You are on page 1of 12

Geotextiles and Geomembranes xxx (xxxx) xxx

Contents lists available at ScienceDirect

Geotextiles and Geomembranes


journal homepage: www.elsevier.com/locate/geotexmem

Experimental study of the performance of geosynthetics-reinforced soil


walls under differential settlements
Liang Lu a, b, c, *, Shuwen Ma b, e, Zongjian Wang d, Yi Zhang d
a
Key Laboratory of New Technology for Construction of Cities in Mountain Area, Ministry of Education, School of Civil Engineering, Chongqing University, Chongqing,
400045, China
b
School of Civil Engineering, Chongqing University, Chongqing, 400045, China
c
Key Laboratory for Special Area High Way Engineering of the Ministry of Education, Chang’an University, Xian, 710064, China
d
Institute of Geotechnical Engineering, Chongqing Jiaotong University, Chongqing, 400074, China
e
National Joint Engineering Research Center for Prevention and Control of Environmental for Prevention and Control of Environmental Geological Hazards in the TGR
Area, 400045, China

A R T I C L E I N F O A B S T R A C T

Keywords: The deformation performance and settlement failure mechanism of geosynthetics-reinforced soil (GRS) walls are
Geosynthetics the two key points of engineering design under the differential settlement. This paper presents model tests of
GRS wall deformation performance and failure mechanism of the GRS wall with and without lateral restriction under
Differential settlement
differential settlement conditions. The observation and measurement results, including force and vertical
Deformation performance
displacement of geosynthetics and lateral deformation of facing panels, indicate good settlement control per­
Failure mechanism
formance of GRS wall during construction and under differential settlement. Results indicate that the influence of
the stress state of facing panels on the settlement control performance of GRS wall cannot be ignored. And the
differential settlement failure of GRS wall is likely to occur in the joint of facing panels and geosynthetics. For
good illustrations, two analytical approaches about deformation and stress of geosynthetics were proposed based
on elastic cable theory, in GRS wall with and without lateral restriction. The expressions exclude the necessity to
carry out sophisticated numerical analyses to stress and deformation and may help to develop the design
guidelines for such GRS wall.

were established to simulate the stress and deformation condition of the


1. Introduction multilayer geosynthetics, it found that the constitutive models were in
good agreement with the measured values (Zou et al., 2016; Yu et al.,
As a composite structure, the GRS wall is increasingly used in a wide 2016a; Ambauen et al., 2016). Numerical simulations were conducted to
range of industries (e.g., highway, bridge, sub-grade and slope) for their analysis the stress and deformation trend of the GRS wall (Skinner and
high strength and stiffness to weight ratio (Han et al., 2018; Tatsuoka Rowe, 2005; Yu et al., 2016b; Han and Leshchinsky, 2010; Huang et al.,
et al., 2009; Shin et al., 2011; Koerner and Soong, 2001). It has brought 2013). Results showed that the GRS wall can reduce the differential
much economic benefits as well as the better engineering performance settlement effect at the top of the wall. The importance of GRS wall has
in the past years in reducing the differential settlement such as “bridge been widely recognized in the application of differential settlement
bump” (Garga and Oshaughnessy, 2000; Portelinha, 2012; Boler et al., control. However, differential settlement can become excessive for
2019), which often occurs in rainfall and earthquake prone areas. As a specific reasons such as excessive height of wall, lack of stability of wall,
result, related research has emerged endlessly. Model tests of GRS walls insufficient strength of connection between facing panel and geo­
conducted in the past have shown that the maximum values of tensile synthetics, insufficient compaction nearby facing panel, consolidation of
forces of the geosynthetics occurred in the stress concentration zone at foundation soil, poor drainage, and compression of soft ground (Wang
the end of construction, besides, different connections between the et al., 2009; Berg et al., 2009; Kim et al., 2010; Leshchinsky and Ling,
geosynthetics and facing panels leaded to various bearing capacity of 2010; Sadat et al., 2018). According to Chinese records (Wei et al.,
GRS walls (Bourgeois et al., 2011; Xiao et al., 2016). Constitutive models 2004), a GRS wall of the Three Gorges project with a total height of 57 m

* Corresponding author. Key Laboratory of New Technology for Construction of Cities in Mountain Area, Ministry of Education, School of Civil Engineering,
Chongqing University, China.
E-mail addresses: luliangsky@163.com, luliangsky@cqu.edu.cn (L. Lu).

https://doi.org/10.1016/j.geotexmem.2020.09.007
Received 11 May 2020; Received in revised form 7 August 2020; Accepted 20 September 2020
0266-1144/© 2020 Published by Elsevier Ltd.

Please cite this article as: Liang Lu, Geotextiles and Geomembranes, https://doi.org/10.1016/j.geotexmem.2020.09.007
L. Lu et al. Geotextiles and Geomembranes xxx (xxxx) xxx

Notation σ stress tensor


Ean horizontal earth pressure of the nth-layer geosynthetics
h differential settlement Kn earth pressure coefficient
l horizontal length of deformation section Wn normal pressure of the nth-layer geosynthetics
E Young modulus Sx Sy length and width of the facing panels
A cross-sectional area of the geosynthetics γ specific gravity
q equivalent uniform load lp theoretical horizontal length prediction of geosynthetics in
h2 vertical distance between two adjacent geosynthetics deformation section
h1 h2/2 lt horizontal length of geosynthetics in deformation section
G0 overburden weight of the first-layer geosynthetics obtained by laboratory test
Gn weight of nth layer reinforced soil at h2 height (n ≥ 1) lm horizontal length of geosynthetics in deformation section
FN support force from lower filler obtained by numerical model
θ vertical inclination angle of the geosynthetics μt μm errors

collapsed after one year of operation. It was found that the settlement on connection of facing panel and geosynthetics (Berg et al., 2009; Bernardi
the top of the facing panel is 70 mm, while the settlement on the top of et al., 2009). The analysis of the local failure of GRS wall under differ­
the soil reinforcement near the facing panel is 350 mm, the difference ential settlements, especially the stress and deformation behaviour of
between them is 280 mm. We monitored a GRS wall from Iwate/Miyagi the geosynthetics near the facing panels, are rarely studied at present.
Inland Earthquake for nearly 3 years and found out that the maximum On the basis of the above-mentioned works, this paper aims to
settlement difference between the reinforced area and the facing panel explore the deformation performance and possible failure mechanism of
was 300 mm. These excessive differential settlements caused the broken the GRS wall under the differential settlement, and to investigate the
of geosynthetics near the panel, separation between the panel and soil influence of the force state of facing panels. To achieve these objectives,
reinforcement, and large-scale instability of the facing panel. Therefore, GRS wall models with and without lateral restriction (for facing panels)
it is urgent to study the performance and failure mechanism of GRS walls were constructed and tested in the laboratory under the differential
under differential settlements. settlement. For understanding the deformation performance and failure
Under differential settlements, the research of GRS wall was mainly model of GRS wall subjected to differential settlement more intuitively,
focused on the interaction mechanism between the geosynthetics and the finite element model was established by the finite element software-
filler (Hu et al., 2014), the failure mechanisms and stability analysis Midas/GTS. We chose the same number and value of parameters as the
(Song et al., 2017), the mechanical and permeability properties of laboratory model tests so that the complexity of the model and the
geosynthetics (Assarar et al., 2011; Vahedifard et al., 2014; Raisinghani comparative study would be reasonable. In order to analyze the exper­
and Viswanadham, 2010) and the settlement observation (Cui and Tan, imental phenomena quantitatively, two analytical approaches based on
2015). Sadat et al. (2018) found the phenomenon of "stress elastic cable theory for calculating deformation and stress of geo­
concentration-fracture-separation" is easy to occur at the connection of synthetics were proposed, in GRS wall with and without lateral re­
the facing panel and the geosynthetics under the condition of differential striction. The deformation expression proposed can be used to calculate
settlement. In addition, Wang et al. (2018) found that the phenomenon and predict the flexural deformation and deformation range of the
of "stress concentration-fracture-separation" would lead to the failure geosynthetics, which can provide reference for early warning of engi­
between the panel and geosynthetics. Similarly, Yoshida et al. (2003) neering settlement. The stress expression enables to calculate the stress
also discovered this type of destruction of the GRS wall under differ­ of geosynthetics under differential settlements and predict settlement
ential settlements, as shown in Fig. 1. This is not exactly consistent with stability of GRS wall, especially applicable for practical engineering
the conventional engineering design where maximum stress of geo­ analysis of walls where settlement is observed and for practical engi­
synthetics always appeared near the rupture surface of sliding body. neering design for very tall walls.
Meanwhile, for GRS wall, it appears that no standard guidelines are
available for deformation and tensile strength-based design near the

Fig. 1. Deformation and failure of GRS wall: (a) 3D (modified from Yoshida et al., 2003), (b) 2D.

2
L. Lu et al. Geotextiles and Geomembranes xxx (xxxx) xxx

2. Laboratory model test distribution curve of the sand is shown in Fig. 3.


The sand was used as filler for GRS wall and was prepared based on
This section focuses on the deformation characteristics and failure dense-graded gradation with nominal maximum particle size of 4 mm.
mechanism of GRS wall under the differential settlement. A small-scale 4 The void content and the angularity of sand aggregate are 35% and 2
layer reinforced-earth retaining wall was built in a test chamber. And respectively. The coefficient of curvature (Cc) and the uniformity coef­
then the wall was forced to move vertically downward relative to the ficient (Cu) were 1.44 and 16 respectively with D60, D30 and D10 of 4,
facing panels through applying difference settlement artificially. The 1.2 and 0.25 mm, respectively. In the model test, the sand needed to be
details of the model test are as follows: controlled in a target relative density of 50% (Wang et al., 2015). Ac­
cording to consolidated drained triaxial tests (ASTM, 2011b), the
2.1. Test device cohesion and the peak friction angle of the sand at 50% relative density
were found to be 0 kPa and 42◦ , respectively.
The GRS wall was constructed and tested in a fabricated test chamber
that was composed of frame and Plexiglas, as shown in Fig. 2(a). The 2.3. Facing panel
Plexiglas permitted a transparent view of the structure deformation
during the test as indicated in the figure. The interior side of the fabri­ Individual model facing panels (prepared by cutting wooden plates)
cated test chamber was greased to minimize the friction. In the model with compressive strength of 130 MPa were selected as the facing panels
test, the bottom of the GRS wall model was surrounded by an imper­
meable membrane, which could be used to reduce sand penetration from
the sidewall of model groove. The vertical red-dyed sand on the surface
of Plexiglas was distributed regularly in the two-thirds area near the
facing panel, with a horizontal spacing of 10 mm. The horizontal red-
dyed sand was laid along with the layered laying of backfill, with a
vertical spacing of 10 mm. This way has a tracking function of geo­
synthetics deformation. The completed GRS wall and the dimensions
were 350 mm wide, 720 mm deep and 75 mm long as shown in Fig. 2(b).

2.2. Backfill soil

The clean sand was selected to simulate reduced-size and strength


filler for lab-scale model tests. The sieve analysis tests of the clean sand
Fig. 3. Grain size distribution curve for the sand.
were conducted following ASTM D2487-11(2011a) and the grain size

Fig. 2. Laboratory GRS wall test. (unit: mm).

3
L. Lu et al. Geotextiles and Geomembranes xxx (xxxx) xxx

of GRS wall, which were 10 mm thick, 150 mm high and 350 mm long,
as shown in Fig. 4(b). The common block sizes of GRS wall has a
dimension of 200 mm (thickness), 150 mm (height), and 600 mm
(length), whose ratio is approximately 1:5 compared with the field block
size. Two geosynthetics were fixed on the height direction of each facing
panel. The geosynthetics were spaced at 75 mm vertically, starting from
37.5 mm at the top of facing panel based on the consideration of uniform
stress range (Fig. 4(b)).

2.4. Geosynthetics

In the model test, nylon yarn netting with low strength was selected
as the geosynthetics, as shown in Fig. 4(a). The parameters of nylon yarn
Fig. 5. Tensile force vs. strain of nylon yarn netting.
netting are aperture (1.0 mm), mesh number (18), wire diameter (0.14
mm) and weight (120 g/m2), which are interlocked with backfill soil at
the aperture. Tensile test of the nylon yarn netting was carried out that the mechanical connection is more favourable to play the tensile
following ASTM D6637 (2001). The test results shown that the ultimate properties of the geosynthetics. In this study, mechanical connection
strength of nylon yarn netting is 30 kN/m when the strain reached 5%. mode was selected and modelled as shown in Figs. 4(b) and Fig. 6. The
And the tensile force of nylon yarn netting was 16.7 kN/m when the mechanical connection was created by driving many fine steel nails into
strain reached 2%, which good consistent with similitude rule (Wang the aperture of the geosynthetics and fixing it with strong glue. Where,
et al., 2015). In addition, Fig. 5 shows the tensile force-strain relation­ the edge of geosynthetics was wrapped with a distance of 100 mm to
ship of the nylon yarn netting, which was used to determine the elastic prevent the geosynthetics sliding from the fixed position of steel nails. In
modulus (E) as a function of the tensile stress. Based on the similarity this condition, the geosynthetics would not be pulled out from the facing
ratio (1:5) and the typical tensile strengths range of geosynthetics from panels.
58 to 210 kN/m (Tensar International, 2012), the tensile strength of the
nylon yarn netting is within the range of 12–42 kN/m (Xiao et al., 2016). 2.6. Simulation of differential settlement
It can be obtained from Fig. 5 and Eq. (1) that the tensile strength of
nylon yarn netting is 33 kN/m, which satisfies the model scale effect In reality, the differential settlement of the facing panel is much
(ratio is approximately 1:5). smaller than that of the reinforcement soil near the panel (Soltanijigheh
F F et al., 2019), therefore the differential settlement between them occurs.
E= = (1) In order to simulate the differential settlement, the model test was
Aεa btεa
conducted by the method of limiting the vertical movement of the panel
Where F is the tension; b is the width of the nylon yarn netting, which is and allowing the vertical movement of the soil reinforcement. In the
50 mm; t is the thickness of the nylon yarn netting, which is 0.66 mm; εa model test, the displacement control linkage hydraulic jack with a
is the actual strain of the nylon yarn netting, which is given by tensile bearing capacity of 5000 kg and the maximum drop of 150 mm, is
testing machine. selected as the deformation device of GRS wall model. The differential
settlement deformation of wall can be realized by controlling the in­
2.5. Connection modes between facing panel and geosynthetics crease and decrease of the oil pressure artificially. The linkage hydraulic
jack, forcing the GRS wall model to move downward relatively to create
The importance of geosynthetics in the application of GRS walls has differential settlements, was positioned at the center of the beam plates.
been widely recognized. However, the effect of the exertion degree of It is necessary to set a balance plate with the thickness of 30 mm be­
tensile properties of geosynthetics on the settlement control perfor­ tween the linkage hydraulic jack and the GRS wall model for main­
mance of GRS walls has been rarely investigated. Both mechanical and taining the integrity of the model. Thus, the allowed maximum value of
frictional connections are the key connection modes of geosynthetics differential settlement is 120 mm (150 mm–30 mm), and the differential
and facing panels in GRS walls (Simac, 1990; Xiao et al., 2016). settlement value for stage measurement can be set as 30 mm, 60 mm, 90
Compared with frictional connection, the engineering practice proves mm, and 120 mm. The jacks were arranged as triangular to avoid

Fig. 4. Fixed nylon yarn netting into facing panel (mechanical connection) and installed strain gauges.

4
L. Lu et al. Geotextiles and Geomembranes xxx (xxxx) xxx

Fig. 6. Mechanical connection diagrams between panel and geosynthetics.

rotational movements or differential settlements in the longitudinal di­


Table 1
rection, as shown in Fig. 7. This procedure can simulate differential
Comparison scheme of GRS wall.
settlements of the GRS wall.
case I II

Panel constraint with lateral restriction without lateral restriction


2.7. Model design type
Legend
A set of comparative tests relative to GRS wall with and without
lateral restriction were installed to analyze the influence of lateral earth
pressure on the stability of differential settlement of GRS wall. The
lateral earth pressure can be offset by the lateral restriction of GRS wall
model, so as to achieve the purpose of comparison with the test whose
lateral earth pressure is not zero. The main reason that the lateral earth and the instrumentation readings were taken at 30, 60, 90 and 120 mm
pressure can be offset by lateral restriction is that the facing panel is not of differential settlements. The required time for each differential set­
allowed by the lateral restriction to tilt and flip to the outside of GRS tlement increment was determined based on the preparatory experi­
wall. The design details of the comparative test are shown in Table 1. ment. Each differential settlement after application was maintained for
The main construction steps of the small-scale laboratory test are as 60 min, during which the vertical displacements of the nylon yarn net­
follows (Fig. 8): tings and the lateral displacements of facing panels and were recorded
every 15 min. Then the mean value was taken as the measurement
1. Putting the displacement control linkage hydraulic jack at the bot­ result. Vertical displacements and lateral displacements occurred
tom of the fabricated test chamber. immediately after the differential settlement, and then slowly increase at
2 Placing the balance plate on the top of linkage hydraulic jack. a maximum speed of 15 min. After that, further increases of the vertical
3 . Laying the two facing panels connected with nylon yarn nettings in displacements and the lateral displacements could be ignored because of
a superposition pattern in the direction of height. The facing panel of little progress. Therefore, 15 min for each differential settlement
the GRS wall model consisted of 2 courses superposition of wooden increment were selected and used in all subsequent tests. The measured
plates, allowing the flexibility of lateral movements, except for the strain (εm) by the strain gauge may be different from the actual strain
comparative test (not allow lateral movements of facing panel). (εa) of the geosynthetics. Under the laboratory conditions, it is difficult
4 Filling sand and compacting to 50% of its relative density. for strain gauge readings to be exactly consistent with the actual strain.
5. Laying nylon yarn nettings and sand alternately according to Fig. 2 The protective technology of strain gauge has significant influence on
until the construction of GRS wall is completed. the measured data (Zuo et al., 2014). And it is very important to reduce
6. Once the GRS wall was completed, a set of strain values and photo­ the influence of strain gauge reading error caused by improper protec­
graphs were taken as the initial value of the measurement targets. tion on test results. Therefore, the data of the strain gauge (εm) was
corrected by summarizing the relationship between the actual strain (εa)
The allowed maximum differential settlement created was 120 mm, and the measured strain (εm) that obtained by the ruler and the strain
gauge. It can be seen from Fig. 9 that the relationship of εm and εa can be
represented by εa = 17.15εm.

2.8. Monitoring

The vertical displacements of the nylon yarn nettings of layers 1 to 4


were measured through the glass and an appropriate experimental de­
vice under the differential settlement. The stresses were measured by
attaching the strain gauges to the each nylon yarn netting with double-
sided pasted to eliminate the influence of stretching and bending
moment. The strain gauge were located at 20 mm, 40 mm, 80 mm, 160
Fig. 7. Installation of instrumentation: installed balance plate and linkage mm and 320 mm respectively from the facing panels, as shown in Fig. 4
hydraulic jack at the bottom. (b). The tensile forces and vertical displacements of nylon yarn nettings,

5
L. Lu et al. Geotextiles and Geomembranes xxx (xxxx) xxx

Fig. 8. Construction flow-process diagram of GRS wall model.

established to verify the reliability of laboratory model test results and to


observe the deformation performance of GRS wall model more intui­
tively. This detailed model is built based on the experimental configu­
ration, including the facing panels, geosynthetics, filler and contact
areas between them, are shown in Fig. 10. Moreover, this figure also
shows the boundary and loading conditions applied to the finite element
models. In the numerical simulation, the 8-node 6-hedral isoparametric
solid elements were used for filler and facing panels, 2-D geogrid planar
elements were used for geosynthetics. In the Midas/GTS, the filler was
simulated by the Mohr-Coulomb model (M-C model), the facing panels
and geosynthetics were simulated by linear elasticity model. In addition,
the slip/open behaviors between the facing panels and filler, the geo­
synthetics and filler were modelled using Goodman contact elements,
the connection between the facing panels and geosynthetics were
Fig. 9. The relationship between εa and εm. modelled using continuous elements. The material properties of the GRS
wall model used in this study are shown in Table 2 and Table 3
and the horizontal displacements of the facing panels were measured by respectively.
model tests respectively.
3.2. Deformation curves of geosynthetics under the differential settlement
3. Results and analysis
Test found that the deformation of GRS wall under differential set­
The experimental procedure for simulating the behavior of GRS wall tlement has certain regularity, as shown in Fig. 11. The deformation
model under the differential settlement conditions were detailed
description in Sec. 2. In order to understand the behavior of GRS wall
more intuitively, two finite element models of GRS wall with and
without lateral restriction are established in this section, and the model
parameters are determined according to the completed lab-scale GRS
wall test. The deformation characteristics and failure mode of GRS walls
are analyzed according the results of both laboratory model tests and
finite element models under the differential settlements. In addition,
two analytical approaches about deformation and stress of geosynthetics
are proposed and compared to the numerical analysis and model test
results. All the reported results in this paper are the increases due to
differential settlements, i.e., not including all parameters prior to dif­
ferential settlements (Assarar et al., 2011).

3.1. Finite element models based on the laboratory model

Midas/GTS (Geotechnical and Tunnel analysis System) software was


used in this study to achieve the 4-layer GRS wall in 3D model. Different
constitutive models were used for the filler, facing panels, geosynthetics Fig. 10. Finite element model of GRS wall based on the lab-scale model test.
and materials interfaces. Two finite element comparison models are (unit: mm).

6
L. Lu et al. Geotextiles and Geomembranes xxx (xxxx) xxx

Table 2
Filler and facing panel properties.
Parameter

Material E (kN/m2) γ (kN/m3) ν K (kN/m2) G (kN/m2) c (kPa) ϕ (◦ )

Filler 90000 26.1 0.19 48387.096 37815.13 18.7 34.7


Facing panel 130000 24 0.3

is h. For each layer of geosynthetics, the equivalent uniform load (q) is


Table 3
related to the upper soil weight and the supporting force of the lower
Nylon yarn netting properties.
structure, that is, the equivalent uniform load of each layer geo­
Parameter Et Tensile Thickness Maximum ν synthetics is different. The key to this section is to solve the equivalent
(kN/ strength (kN/ (mm) tensile load/N
uniform load (q1~qn) of each layer of geosynthetics. Therefore, the
m) m)
calculated horizontal length (l) of geosynthetics with flexural deforma­
Values 33 16.7 0.66 375 0.23
tion will vary under the different reinforcement layers and different
kN/m
settlement conditions. In the mean time, the end location of deformation
section changes with each calculation.
Fig. 12 shows the GRS wall description. Where G0 is the overburden
weight of the first layer geosynthetics at h1 height, G1, G2 until Gn-1 are
the weight of second, third until nth-layer soil at h2 height respectively, h
has the same meaning as Eq. (3).Where A is the connection point be­
tween the facing panel and the geosynthetics, B is the maximum set­
tlement point of settlement curve, in reality, however, the settlement of
geosynthetics from B to the end far away from the facing panel is the
same, so this section is not discussed any more.
It can be seen from Fig. 12 that the settlement curve includes two
parts: deformation section and non-deformation section. The geo­
synthetics of non-deformation section is not the focus of this paper
because only static friction, no sliding friction and obvious tensile
deformation between geosynthetics and backfill soil. In this paper, the
settlement curve of deformation section was selected as the object of
deformation analysis.
To simplify the calculation, the following hypothesis is proposed:

1. The facing panels are not subjected to force.


2. The variation of the cross-sectional area of the geosynthetics under
Fig. 11. Differential settlement of GRS wall model. external loads is considered negligible.
3. The friction between the filler and the facing panel is considered
trend of nylon yarn netting, including nonlinear deflection and separa­ negligible.
tion of reinforcement and soil, was consistent with that of geosynthetics 4. The friction between the GRS wall and the groove wall is considered
in Fig. 1. The deformation curves prediction of geosynthetics under the negligible.
differential settlement is the key to performance study of GRS wall. 5. The shear and flexural rigidity of geosynthetics are considered
Stress and deformation of single-layer geosynthetics of GRS wall negligible, and only uniform distributions of tensile stresses and
were presented using the elastic cable theory (Wang et al., 2018). The extensional strains are admitted in the geosynthetics cross-section.
results show that the elastic cable theory has good applicability in
behavioural analysis of GRS wall. Thus, the curve equation of geo­ Assuming that the specific gravity of the filler is γ, Gn (n = 0, 1, 2, 3)
synthetics of GRS wall can be written as: shown in Fig. 12 are defined in Eq. (4).
}
h 2 2h G0 = γh1
y= − x + x (2) (4)
l2 l G1 = G2 = G3 = γh2

Where h denotes the differential settlement, l denotes the length of the


part where the geosynthetics has deformation (i.e., deformation section,
as shown in Fig. 1), y and x denote vertical displacement and horizontal
position of geosynthetics respectively.
In Eq. (2), the known parameter is h (obtained by engineering design
or observation), the unknown parameter is l. l can be calculated by Eq.
(3).
√̅̅̅̅̅̅̅̅̅̅̅̅̅
3
4 4EAh
l= (3)
3q

Where E denotes Young modulus, A denotes cross-sectional area of the


geosynthetics, and q denotes the equivalent uniform load at the top of
each layer geosynthetics.
In Eq. (3), the independent variable is q and the controllable variable
Fig. 12. Schematic diagram of GRS wall.

7
L. Lu et al. Geotextiles and Geomembranes xxx (xxxx) xxx

The forces of first layer geosynthetics unit mainly includes: the ten­ Table 4
sion (T1), the friction (ω1), the overburden weight (G0), and the support Comparisons among theoretical predictions, laboratory tests and numerical
force from lower filler (FN1). Fig. 13 shows the forces description of the analyses.
first layer geosynthetics unit. Layer h/mm lp/mm lt/mm lm/mm μt/% μm/%
According to the mechanical equilibrium, the following relationships Layer1 30 240 194 297 23.7 19.2
are obtained. 60 369 327 417 12.8 11.4
90 471 443 501 6.32 5.9
FN1 = G0 sin θ1 = γh1 sin θ1 (5) 120 553 550 556 0.55 0.5
Layer2 30 194 163 231 19.0 15.9
q1 = uG0 (6) 60 303 274 335 10.6 9.5
90 391 370 413 5.7 5.4
Where θ is the dip angle of geosynthetics, q1 is the upper uniform load on 120 460 458 462 0.4 0.4
Layer3 30 191 162 225 17.9 15.2
the first layer of geosynthetics, u is the width of the geosynthetics.
60 301 272 333 10.7 9.6
The forces acting of second layer geosynthetics unit mainly includes: 90 386 366 407 5.5 5.2
the tension (T2), the friction (ω2), the weight between the first and 120 451 449 453 0.4 0.4
second layer geosynthetics (G1), and the support force from lower filler Layer4 30 178 162 196 9.9 9
(FN2), as shown in Fig. 13. 60 289 270 309 7.0 6.6
90 376 362 390 3.9 3.7
According to the mechanical equilibrium, the following relationships
120 442 441 443 0.2 0.2
are obtained.
a
p indicate theoretical prediction, t indicate laboratory test, m indicate finite
FN2 = FN1 cos(θ2 − θ1 ) + G1 sin θ2 (7) element model.
b
( ) μt= (lp-lt)/lt×100%.
q2 = u(FN1 sin θ1 + G1 ) = u G0 sin 2 θ1 + G1 (8) c
μm= (lm-lp)/lm×100%.

According to the mechanical balance, the equivalent uniform load


larger than that at the vertical distance of 90 mm–120 mm from the top
(qn) of nth-layer (n ≥ 2) geosynthetics is calculated by FNn-1 and G1.
of structure. Obviously, a good agreement is observed among y(x) pre­
Therefore, the following relationships are obtained.
dictions, test results and numerical analyses of geosynthetics with the
FN3 = FN2 cos(θ3 − θ2 ) + G1 sin θ3 (9) increase of stress level. The main reason for the slight difference is the
boundary effect and the redistribution of stress between geosynthetic
FNn = FN(n− 1) cos(θn − θn− 1 ) + G1 sin θn (10) and soil under the differential settlements are considered in laboratory
model test, but without consideration in theoretical prediction and nu­
q3 = u(FN2 sin θ2 + G1 ) (11) merical analysis. It is found that noticeable stress redistribution devel­
oped in the relatively low stress levels (i.e., two layers of geosynthetics
q4 = u(FN3 sin θ3 + G1 ) (12) near the top of the wall) during the initial stage of differential settle­
ment. With the increase of stress level, that is, the distance between
qn = u(FN(n− sin θn− 1 + G1 ) (13)
geosynthetics and the top of retaining wall, the effect of stress redistri­
1)

Eq. (13) shows the relationship of qn under the same differential bution gradually weakens. At the same time, the experimental results
settlement: q1< q2< ……< qn. In this section, l can be obtained by gradually approach the y(x) predictions and numerical simulation
substituting qn into Eq. (3), and l is inversely proportional to qn. y(x) can results.
be obtained by substituting Eq. (3) into Eq. (2). The deformation curves of all geosynthetics calculated by Eq. (2),
No obvious difference can be found among the deformation section which located in deformation section, were compared with those ob­
curvature of all geosynthetics in the small-scale laboratory test under the tained from both model test and numerical simulation due to differential
same settlement, thus, assuming that θn is fixed in theoretical defor­ settlements. Under different reinforced zone movements, the calculated
mation predictions. Table 4 summarizes the comparison of horizontal vertical displacement of each layer geosynthetics was in a good agree­
length among y(x) predictions (obtained by Eq. (2)), model test results ment with those obtained from both model test and numerical simula­
and numerical analyses for four layer geosynthetics under differential tion, as shown in Fig. 14. For geosynthetics under the differential
settlements. It can be clearly seen at the vertical distance of 90 mm–120 settlement, the deformation curve is nonlinear until the vertical
mm from the top of structure, both errors μt and μm decreased to 5% of displacement reached design targets (with the corresponding design
that at the upper structure with the high stress level. Errors μt and μm displacement of 30 mm, 60 mm, 90 mm and 120 mm respectively). The
(greater than 10% or even 20%) at the vertical distance of 30 mm–60 figures show that the settlement increases gradually until passing the
mm (i.e., in the low stress level) from the top of structure are much critical line located at deformation section, herein the concave curve can

Fig. 13. Schematic diagram of the forces of first layer reinforced material unit.

8
L. Lu et al. Geotextiles and Geomembranes xxx (xxxx) xxx

Fig. 14. Settlement deformation curves of four-layer geosynthetics after each settlement.

be approximated by Eq. (2). After that, it gradually keeps in vertical above is borne by the corresponding nylon yarn netting within the
design displacement unchanged in the direction of leaving facing panels. bearing range, the backfill below cannot fill the increased space. After
The expression proposed can be used to calculate and predict the flex­ comparison, the settlement control performance of GRS wall with lateral
ural deformation and deformation range of the geosynthetics, which can restriction is better than the GRS wall without lateral restriction. In
provide reference for early warning of engineering settlement. The short, the influence of facing panels on the settlement control perfor­
above analysis results are applicable to GRS wall, that is, the settlement mance of GRS wall cannot be ignored.
performance of geosynthetics can reflect that of GRS wall.
We found that the curvature of the flexural deformation increases
with the increase of the buried depth of geosynthetics under the same 3.4. Failure mechanism of GRS walls
differential settlement. The reason may be that the upper load of geo­
synthetics increases with the increase of buried depth, i.e., q1< q2< q3< Test found that there was no sign of damage when the maximum
q4. l gradually decreases with the increase of q, which can be verified by settlement (120 mm) of the settlement device was reached for the GRS
Eq. (3). Thus, this phenomenon can be explained clearly by trigono­ wall with and without lateral restriction, indicate good performance of
metric function (i.e., tan θ = h/l). GRS wall during construction and in-service. In order to observe the
differential settlement failure mode of GRS wall, the load was applied
uniformly on the top of model until the GRS wall failed. It is found that
3.3. Comparison of GRS wall with and without lateral restriction
the differential settlement instability mode of GRS wall is tensile failure
of the nylon yarn nettings, and the failure occurs at the joint of the facing
Fig. 15 shows the comparison of the test results of with and without
panel and the nylon yarn nettings, as shown in Fig. 16. Thus, the failure
lateral restriction of GRS wall at differential settlement. The test shows
modes of GRS wall with and without lateral restriction under differential
that it has no obvious deflection in facing panels with lateral restriction
settlements can be revealed by the stress tensor of geosynthetics.
of GRS wall under differential settlement (Fig. 14(a)), but has obvious
The stress tensor of geosynthetics (nylon yarn nettings) of GRS wall
deflection in facing panels without lateral restriction of GRS wall under
with lateral restriction can be obtained by following expression:
the same settlement (Fig. 14(b)). Similarly, two larger separations of
nylon yarn nettings and backfill soil occur in the GRS wall without 2Eh2 √̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅
̅
lateral restriction, but only a small separation of that occur in the one σ= 4h2 x2 − 8h2 lx + 4h2 l2 + l4 (14)
3l4
with lateral restriction. Those results can be explained that it has not any
effect on facing panels due to lateral restraint counteracts the earth Where σ is dependent variable, x is independent variable.
pressure acting on that panels. So that the space between the nylon yarn Similarly, the stress tensor of geosynthetics of GRS wall without
nettings and backfill soil below after settlement is almost the same as lateral restriction can be obtained by following expression:
before, and it is difficult to separate them due to the friction. However, ( )
2Eh2 Kn Wn Sx Sy √̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅
̅
the separation of the nylon yarn nettings and backfill soil below σ= 4
+ 2
4h2 x2 − 8h2 lx + 4h2 l2 + l4 (15)
3l Al
occurred in the fact that the deflection of facing panels of the GRS wall
without lateral restriction. Probably because the weight of backfill soil In which,

9
L. Lu et al. Geotextiles and Geomembranes xxx (xxxx) xxx

Fig. 15. Comparison of GRS walls with two restraint forms after differential settlement.

Kn Wn Sx Sy = Ean (16) length and width of the facing panels respectively.


Taking the GRS wall without lateral restriction as an example, the
Where Ean is the horizontal earth pressure behind the facing panels of stresses distribution of geosynthetics in deformation section calculated
the nth-layer geosynthetics, Kn is the earth pressure coefficient, Wn is from Eq. (15) was compared with those obtained from both model test
the normal pressure of the nth-layer geosynthetics, Sx and Sy are the and numerical simulation due to the differential settlements. Under

10
L. Lu et al. Geotextiles and Geomembranes xxx (xxxx) xxx

facing panels and geosynthetics, decreases gradually along the extension


direction of the geosynthetics, and finally reaches the minimum at the
critical line. This behavior indicates that the failure of GRS wall without
lateral restriction is likely to occur in the joint of facing panels and
geosynthetics, which is highly consistent with the differential settlement
failure of actual GRS wall (including model test). This conclusion is
basically consistent with the phenomenon of “stress concentration-
fracture-separation” in the joint of facing panels and geosynthetics,
which located at the same deformation section mentioned above. This
behavior is due to a stiffness difference between the geosynthetics and
facing panels, which should not be neglected in design of GRS wall. After
that, the stress of each layer geosynthetics gradually remains in the
minimum value along the distance increase from geosynthetics to facing
panels. A very high consistency and obvious regularity of the stress
prediction of geosynthetics can also be noted by comparing theoretical
calculation and FEM during the differential settlement (Fig. 17). How­
ever, the results of physical model experiments are quite different from
those of theoretical calculation and numerical simulation. These trends
of behavior could be due to the slip of the measuring device and the
Fig. 16. Comparison of GRS walls with tw. Notation restraint forms after dif­ boundary effect during the differential settlement.
ferential settlement.
4. Conclusions
different reinforced zone movements, the calculated stress of each layer
geosynthetics was basically consistent with those obtained from both In this study, laboratory model test and numerical simulation were
model test and numerical simulation, as shown in Fig. 17. The results presented to investigate the behavior of GRS wall with and without
show that the stress expression (Eq. (15)) enables to calculate the stress lateral restriction of facing panels under the differential settlement. The
of geosynthetics under differential settlements and predict differential major conclusions from this study can be summarized as follows:
settlement stability of GRS wall. For example, with uniform loading, if
the calculated stress to be borne by the geosynthetics according to the (1) First, the test found that the deformation of geosynthetics has the
allowable settlement control value larger than actual tensile strength of nonlinear characteristic of concave curve under the differential
that, the design parameters of the geosynthetics need to be adjusted to settlement, i.e., the deformation of geosynthetics in the defor­
meet the design requirements. It can be seen from the deformation mation zone is gradually decreases far away from the facing
section in Fig. 17 that the maximum stress is concentrated at the joint of panels. Then, a theoretical expression is proposed to calculate

Fig. 17. Stress distribution of geosynthetics after each settlement.

11
L. Lu et al. Geotextiles and Geomembranes xxx (xxxx) xxx

and predict the deformation and deformation range of the geo­ Cui, Z.D., Tan, J., 2015. Analysis of long-term settlements of Shanghai Subway Line 1
based on the in situ monitoring data. Nat. Hazards. (Dordr). 75 (1), 465–472.
synthetics, based on which the deformation of GRS wall can be
Garga, V.K., Oshaughnessy, V., 2000. Tire-reinforced earth fill. Part 1: construction of a
obtained. The theoretical deformation curves match the test re­ test fill, performance, and retaining wall design. Can. Geotech. J. 37 (1), 75–96.
sults reasonably well. Han, J., Leshchinsky, D., 2010. Analysis of back-to-back mechanically stabilized earth
(2) In order to predict the stress distribution of geosynthetics under walls. Geotext. Geomembranes 28 (3), 262–267.
Han, J., Jiang, Y., Xu, C., 2018. Recent advances in geosynthetics-reinforced retaining
the differential settlement, a stress expression was proposed in walls for highway applications. Front. Struct. Civ. Eng. 12 (2), 239–247.
this paper. The calculated maximum stress is concentrated at the Huang, J., Han, J., Parsons, R.L., Pierson, M., 2013. Refined numerical modeling of
joint of facing panels and geosynthetics, which is highly consis­ alaterally-loaded drilled shaft in an MSE wall. Geotext. Geomembranes 37, 61–73.
Hu, K., Kulkarni, D.D., Choi, I., Tsukruk, V.V., 2014. Graphene-polymer nanocomposites
tent with the actual failure of GRS wall under the differential for structural and functional applications. Prog. Polym. Sci. 39 (11), 1934–1972.
settlement caused by "stress concentration - fracture - separation" Kim, D., Bhowmik, S., Willmer, J., 2010. A Case History of MSE Wall Failure: Finite
of geosynthetics. This behavior is due to a stiffness difference Element Modeling and Evaluation. GSP 199, GeoFlorida 2010. American Society of
Civil Engineers, pp. 2232–2242.
between the geosynthetics and facing panels, which should not be Koerner, R.M., Soong, T.Y., 2001. Geosynthetics reinforced segmental retaining walls.
neglected in design of GRS wall. Geotext. Geomembranes 19 (6), 359–386.
(3) The test found that obvious deflections of facing panels and larger Leshchinsky, D., Ling, H.I., 2010. US and Australia-Collapse examples and lessons of
reinforced soil retaining walls, special feature: reliability of reinforced soil walls.
separations between geosynthetics and backfill soil are appeared October issue Mag. Basic Eng. 24–28 (In Japanese).
in the GRS wall without lateral restriction, which reflects the Portelinha, F.H.M., Bueno, B.S., Zornberg, J.G., 2012. Performance of geotextile
weak settlement control performance. For the lateral restriction reinforced earth wall in unsaturated poorly draining backfill earth conditions. Proc.
5th Eur. Geosynth. Congr. 455–465.
can counteracts the earth pressure, the settlement control per­
Raisinghani, D.V., Viswanadham, B.V.S., 2010. Evaluation of permeability
formance of GRS wall with lateral restriction is better than that characteristics of a geosynthetics-reinforced soil through laboratory tests. Geotext.
without one. Therefore, the effect of lateral restriction on the Geomembranes 28 (6), 579–588.
settlement control performance of GRS wall cannot be ignored. Sadat, M.R., Huang, J., Bin-Shafique, S., Rezaeimalek, S., 2018. Study of the behavior of
mechanically stabilized earth (MSE) walls subjected to differential settlements.
Geotext. Geomembranes 46 (1), 77–90.
The research results of this paper exclude the necessity to carry out Shin, E.C., Cho, S.D., Lee, K.W., 2011. Case study of reinforced earth wall failure during
sophisticated numerical calculations to stress and deformation and may extreme rainfall. Int. Symp. Backwards Probl. Geotech. Eng. Monit. Geo-Constr.
146–153.
help to develop the design of very tall GRS walls. Simac, M.R., 1990. Connections for geogrid systems. Geotext. Geomembranes 537–546.
Skinner, G.D., Rowe, R.K., 2005. Design and behaviour of a geosynthetics reinforced
retaining wall and bridge abutment on a yielding foundation. Geotext.
Declaration of competing interest Geomembranes 23 (3), 234–260.
Soltanijigheh, H., Bagheri, M., Amanighadim, A.R., 2019. Use of hydrophilic polymeric
stabilizer to improve strength and durability of fine-grained soils. Cold Reg. Sci.
None. Technol. 187–195.
Song, F., Liu, H., Chai, H., Chen, J., 2017. Stability analysis of geocell-reinforced
Acknowledgements retaining walls. Geosynth. Int. 24 (5), 442–450.
Tatsuoka, F., Hirakawa, D., Nojiri, M., Aizawa, H., Nishikiori, H., Soma, R.,
Tateyama, M., Watanabe, K., 2009. A new type of integral bridge comprising
This study was funded by the National Natural Science Foundation of geosynthetics-reinforced soil walls. Geosynth. Int. 16 (4), 301–326.
China (grant number 51679018); the Chongqing Research Program of Tensar International, 2012. Tensar Uniaxial Geogrids for Soil Reinforcement. Technical
Flyer.
Basic Research and Frontier Technology (grant number Vahedifard, F., Leshchinsky, B.A., Sehat, S., Leshchinsky, D., 2014. Impact of cohesion on
cstc2017jcyjA1410); the Graduate Scientific Research and Innovation seismic design of geosynthetics-reinforced earth structures. J. Geotech. Geoenviron.
foundation of Chongqing (grant number CYB19017); the special Fund 140 (6), 1–12.
Wang, L., Chen, G., Chen, S., 2015. Experimental study on seismic response of geogrid
for Basic Scientific Research of Central Colleges, Chang’an University reinforced rigid retaining walls with saturated backfill sand. Geotext.
(grant number 300102219514). Geomembranes 43 (1), 35–45.
Wang, Z.J., Ma, S.W., Tang, X.S., Wu, J.M., Zhi, X.P., Lu, L., 2018. Application of elastic
cable theory in design of reinforced earth structure, 01 Chin. J. Geotech. Eng. 40,
References 122–129 (In Chinese).
Wang, Z., Tatta, N., Hattori, H., Tsuji, S., Ohta, H., 2009. Deformation of a reinforced
Ambauen, S., Leshchinsky, B., Xie, Y., Rayamajhi, D., 2016. Service-state behavior of earth double structure wall due to an earthquake and its repairworks.
reinforced earth walls supporting spread footings: a parametric study using finite- Jioshinsetikkusu Rombunshu (Geosynth. Eng. J.) 125–130.
element analysis. Geosynth. Int. 23 (3), 156–170. Wei, L.m., Zhang, J.L., He, Q., 2004. Effect of differential settlement on the tension force
Assarar, M., Scida, D., ElMahi, A., Poilane, C., Ayad, R., 2011. Influence of water ageing and deformation behavior of reinforcements for reinforced earth retaining wall, 05
on mechanical properties and damage events of two reinforced composite materials: China Railw. Sci. 91–95 (In Chinese).
flax-fibres and glass-fibres. Mater. Des. 32 (2), 788–795. Xiao, C., Han, J., Zhang, Z., 2016. Experimental study on performance of geosynthetics-
ASTM, 2011a. Standard Practice for Classification of Soils for Engineering Purposes reinforced soil model walls on rigid foundations subjected to static footing loading.
(Unified Soil Classification System). D2487–11, West Conshohocken, PA. Geotext. Geomembranes 44 (1), 81–94.
ASTM, 2011b. Consolidated Drained Triaxial Compression Test for Soils, D7181–11. Yoshida, K., Kubota, K., Yokota, Y., Tatta, N., Arai, K., 2003. Measurement of reinforced
West Conshohocken, PA. soil retaining wall of the double wall structure. Geosynth. Eng. J. 18, 125–130 (In
ASTM D 6637, 2001. Standard Test Method for Determining Tensile Properties of Japanese).
Geogrids by the Single or Multi-Rib Tensile Method. ASTM International, West Yu, Y., Bathurst, R.J., Allen, T.M., 2016b. Numerical modeling of the SR-18 geogrid
Conshohocken, Pennsylvania, USA. reinforced modular block retaining walls. J. Geoteh. Geoenviron. 142 (5),
Berg, R.R., Christopher, B.R., Samtani, N.C., 2009. Design and Construction of 04016003.
Mechanically Stabilized Earth Walls and Reinforced Soil Slopes. Federal Highway Yu, Y., Bathurst, R.J., Allen, T.M., Nelson, R., 2016a. Physical and numerical modelling
Administration. Publication No. FHWA-NHI-10-024. of a geogrid-reinforced incremental concrete panel retaining wall. Can. Geotech. J.
Bernardi, M., Collin, J.G., Leschinsky, D., 2009. Design Manual for Segmental Retaining 53 (12), 1883–1901.
Walls, third ed. National Concrete Masonry Association, Herndon, VA, p. 281. Zou, C., Wang, Y., Lin, J., Chen, Y., 2016. Creep behaviors and constitutive model for
Boler, H., Mishra, D., Tutumluer, E., Chrismer, S., Hyslip, J.P., 2019. Stone blowing as a high density polyethylene geogrid and its application to reinforced soil retaining
remedial measure to mitigate differential movement problems at railroad bridge wall on soft soil foundation. Construct. Build. Mater. 114, 763–771.
approaches. Proc. Inst. Mech. Eng. - Part F J. Rail Rapid Transit 233 (1), 63–72. Zuo, J., Meng, L., Han, K., Cui, M., 2014. Influence of the strain gauge protection
Bourgeois, E., Soyez, L., Le Kouby, A., 2011. Experimental and numerical study of the technology of composite structures on the measurement data, 000(004). Struct.
behavior of a reinforced earth wall subjected to a local load. Comput. Geotech. 38 Strength Res. 29–33 (In Chinese).
(4), 515–525.

12

You might also like