ARTICLE IN PRESS

Geotextiles and Geomembranes 24 (2006) 52–57

www.elsevier.com/locate/geotexmem

Technical Note

Numerical procedures for deformation calculations in the

reinforced soil walls
O. Al Hattamleha,, B. Muhunthanb
a
Civil Engineering Department, The Hashemite University, P.O. Box 150459, Zarqa 13115, Jordan
b
Civil and Environmental Engineering Department, Washington State University, Pullman, WA 99164, USA
Received 28 January 2005; received in revised form 10 July 2005; accepted 16 July 2005
Available online 19 October 2005

Abstract

This study presents a membrane analogy method to evaluate the deflection of fabric-reinforced earth walls. The resulting equations
were solved using a finite difference scheme to obtain the deflection. The numerical results were compared with a full-scale study. The
comparisons show good performance of the model.
r 2005 Elsevier Ltd. All rights reserved.

Keywords: Geosynthetic; Soil wall; Membrane analogy; Numerical analysis; Case study

1. Background
Lower cost, lightweight, improved durability, high
frictional characteristics, and the relative ease associated
with handling and transportation have contributed to the
rapid growth of polymeric fabrics in reinforced earth wall
technology over the last few decades. However, because of
their lower moduli much greater strains are induced in
fabric-reinforced earth walls. Therefore, deformations need
to be considered to ensure that the deflections at the face of
the wall are tolerable.
Field measurements of tensile force distribution have
shown it to be nonlinear with the maximum occurring at a
distance away from the face of the wall (Floss and Thamm,
1979). Several researchers have accounted for the non-
linear tensile stress distribution using a membrane analogy
(Love et al., 1987; Bourdeau, 1989; Espinoza, 1994; Shukla
and Chandra, 1994, 1995; Yin, 1997a, b). The resulting
governing equations have been solved using finite differ-
ence as well as finite element numerical schemes. These
methods have focused on the applications on the sub-grade
deflection.
This study used finite difference scheme of a membrane
analogy to calculate the short-range deflections in the
fabric and in the face of the reinforced earth wall. The
results are compared with a full-scale test results reported
by Public Works Research Institute in Japan (PWRI), 1997
and those predicted by Rowe and Skinner (2001) using
finite element method.
2. Deformation of the fabric based on membrane analogy
The flexibility of the fabric along with the conditions of
the load and support conditions leads to the development
of a membrane type behavior. This had prompted a
number of researchers to apply the membrane analogy to
study the deflection of highway subgrades reinforced with
fabric (Love et al., 1987; Bourdeau, 1989, Espinoza, 1994;
Shukla and Chandra, 1994, 1995; Yin, 1997a, b). The
membrane analogy is applied here to the case of wall
deflection.
Fig. 1 shows a two-dimensional plane strain of the static
equilibrium of an elastic membrane. Using the stochastic
stress diffusion theory (Sergeev, 1969; Harr, 1977) for two-
dimensional plane strain conditions, the expected vertical
stress s̄v ðx; zÞ at a point (defined by the coordinate x and z)
given as (Bourdeau, 1989)

Corresponding author. Tel.: +962(5)3903333x5004. qs̄z q2 s̄z

¼D 2 , (1)
E-mail address: hattamleh@gmail.com (O. Al Hattamleh). qz qx

0266-1144/$ - see front matter r 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.geotexmem.2005.07.001
 ARTICLE IN PRESS
O. Al Hattamleh, B. Muhunthan / Geotextiles and Geomembranes 24 (2006) 52–57 53

where D is the coefficient of diffusion, which governs the where s̄z;2 is the vertical stress at the fabric–lower soil layer
rate at which the upper soil layer spreads the applied interface, oðxÞ the membrane deflection, and ks the
surface load. It can be assumed to be related to, K0, the coefficient of subgrade reaction.
coefficient of earth pressure at rest and the depth as The interface frictional stress at the sand fabric interface
(Bourdeau, 1989) is given by the Mohr Coulomb as:
D ¼ K 0 z. (2) tðxÞ ¼ mðs̄z;1 ðxÞ þ gH 1 Þ. (6)
Using Gaussian distribution (Harr, 1977), the expected where s̄z ðxÞ is the vertical stress at the backfill underneath
vertical stress under an applied pressure, p, uniformly fabric interface, g the unit weight of the backfill upper
distributed over a strip of width 2a, can be evaluated as layer, H1 the thickness of the soil column above the fabric,
    and m the interface friction coefficient.
s̄z xþa x
a The force acting on the deflected membrane is as shown
¼ c pffiffiffiffiffiffi
c pffiffiffiffiffiffi , (3)
p z K0 z K0 in Fig. 2. The equilibrium of forces in the horizontal
where c is the cumulative Gaussian distribution function: direction results in (Fig. 2):
Z x Z x
cðxÞ ¼
2
e
t =2 dt. (4) T H ðxÞ þ tH ðxÞ dx ¼ T 0 , (7)
0
0

The compressible soil is assumed to offer a reaction to where tH ðxÞ is the horizontal component of the friction
the loading pressure proportional to its deflection as in stress at the interface, T H ðxÞ the horizontal component of
Winkler model as the tensile force in the membrane, and T0 the horizontal
tensile force, at the origin of coordinate (i.e., section A–A
s̄z;2 ¼ ks oðxÞ, (5) in Fig. 1).
Equilibrium in vertical direction is written as
Z x Z x
T V ðxÞ
ðs̄z;1
s̄z;2 Þ dx þ tV ðxÞ dx ¼ T 0 , (8)
0 0

where tV is the vertical component of the frictional stress.

Taking the derivative of Eq. (8) and using Eq. (5) results
in
dT V
þ tV ðxÞ þ ks oðxÞ ¼ s̄z;1 . (9)
dx
From the geometry of forces and deflection (Fig. 2):
T V do
¼ . (10)
T H dx
Rewriting Eq. (10) and taking the implicit derivative
with respect to x:

Fig. 1. Two-dimensional plane strain model of static equilibrium of an dT V dðT H doÞ dT H do d2 o

¼ ¼ þ TH 2 . (11)
elastic membrane (modified after Bourdeau, 1989). dx dx dx dx dx

Fig. 2. Forces acting on the deflected membrane (redraw after Bourdeau, 1989).
 ARTICLE IN PRESS
54 O. Al Hattamleh, B. Muhunthan / Geotextiles and Geomembranes 24 (2006) 52–57

Substitution of Eq. (11) into Eq. (9) results in Given

µ, L, P, ν, ks, γ, H
dT H do d2 o do
þ T H 2 þ tH þ ks o ¼ s̄z;1 . (12)
dx dx dx dx
At each nodal point,
Finally, substitution of the derivative of Eq. (7) into get σz(x) from Eq. 3
Eq. (12) results in
d2 oðxÞ At each nodal point,
T H ðxÞ þ ks oðxÞ ¼ s̄z;1 ðxÞ. (13) evaluate τ(x) from Eq. (6)
dx2
The total deformation condition of the reinforcement
fabric can be given as (Bourdeau 1989): Assign minute value to
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
 ffi sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
 ffi Th(x) identical to all node
Z L Z
doðxÞ 2 1 L doðxÞ 2
dL ¼ 1þ dx
L ¼ T H ðxÞ 1 þ dx,
0 dx E 0 dx
Form Matrix [B]
(14)
and solve Eq. (17)
where E is the elongation modulus of the fabric of total
length 2L.

Update
The boundary conditions are NO
Is Eq. (14) satisfied
do
¼ 0 for x ¼ 0 ðx ¼ 0 located in section A2A ðFig: 1ÞÞ, yes
dx
(15a) Get the value of
Th(x) from Eq. (7)
T H ¼ 0 for x ¼ L. (15b)
The membrane equation (Eq. (13)) can be solved using a
finite difference schemes. Assuming that the reinforcement Is the difference of ω (0)
is discretized in intervals of length Dx, Eq. (13) becomes NO
from step n and n-1 < the
tolerance (1.0 x 10-6)
T H;i oi
1 þ 2ðDx2 ks
T H;i Þoi þ T H;i oiþ1 ¼ 2Dx2 s̄z;1 ,
(16)
yes
where the subscript i denotes the node number ranging
from 1 to N. Compute the T(x)
The above equation can be written in a matrix form as
½B  ðoÞ ¼ ðs̄Þ (17)
End
where [B] is a tridiagonal matrix, and ðoÞ and ðs̄Þ are the
nodal vectors of displacement and tension, respectively.
This finite difference scheme can be solved subject to the Fig. 3. Flow chart to solve the iterative finite difference scheme.
constraints imposed by Eqs. (7) and (12) and the boundary
conditions (Eqs. (15a) and (15b)). The flow chart in Fig. 3
depicts the iterative procedure to solve the linear system of layers of geosynthetic grid reinforcement. Eleven are 6.0 m
equation presented in Eq. (17). It is worthy to note that the long and five are short, 1.0 m long, used to improve the
minimum number of node was 40 along the geosynthetic local stability of the facing blocks. The backfill used in the
grid. Refine beyond this discretization gave the same result. construction was sand with about 30% of silty clay as fine
materials. The detailed description of the foundation soil
3. Case study layers can be found in the Rowe and Skinner (2001).

The developed finite difference scheme for the membrane
analogy model is examined using the results of a full-scale
test study reported by the Public Works and Research
Institute (PWRI) in Japan. It involves a test wall
constructed in the Kanto region of Japan by PWRI
(Ochiai and Fukuda, 1996; Nakajima et al., 1996; PWRI,
1997; Tsukada et al., 1998; Miyata, 1996).
Fig. 4 shows the cross-section details of the wall,
reinforced fabric, and the foundation soil layers. The wall
is 8 m in high, concrete facing block and reinforced with 16
4. Model parameters
The model parameters required are deduced from both
measured data reported by PWRI (1997) and those used by
Rowe and Skinner (2001). The backfill material friction
angle is taken to be 291, which will affect both the subgrade
reaction and lateral earth pressure moduli. The subgrade
reaction modulus is given by Scott (1981) as
k30 ¼ 18N cor , (18)
 ARTICLE IN PRESS
O. Al Hattamleh, B. Muhunthan / Geotextiles and Geomembranes 24 (2006) 52–57
Fig. 4. Cross-section of the wall (redraw after Rowe and Skinner, 2001).
where k30 is the 0.3 m  0.3 m square plate subgrade
reaction and Ncor is the corrected standard penetrations
resistance. The corrected standard penetrations resistance
Ncor is correlated to soil friction angle as (Hatanaka and
Uchida, 1996):
pffiffiffiffiffiffiffiffiffiffiffiffiffiffi
f ¼ 20N cor þ 20. (19)
The subgrade reaction k30 for backfill soil based on Eq.
(18) will be 73.0 MN/m3. Das (1999) gave the following
correction factor for the subgrade reaction of length L and
width B:
 
kBB 1 þ 0:5 BL
k¼ . (20)
1:5
Eq. (20) indicates that the value of k of a very long
foundation with a width of B is approximately 0.67 kB  B,
which will be the case here, since we calculated the
deformation along the length L and strip width of 1.0 m.
Therefore, k will be 48.6 MN/m3. The reported young
55
modulus for the used geosynthetic grid materials and the
allowable tension strength are 980.0 and 29.4 kN/m,
respectively.
5. Comparison of observed and calculated strain in
reinforcement
The measurement of strain on the reinforcement was
taken in five levels above the base. The strain measured
showed an increase trend toward the wall face. Similar
trends were reported by several investigators, Floss and
Thamm (1979), Bathurst and Simac (1994). All measured
strains were below 1.5%, therefore, they are below 3.0%
calculated directly from the reported allowable design
strength and young modulus. The observed and calculated
results are shown in Fig. 5. The calculated strains show
good agreement especially along the bottom three layers.
The predictions appear to be slightly on the higher side
for the top two layers. The observed and calculated
 ARTICLE IN PRESS
56 O. Al Hattamleh, B. Muhunthan / Geotextiles and Geomembranes 24 (2006) 52–57

1.0 8
0.8 Height=7.65 m
7
0.6
0.4 Observed (PWRI,1997) 6

Wall Elevation (m)

0.2 Finite Difference Model 5
Model
4 Observed (PWRI, 1997)
0.8 Height=5.65 m Rowe and Skinner (2001)

0.6 3

0.4 2
0.2
1

0.8 Height=3.65 m 0
0 20 40 60 80 100
Strain (%)

0.6 Deformation (mm)

0.4
Fig. 6. Comparison of observed and calculated deformation in the wall.
0.2

0.8 Height=1.65 m Acknowledgments

0.6
0.4
0.2
0.8 Height=0.65 m
0.6
0.4
0.2
0.0
0 1 2 3 4 5
Distance from the Face of the Wall (m)
Fig. 5. Comparison of observed and calculated reinforcement strain.
deformation in the face of the wall and those predicted by
Rowe and Skinner (2001) using finite element are shown in
Fig. 6. The results show a reasonable agreement in the
lower part of the wall. However, the predicted deforma-
tions in the upper part are slightly a way from the reported
values. This might be due to the fact that the membrane
affect is not functioning for the short reinforcement and
thus the reported finite element by Rowe and Skinner
(2001) show superiority in this regard.
6. Conclusions
The deflection in the face of the reinforced soil and in the
geosynthetic fabric was predicted using a finite difference
scheme adopt a membrane analogy. The calculated
deformations from this scheme were compared with full-
scale test data. The comparison was found to be reasonably
good.
The authors thank Professor P. L. Bourdeau at Purdue
University for sharing his literature on membrane analogy
on reinforced earth. The authors also thank Professor C. T.
Crowe for the Tridiagonal Matrix algorithm (TDMA) used
in the solution of the finite difference equation.
References
Bathurst, R.J., Simac, M.R., 1994. Geosynthetic reinforced segmental
retaining wall structures in North America. In: Fifth International
6 7 Conference on Geotextiles, Geomembranes and Related Products,
Singapore, 5–9 September 1994. Southeast Asia Chapter of the
International Geotextiles Society, pp. 1275–1298.
Bourdeau, P.L., 1989. Modelling of membrane action in a two-layer
reinforced soil system. Computers and Geotechnics 7 (1–2), 19–36.
Das, B.M., 1999. Principle of Foundation Engineering. PWS Publishing,
Pacific Grove, CA.
Espinoza, R.D., 1994. Soil–geotextile interaction: evaluation of membrane
support. Geotextiles and Geomembranes 13 (5), 281–293.
Floss, R., Thamm, B.R., 1979. Field measurement of a reinforced earth
retaining wall under static and dynamic loading. In: International
Conference Soil Renforcement vol. III, Paris, pp. 183–188.
Harr, M.E., 1977. Mechanics of Particulate Media: A Probabilistic
Approach. McGraw-Hill, New York.
Hatanaka, M., Uchida, A., 1996. Empirical correlation between penetra-
tion resistance and internal friction angle of sandy soil. Soils and
Foundations 36 (4), 1–10.
Love, J.P., Burd, H.J., Milligan, G.W.E., Houlsby, G.T., 1987. Analytical
and model studies of reinforcement of a layer of granular fill on a soft
clay subgrade. Canadian Geotechnical Journal 24 (4), 611–622.
Miyata, K., 1996. Wall reinforced with fiber reinforced plastic geogrids in
Japan. Geosynthetics International 3 (1), 1–11.
Nakajima, T., Toriumi, N., Shintani, H., Miyataka, H., Dobahi, K., 1996.
Field performance of a geotextile reinforced soil wall with concrete
facing blocks. In: Ochiai, H., Yasufuku, N., Omine, K. (Eds.), Earth
Reinforcement. Balkema, Rotterdam/Brook, pp. 427–432.
Ochiai, Y., Fukuda, N., 1996. Experimental study on geotextile-reinforced
soil walls with different facing. In: de Groot, M.B., de Hoedt, G.,
Termaat, R.J. (Eds.), Symposium on Geosynthetics: Application,
Design and Construction. Balkema, Rotterdam/Brook, pp. 113–120.
 ARTICLE IN PRESS
O. Al Hattamleh, B. Muhunthan / Geotextiles and Geomembranes 24 (2006) 52–57
Public Works Research Institute (PWRI), 1997. Observation of behaviour
of full scale model test on reinforced retaining wall using geotextile.
Report #3487, PWRI, Japan.
Rowe, R.K., Skinner, G.D., 2001. Numerical analysis of geosynthetic
reinforced retaining wall constructed on a layered soil foundation.
Geotextiles and Geomembranes 19, 387–412.
Scott, R.F., 1981. Foundation Analysis. Prentice-Hall, Englewood Cliffs, NJ.
Sergeev, I.J., 1969. The application of probability—process equations to
the theory of stress distribution in non-cohesive soil foundation beds.
Soil Mechanics and Foundation Engineering 2, 84–88.
Shukla, S.K., Chandra, S., 1994. A generalized mechanical model for
geosynthetic-reinforced foundation soil. Geotextiles and Geomem-
branes 13 (12), 813–825.
57
Shukla, S.K., Chandra, S., 1995. Modelling of geosynthetic-reinforced
engineered granular fill on soft soil. Geosynthetics International 2 (3),
603–618.
Tsukada, Y., Ochiai, Y., Miyataka, H., Tajiri, 1998. Field performance of
a geosynthetic-reinforced soil wall with rigid facing. In: Sixth
International Conference on Geosynthetics, Atlanta, USA,
25–29 March 1998. Industrial Fabrics Association International,
pp. 577–580.
Yin, J.H., 1997a. A nonlinear model of geosynthetic reinforced
granular fill over soft soil. Geosynthetics International 4 (5),
523–537.
Yin, J.H., 1997b. Modeling of geosynthetic reinforced granular fill over
soft soil. Geosynthetics International 4 (2), 165–185.