applied in 3D numerical analyses were parallel to the construction
applied in the 2D numerical analyses.
6.3.4 Input (pre-processing) program
The three-dimensional simulation program consists of two different modes are that
model mode and calculation mode. The model mode has many tools function for a
geometry modeling such as embankment, soil layers, and Kenaf LLGs model.
Moreover, generate mesh function was use for element distribution (very coarse,
70
coarse, medium, fine, and very fine). The calculation mode consists of all facilities to
define calculation phases including the initial phase and construction phases.
- Graphical input: 15-node elements were used in the numerical simulation model.
The geometry and boundary conditions as plan view of full scale embankment test.
- Work planes: Work planes are two-dimensional in vertical cross-section planes (xy
direction) and can create the horizontal planes (z-direction). The outer boundaries of
the work planes are based on the initial setting of the x
min
, x
max
, y
min
, y
max
, z
min
and
z
max
parameters which defined in the General settings. Note that, all work planes have
the same outer boundaries
- Soil profile: Bore holes are used to define the soil layers and ground water level.
Multiple bore holes can be placed in the geometry to define a non-horizontal soil or
inclined ground surface. The soil profile can be divided into four layer.
- Material models: The material model was assigned two differences model that are
Mohr-Coulomb and Soft soil models. Type of material behavior can be divided into
two types that are undrained and drained behavior.
- Geotextile: Kenaf LLGs reinforcement was modeled as geogrid in the software. This
element needs only an axial stiffness, EA, where E is the modulus of elasticity of the
material and A is the cross-sectional area per unit width in out of plane.
- Interfaces: Interfaces are composed of sixteen node interface elements which are
eight pairs of nodes, compatible with the 8-nodes quadrilateral side of a soil element.
Interface is assigned a virtual thickness which is an imaginary dimension.
6.3.5 Calculations
The numerical calculation can be divided into many calculation phases, it depend on
stage of construction. On the other hand, the first phase of calculation mode has to
assign initial stress (initial phase). After the initial phase, calculation phase design by
user. Staged construction can be simulation of full scale embankment construction
stage by activating clusters of elements.
The types of calculations can be divided into four basic type of calculation: Plastic
calculation, Consolidation analysis, Gravity loading and K
0
procedure but for latter
two types are only used for initial stage. A plastic calculation is used to carry out and
elastic-plastic deformation analysis according to small deformation theory whereas
consolidation analysis is usually preformed when it is necessary to analyze the
development and dissipation of excess pore pressures in a saturated clay as a function
of time. Gravity loading is a type of plastic calculation, in which initial stresses are
generated base on the volumetric weight of soil. And the last one is K
0
procedure; it is
a special calculation method which can be used to define the initial stress for the
model.
6.3.6 Output (post-processing)
The Output program provides all facilities to view and list in table of the results after
generating and also 3D finite element calculations such as displacements, pore-presser
71
and the stresses at any points. Moreover, it is possible to view output in individual
work plane.
The output program provides several sub-menus to illustrated total deformations and
deformation in x, y and z directions. Stress and strain can also show in many type
including total stress and effective stress. Moreover, it can show the plastic points
which are the stress points in the plastic state. The plastic points are indicated by
small symbols than have different shapes and colors. A red hollow square (Mohr-
Coulomb point) indicates that the stresses lie on the surface of failure envelope. A
white solid square (Tension cut-off point) indicates that the clusters failure by tensile
strength.
6.4 Results and Discussions
6.4.1 Settlement versus time curve
The results were observed at the center point of coated and non-coated sections. The
field settlement data were measured at surface and subsurface settlement plates (0.2m,
3m and 6m depth below the ground surface). The settlement rates slowly increase
during the construction, and immediate elastic settlements occurred.
After construction period, the rates of settlement rapidly increased until 250 days after
the end of the construction. 250 days after the end of the construction, the settlements
rate slowly increased until nearly constant. The settlement rate on the coated Kenaf
LLGs section at the surface, 3m and 6m depths were 295mm, 179mm and 79mm,
respectively. Moreover, The corresponding settlement result on the non-coated Kenaf
LLGs sections at the surface, 3m and 6m depths were 298mm, 183mm and 82mm,
respectively as illustrated in Figs. 6.2 to 6.4. The simulated settlement result by FEM
3D were compared with field data result at the surface, subsurface 3m and 6m depths
as illustrated in Fig. 6.5. The settlement result of coated and non-coated Kenaf LLGs
sections were similar magnitudes of settlement rate were obtained. In addition,
settlement rate of 6m depth is lower than the rate of settlement at 3m depth and
surface settlement, respectively. Settlement results from FEM 3D output programs are
shown in Fig. 6.6. The results from output program are shown the settlement at 14
days, and 250 days after construction.
6.4.2 Comparison of observed settlement with different prediction methods
Moreover, the consolidation settlement prediction was made using the consolidation
coefficient derived from Asaoka (1978) and Terzaghis one-dimensional
consolidation method. Referring to Fig. 6.7, the rates of settlement results from using
Asaokas method, observed field data, FEM 2D and FEM 3D simulation are plotted
together.
The simulated results of observed and predicted values of surface settlements obtained
from FEM 3D analyses were closer to observed field data than the predicted results
from FEM 2D analyses. The observed and predicted from using Asaokas method,
observed field data, FEM 2D and FEM 3D simulation at subsurface settlements (3m
and 6m depth below ground surface) are illustrated in Figs. 6.8 and 6.9, respectively.
The observed and predicted values using Asaokas method agreed well while the
corresponding predicted values from Terzaghi one-dimensional consolidation theory
72
overpredicted the observed field data. Additionally, the rates of settlement from FEM
2D method were higher than observed settlement data while the FEM 3D predictions
generally agreed with observed settlements at the surface, 3m and 6m depth due to the
fact that the three-dimensional loading condition is closer to the field condition. The
predicted values of surface and subsurfasce settlements from FEM simulations were
influenced by the boundary value problem (Auvinet and Gonzalez, 2000).
Furthermore, the observed settlement pattern apparently more related to 3D than 2D
conditions because of its symmetry and plan dimensions of the test embankment
(Teerawattanasuk, 2008). Therefore, the effect of boundary condition (2D and 3D)
applied in numerical analysis can be considered to be an important factor that
influence the predicted results due to the differences of schematic plan view
dimensions.
According to Skempton and Bjerrum(1957) method, the difference between FEM 2D
method and FEM 3D methods can be explained by using Skempton and
Bjerrum(1957) method. A correction factor () should be applied to the settlement
calculated on the basis of odometers test. The correction factor () decreases with
increasing overconsolidation ratio (OCR). The correction factor () at the surface,
3m and 6m depths were 0.72, 0.73 and 0.75, respectively. Similarly, the difference
between FEM 2D and FEM 3D methods at the surface, 3m and 6m depths were 28%,
27% and 25%, respectively. Therefore, the geometric effects should be considered as
important factors that can affect the results of the numerical simulations, which are
consistent with the current settlement predictions with Skempton and Bjerrum
corrections (Bergado and Teerawattanasuk, 2008).
6.4.3 Excess pore water pressure versus time curve
The comparison of excess pore water pressures between the observed field data and
the FEM 3D simulation variations with time at 3m and 6m depth, during and after the
construction, are shown in Figs. 6.10 to 6.11. In reality, the excess pore pressures in
the soft clay started to dissipate during construction. In addition, the excess pore
pressures were measured from slow responding standpipe piezometer and contributed
to the low values of excess pore pressure during the early rapid construction in stages.
The excess pore water pressures results from FEM 3D output programs are shown in
Fig. 6.12. The results from output program are shown the excess pore water pressures
at 8 days, 15 days, 30 days, and 180 days after construction.
In case of simulated and observed field data, the maximum pore water pressures at 3m
depth were 36 kPa for FEM 3D. From numerical simulation results on 6m depth, the
maximum pore water pressures were 32 kPa, for FEM 3D. The maximum pore water
pressures from observed field data were slightly higher than FEM 3D and tend to
decrease and space closely followed the FEM 3D.
6.4.4 Comparison of observed excess pore water pressure with different
prediction methods
The predicted excess pore pressure from observed field data, FEM 2D and FEM 3D
simulation are plotted together as shown in Figs. 6.13 to 6.14 for 3m and 6m depths,
respectively. However, the excess pore water pressures from FEM 3D simulation
yielded satisfactory agreement with the observed data. The reasons of difference
between the finite element software and the observed field data is that, the
73
groundwater table in finite element must be specified in calculations method.
Consequently, the rate of dissipation for 2D simulation is lower than FEM 3D
simulations. The FEM 2D simulation, the excess pore water pressure can dissipate in
two-direction (x and y directions). While, the excess pore water pressure of FEM 3D
can conveniently dissipate in three-direction (x, y, and z directions). Therefore, the
actual excess pore water pressures for this embankment agreed closely with the 3D
conditions.
The predicted excess pore pressure from Terzaghis one-dimensional, Skempton and
Bjerrums three-dimensional method, observed field data, FEM 2D and FEM 3D
simulation are plotted together as shown in Figs. 6.15 to 6.16 for 3m and 6m depth,
respectively. The excess pore water pressures by Terzaghis one-dimensional and
FEM 2D were overpredicted. The prediction, the Skempton and Bjerrums three-
dimensional method and FEM 3D simulation agreed well with the observed data at
3m and 6m depths.
6.4.5 Deformation of reinforcement versus time curve
Compared to FEM 3D simulations and observed field data for the deformation Kenaf
LLGs at 240 days after embankment construction are presented in Fig. 6.17. It can be
observed that FEM 3D simulations agreed with the observed field data.
On the other hand, the predicted deformation from observed field data, FEM 2D and
FEM 3D simulation were plotted together in Fig. 6.18. The predicted deformation
distribution also shows consistent results in which the deformation of FEM 2D
simulation are greater than that of FEM 3D simulation due to overestimated
settlements of the former than the latter. However, the predicted results from FEM 3D
simulation gave closer deformation of Kenaf LLGs reinforced than the FEM 2D
simulation because the predicted settlement and lateral movement from FEM 2D
numerical analysis predominantly overestimated the observed field data.
6.4.6 Stresses of reinforcement versus time curve
The deformations were obtained from the observed field data. The wire extensometer
were installed in 6 layers at elevation 0.00, +0.50, +1.00, +1.50, +2.00 and +2.50 m
above original ground surface. The stresses in Kenaf LLGs reinforced from the
observed field data were calculated at 14 and 250 days after construction base on
pullout test results as shown in Fig. 4.30. The stresses of Kenaf LLGs reinforced at
the bottom of full scale test embankment have higher than at the top. In addition, the
stresses of Kenaf LLGs reinforced at the middle have higher than the stresses at the
side because of overburden pressure at the bottom layer are larger than at the top
layer.
The stresses in Kenaf LLGs reinforced from FEM 3D simulations have compared
with observed field data as shown in Fig. 6.19. The stresses in Kenaf LLGs reinforced
from FEM 3D simulations were agreed well with observed field data.
The stresses in Kenaf LLGs reinforced from observed field data, FEM 2D and FEM
3D simulations were plotted together in Fig. 6.20. The stresses in Kenaf LLGs
reinforced from FEM 3D simulations have similarly trend with FEM 2D simulations.
In contrast, the stresses in Kenaf LLGs reinforced from FEM 2D simulations have
74
larger stresses with FEM 3D simulations. At the top layer of full scale reinforced
embankment, the stresses in Kenaf LLGs reinforced from FEM 2D simulations were
close to the value of FEM 3D simulations. However, the stresses of reinforcement
results from FEM 2D simulations were higher than FEM 3D because of overpredicted
settlement in FEM 2D simulations.
6.5 Back Analysis of Parameters
The parameters from previous studies were analyzed and revised based on the FEM
2D, FEM 3D simulations of observed field data from Kenaf LLGs reinforced
embankment in order to back analyze the parameters. Figures 6.21 to 6.26 show the
settlement results of the FEM 2D and FEM 3D simulations using the parameters from
Jamsawang (2009) at the surface and subsurface settlements, respectively.
The FEM 2D simulations settlement surface and subsurface settlement at 3m and 6m
depths, respectively of Jamsawang (2009) has closed to FEM 2D simulations for
Kenaf LLGs embankment as shown in Figs. 6.21 to 6.23. In contrast, the FEM 3D
simulations settlement of Jamsawang (2009) has smaller settlement than observed
field data due to the OCR values is 1.5 at all soft clay layers while in this study the
OCR values of 1.55, 1.4, and 1.3, for soft clay layer 1, 2, and 3, respectively. From the
Tanchaisawat (2008), the drain condition was used to analyses that cause high
settlements during construction become constant after construction. The trend FEM
3D simulations settlements were similar for all surface and subsurface settlement at
3m and 6m depths, respectively as shown in Figs. 6.24 to 6.26. It can be seen that the
settlements predictions of the soft clay foundation mostly depended on the assumed
thickness of the uppermost weather crust layer and the overconsolidation ratios of the
soft clay layer.
The FEM 2D and FEM 3D simulations excess pore water pressures at subsurface 3m
and 6m depths from Jamsawang (2009) were plotted and compared with the observed
field data from Kenaf LLGs reinforced embankment. The FEM 2D simulations, the
excess pore water pressures from Jamsawang (2009) were slightly less than observed
field data but the excess pore water pressures decreased after construction as shown in
Figs. 6.27 to 6.28. The FEM 3D simulations of excess pore water pressures from
Jamsawang (2009) were close with observed field data at 3m depth while the excess
pore water pressures were less than observed field data then it was decreased after
construction due to the different of cohesion and OCR values of the soft clay layer.
The predicted excess pore water pressures were sensitive to the OCR values of the
soft clay layer as shown in Figs. 6.29 to 6.30.
6.6 Stability Analyses of the Embankment by Finite Element
PLAXIS software based on phi-c reduction was used to calculate factor of safety at
time after construction and at time more than one year after construction. FEM 3D
simulation were performed in order to compare the stability of the test embankment
reinforced with and without Kenaf LLGs.
The factor of safety from FEM 3D analysis were decreased from 3.3 to 2.35 then
increased to 2.8 for without Kenaf LLGs while the factor of safety were decreased
from 3.3 to 2.65 then increased to 3.1 for improved with Kenaf LLGs as shown in Fig.
75
6.31. Figure 6.32 shows a comparison of the test embankment in the stability analyses
using numerical finite element for stability by analyses using the factor of safety.
Based on FEM analyses during the construction period, the undrained condition
analyses seem to be the most critical condition for its stability. The factors of safety
were reduced during construction period then increase afterwards. In addition, it was
found that the two dimensional FEM 2D analyses have significant differences in the
factors of safety from FEM 3D analysis. The factor of safety by two dimensional
(FEM 2D) analyses were lower factor of safety values that obtained by three
dimensional (FEM 3D) analysis on both with and without Kenaf LLGs reinforcement
because the plain strain analysis of FEM 2D does not consider the boundary resistance
in three-directions. Several studies (Seed et al. 1990; Hovland 1977; and Li et al.
2009) have indicated that the factor of safety from a 3D analysis can be greater than
that the corresponding values from a 2D analysis. Nevertheless, using FEM 2D
analysis is conservative for design and non-conservative when determining the
strength parameters. In addition, the locations and depths of the failure surface are
highly influenced by the types of analyses. The zones of failure in stability analyses
using 3D are shown in Fig. 6.33 to Figs. 6.34. The failure surfaces of FEM 2D
analyses tented to be deeper than the FEM 3D analyses. However, the failure surfaces
of the FEM 3D seems to be wider in extent than the FEM 2D analyses.
6.7 Summary and Concluding Remarks
The observed field data and the simulations of Kenaf LLGs reinforced embankment
were analyzed to investigate the influence of the boundaries conditions in three-
dimensional (FEM 3D) using PLAXIS finite element software. The predicted data of
surface and subsurface settlements obtained from FEM 3D analyses agreed well to the
observed data due to the geometric effects of the embankment with length to width
ratio (L/B) of 1.0. The FEM 3D settlement predictions were consistent with the
observed data due to lower pore pressures and lower deformations in the 3D
conditions. The simulations of excess pore water pressures from FEM 3D simulations
agreed well with the observed data at 3m and 6m depths due to the similarities of the
3D boundary conditions in the field. The predicted results from FEM 3D simulations
yield closer deformations and stresses of Kenaf LLGs reinforcements observed data.
Thus, The FEM 3D simulations captured the overall behavior of the Kenaf LLGs
reinforced embankment with more reasonable agreement between the field
observations and the predicted values compared to the FEM 2D simulation due to the
geometric effects and plan dimensions of the test embankment. Therefore, the
geometric effects should be considered as important factors that can affect the results
of the numerical simulations. The Kenaf LLGs can be applied for short term
reinforcements in geotechnical engineering application to improve the stability of
embankment on soft clay foundation immediately after construction.
76
CHAPTER 7
FIELD EROSION CONTROL TESTS
7.1 General
The full-scale test embankment was constructed using silty sand backfill materials and
reinforced by woven Kenaf LLGs. On the top layer of embankment was coverd by
weather crust clay. The materials used for soil erosion control were water hyacinth
natural fiber yarns fabricated in plain pattern. Based on laboratory of water hyacinth
LLGs test results in Chapter 3, the Water hyacinth LLGs can be applied on the test
embankment for erosion control. In this study, the water hyacinth yarns were
fabricated to woven geotextile called limited life geotextiles to help stabilize the side
slope and improved the growth of vegetation for erosion control in geotechnical
application. The study was aimed to access the effectiveness of the new material for
geotextile made from water hyacinth to control soil erosion in a representative slope at
different plots of tests. The erosion control test was performed by using the artificial
rainfall. The flow rates of water runoff and amount of soil loss were investigated after
growing period of grass 4, 6, 8 and 10 weeks.
7.2 Erosion Control Embankment Test
The experiment was conducted on the test embankment which was constructed at the
northern part of the campus of the Asian Institute of Technology (AIT) in
KlongLuang, Pathumthani, Thailand in February, 2011. The height of embankment is
four meters. The embankment was constructed by using silty sand backfill until 3 m
height with six layers of plain pattern of Kenaf LLGs. The vertical spacing is 0.5 m to
reinforce the embankment, and weather crust of soft Bangkok clay 1 m height was
covered the silty sand backfill embankment. Moreover, the side slope consisted of 1
vertical to 1.5 horizontal and back slope consisted of 1 vertical to 1 horizontal. The
experiments were conducted at slope 1 vertical to 1.5 horizontal at the east and west
side of embankment. The embankment for monitoring is shown in Fig. 7.1.
7.3 Water Hayacinth Woven LLGs
Water Hayacinth (Eichhornia crassipes) woven LLGs are made from 100% water
hyacinth fiber twine woven into the pattern. In this study, The woven water hyacinth
LLGs with two different opening sizes of 8x8 mm and 12x12 mm were selected to
investigate their performance which are including coated and non-coated with
polyurethane to reduce water absorption and effect of weathering on the water
hyacinth LLGs increasing their life time. The properties of woven water hyacinth
LLGs are tabulated in Table 3.1. The woven water hyacinth LLGs (Fig. 7.2) were
spread on weather crust of soft Bangkok clay at side slope 1 vertical to 1.5 horizontal
of the embankment. The water hyacinth LLGs dimension was 1 m width and 5 m
length. The opening size of 12x12 mm of woven water hyacinth LLGs was used at east
side and the opening size of 8x8 mm of woven water hyacinth LLGs was used at west
side of the embankment as shown in Fig. 7.3 and 7.4.
7.4 Ruzi Grass
Ruzi grass is called in scientific name Brachiaria ruziziensis. Ruzi grass has been
promoted as a good grass for improving pastures for cattle in the tropics. The seed can
77
easily be planted in high rainfall area. Ruzi grass is extensively used in Thailand, its
acceptance by farmers depends on the availability of a cheap and high quality seed
supply as shown in Fig. 7.5. The seed of Ruzi grass was spread on the soil 60 g/m2
before installed woven water hyacinth LLGs as shown in Fig. 7.6. The Ruzi grass can
grow up through the water hyacinth LLGs sample opening size 8x8 mm and 12x12
mm as shown in Fig. 7.7 and 7.8.
7.5 Set Up Erosion Control Embankment Test
The field apparatus for both side of test embankment consists of five plots, divided by
pinewood sheets to prevent water running on to them from up slope and contamination
from the surrounding plots. Before the field site facilities could be set up, the funnels
and dividers had to be installed. To remedy this situation, relief trenches were dug into
the downhill side of the bucket holes to an approximate depth of 700 mm. The seed of
Ruzi grass was spread on the soil 60 g/m
2
at grassed control plots before installed
woven water hyacinth LLGs. The opening size of 12x12 mm of woven water hyacinth
was performed at the East of embankment with 5 plots including polyurethane coated
water hyacinth LLGs with Ruzi grass, non polyurethane coated water hyacinth LLGs
with Ruzi grass, Ruzi grass only, polyurethane coated water hyacinth LLGs without
Ruzi grass and polyurethane coated water hyacinth LLGs without Ruzi grass. The
opening size of 8x8 mm of woven water hyacinth was performed at the West of
embankment with 5 plots including polyurethane coated water hyacinth LLGs with
Ruzi grass, non polyurethane coated water hyacinth LLGs with Ruzi grass, bare soil,
polyurethane coated water hyacinth LLGs without Ruzi grass and polyurethane coated
water hyacinth LLGs without Ruzi grass. Experimental set up at the test embankment
is shown in Fig. 7.9.
7.6 Rainfall Simulation
In this study, the erosion control test was performed by using the Artificial Rainfall to
assess the effectiveness of water hyacinth woven LLGs in erosion control and water
run-off. The rainfall simulator is capable of creating uniform rain drops and intensities
over the entire area of the specimen. In this study, the rainfall intensity of 120mm/h
which normal occurs in Thailand with return period ranging from 2 to 200 years during
90 minute duration of rainfall was obtained to investigate the runoff effect. The
Artificial Rainfall was installed by using pump and sprinkler as shown in Fig. 7.10.
The water motor pump was used to connect with the reservoir near the test
embankment (Fig. 7.11). The water delivery system of water supply using PVC pipe
line connected 1 inch PVC pipes and it is scaled down to inch diameter and
controlled by two points of pressure control. The nine sprinklers were installed in both
side of the embankment at the height of 1.0 m along the edge and middle.
The rain gauge made from PVC pipe with 4 inches diameter and 32 cm long as shown
in Fig 7.12. The 100 ml cylinder were put inside and 15 mm diameter of cone was put
on top the PVC pipe to measure the rainfall intensity and the standard size of the
rainfall is recorded continuously. The rainfall at the first minute was collected to
measure the intensity by pouring the cylinder to read the volume and the volumes were
calculated by a formula to determine the intensity of rainfall. The required rainfall
intensities in this study were 120 mm/hr per two hours. The windshields were used to
protect the wind at the side of embankment.
78
7.7 Monitoring Soil Erosion
Simulated rainfall erosion tests were performed with rainfall intensity of 120 mm/hr on
the both side of the embankment slope (1V:1.5H) after growing period of grass 4
weeks. The rain gauges were used to collected and measure the intensity by pouring
the cylinder to read the volume. The flow rates of water runoff that can pass on the test
slope from each test plots were recorded by flow of surface water through quantity of
ow measured versus time each plots. Runoff samples were taken by using plastic
containers (Fig. 7.13) after the flow rates were constant. The water runoff sample in
plastic bag is shown in Fig. 7.14. Then, Runoff samples were kept inside plastic
container at the laboratory until sediments were deposited by gravity. Thereafter, the
clear water was removed and the sludge was dried by using an oven at 105C. The
dried soil in each plots were weighed, then the eroded sediment and runoff volume
were calculated. The following procedure was observed for the collection of field site
data after growing period of grass 4, 6, 8 and 10 weeks.
7.7.1 Runoff rate
The effect of the new material for geotextile made from water hyacinth to control soil
erosion in a representative slope (1V:1.5H) at different plots of tests at the growing
periods of Ruzi grass of 4, 6, 8, and 10 weeks. For the opening size of 12 mm of
woven water hyacinth LLGs was used at the east side of embankment (Fig. 7.15), the
results show that the polyurethane coated and non-coated woven water hyacinth LLGs
have higher amount of runoff rate followed by Ruzi grass only, polyurethane coated
and non-coated woven water hyacinth LLGs with Ruzi grass, respectively, as shown in
Fig. 7.16.
At the west side of embankment (Fig. 7.17), the results of woven water hyacinth LLGs
8 mm opening size is shown in Fig. 7.18. The total soil loss on bare soil was much
higher than polyurethane coated and non-coated woven water hyacinth LLGs and the
runoff tend to increase with time. Conversely, the runoff decreased with increasing
growing period of Ruzi grass. The runoff rates were significantly reduced with woven
water hyacinth LLGs and Ruzi grass covers. There was not much statistical difference
in runoff between coated and non-coated woven water hyacinth LLGs.
From the results, it can be seen that combination of water hyacinth LLGs with Ruzi
grass reduced the amount of runoff from the rainfall. The cover of the LLGs can be
reduced the impact of raindrops on the surface of the soil, and the barrier between the
rain and the soil delayed the development of runoff. Runoff rate has 50 percent
decreased by using water hyacinth LLGs covered soil surface. Runoff rate decreased
with increasing grass cover and the grasses had decreased of the flow rate by 50%
compared to the bare soil. In contrast, combination of soil with LLGs reduced the
amount of runoff from the rainfall. The high coverage of the LLGs reduced the impact
of raindrops on the surface of the soil, and the barrier between the rain and the soil
delayed the development of runoff. Therefore, LLGs made from water hyacinth makes
more effective at suppressing soil erosion because it reduces the impact of rainfall and
delays runoff.
79
7.7.2 Soil loss
This study evaluated the runoff rate and soil loss at different plots of test with ruzi
grass cover at the growing periods of 4, 6, 8, and 10 weeks. The eroded sediment was
collect from the field tested embankment and it was dried by using an oven in the
laboratory. Figure 7.19 shows the total soil loss with varying time from the East of
embankment by using 12 mm opening size of water hyacinth LLGs. From the results
show that very low soil loss were found in 4 to 10 weeks from polyurethane coated and
non-coated water hyacinth LLGs combined with Ruzi grass. The amount of soil loss
from Ruzi grass only (0.25 kg/m2) plot is less than the amount of soil loss from
covered soil surface by water hyacinth LLGs only about 50 percent. The amount of
soil loss tends to be decreased when increasing time.
For the west of embankment by using 8 mm opening size of water hyacinth LLGs, the
results shows that bare soil plot has highest amount of soil loss about 2.2 kg/m2 (Fig.
7.20). The polyurethane coated and non-coated water hyacinth LLGs can be reduced
the amount of soil loss about 70% of bare soil. Moreover, very low amount of soil loss
has found from the polyurethane coated and non-coated water hyacinth LLGs
combined with Ruzi grass plots due to the root of Ruzi grass.
Related with bare soil, the LLGs with opening size of 8x8 mm had the lowest amount
of soil loss than bare soil. The total soil loss on bare soil was much higher than with
LLGs and combined with Ruzi grass cover. Therefore, the smallest opening size of
water hyacinth LLGs with 8x8 mm has better erosion resistance than 12x12 mm LLGs.
The results indicated that with increasing coverage of Ruzi grass at 4, 6, 8 and 10
weeks growing period of grasses and LLGs the soil loss was greatly reduced.
7.9 Effect of Water Hyacinth LLGs Coated and Non-coated with Polyurethane
From the results of this study, the polyurethane coating was not effect to the
embankment runoff rate and soil loss at different plot of tests. The results of runoff rate
and soil loss were similar in both coated and non-coated water hyacinth LLGs. During
the experiment, the coated water hyacinth LLGs were not absorbed the water from the
artificial rain and runoff because it was affected by the coating polyurethane on the
surface as shown in Fig. 7.21. Thereafter, the non-coated water hyacinth LLGs was
degraded by moisture and UV light from the sun after 3 months while the non-coated
water hyacinth LLGs degraded after 4 months. It means that the resistance of coated
water hyacinth LLGs can increase the life time about 25 percent.
7.10 Summary and Concluding Remarks
The woven water hyacinth LLGs with two different opening sizes dimension of 8x8
mm and 12x12 mm were selected to investigate the performance. The opening size of
12x12 mm of woven water hyacinth was performed at the east of embankment with 5
plots and opening size of 8x8 mm of woven water hyacinth was performed at the west
of embankment This study evaluated the runoff rate and soil loss at different plots of
test with woven water hyacinth LLGs in combination with Ruzi grass cover at the
growing periods of 4, 6, 8, and 10 weeks.
80
The limited life geotextiles made from water hyacinth LLGs can reduce the amount of
runoff. The soil cover with LLGs with 8x8 mm opening size has 17% decreased the
amount of soil loss rather than LLGs with 12x12 mm opening size while the runoff
were similar for both opening size. After 4, 6, 8 and 10 weeks growing periods of Ruzi
grasses the amount of soil loss was reduced because of increasing coverage to the soil.
From the measurements of flow rate of rainfall and amount of soil loss, better results
were found when growing Ruzi grass was combined with woven water hyacinth LLGs.
Moreover, the coated water hyacinth LLGs can increase the life time about 25 % of
non-coated LLGs but there were no effect to the runoff and soil loss. Therefore, woven
water hyacinth LLGs can be effective to control soil erosion because it reduced the
impact of rainfall and reduced the runoff. Thus, plain pattern of water hyacinth can be
applied for soil erosion control in geotechnical field. It should be noted that in addition
to the properties of these vegetation covers and LLGs, the efficiency of erosion control
depends on the soil properties, the slope of the land, and rainfall intensity.
81
CHAPTER 8
CONCLUSIONS AND RECOMMENDATION
8.1 Conclusions
8.1.1 Laboratory tests
Natural fiber woven LLGs made from Kenaf and water hyacinth were investigated to
find the suitable pattern for geotextile application. Three patterns of weaving LLGs
were obtained, namely: plain, knot-plain, and hexagonal patterns. Tensile strength
tests were conducted in warp and weft directions to select the most appropriate pattern
for geotextile application. All weaving patterns of Kenaf LLGs resulted in better
tensile strength than water hyacinth. The plain pattern of woven Kenaf LLGs was
found to be the most appropriate pattern for soil reinforcement due to its high tensile
strength. Water hyacinth can be used to apply in function of soil erosion control in the
geotechnical field application due to its high moisture absorption properties.
The interaction coeffients of Kenaf LLGs and backfill soils were investigated. The
backfill soil can be classified as silty sand from grain size distribution curve which
has 18.1 kN/m
3
maximun dry unit weight and 10% of optimum moisture content from
standard Proctor compaction test. Moreover, large scale direct shear tests were carried
out to find the interface strength of Kenaf woven LLGs with sand backfill material as
well as the sand backfill materials under different normal confining pressures of 40,
80, and 120 kPa. The important variables for LLGs reinforcement structure design
and analysis were investigated. The interaction coefficient at direct shear interfaces is
found to be 0.81. Furthermore, pullout tests were performed using normal confining
pressures of 20, 40 and 60 kPa which cover the range of possible applied confining
pressures in the field applications. Slippage failure occurred below the normal
confining pressure of 40 kPa and tensile failure occurred at higher normal confining
pressures. The pullout interface coefficients for Kenaf LLGs were 1.11, 1.07 and 0.88
corresponding to applied normal pressures of 20 kPa, 40 kPa and 60 kPa. Since the
data from large scale direct shear test was lower, its interface coefficient 0.8 was used
for subsequent analyses. Based on the results of the interaction between the Kenaf
LLGs reinforcement and backfill soil, the direct shear and pullout resistances indicate
that the Kenaf LLGs can be used for temporary soil reinforcement applications.
Woven limited life geotextile made from water hyacinth was used to investigate
erosion control application. The two different opening size dimensions of 8x8 mm and
12x12 mm in plain pattern were obtained. The tensile strength of 8 mm opening size
of water hyacinth LLGs has higher strength than 12x12 mm opening size. Hydraulic
transmissivity test were performed to get flow rate per unit width within the plane of
geotextile. The averages of transmissivity has decreased when applied more pressure
for both 8x8 mm and 12x12 mm opening size of woven water hyacinth LLGs as
expected. In addition, permittivity test were performed the get volumetric flow rate of
water per unit cross-sectional area per unit head. The average flow rate on woven
water hyacinth LLGs has increased at higher hydraulic gradient in permittivity test.
82
8.1.2 Performance of full scale test embankment
The full-scale test embankment was constructed at the AIT campus using silty sand
silty sand backfill materials reinforced with woven Kenaf LLGs to evaluate the
performance of the Kenaf LLGs natural fibers geotextiles on soft Bangkok clay. The
embankment construction February 6, 2011 and completed after 10 days on February
15, 2011. The monitoring period during construction was measured everyday day and
after construction was measured weekly for 2 months thereafter, every months until
250 days after construction due to flooding crisis in Thailand in 2011.
The embankment was instrumented with settlement plates, piezometers, and wire
extensometers. The instrumentations in the subsoil that were installed prior to the
construction of the woven Kenaf LLGs reinforced embankment. Consequently, the
embankment behavior was monitored concerning its surface settlement, subsurface
settlement, excess pore water pressure and deformation of reinforcements. After
construction, the rates of settlement rapidly increased until 250 days from the end of
construction. Rapid dissipation of excess pore water pressures were observed at the
end of construction. The excess pore water pressures slowly dissipated and tend to be
constant after 250 days. Moreover, the time to reach 90% average degree of
consolidation (t
90
) based on the observed data was 250 days. Thus, the embankment
had reinforcements until 250 days at degree for consolidation (t
90
) which was less than
the life of Kenaf LLGs estimated at one year (365 days).
In addition, at middle of the embankment, the Kenaf LLGs reinforcement had higher
deformations in both coated and non-coated Kenaf LLGs and the deformations
increased with time. The mobilized tensile force (T
mob
) was 20 kN/m at the estimated
4% of strain from wide width tensile test and and 22 kN/m from pullout test. Using
the mobilized tensile force of 20 kN/m in the reinforcement, the factor of safety
increased by 17 %. The polyurethane coating did not affect to the embankment
behavior as its purpose is to extend the life of Kenaf LLGs.
8.1.3 Numerical analyses of full scale test embankment
FEM 2D and 3D numerical simulations were carried out to study the behavior of a full
scale test embankment reinforced with Kenaf LLGs on soft Bangkok clay. The rates
of settlement from FEM 2D method overestimated the observed settlements data
while the FEM 3D predictions agreed with observed settlements due to the three-
dimensional geometrical loading of the embankment with length to width ratio (L/B)
of 1.0. Regarding the maximum excess pore-water pressures at the locations of 3m
and 6m depth, the FEM 2D analyses overestimated while the FEM 3D simulation
yielded satisfactory agreement with the observed data. The reinforcement
deformations and stresses in both coated and non-coated Kenaf LLGs reinforcement
have higher values at the middle portions of the embankment and the predicted results
from FEM 3D simulation yielded closer deformations of Kenaf LLGs reinforced than
the FEM 2D simulation. Consequently, FEM 3D simulation captured the overall
behavior of the Kenaf LLGs reinforced embankment with more reasonable agreement
between the field observations and the predicted values compared to the FEM 2D
simulation. The behavior of the sections on coated and non-coated LLGs were similar.
Consequently, the Kenaf LLGs can be applied for short term embankment
reinforcement in order to improve the stability of embankment on soft clay.
83
8.1.4 Erosion control test of full scale test embankment
The water hyacinth yarns were fabricated to woven geotextile to help stabilize the
slope and improved the growth of vegetation for erosion control application. Two
types of opening sizes (8x8 and 12x12 mm) were obtained. The study was aimed to
access the effectiveness of the new material for geotextile made from water hyacinth
to control soil erosion in a representative slope at different plots of tests. The erosion
control test was performed by using the artificial rainfall. The flow rates of water
runoff and amount of soil loss were investigated. From results of flow rate of the
runoff and amount of soil loss, better results were found when growing Ruzi grass
with woven water hyacinth LLGs. The 4, 6, 8 and 10 weeks growing period of Ruzi
grasses reduced the amount of soil loss because of increasing coverage of Ruzi grass.
The data indicated that soil cover with LLGs with 8x8 mm opening size has
significantly reduced the amount of surface runoff rather than LLGs with 12x12 mm
opening size. The coated water hyacinth LLGs can increase the life time about 25% of
non-coated LLGs and there were no effect to the runoff and soil loss. Consequently,
the limited life geotextile made from water hyacinth LLGs combined with Ruzi grass
can reduce the flow rate of runoff and amount of soil loss. Woven water hyacinth
LLGs can makes effectiveness to control soil erosion. Thus, plain pattern of water
hyacinth can be applied for soil erosion control in geotechnical field.
8.2 Recommendations for Further Researches
1. Numerical analysis of laboratory pullout and direct shear test should be
conducted to study the appropriate model for simulation backfill material and
Kenaf LLGs reinforcement.
2. Dynamic or moving load analyses of full scale test embankment should be
investigated to study its behavior in seismic area or used as the roadway
embankment structure.
3. To conduct full scale test embankment with varying slope inclinations and
backfill materials in order to investigate its performance and compare with the
simulated results.
84
REFERENCES
Abdullah, C. H. and Edil, T. B. (2007). Behaviour of geogrid-reinforced load transfer
platforms for embankment on rammed aggregate piers. Geosynthetics
International, Vol.14, No. 3, pp. 141153.
Abiera, H. O. (1991). Mechanically Stabilized Earth using TENSAR, Bamboo and
Steel Grid Reinforcements with Lateritic Soil as backfill, M. Eng. Thesis
GT-90-21, Asian Institute of Technology, Bangkok, Thailand.
Alfaro, M. C., (1996). Reinforced soil wall-embankment system on soft foundation
using inextensible and extensible grid reinforcements. Phd. Dissertation,
Saga University, Japan.
Amad F. and Bateni F. (2009). Performance of oil palm fibres on improvement of soil
strength. Proceedings GIGSA GeoAfrica 2009 Conference, September, 2 5,
2009, Cape Town, Africa.
Anand S. (2008). Designer natural fiber geotextiles-A new concept. Indian Journal of
Fiber & Textile Research, Vol. 33, pp. 339-344.
Anderson, L. R. and Neilsen, M. R. (1984). Pullout resistance of wire mats embedded
in soil. Report for the Hilfiker Co., from the Civil and Enviromental
Engineering Department, Utah State University, Logan Utah.
Arai, K. Yoshida K., Tsuji S., and Yokota Y., (2007). Full-scale model test and
numerical analysis of reinforced soil retaining wall. The International
Symposium on Earth Reinforcement (IS Kyushu), Fukuoka in Japan.
Arno P., (1989). Concise Encyclopedia of Wood and Wood-based Materials, 1
st
ed.,
Pergamon Press, New York, US
Asaoka, A., (1978). Observational procedure of settlement prediction, Soils and
Foundations, 18(4), 87-101.
ASTM, (2007). ASTM D422-63. Standard Test Method for Particle- Size Analysis of
Soils. ASTM, West Conshohocken, Pennsylvania.
ASTM, (2007). ASTM D698. Standard Test Methods for Laboratory Compaction
Characteristics of Soil Using Standard Effort. ASTM, West Conshohocken,
Pennsylvania.
ASTM, (2007). ASTM D1102-84. Standard test method for ash in wood. ASTM,
West Conshohocken, Pennsylvania.
ASTM, (2007). ASTM D 2435. Standard Test Method for One- Dimensional
Consolidation Properties of Soils . ASTM, West Conshohocken ,
Pennsylvania.
85
ASTM, (2007). ASTM D2850. Standard Test Method for Unconsolidated-Undrained
TriaxialCompression Test on Cohesive Soils. ASTM, West Conshohocken,
Pennsylvania.
ASTM, (2007). ASTM D3080. Standard Test Method for Direct Shear Test of Soils
Under Consolidated Drained Conditions . ASTM, West Conshohocken ,
Pennsylvania.
ASTM, (2007). ASTM D4491. Standard Test Methods for Water Permeability of
Geotextiles by Permittivity). ASTM, West Conshohocken, Pennsylvania.
ASTM, (2007). ASTM D4716 (Standard Test Method for Determining the (In-plane)
Flow Rate per Unit Width and Hydraulic Transmittivity of a Geosynthetic
Using a Constant Head). ASTM, West Conshohocken, Pennsylvania.
ASTM, (2007). ASTM D5321. Standard Test Method for Determining the Coefficient
of Soil and Geosynthetic or Geosynthetic and Geosynthetic Friction by the
Direct Shear Method. ASTM, West Conshohocken, Pennsylvania.
ASTM, (2009). ASTM D4595. Standard Test Method for Tensile Properties of
Geotextiles by the Wide-width Strip Method. ASTM, West Conshohocken,
Pennsylvania.
Asaoka, A., 1978. Observation procedure of settlement prediction. Soils and
Foundations, Vol. 18, No. 4, pp. 87-101.
Askew, M.F. (2000). Summary ReportFibre Crops. IENICA, Ref. 1495, August,
pp. 2338 (Chapter 5).
Auvinet and Gonzalez. (2000), Three-dimensional reliability analysis of earth slopes,
Computers and Geotechnics, Vol. 26, pp. 247261.
Bathurst, R. J., Allen, T. M. and Walters, D. L. (2005) Reinforcement loads in
geosynthetic walls and the case for a new working stress design method
(Mercer Lecture). Geotextiles and Geomembranes, Vol. 23, No. 4, pp. 287
232.
Balasubramaniam, A.S., Hwang, Z.M., Waheed, U., Chaudhry, A.R., and Li, Y.G.
(1978). Critical state parameters and peak stress envelopes for Bangkok
clays, Quarterly Journal of Engineering Geology, Vol. 11, pp. 219-232.
Basu G., Roy A. N., Bhattacharyya S. K., Ghosh S. K. (2009). Construction of
unpaved rural road using jute-synthetic blended woven geotextile: a case
study. Geotextiles and Geomembranes, Vol. 27, No. 6, pp. 50612.
Bathurst, R. J., Allen, T. M. and Walters, D. L. (2005). Reinforcement loads in
geosynthetic walls and the case for a new working stress design method
(Mercer Lecture). Geotextiles and Geomembranes, Vol. 23, No. 4, pp. 287
232.
86
Beckman, W.K., Mills, W.H. (1957). Cotton fabric reinforced roads, Engineering
News Records, Vol. 115, No. 14, pp. 453455.
Bera, A. K., Ghosh A. and Ghosh A. (2008). Unconfined compressive strength of fly
ash reinforced with jute geotextiles. Proceedings Indian Geotechnical
Conference (IGC - 2008) December 17-19, 2008, Bangalore, India.
Beg M.D.H. and Pickering K.L., (2008). Accelerated weathering of unbleached and
bleached Kraft wood fibre reinforced polypropylene composites, Polymer
Degradation and Stability. Vol. 93, No. 10, pp. 1939-1946.
Bergado, D.T., Teerawattanasuk, C. (2008). 2D and 3D numerical simulations of
reinforced embankments on soft ground. Geotextiles and Geomembranes,
Vol. 26, No.1, pp. 3955.
Bergado, D.T., Long, P.V., Murthy, B.R.S. (2002). A case study of geotexile-
reinforced embankment on soft ground. Geotextiles and Geomembranes,
Vol. 20, No. 6, pp. 343365.
Bergado, D. T., Teerawattanasuk C., Youwai S. and Voottipruex P. (2000). FE
modeling of hexagonal wire reinforced embankment on soft clay. Canadian
Geotechnical Journal, Vol. 37, pp. 1209-1226.
Bergado, D. T., Chai, J. C. and Miura, N., (1995). FE Analysis of grid reinforced
embankment system on soft Bangkok clay. Computers and Geotechnics,
Vol. 17, pp. 447-471.
Bergado, D. T. and Chai J. (1994). Pull-out force/displacement relationship of
extensible grid reinforcements, Geotextiles and Geomembranes, Vol.13,
No.5, pp. 295-316.
Bergado D. T., Daria P.A.M., Sampaco C.L., Alfaro M.C., 1991a. Prediction of
embankment settlement by in-situ test, Geotechnical Testing Journal, Vol.
14, No. 4, pp. 425-439.
Bergado, D. T., Shivashankar, R., Sampaco, C. L., Alfaro, M. C., Anderson, L. R. and
Balasubramaniam, A. S., (1991) Behavior of of welded wire wall with poor
quality cohesive-frictional backfills on soft Bangkok clay (A case study).
Canadian Geotechnical Journal,Vol. 28, pp. 860-880.
Bergado, D. T., Enriques, A. S., Sampaco, C. L. and Alfaro, M. C., 1990. Inverse
analysis of geotechnical parameters on improved soft Bangkok clay. Journal
of Geotechnical Engineering ASCE, Vol. 118, No. 7, pp. 1012-1030.
Bernal, A., Salgado, R., Swan, R. and Lovell, C. (1997). Interaction between tire
shreds, rubber-sand and geosynthetics. Geosynthetics International, Vol. 4,
No. 6, pp. 623-643.
87
Bhattacharyya, R., Fullen, M.A., Davis, K. and Booth, C.A. (2010). Use of palm-mat
geotextiles for rainsplash erosion control, Geomorphology, Vol. 119, pp. 52-
61.
Bhattacharyya, R., Fullen, M.A., Davies, K., Booth, C.A., (2009). Utilizing palm leaf
geotextile mats to conserve a loamy sand soil in the United Kingdom.
Agriculture, Ecosystems & Environment, Vol. 130, pp. 5058.
Bhattarai. R. Kalita, P.K., Yatsu, S., Howard, H.R., and Svendsen, N.G. (2011).
Evaluation of compost blankets for erosion control from disturbed Lands,
Journal of Environmental Management, Vol. 92, pp. 803-812.
Bjerrum, L., (1972). Embankment on soft ground, Proc. ASCE Conf. on Performance
of Earth Support Structure, Purdue University, Lafayette, Indiana, Vol2, pp
1-54.
Biot, M. A., (1941). General theory of three dimensional consolidation, Journal of
Applied Physics, Vol. 12, pp. 155-164.
Brianon, L., Villard, P., (2008). Design of geosynthetic-reinforced platforms
spanning localized sinkholes. Geotextiles and Geomembranes, Vol. 26, No.
5, pp. 416428.
Brink, M. and Escobin, R.P., (2003). Plant Resources of South-East Asia, Fibre
Plants, Backhuys Publishers, Leiden. the Netherlands. No. 17.
Britto, A. M. and Gunn, M. J., (1987). Critical State Soil Mechanics via Finite
Elements. Ellis Horwood, Chichester, U.K.
Burland, J.B. (1990). The compressibility and shear-strength of natural clay,
Geotechnique, Vol, 40, No. 3, pp. 329-378.
Chai, J. C., 1992. Interaction Behavior between Grid Reinforcement and Cohesive
Frictional Soils and Performance of Reinforced Wall/Embankment on Soft
Ground. D. Eng Dissertation No. GT-91-1, Asian Institute of Technology,
Bangkok, Thailand.
Chandrakaran S., Sankar N. and Subaida E.A. (2008). Performance of unpaved road
section reinforced with woven coir geotextile. Proceedings Indian
Geotechnical Conference, December 17-19, 2008, Bangalore, India.
Choudhury P.K., Das A. and Sanyal T. (2008). Bio-engineering approach with jute
geotextile for slope stabilization,
http://www.jute.com/HTML/Paper%20ASIAN%2008.pdf
Christopher, B. R., Gill, S. A., Giroud, J. P., Juran, I., Schlosser, F., Mitchell, J. K.
and Dunnicliff, J., (1989). Design and construction guidelines. Reinforced
Soil Structures, Vol. 1, STS Consultants, Ltd., Northbrook, Illinois.
88
Chattopadhyay B.C. and Chakravarty S. (2009) Application of jute geotextiles as
facilitator in drainage, Geotextiles and Geomembranes, Vol. 27, pp. 156.161.
Chattopadhyay B.C. and Chakravarty S. (1999). River bank protection with geojute,
an efficient cost effective method, Journal of IE(I) 79 (1999), pp. 154157.
Cox, J. B., 1968. A review of the engineering characteristics of the recent marine
clays in southeast asia, Research Report No. 6, Asian Institute of
Technology, Bangkok, Thailand.
Datye, K.R., and Gore, V.N., (1994). Application of natural geotextiles and related
products. Geotextiles and Geomembranes, Vol. 13 No. 6/7, pp. 371388.
D Appolonia, D. J., Poulos, H. G. and Ladd, C. C., 1971. Initial settlements of
structure on clay, Journal of the Soil Mechanics and Foundation Division,
pp. 1359-1376.
Dinu I.and Saska, M. (2007). Production and properties of soil erosion control mats
made from sugarcane bagasse, Journal American Society of Sugar Cane
Technologists, Vol. 27, pp. 36-47.
Dumn, M. J., (1977). Geotechnical observation from a deep bore hole at Rangsit, M.
Eng. Thesis No. 1004, Asian Institute of Technology, Bangkok, Thailand.
Duncan, J. M. and Chang, C. Y., (1970). Non-linear analysis of stress and strain in
soil. Journal of Soil Mechanics and Foundation Engineering Division,
ASCE, Vol. 96, No. 5, pp. 1689-1653.
Dutta, S., Padade, A. H. , and Manda, J. N. (2012). Experimental study on natural
bamboo geogencased stone column. Proceedings 5th Asian Regional
Conference on Geosynthetics, Bangkok, Thailand.
English, B. (1995). Geotextiles A specific application of biofibers, Proceedings of
Conference on Research in Industrial Application of Non Food, Copenhagen,
Denmark, pp. 79-86.
Farrag, K., Acar, Y. and Juran, I. (1993). Pullout resistance of geogrid reinforcements.
Geotextiles and Geomembranes, Vol. 12, pp. 133-159.
Foose, G., Benson, C. and Bosscher, P. (1996). Sand reinforced with shredded waste
tires. Journal of Geotechnical and Geoenvironmental Engineering, ASCE,
Vol. 122, No. 9, pp. 760-767.
Ghosh A. and Bera A. K. (2005). Bearing capacity of square footing on pond ash
reinforced with jute-geotextile, Geotextiles and Geomembranes, Vol. 23, pp.
144-173.
Gunnarsson C.C. and Petersen C.M., (2007). Water hyacinths as a resource in
agriculture and energy production: A literature review. Waste Management
Journal. Vol. 27, pp. 117-129.
89
Gurung, N, Iwao, Y, Madhav, M. R., (1999). Pullout test model for extensible
reinforcement. International Journal for Numerical and Analytical Methods
in Geomechanics, Vol. 23, pp. 1337-1348.
Hasan, S.H., Talat, M. and Rai, S., (2007). Bioresour. Technol. No. 98, pp. 918-
928.
Hinchberger, S. D., and Rowe, R. K., (2003). Geosynthetic reinforced embankments
on soft clay foundations : predicting reinforcement strains at failure.
Geotextiles and Geomembranes, Vol. 21, No. 3, pp. 151-175.
Hird, C. C, and Kwok, C. M., (1989). FE studies of interface behavior in reinforced
embankments on soft ground, Computers and Geotechnics, Vol. 8, pp. 111-
131.
Holtz, R.D., Christopher, B.R. and Berg, R.R. (1997), Geosynthetic Engineering,
BiTech Publishers, Ltd, Pp. 452.
Islam, M. N., Serker, N.H.M., Rasel, H. M. and Matin, I. (2009) Stabilization of soft
organic clayey soil using natural fibres, International Symposium on
Geotechnical Engineering, Ground Improvement and Geosynthetics for
Sustainable Mitigation and Adaptation to Climate Change including Global
Warming, Bangkok, Thailand.
Janbu, N., (1963). Soil compressibility as determined by oedometer and triaxial test,
Proc. Europe Conf. Soil Mech. Found. Eng., Wiesbaden, Germany, Vol. 1,
pp. 19-25
Janbu, N., Bjerrum, L., and Kjaersli, B., (1956). Veiledning ved losning av
fundamenterignsoppaver, N.G.I. Publication, No. 16, pp. 93.
Jankauskas, B., Jankauskiene, G., Fullen, M.A. and Booth, C.A. (2008) Utilizing
palm-leaf geotextiles to control soil erosion on roadside slopes in Lithuania.
Lithuanian Journal of Science (Agricultural Sciences), Vol. 3, pp. 22-28.
Jankauskas, B., Jankauskiene, G. and Fullen, M.A. (2012) Soil conservation on road
embankments using palm-mat geotextiles: Field studies in Lithuania. Soil Use
and Management, Vol. 28, pp. 266275.
Jenkins, S.H., (1930). Biochem. J. 24 pp. 1428-1432.
Jewell, R. A. (1986). The mechanics of reinforced embankments on soft soils,
Proceedings of the Prediction Symposium on a Reinforced Embankment on
Soft Ground, 17-18
th
Sept., 1986, Kings College, Strand, London U.K., pp.
159-163.
Jewell, R. A., Milligan, G. W. E., Sarsby, R. W. and Dubois, D. (1984). Interaction
between soil and geogrids, Proceedings of the Symposium on Polymer Grid
Reinforcement in Civil Eng., Thomas Telford Limited, London, U.K., pp.
19-29.
90
Jeon, H. Y. (2012). Geosynthetics innovation for sustainable engineering.
Proceedings 5th Asian Regional Conference on Geosynthetics, Bangkok,
Thailand.
John, N.W.M., (1987). Geotextiles, Blackie & Son Ltd., London, UK.
Kabiling, M. B., 1997. Pullout capacity of different hexagonal link wire sizes and
configurations on sandy and volcanic ash (lahar) backfills, M. Eng. Thesis
No. GE 96-4, Asian Institute of Technology, Bangkok, Thailand.
Kanjanophas, S., (1969). Compressibility of Bangkok clay in the weathered clay zone,
M. Eng Thesis No. 338, Asian Institute of Technology, Bangkok, Thailand.
Karpurapu, R. and Bathurst, R. J. (1995). Behavior of geosynthetic reinforced soil
retaining walls using the finite element method. Computers and
Geotechnics, Vol. 17, No. 3, 279299.
Kirby, RH., (1963). Vegetable bres. World crops Books. London UK: Leonard Hill
Ltd.;
Koerner, R. M. (1998). Designing with geosynthetics, Prentice-Hall Inc, New Jersey,
USA.
Kondner, R. L and Zelasko, J. S., (1963). A hyperbolic stress-strain formulation of
sand, Proc. of Second Pan American Conf. on Soil Mech. and Foundation
Engg.
Kongkitkul, W. (2001). Numerical and Analytical Modeling of Interaction between
Hexagonal Wire Mesh Reinforcement and Silty Sand Backfill during In-soil
Pullout Tests, M. Eng. Thesis No. GE 00-18, Asian Institute of Technology,
Bangkok, Thailand.
Kumar, A. B., Ambarish, G., and Amalendu, G. (2008). Effect of specimen size on
unconfined compressive strength of fly ash reinforced with jute geotextiles.
Proceedings Indian Geotechnical Conference, December 17-19, 2008,
Bangalore, India.
Lai, Y. P., Bergado, D. T., Lorenzo, G. A., and Duangchan, T. (2006). Full-scale
reinforced embankment on deep jet mixing improved ground. Ground
Improvement, Vol. 10, No. 4, pp. 153-164.
Lekha, K.R. and Kavitha, V. (2006). Coir geotextile reinforced clay dykes for
drainage of low-lying areas, Geotextiles and Geomembranes, Vol.24, No. 1,
pp. 3851.24.
Lekha, K.R. (2004). Field instrumentation and monitoring of soil erosion in coir
geotextile stabilized slopes- A case study. Geotextiles and Geomembranes,
22, No. 5, pp. 399413.
91
Li, A.L., and Rowe, R.K. (2008). Effects of viscous behaviour of geosynthetic
reinforcement and foundation soils on embankment performance. Geotextiles
and Geomembranes, Vol. 26, No. 4, pp. 317334.
Li Y., Mai Y-W., and Ye L. (2000). Sisal fibre and its composites: a review of
recent developments. Composites Science and Technology, Vol. 60 pp. 2037-
2055.
Long, P. V., Bergado, D. T. and Balasubramaniam, A. S. (1996) Stability analysis of
reinforced and unreinforced embankments of soft ground. Geosynthetics
International, Vol. 3, No. 5 pp. 583604.
Maity, J., Chattopadhyay, B.C., and Mukherjee, S.P. (2012). Application of fibers
from sabai grass in construction of subbase of roads in conjunction with sand.
Proceedings 5th Asian Regional Conference on Geosynthetics, Bangkok,
Thailand.
Mandal, J.N., (1989). Geojute. International Geotextile Society News. Vol. 5, No. 3,
November, Canada, pp. 12.
Malik, A., (2007). Environmental challenge vis a vis opportunity: The case of water
hyacinth, Environment International. Vol. 33, No.1, pp. 122-138.
Mesri, G., 1973. Coefficient of secondary compression, Journal of the Soil Mechanic
and Foundations Division, ASCE, Vol. 99, No. SM1, pp. 122-137.
Methacanon, P., Weerawatsophon, U., Sumransin, N., Prahsarn, C., and Bergado,
D.T. (2010). Properties and potential application of the selected natural fibers
as limited life geotextiles. Carbohydrate Polymers, Vol. 82, No. 4, 1090
1096.
Mir, E. N., 1996. Pullout and direct shear test of hexagonal wire mesh reinforcements
in various fill materials including lahar from Mt. pinatubo, philipines. M.
Eng. Thesis No. GE 95-18, Asian Institute of Technology, Bangkok,
Thailand.
Mitchell, D.J., Barton, A.P., Fullen, M.A., Hocking, T.J., Zhi, W.B., Yi, Z, (2003).
Field studies of the effects of jute geotextiles on runoff and erosion in
Shropshire, UK. Soil Use & Management, Vol.19, pp.182184.
Mohanty, A. K., Misra, M. and Hinrichsen, G. (2000). Biofibres, biodegradable
polymers and biocomposites: An overview. Macromolecular Materials and
Engineering, Vol. 276-277, No. 1, pp. 124.
Mwasha, A. and Petersen, A., (2010).Thinking outside the box: The time dependent
behavior of a reinforced embankment on soft soil, Materials and Design,
Vol. 31, pp. 23602367.
Mwasha, A. (2009a). Using environmentally friendly geotextiles for soil
reinforcement: A parametric study, Materials and Design, Vol. 30, pp. 1798-
1803.
92
Mwasha, A. (2009b). Designing bio-based geotextiles for reinforcing an embankment
erected on soft soil, Materials and Design, Vol. 30, No. 7, pp. 26572664.
Mwasha, A.P., Sarsby, R.W., and Karri, R.S. (2002). Potential for using Sisal bres as
geotextiles in East Africa. Proceedings 9
th
Internatioinal Conferences of
International Association for Engineering Geology and the Environment,
Durban, South Africa.
Nurul Islam Md., Younus Ali M.M., Kamrujjaman Seker N.H.M., Ayesha Siddika.
(2010). The Seventeenth Southeast Asian Geotechnical Conference, May 10-
13, 2010, Taipei, Taiwan.
Nurul Islam Md., Rasel H.M, Kamrujjaman Serker N.H.M. and Matin I., (2009).
Stabilization of soft organic clayey soil using natural fibers. International
Symposium on Geotechnical Engineering, Ground Improvement and
Geosynthetics for Sustainable Mitigation and Adaptation to Climate Change
including Global Warming, December 3-4, 2009, Bangkok, Thailand
Park, J. H., Yuu, J. J., and Jeon, H. Y. (2010). New suggestion of testing and
regulation for green geosynthetics by bio-degradable mMaterials. The 1st
International GSI-Asia Geosynthetics Conference Taichung, Taiwan, Nov.
16-18, 63-65.
Pilkey O.H., Morton R.A., Kelley, J.T. and Penland, S. (1989). Coastal Land Loss..
American Physical Union, Washington, D.C. U.S.A., p.73
Prempramote, S., Interaction between geogrid reinforcement and tire chip-sand
mixture, M. Eng. Thesis No. GE-04-12, Asian Institute of Technology,
Bangkok, Thailand
Prichard M. (1999). Vegetable Fibre Geotextiles. PhD thesis, Bolton Institute, UK.
Rajagopal, K., and Sanyal, T. (2012). Sustainable infrastructure development
including limited life Geosynthetics. Proceedings 5th Asian Regional
Conference on Geosynthetics, Bangkok, Thailand.
Ramesh H.N., Manoj Krishna K.V., and Mamatha H.V. (2010). Compaction and
strength behavior of lime-coir fiber treated Black Cotton soil. Geomechanics
and Engineering, Vol. 2, No. 1, pp. 19-28.
Ranganathan S.R. (1994). Development and potential of jute geotextiles, Geotextiles
and Geomembranes, Vol.13, pp. 421-433.
Rawal A. and Anandjiwala R. (2007). Comparative study between needlepunched
nonwoven geotextile structures made from flax and polyester fibres,
Geotextiles and Geomembranes, Vol. 25, pp. 61-65.
Reddy, N. and Yang Y., (2005). Biofibers from agricultural byproducts for industrial
applications, Trends Biotechnol, Vol. 23, pp. 2227.
93
Reddy, N. and Yang, Y. (2005). Structure and properties of high quality natural
cellulose fibers from cornstalks, Polymer, Vol. 46, pp. 5494-5500.
Reddy, N. and Yang, Y., (2007). Natural cellulose fibers from switchgrass with
tensile properties similar to cotton and linen, Biotechnology and
Bioengineering, Vol. 97, No. 5, 1021-1027.
Rickson, R.J., (1988). The use of geotextiles in soil erosion control: comparison of
performance on two soils. In: Rimwanich, S. (Ed.), Land Conservation for
Future Generations Proc. of Conference V. International Soil Conservation,
January1829, Bangkok. Department of Land Development, Ministry of
Agriculture and Cooperatives, Bangkhen, Bangkok, pp. 961970.
Roscoe, K. H. and Burland, J. B., (1968). On the generalized stress-strain behavior of
wet clays, Proceeding of Engineering Plasticity, Cambridge University
Press, pp. 539-609.
Rowe, R. K. and Ho, S. K., (1997). Continuous panel reinforced soil walls on rigid
foundations. Journal of Geotechnical and Geoenvironmental Engineering,
ASCE, Vol. 123, No. 10, pp. 912-920.
Rowe, R. K. and Li, A. L. (2002) Behavior of reinforced embankments on soft rate-
sensitive soils. Geotechnique, Vol. 52, No. 1, pp. 2940.
Rowe R.K. and Taechakumthorn C. 2008. Combined effect of PVDs and
reinforcement on embankments over rate-sensitive soils. Geotextiles and
Geomembranes, Vol. 26 No. 3, pp. 239-249.
Sahaa, P., Roy, D., Mannaa, S., Adhikari, B., Senc, R., and Roy, S. (2012). Durability
of transesteried jute geotextiles. Geotextiles and Geomembranes. Vol.35,
pp.69-75.
Sarsby R.W. (2007). Use of Limited Life Geotextiles (LLGs) for basal reinforcement
of embankments built on soft clay, Geotextiles and Geomembranes, Vol. 25,
302-310.
Sarsby R.W. (2005). Limited life geotextiles (LLGs) for soil reinforcement,
Proceedings International Symposium on Tsunami Reconstruction with
Geosynthetics, Bangkok, Thailand.
Sarsby, R.W., Ali, M., Alwis, R., de Khaffaf, J.H., and McDougall, J.M. (1992). The
use of natural and low cost soil reinforcement, International Proceeding of
International Conference on Nonwoven. The Textile Institute, New Delhi,
December, pp. 297310.
Schuhe, (1999). Fibre crops. www.canuc.ca/Expert%20Network.nsf/52c0f5cd1a56224
Shivashankar, R., (1991). Behavior of Mechanically Stabilized Earth (MSE)
Embankment with Poor Quality Backfills on Soft Clay Deposits, Including a
Study of the Pullout Resistances. D. Eng. Dissertation, Asian Institute of
Technology, Bangkok, Thailand.
94
Shukla, S., and Kumar, R., (2008) Overall slope stability of prestressed geosynthetic-
reinforced embankments on soft ground. Geosynthetics International, Vol.
15, No. 2, pp. 165-171.
Sinsakul S., Tiyapairat S., Chaimanee N. and Aramprayoon, A. (2002). The changing
coastal area at the Gulf seacoast of Thailand, Technical report, Department of
Mineral Resources, pp. 173.
Sinsakul S., Tiyapairat S., Chaimanee N. and Aramprayoon, A. (1999). The Changing
Coastal Area at the Andaman Seacoast of Thailand, Technical Report,
Department of Mineral Resources, pp. 58.
Siripong A. (2008). The Beaches are Disappearing in Thailand. Proceedings of the
International Symposium on Geoscience Resourses and Environments of Asian
Terranes (GREAT 2008), November 24-28, 2008, Bangkok, Thailand.
Skempton, A. W. and Bjerrum, L., (1957). A contribution to the settlement analysis of
foundations on clay. Geotechnique, Vol. 7, No. 4, pp. 168-178.
Skinner, G. D. and Rowe R. K. (2003). Design and behaviour of geosynthetic
reinforced soil walls constructed on yielding foundations. Geosynthetics
International, Vol. 10, No. 6, pp. 200214.
Skinner, G. D. and Rowe, R. K. (2005). A novel approach to estimating the bearing
capacity stability of geosynthetic reinforced retaining walls constructed on
yielding foundations. Canadian Geotechnical Journal, Vol. 42, No. 3, pp.
763779.
Smets T., and Poesen J. (2009). Impacts of soil tilth on the effectiveness of biological
geotextiles in reducing runoff and interrill erosion. Soil & Tillage Research,
Vol. 103, pp. 356363.
Smith, I. M., Su, N., (1997) .Three-dimensional FE analysis of a nailed soil wall
curved in plan. International Journal for Numerical and Analytical Methods
in Geomechanics, Vol. 21, pp. 583-599.
Smith I. M. and Griffith D. V., (1988). Programming the Finite Element Method,
Second Edition, John Willy and Sons, Chichester, U. K.
Srikongsri, A. (1999). Analytical model for Interaction between Hexagonal Wire
Mesh and Silty Sand Backfill. M. Eng. Thesis No. GE-98-14, Asian Institute of
Technology, Bangkok Thailand.
Subaida E. A., Chandrakaran S., Sankar N. (2009). Laboratory performance of
unpaved roads reinforced with woven coir geotextiles. Geotextiles and
Geomembranes, Vol. 27, No. 3, pp. 20410.
Subaida E.A., Chandrakaran S. and Sankar N. (2008). Experimental investigations on
tensile and pullout behaviour of woven coir geotextiles, Geotextiles and
Geomembranes, Vol. 26, pp. 384-392.
95
Sukumaran, K., Satyanarayana, K. G., & Pillai, S. G. K. (2001). Structure, physical
and mechanical properties of plant fibers of Kerala. Metals Materials and
Processes, 13, 121136.
Tanchaisawat T., Bergado D. T., and Voottipruex P. (2009). 2D and 3D simulation of
geogrid-reinforced geocomposite material embankment on soft Bangkok
clay. Geosynthetics International, Vol. 16, No. 6, pp. 420-432.
Tanchaisawat. (2008). Interactions and performances of geogrid reinforced tire chips-
sand lightweight embankment on soft ground. Doctoral Dissertation, GE-07-
01. Asian Institute of Technology, Thailand.
Tanchaisawat, T., Bergado, D.T. and Kanjananak T., (2007). Lightweight
geomaterials for bridge approach utilization. Proceeding 16
th
Southeast Asian
Geotechincal Conference, Subang Jaya, Malaysia, pp. 8-11.
TAPPI Test Methods (1996). TAPPI T222 om-88 Klason Lignin.
Tatlisoz, N., Edil T. B. and Benson, C. H. (1998). Interaction between reinforcing
geosynthetics and soil-tire chip mixtures. Journal of Geotechnical and
Geoenvironmental Engineering, ASCE, Vol. 124, No. 11, pp. 1109-1119.
Tatlisoz, N., Benson, C. H., and Edil, T. B., 1997. Effect of fines on mechanical
properties of soil-tire chip mixtures. Testing Soil Mixed with Waste or
Recycled Materials, ASTM STP 1275, pp. 93-108.
Teerawattanasuk, C., (1997). Interaction and deformation behavior of hexagonal wire
mesh reinforcement at vicinity of shear surface on sand and volcanic ash
(lahar) backfills, M. Eng. Thesis GE-96-14, Asian Institute of Technology,
Bangkok, Thailand.
Terzaghi, K. and Peck, R. B., (1967). Soil Mechanics in Engineering Practice, 2
nd
edition, New York, Wiley, p 729.
Tin, N. (2009). Factors affecting the kinked steel grid reinforcement and modification
of K- stiffness method in MSE structure on soft ground. M. Eng. Thesis No.
GE 08-03, Asian Institute of Technology, Bangkok, Thailand.
Varsuo, R. J., Grieshaber, J. B. and Nataraj, M. S. (2005). Geosynthetic reinforced
levee test section on soft normally consolidated clays. Geotextiles and
Geomembranes, Vol. 23, No. 4, pp. 362383.
Vermeer, P. A. and Brinkgreve, R. B. J. (Editor), (1995). Finite Element Code for Soil
and Rock Analysis, A. A. Balkema, Rotterdam, Netherland.
Vinod, P. and Bhaskar, A.B. (2012) Model studies on woven coir geotextile-
reinforced sand bed. Ground Improvement, Vol. 165, No. 1, pp. 53 57.
Voottipruex, P., 2000. Interaction of hexagonal wire reinforcement with silty sand
backfill soil and behavior of full scale embankment reinforced with
hexagonal wire. D. Eng. Dissertation No. GE-99-1, Asian Institute of
Technology, Bangkok, Thailand.
96
Wongsawanon, T. (1998). Interaction between Hexagonal Wire Mesh Reinforcement
and Silty Sand Backfill. M. Eng. Thesis No. GE 97-14, Asian Institute of
Technology, Bangkok, Thailand.
Youwai, S., Bergado, D. T. and Supawiwat, N. (2004). Interaction between hexagonal
wire reinforcement and rubber tire chips with and without sand mixtures.
Geotechnical Testing Journal, Vol. 27, No. 3, pp. 1-9.
Youwai S. and Bergado D. T., (2003). Strength and deformation charactereristics of
shredded rubber tire-sand mixtures. Canadian Geotechnical Journal, Vol.
40, pp. 254-264.
97
Table 2.1 Technical characteristics to geotextiles made from man-made bres (Sarsby, 2007)
Table 2.2 The onset, weight loss, char and ash values for the pure compounds and treated and
untreated bagasse samples (Dinu and Saska, 2007)
Table 2.3 Details of the four commercial erosion mats used for comparison with the bagasse
products (Dinu and Saska, 2007)
98
Table 2.4 Product specifications (Dinu and Saska, 2007)
Table 2.5 Effectiveness of palm-mat geotextiles on soil splash erosion for the
plots (n=22 sets of measurements; total precipitation=919.2 mm)
(Bhattacharyya et al., 2010)
99
Table 2.6 Normalized chemical composition (% wt on dry basis) (Methacanon et al., 2010)
100
Table 2.7 Types and functions of Geosynthetics
Function
Types
R
e
i
n
f
o
r
c
e
m
e
n
t
S
e
p
a
r
a
t
i
o
n
C
u
s
h
i
o
n
i
n
g
F
i
l
t
r
a
t
i
o
n
T
r
a
n
s
m
i
s
s
i
o
n
I
s
o
l
a
t
i
o
n
Geotextile
Geogrid
Geonet
Geomembrane
Geocomposite
Geopipe
Geosynthetic
Clay Liners
Note: a geocomposite consists of a combination of a geotextile and a geonet; a geotextile
and a geogrid; a geogrid and a geomembrane; or a geotextile, a geonet, a geogrid, and
a geomembrane or anyone of these four materials with another material.
Table 2.8 Summary of ratios between field and laboratory derived coefficient of
consolidation (Bergado et al, 1995)
101
Table 3.1 Mass per unit areaof plain pattern
Natural Fibers Mass per Unit Area (g/m
2
)
Kenaf LLGs 1157
Water hyacinth LLGs 1469
Table 3.3 Properties of woven water hyacinth LLGs for erosion control
Item Test method 8x8 mm 12x12 mm
Thickness (mm) ASTM D5199 6.96 6.94
Mass per unit area (g/m
2
) ASTM D5261 854 648
Wide width tensile strength (kN/m) ASTM D4595 6.6 4.2
Elongation at break (%) ASTM D4595 16 13
102
Table 4.1 Properties of soil at 3m and 6m depths
22%
78%
50%
112%
30%
55%
42%
60%
Plastic Limit
Natural water content
Plasticity Index
Liquidity Index
Test results
Depth of sampling (m)
Liquid Limit
3m 6m
72% 72%
Properties
Table 4.2 Soil and parameters used in Slide 5.0 software for stability of embankment
Materials
(kN/m
3
)
c
(kPa)
(deg)
Subsoil Depth(m)
Weathered crust 0 to 2.0 17 10 23
Soft clay 2.0 to 8.0 15 4 23
Medium stiff clay 8.0 to 10.0 18 10 25
Stiff clay 10.0 to 30.0 19 30 26
Embankment fill
Weathered clay 16 17 26
Silty sand 20 10 30
100
Table 5.1 Soil models and parameters used in FEM simulation on full-scale embankment test
Materials Depth Model Material
sat
unsat
k
x
= k
z
k
y
E'ref ' * * c' ' OCR R
in
(m) behavior (kN/m
3
) (kN/m
3
) (m/day) (m/day) (kPa) (kPa) (deg)
Subsoil
Weathered crust 0 - 2 MCM undrained 17 15 0.002 0.001 3000 0.25 10 23
Soft clay 1 2 - 4 SSM undrained 15 13 0.0008 0.0004 0.14 0.028 3 23 1.55
Soft clay 2 4 - 6 SSM undrained 15 13 0.0008 0.0004 0.14 0.028 3 23 1.4
Soft clay 3 6 - 8 SSM undrained 15 13 0.0008 0.0004 0.14 0.028 3 23 1.3
Medium stiff 8 -10 MCM undrained 17 15 0.0004 0.0002 5000 0.25 10 25
Stiff clay 10-30 MCM undrained 19 17 0.004 0.002 9000 0.25 30 26
Embankment
Sand MCM drained 20 18 1 1 7500 0.3 11.3 35.63 0.8
Clay MCM drained 16 15 0.002 0.001 3000 0.25 10 26
Rin = Interface Coefficient
MCM: Mohr-Coulomb model; SSM: soft soil model; MCM: Mohr-Coulomb model.
1
0
3
104
Table 5.2 Properties of Kenaf LLGs
Properties of Kenaf LLGs Natural Fibers
Mass per unit area (g/m
2
) 1157
Thickness (mm) 5.27
Tensile strength (kN/m) 20
Elongation (%) 4
Toughness (kN/m) 1.5
Axial stiffness (kN/m) 400
105
S
t
i
f
f
c
l
a
y
F
i
g
.
2
.
1
S
o
i
l
p
r
o
f
i
l
e
a
n
d
p
r
o
p
e
r
t
i
e
s
o
f
t
h
e
s
u
b
s
o
i
l
a
t
t
h
e
A
s
i
a
n
I
n
s
t
i
t
u
t
e
o
f
T
e
c
h
n
o
l
o
g
y
c
a
m
p
u
s
(
B
e
r
g
a
d
o
e
t
a
l
.
,
2
0
0
0
)
106
Fig. 2.2 Classification of natural fibers (Mohanty et al., 2000)
Fig. 2.3 Stress-strain properties of various vegetable fiber yarns (Anand, 2008)
Natural fibers
Plant Animal Mineral fibers
Seed Fruit Bast Leaf Wood Stalk
Wool Silk
Asbestos
Cotton
Kapok
Coir Flax
Hemp
Jute
Kenaf
Ramie
Pineapple
Sisal
Henequen
Abaca
Wheat
Rice
Maize
107
Fig. 2.4 Axial stress versus axial strain curve for unreinforced and
reinforced fly ash (Bera et al., 2008)
Fig. 2.5 Typical stress-strain curves obtained from triaxial CD test
on soil reinforced with OPEFB fibers (Amad and Bateni, 2009)
108
Fig. 2.6 Typical stress-strain curves obtained from triaxial CD test on effect of fiber
length of 15, 30 and 45mm (Amad and Bateni, 2009)
Fig. 2.7 Effect on friction angle between coated and uncoated OPEFB fiber
under triaxial CD test (Amad and Bateni, 2009)
109
Fig. 2.8 Unconfined compressive strength of optimum combination of Black Cotton
soil- lime with various percentage and aspect ratios of Coir Fiber (CF)
(Ramesh et al., 2010)
Fig. 2.9 Schematics of an example for green geosynthetics application
(Park et al., 2010)
110
Fig. 2.10 Short-term applications for geotextiles in embankments:
(a) primary embankment (b) splitting failure (c) circular failure
(d) basal failure (e) no failure (Anand, 2008)
Fig. 2.11 Variation in the ultimate tensile strength of the coir net after eld
application (Lekha, 2004)
111
Fig. 2.12 Drum arrangement for collection of runoff and erosion debris
(Lekha, 2004)
Fig. 2.13 Sketch of interrill erosion plot: (a) plan view, and (b) profile view
(Smets and Poesen, 2009)
112
Fig. 2.14 Soil loss erosion tested by using rainsplash simulator
(Jennifer, 2010)
Fig. 2.15 Comparison of soil losses and C- factors for six different RECP
(Jennifer, 2010)
113
Fig. 2.16 Layout of experimental plots at Urbana,Illinois, U.S.A.
(Bhattarai et.al, 2011)
Fig.2.17 Bare (left and compost cover (right) compartment after rainfall simulator
experiment with 16% slope condition (Bhattarai et.al, 2011)
114
Fig. 2.18 Design life envelope for an embankment reinforcement geotextile
(Sarsby, 2007)
Fig. 2.19 Comparison of time-settlement responses in the protected and non-
protected dykes in 1 year (Lekha and Kavitha, 2006)
115
Fig. 2.20 Photographs of woven coir geotextiles used for the study.
(a) MMA1, (b) MMA2, (c) MMB1, (d) MMV1 and (e) MMV2
(Subaida et al., 2008)
116
Fig. 2.21 Results of tension tests on geotextiles in warp direction
(Subaida et al., 2008)
Fig. 2.22 Direct shear test results of geotextiles in fine sand at 100 kPa
(Subaida et al., 2008)
117
(a) (b)
Fig. 2.23 Pull out resistance at 10 kPa (a) MMB1 (b) MMA2
(Subaida et al., 2008)
Fig. 2.24 Photographs of woven coir geotextiles (Subaida et al., 2009)
118
Fig. 2.25 Experimental set-up. (a) Loading arrangement.
(b) Arrangement of dial gauges. (Subaida et al., 2009)
119
Fig. 2.26 Plastic surface deformation of 267 mm thick base (U/D 0.3)
(Subaida et al., 2009)
Fig. 2.27 Photograph of juteHDPE union fabric (MD HDPE,
and CD jute) (Basu et al., 2009)
120
Fig. 2.28 Force-CBR probe displacement curve of juteHDPE union fabric
and 100% HDPE fabric. (Basu et al., 2009)
Fig. 2.29 Schematic diagram showing the effects of consolidation at different part of a
typical embankment erected on the soft soil (Mwasha, 2009b)
121
Fig. 2.30 Embankment constructed within a box (Mwasha and Pertersen, 2010)
Fig. 2.31 The condition of geotextile after each loading
(Mwasha and Pertersen, 2010)
122
Fig. 2.32 Graph of pull out against vertical load for aggregate samples
(Mwasha and Pertersen, 2010)
(a) (b)
Fig. 2.33 (a) Mould with square footing (b) Loading the mould in universal
testing machine (Nurul Islam Md. et al., 2010)
(a) (b)
Fig. 2.34 Comparison between geojutes and geotextiles (Nurul Islam Md. et al., 2010)
123
Fig.2.35 Surface and cross-section morphology (Methacanon et al., 2010)
Fig. 2.36 Moisture absorption of the natural fibers at 95% RH, 23 C
(Methacanon et al., 2010)
124
Fig. 2.37 TGA thermograms of the four studied natural fibers
(Methacanon et al., 2010)
0
50
100
150
200
250
sisal roselle reed water hy acinth
T
e
n
s
i
l
e
s
t
r
e
s
s
a
t
m
a
x
i
m
u
m
l
o
a
d
(
M
P
a
) dry
wet
(a)
0
5
10
15
20
25
30
sisal roselle reed water hy acinth
E
l
o
n
g
a
t
i
o
n
a
t
b
r
e
a
k
(
%
)
dry
wet
(b)
Fig. 2.38 Natural fibers in wet and dry states
(a)Tensile strength at break (b) Elongation at break
(Methacanon et al., 2010)
0
10
20
30
40
50
60
70
80
90
100
0 100 200 300 400 500 600 700 800
Temperature (C)
%
W
e
i
g
h
t
l
o
s
s
Reed
Waterhyacinth
Roselle
Sisal
125
Fig . 2.39 Accelerated condition for various time
(a)Tensile strength at break (b) Elongation at break
(Methacanon et al., 2010)
126
Fig. 2.40 Potential external failure mechanism of MSE structures
Fig. 2.41 Interactions between soil and reinforcement
127
Fig. 2.42 Components of the direct shear resistance of grid reinforcement
(Jewell et al., 1984)
Fig. 2.43 Friction and bearing resistances on reinforcement surfaces
(Bergado and Chai 1994)
a) Shear resistance between soil and reinforcement plane surface area
b) Soil to soil shear resistance
128
Fig. 2.44 Components of pullout resistance for geogrid reinforcement
(Bergado and Chai 1994)
Fig. 2.25
Fig. 2.45 Typical comparison of pullout displacement along Tensar geogrid (SR 80)
(Chai, 1992)
129
Fig. 2.46 Chart of value settlement ration (D Appolonia et al., 1971)
130
Fig. 2.47 Diagrams for the factor m0 andm1 used in the calculation of immediate
average settlement of uniformly flexible areas on homogeneous isotropic
saturated clay (Janbu et al., 1956)
Fig. 2.48 Correlation factor for pore pressures set up under a foundation
(Skempton and Bjerrum, 1957)
131
Fig. 2.49 C
146
Fig. 3.1 Natural fibers used for this study
Water hyacinth
147
(a)
(b)
(c)
Fig. 3.2 Water hyacinth patterns for tensile test (a) Knot-plain (b) Hexagonal (c) Plain
148
(a)
(b)
(c)
Fig. 3.3 Roselle or Thai Kenaf patterns for tensile test (a) Knot-plain (b) Hexagonal (c) Plain
149
Fig. 3.4 Tensile test machine
Fig. 3.5 Tensile strength test
150
0.00
2.00
4.00
6.00
8.00
10.00
12.00
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150
Strain (%)
T
e
n
s
i
l
e
(
k
N
/
m
)
Plain1 (Machine)
Plain2 (Machine)
Plain3 (Machine)
Plain4 (Cross)
Plain5 (Cross)
Hexagonal1 (Machine)
Hexagonal2 (Machine)
Hexagonal3 (Machine)
Hexagonal4 (Cross)
Hexagonal5 (Cross)
Knot-Plain1 (Machine)
Knot-Plain2 (Machine)
Knot-Plain3 (Machine)
Knot-Plain4 (Cross)
Knot-Plain5 (Cross)
Fig. 3.6 Comparison of tensile strength of water hyacinth LLGs for all patterns
0.00
5.00
10.00
15.00
20.00
25.00
30.00
0 20 40 60 80 100 120 140 160 180 200
Strain (%)
T
e
n
s
i
l
e
(
k
N
/
m
)
Plain1 (Machine)
Plain2 (Machine)
Plain3 (Machine)
Plain4 (Cross)
Plain5 (Cross)
Hexagonal1 (Machine)
Hexagonal2 (Machine)
Hexagonal3 (Machine)
Hexagonal4 (Cross)
Hexagonal5 (Cross)
Knot-Plain1 (Machine)
Knot-Plain2 (Machine)
Knot-Plain3 (Machine)
Knot-Plain4 (Cross)
Knot-Plain5 (Cross)
Fig. 3.7 Comparison of tensile strength of Kenaf LLGs for all pattern
151
0.00
5.00
10.00
15.00
20.00
25.00
30.00
0 5 10 15 20 25 30 35
Strain (%)
T
e
n
s
i
l
e
(
k
N
/
m
)
Kenaf1 (Machine)
Kenaf2 (Machine)
Kenaf3 (Machine)
Kenaf4 (Cross)
Kenaf5 (Cross)
Water Hyacinth1 (Machine)
Water Hyacinth2 (Machine)
Water Hyacinth3 (Machine)
Water Hyacinth4 (Cross)
Water Hyacinth5 (Cross)
Fig. 3.8 Comparison of tensile strength of water hyacinth and KenafLLGs for plain pattern
(a) (b)
Fig 3.9 Water hyacinth LLGs: (a) spacing 88mm (b) spacing 1212mm
152
0
1
2
3
4
5
6
7
8
0 5 10 15 20
Strain (%)
L
o
a
d
(
k
N
/
m
)
Sample 1
Sample 2
Sample 3
Fig. 3.10 Tensile strength of water hyacinth 8 mm opening size for plain pattern
0
1
2
3
4
5
6
0 5 10 15 20
Strain (%)
L
o
a
d
(
k
N
/
m
)
Sample 1
Sample 2
Sample 3
Fig. 3.11 Tensile strength of water hyacinth 12 mm opening size for plain pattern
153
0
1
2
3
4
5
6
7
8
0 5 10 15 20
Strain (%)
L
o
a
d
(
k
N
/
m
)
Sample 1 (8 mmopening size)
Sample 2 (8 mmopening size)
Sample 3 (8 mmopening size)
Sample 4 (12 mmopening size)
Sample 5 (12 mmopening size)
Sample 6 (12 mmopening size)
Fig. 3.12 Comparison of tensile strength of water hyacinth 8 mm and 12 mm for plain pattern
Fig. 3.13 Permittivity equipment
154
Fig. 3.14 Specimen with permittivity apparatus
Flow Rate
100
150
200
250
300
350
400
A B C D
F
l
o
w
R
a
t
e
(
l
/
m
2
-
s
)
Spacing 12 mmwith 100 mmhead
Spacing 12 mmwith 50 mmhead
Spacing 8 mmwith 100 mmhead
Spacing 8 mmwith 50 mmhead
Fig. 3.15 Comparison of flow rate
155
156
0.000000
0.000001
0.000002
0.000003
0.000004
0.000005
0.000006
0.000007
0.000008
0.000009
0.000010
0 0.2 0.4 0.6 0.8 1 1.2
Gradi ent
T
r
a
n
s
m
i
s
s
v
i
t
y
,
c
m
/
s
100 kPa
200 kPa
300 kPa
Fig. 3.18 Transmissvity of specimen opening size 8 mm
0.00000
0.00001
0.00002
0.00003
0.00004
0.00005
0.00006
0.00007
0.00008
0.00009
0.00010
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1
Gradi ent
T
r
a
n
s
m
i
s
s
v
i
t
y
,
c
m
/
s
100 kPa
200 kPa
300 kPa
Fig. 3.19 Transmissvity of specimen opening size 12 mm
157
0
10
20
30
40
50
60
70
80
90
100
0.01 0.1 1 10
Grain Size ( mm )
P
e
r
c
e
n
t
F
i
n
e
r
(
%
)
Fig. 3.20 Grain size distribution curve for backfill sand
1.750
1.760
1.770
1.780
1.790
1.800
1.810
1.820
1.830
1.840
1.850
3 4 5 6 7 8 9 10 11 12 13 14 15
Moisture content (%)
D
r
y
U
n
i
t
W
e
i
g
h
t
(
t
o
n
/
m
3
)
Max. Dry Unit Weight =1.810 ton/m
3
, OptimumMoisture Content =10.0 %
Fig. 3.21 Relationship between moisture content and maximum dry unit weight
158
Fig. 3.22 Kenaf LLGs with spacing 4x4 mm
Fig. 3.23 Dimensions of pullout test apparatus (Bergado et al., 1996)
159
Fig. 3.24 Schematic diagram of clamping system used for pullout test (Kongkitkul, 2001)
160
Fig. 3.26 LVDT and dial gage for measuring the horizontal displacement in front
161
Fig. 3.28 Pullout displacement of Kenaf LLGs at different normal pressures
Fig. 3.29 The relationships between maximum pullout resistances and normal
pressures during pullout tests
162
Fig. 3.31 Kenaf LLG was folded with sand backfill to shear
163
Fig. 3.32 The relationship between shear stress and horizontal displacement
from direct shear tests
Fig. 3.33 The relationship between shear stress and normal stress from direct shear tests
164
Fig. 4.1 Layout of test embankment location and locations of field vane test
and borehole
14 16 18 20
Unit weight: kN/m
3
10
9
8
7
6
5
4
3
2
1
0
D
e
p
t
h
:
m
2.6 2.65 2.7
G
s
0 40 80 120
PL, W
N
, LL
0 20 40 60
Corrected S
u
(kPa)
from vane shear test
0 50 100 150
P'
o
and P'
max
0 1 2 3 4 5 6
OCR
0 0.1 0.2 0.3 0.4
CR
0 0.04 0.08 0.12
RR
Weathered crust
Soft clay
Medium stiff clay Unit weight
G
s
PL W
N
LL
P'
o
P'
max
Weathered crust
14 16 18 20
Unit weight: kN/m
3
10
9
8
7
6
5
4
3
2
1
0
D
e
p
t
h
:
m
2.6 2.65 2.7
G
s
0 40 80 120
PL, W
N
, LL
0 20 40 60
Corrected S
u
(kPa)
from vane shear test
0 50 100 150
P'
o
and P'
max
0 1 2 3 4 5 6
OCR
0 0.1 0.2 0.3 0.4
CR
0 0.04 0.08 0.12
RR
Weathered crust
Soft clay
Medium stiff clay Unit weight
G
s
PL W
N
LL
P'
o
P'
max
Weathered crust Weathered crust
Fig. 4.2 Soil profile and soil properties at the site
165
Fig. 4.3 Vane shear test in the field
0.9
1
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
1.9
2
1.0 10.0 100.0 1000.0
Applied Pressure (kN/m
2
)
V
o
i
d
R
a
t
i
o
,
e
Fig. 4.4 Relationship between applied pressure and void ratio at 3m depth
P
max
= 73kPa, OCR=1.55
166
0 . 1 0
1 . 0 0
1 0 . 0 0
1 1 0 1 0 0 1 0 0 0
Applied Pressure (kN/m
2
)
C
o
e
f
f
i
c
i
e
n
o
f
C
o
n
s
o
l
i
d
a
t
i
o
n
,
C
v
(
m
2
/
y
e
a
r
)
Cv(t90)
Cv(t50)
Fig.4.5 Relation between applied pressure and coefficient of consolidation
at 3 m depth
0.9
1
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
1.9
2
1.0 10.0 100.0 1000.0
Applied Pressure (kN/m
2
)
V
o
i
d
R
a
t
i
o
,
e
Fig. 4.6 Relationship between applied pressure and void ratio at 6 m depth
P
max
=85kPa, OCR= 1
167
0.10
1.00
10.00
1 1 0 1 0 0 1 0 0 0
Applied Pressure (kN/m
2
)
C
o
e
f
f
i
c
i
e
n
o
f
C
o
n
s
o
l
i
d
a
t
i
o
n
,
C
v
(
m
2
/
y
e
a
r
)
Cv(t90)
Cv(t50)
Fig. 4.7 Relation between applied pressure and coefficient of consolidation
at 6 m depth
168
Fig. 4.8 Plan view of test embankment
169
Fig. 4.9 Section A-A view of test embankment
170
Fig. 4.10 Section B-B view of test embankment
171
Fig. 4.11 Details of settlement points (Voottipruex, 2000)
Fig. 4.12 Details of AIT type piezometer (Voottipruex, 2000)
172
Fig. 4.13 Slope stability analysis at front slope 1V:1H without Kenaf LLGs
Fig. 4.14 Slope stability analysis at front slope 1V:1H with Kenaf LLGs
173
Fig. 4.15 Slope stability analysis at front slope 1V:1H without Kenaf LLGs
when applying load
Fig. 4.16 Slope stability analysis at front slope 1V:1H with Kenaf LLGs when applying load
174
Fig. 4.17 First layer of Kenaf LLGs reinforcement with high strength wire
extensometers
Fig. 4.18 Completed embankment construction
175
Fig. 4.19 Comparison of observed settlements at the surface as well as at 3m
and 6m depths
Fig. 4.20 Comparison of observed and predicted surface settlements
176
Fig. 4.21 Comparison of observed and predicted subsurface settlements at 3 m depth
Fig. 4.22 Comparison of observed and predicted subsurface settlements at 6 m depth
177
Fig. 4.23 Observed and predicted average excess pore pressures at 3 m depth
Fig. 4.24 Observed and predicted average excess pore pressures at 6 m depth
178
K
e
n
a
f
L
L
G
s
D
e
f
o
r
m
a
t
i
o
n
(
m
m
)
Kenaf LLGs Length (m)
K
e
n
a
f
L
L
G
s
D
e
f
o
r
m
a
t
i
o
n
(
m
m
)
Kenaf LLGs Length (m)
Fig. 4.25 Deformations of observed in the Kenaf LLGs reinforcements
179
Fig. 4.26 Observed stress in the Kenaf LLGs reinforcements
180
Fig. 5.1 The simulation model in PLAXIS 2D
181
Fig. 5.2 The stage of construction for finite element simulation
182
Fig. 5.3 Flow chart on the overall process involved in running PLAXIS software
183
Fig. 5.4 Comparison of observed and predicted surface settlement
Fig. 5.5 Comparison of observed and predicted subsurface settlement at at 3 m depth
184
Fig. 5.6 Comparison of observed and predicted subsurface settlement at at 6 m depth
Fig. 5.7 Observed and FEM 2D simulations of surface and subsurface settlements
at 3 m and 6 m depths
185
Fig. 5.8 The vertical deformations from PLAXIS 2D simulation
186
Fig. 5.9 Observed and FEM 2D simulations of average excess pore pressure at 3 m depth
Fig. 5.10 Observed and FEM 2D simulations of average excess pore pressure at 6 m depth
187
Fig. 5.11 The excess pore water pressures from PLAXIS 2D simulation
188
Fig. 5.12 Deformation of observed and FEM 2D in the Kenaf LLGs reinforcement
189
Fig. 5.13 Stresses of observed and FEM 2D in the Kenaf LLGs reinforcement
190
Fig. 5.14 Comparison of the test embankment in the stability analyses using numerical
finite element to produce stability by investigating the factor of safety by FEM
2D
Fig. 5.15 The zones of failure in stability analyses using FEM 2D without Kenaf LLGs
reinforced
Fig. 5.16 The zones of failure in stability analyses using FEM 2D with Kenaf LLGs
reinforced
FS = 1.32
FS = 1.44
191
Fig. 6.1 The simulation model in PLAXIS 3D
192
Fig. 6.2 Comparison of observed and FEM 3D at surface settlement
Fig. 6.3 Comparison of observed and FEM 3D at subsurface settlement (3m depth)
193
Fig. 6.4 Comparison of observed and FEM 3D at subsurface settlement (6m depth)
Fig. 6.5 Observed and FEM 3D at surface settlement, subsurface settlement at 3 m and 6 m
depths
194
Fig. 6.6 The vertical deformations from PLAXIS 3D simulation
195
Fig. 6.7 Observed, predicted and finite element surface settlements
Fig. 6.8 Observed, predicted and finite element subsurface settlements at 3 m depth
196
Fig. 6.9 Observed, predicted and finite element subsurface settlements at 6 m depth
Fig. 6.10 Comparison of observed and FEM 3D in excess pore water pressure at 3m depth
197
Fig. 6.11 Comparison of observed and FEM 3D in excess pore water pressure at 6m depth
198
Fig. 6.12 The excess pore water pressures from PLAXIS 3D simulation
199
Fig. 6.13 Observed, FEM 2D and FEM 3D method average excess pore pressure
at 3 m depth
Fig. 6.14 Observed, FEM 2D and FEM 3D method average excess pore pressure
at 6 m depth
200
Fig. 6.15 Observed and predicted average excess pore pressure at 3 m depth
Fig. 6.16 Observed and predicted average excess pore pressure at 6 m depth
201
Kenaf LLGs Length (m)
K
e
n
a
f
L
L
G
s
D
e
f
o
r
m
a
t
i
o
n
(
m
m
)
Kenaf LLGs Length (m)
K
e
n
a
f
L
L
G
s
D
e
f
o
r
m
a
t
i
o
n
(
m
m
)
Fig. 6.17 Deformation of observed and FEM 3D in the Kenaf LLGs reinforcement
202
Fig. 6.18 Deformations from observed and FEM simulations in the Kenaf LLGs
reinforcement
203
Fig. 6.19 Stresses of observed and FEM 3D in the Kenaf LLGs reinforcement
204
Fig. 6.20 Tensile stress from observed and FEM simulations in the Kenaf LLGs
reinforcement
205
Fig. 6.21 Comparison of observed and previous results in FEM 2D at surface settlement
Fig. 6.22 Comparison of observed and previous results in FEM 2D subsurface settlement
at 3m depth
206
Fig. 6.23 Comparison of observed and previous results in FEM 2D subsurface settlement
at 6m depth
Fig. 6.24 Comparison of observed and previous results in FEM 3D at surface settlement
207
Fig. 6.25 Comparison of observed and previous results in FEM 3D subsurface settlement
at 3m depth
Fig. 6.26 Comparison of observed and previous results in FEM 3D subsurface settlement
at 6m depth
208
Fig. 6.27 Comparison of observed and previous results in FEM 2D excess pore water
pressure at 3m depth
Fig. 6.28 Comparison of observed and previous results in FEM 2D excess pore water
pressure at 6m depth
209
Fig. 6.29 Comparison of observed and previous results in FEM 3D excess pore water
pressure at 3m depth
Fig. 6.30 Comparison of observed and previous results in FEM 3D excess pore water
pressure at 6m depth
210
1.00
1.50
2.00
2.50
3.00
3.50
0 100 200 300 400 500 600
3D with Kenaf LLGs Reinforcement
3D without Kenaf LLGs Reinforcement
Time (days)
F
a
c
t
o
r
o
f
S
a
f
e
t
y
Fig. 6.31 Comparison of the values of factor of safety using FEM 3D analyses with and
without Kenaf LLGs reinforcement
1.00
1.50
2.00
2.50
3.00
3.50
0 100 200 300 400 500 600
3D with Kenaf LLGs Reinforcement
3D without Kenaf LLGs Reinforcement
2D with Kenaf LLGs Reinforcement
2D without Kenaf LLGs Reinforcement
Time (days)
F
a
c
t
o
r
o
f
S
a
f
e
t
y
Fig. 6.32 Comparison of the values of factor of safety of the test embankment stability
analyses using FEM2D and FEM3D with and without Kenaf LLGs reinforcement
211
Fig. 6.33 The zones of failure in stability analyses using FEM 3D without
Kenaf LLGs reinforced
Fig. 6.34 The zones of failure in stability analyses using FEM 3D with
Kenaf LLGs reinforced
FS = 1.64
FS = 1.36
212
Fig. 7.1The construction of embankment with slope of 1V:1.5H
Fig. 7.2 Woven water hyacinth LLGs
213
Fig. 7.3 Plan view of test embankment
214
Fig. 7.4 Section view of test embankment
215
Fig. 7.5 Ruzi grass
Fig. 7.6 Spread the seed of Ruzi grass on the soil
216
Fig. 7.7 Ruzi grass grow up through sample 8 mm opening size
Fig. 7.8 Ruzi grass grow up through sample 12 mm opening size
217
Fig. 7.9 Experimental set up at the test embankment
Fig. 7.10 Set up artificial rainfall
218
Fig. 7.11 Water motor pump
Fig. 7.12 Rain gauge
219
Fig. 7.13 Runoff samples from the test
Fig. 7.14 Runoff samples from plots of test
220
Fig. 7.15 The embankment at east side
0.00
2.00
4.00
6.00
8.00
10.00
12.00
14.00
16.00
18.00
20.00
Coated water
hyacinth LLGs +
Ruzi grass
Non-coated water
hyacinth LLGs +
Ruzi grass
Ruzi Grass Coated water
hyacinth LLGs
Non-coated water
hyacinth LLGs
R
u
n
o
f
f
R
a
t
e
(
l
/
m
-
s
)
4 weeks
6 weeks
8 weeks
10 weeks
Fig. 7.16 Runoff rate at east side of embankment (12x12 mm opening size)
221
Fig. 7.17 The embankment at west side
0.00
2.00
4.00
6.00
8.00
10.00
12.00
14.00
16.00
18.00
20.00
Coated water
hyacinth LLGs +
Ruzi grass
Non-coated water
hyacinth LLGs +
Ruzi grass
Bare Soil Coated water
hyacinth LLGs
Non-coated water
hyacinth LLGs
R
u
n
o
f
f
R
a
t
e
(
l
/
m
-
s
)
4 weeks
6 weeks
8 weeks
10 weeks
Fig. 7.18 Runoff rate at west side of embankment (8x8 mm opening size)
222
0.00
0.50
1.00
1.50
2.00
2.50
Coated water
hyacinth LLGs +
Ruzi grass
Non-coated
water hyacinth
LLGs +Ruzi
grass
Ruzi Grass Coated water
hyacinth LLGs
Non-coated
water hyacinth
LLGs
S
o
i
l
L
o
s
s
,
k
g
/
m
2
4 weeks
6 weeks
8 weeks
10 weeks
Fig. 7.19 Soil loss at east side of embankment (12x12 mm opening size)
0.00
0.50
1.00
1.50
2.00
2.50
Coated water
hyacinth LLGs +
Ruzi grass
Non-coated
water hyacinth
LLGs +Ruzi
grass
Bare Soil Coated water
hyacinth LLGs
Non-coated
water hyacinth
LLGs
S
o
i
l
L
o
s
s
,
k
g
/
m
2
4 weeks
6 weeks
8 weeks
10 weeks
Fig. 7.20 Soil loss at west side of embankment (8x8 mm opening size)
223
Fig. 7.21 Polyurethane coated water hyacinth LLGs
224
APPENDIX
PUBLICATIONS
Journals
1. Artidteang, S., Bergado, D.T., Tanchaisawat, T., and Saowapakpiboon, J. (2012).
Investigation of tensile and soil-geotextile interface strength of Kenaf woven Limited
Life Geotextiles (LLGs). Lowland Technology International Journal, Vol. 14, No. 2,
pp. 18.
2. Artidteang, S., Bergado, D.T., Chaiyaput S., and Tanchaisawat, T. (2013).
Performance of Kenaf LLGs reinforced embankment on soft clay, Ground
Improvement (accepted).
3. Artidteang, S., Bergado, D.T., Chaiyaput S., and Tanchaisawat, T. (2013). FEM 2D
and 3D Numerical Simulation of Kenaf Limited Life Geosynthetics (LLGs)
Reinforced Test Embankment on Soft Clay, Geotextile and Geomembranes (review
comments received, encourage to revise and resubmit).
Conferences
1. Artidteang, S., Bergado, D.T., Methacaonon, P., and Prahsarn, C. (2011). Properties of
limited life woven geotextiles (LLGs) for soil reinforcement and soil erosion control
applications, Proc. 14
th
Asian Regional Conference on Soil Mechanics and Geotechnical
Engineering, Hong Kong, China.
2. Tanchaisawat, T., Bergado, D.T., Artidteang, S. (2012). Interaction between Kenaf
geogrid (LLGs) and compacted sand by numerical simulations, Intl. Symposium on
Sustainable Geosynthetics and Green Technology for Climate Change, Bangkok,
Thailand.
3. Artidteang, S., Bergado, D.T., Tanchaisawat, T., and Chaiyaput, S. (2012). Kenaf
woven limited life geotextiles (LLGs) reinforcement interaction by pullout and direct
shear tests, Proc. 8th Intl. Symp. on Lowland Technology, Bali, Indonesia.
4. Artidteang, S., Bergado, D.T., Chaiyaput, S., and Saowapakpiboon, J. (2012).
Investigation of water hyacinth woven limited life geotextiles (LLGs) for soil erosion
control applications, Proc. 5th Asian Regional Conference on Geosynthetics, Bangkok,
Thailand.
5. Chaiyaput, S., Bergado, D.T., and Artidteang, S. (2012). FEM 2D numerical
simulations reinforced embankment on soft ground by limited life Geosynthetics
(LLGs), Proc. 5th Asian Regional Conference on Geosynthetics, Bangkok, Thailand.
6. Artidteang, S., Bergado, D.T., and Chaiyaput, S. (2013). Stability analyses of
embankment with limited life woven geosynthetics (LLGs) reinforced on soft clay, Proc.
18th Southeast Asian Geotechnical & Inaugural AGSSEA Conference, Singapore.