Geotextiles and Geomembranes 50 (2022) 994–1003

Contents lists available at ScienceDirect

Geotextiles and Geomembranes

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

Experimental study on uplift behavior of shallow anchor plates in

geogrid-reinforced soil
Yu-Xin Gao a, Hong-Hu Zhu a, b, *, Yu-Fei Ni a, Chao Wei a, Bin Shi a
a
School of Earth Sciences and Engineering, Nanjing University, Nanjing, Jiangsu, 210023, China
b
Nanjing University High-Tech Institute at Suzhou, Suzhou, 215123, China

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

Keywords: Geogrid reinforcement can significantly improve the uplift bearing capacity of anchor plates. However, the
Anchor plate failure mechanism of anchor plates in reinforced soil and the contribution of geogrids need further investigation.
Geogrids This paper presents an experimental study on the anchor uplift behavior in geogrid-reinforced soil using particle
Failure modes
image velocimetry (PIV) and the high-resolution optical frequency domain reflectometry (OFDR). A series of
Particle image velocimetry (PIV)
Fiber optic sensing
model tests were performed to identify the relationship between the failure mechanism and various factors, such
Interface behavior as anchor embedment ratio, number of geogrid layers, and their location. The test results indicate that soil
deformation and the uplift resistance of anchor plates are substantially influenced by anchor embedment ratio
and location of geogrids, whereas the number of geogrid layers has limited influence. In reinforced soil,
increasing the embedment ratio greatly improves the ultimate bearing capacities of anchor plates and affects the
interlock between the soil and geogrids. As the embedment depth increases, the failure surfaces gradually change
from a vertical slip surface to a bulb-shaped surface that is limited within the soil. The strain monitoring data
shows that the deformations of geogrids are symmetrical, and the peak strains of geogrids can characterize the
reinforcing effects.

1. Introduction budget, difficulties in backfill compaction, and constraints in site spaces

in many cases (Yunkul and Gurbuz, 2020). In this situation, enhancing
Anchor plates are frequently used in the construction of foundations the strength of in-situ soil is a realistic way to improve the ultimate uplift
to provide great uplift resistance for offshore structures and man-made resistance.
slopes (Ilamparuthi et al., 2002; Das and Shukla, 2013). Different ap­ Geosynthetics have been widely adopted for soil reinforcement ap­
proaches have been adopted to analyze the pullout response of anchor plications due to their reinforcement advantages and cost-effectiveness
plates embedded in soil and the associated failure patterns, such as (Vahedifard et al., 2016; Rahimi et al., 2018). They are manufactured
theoretical analyses (Ganesh and Sahoo, 2017; Perazzelli and Ana­ in planar forms, such as geotextiles, geogrids, geomembranes, and
gnostou, 2017; Sharma et al., 2021), laboratory model tests (Murray and three-dimensional forms, such as geocells. The results of previous
Geddes, 1987; El Sawwaf and Nazir, 2006; Sakai and Tanaka, 2007; studies demonstrate the superior performance of geogrids and geo­
Riyad et al., 2020), field tests (Tian et al., 2013; Zhu et al., 2014), and textiles in reinforcing in-situ soils. When geogrids are used in pave­
numerical simulations (Merifield et al., 2006; Choudhary et al., 2018; ments, they not only provide transverse restraint and membrane support
Perez et al., 2018; Kanitz et al., 2019; Evans and Zhang, 2019). There are for base and subbase materials but also play a vital role in reducing
various factors affecting the uplift capacity of anchors, such as the permanent deformation and increasing the bearing capacity (Baadiga
anchorage geometry, soil type and density, and embedment depth. et al., 2021). Geogrids can also reduce slope deformation in the hori­
Increasing the anchorage dimensions or embedment depth can effec­ zontal direction mainly by friction, so as to improve the bearing capacity
tively improve the uplift capacity of anchor plates (Dickin and Leung, of slopes (Bhattacherjee and Viswanadham, 2019). Understanding the
1983; Sakai and Tanaka, 1998), but it is limited by the construction interaction between soil and reinforcement and improving the

* Corresponding author. School of Earth Sciences and Engineering, Nanjing University, 362 Zhugongshan Bldg, 163 Xianlin Ave, Qixia District, Nanjing, Jiangsu,
210023, China.
E-mail addresses: [email protected] (Y.-X. Gao), [email protected] (H.-H. Zhu), [email protected] (Y.-F. Ni), [email protected] (C. Wei),
[email protected] (B. Shi).

https://doi.org/10.1016/j.geotexmem.2022.06.006
Received 26 January 2022; Received in revised form 12 June 2022; Accepted 27 June 2022
Available online 9 July 2022
0266-1144/© 2022 Elsevier Ltd. All rights reserved.
Y.-X. Gao et al. Geotextiles and Geomembranes 50 (2022) 994–1003

Parashar (1994) demonstrated the enhancement in the uplift perfor­

mance of small-scale anchor plates with geosynthetics. Based on model
tests, Ilamparuthi et al. (2008) analyzed the influence of cyclic loading
on the reinforcement performance of anchor plates in geogrid-reinforced
soil. Badakhshan et al. (2018) proposed granular-geosynthetic anchors
to improve the soil strength. Choudhary et al. (2019b) investigated the
behavior of groups of anchor plates in reinforced soil mass through a
series of laboratory model tests and three-dimensional numerical anal­
ysis. These studies demonstrate the beneficial use of reinforcement in
improving the performance of horizontal and vertical anchor plates.
Furthermore, the pullout behavior of inclined anchors embedded in
unreinforced and geogrid-reinforced soils was compared in previous
studies (Mukherjee et al., 2021). Nevertheless, these studies neither
considered the strain variations of geosynthetics in the uplift process nor
obtained theoretical solutions for the uplift bearing capacity of anchor
plates in reinforced soil.
The main objective of the present study is to investigate the strain
distributions of geogrids and the propagation of failure surfaces during
anchor plate uplift tests. The particle image velocimetry (PIV) and fiber
Fig. 1. Grain size distribution of the test soil. optic sensing technologies were used for strain and displacement mea­
surements in the tests. The effects of some factors on the uplift perfor­
composite performance is very important for optimizing engineering mance have been analyzed in detail, such as the embedment depth and
reinforcement schemes (Liu et al., 2009; Rajesh and Viswanadham, layouts of geogrids. In addition, the theoretical solutions to the bearing
2012). capacity of anchor plates in unreinforced and reinforced soil are derived
The beneficial effects of geosynthetics largely depend on their in this paper.
structural type (Latha and Murthy, 2007; Latha and Somwanshi, 2009).
Uniaxial and biaxial geogrids have been globally used to strengthen 2. Methodology
soils. The uniaxial geogrids are only suitable for situations where acci­
dental failure is likely to occur in a predictable direction (Qian et al., 2.1. Materials
2013). As for biaxial geogrids, they can be subjected to tension in the
longitudinal and transverse directions (Al-Naddaf et al., 2019; Zhang The soil used in this experimental investigation was dry and angular
et al., 2021). During the uplift process of anchor plates, the quartz sand. The particle size distribution of the soil is shown in Fig. 1.
soil-embedded geogrid is not in the one-dimensional stress state. Hence, The coefficient of uniformity (Cu) was 1.8, and the coefficient of cur­
biaxial geogrids possessing equal tensile strength in both longitudinal vature (Cc) was 0.91. The soil was classified as poorly graded sand (SP)
and transverse directions were used in this paper. according to the Unified Soil Classification System (ASTM D2487-17a).
In recent years, extensive studies have focused on combining geo­ Based on the results of direct shear tests (ASTM D6528-17b), the friction
synthetics and anchorage technologies(Ghosh and Bera, 2010; Boush­ angle (φ) of the sand at 59% relative density (Dr) was 35◦ .
ehrian et al., 2011; Niroumand et al., 2013; Bhowmik et al., 2019; The biaxial geogrids selected in the experiment are made of basalt
Choudhary et al., 2019a; Yunkul et al., 2021). Krishnaswamy and fiber, which has the advantages of high strength, acid and alkali

Fig. 2. Experimental setup: (a) Photograph of anchor plate uplifting in unreinforced soil; (b) Instrumentation of anchor plate uplift tests in reinforced soil; (b) Layout
of geogrids; (c) Photograph and schematic diagram of the geogrid fixing device (units in × 10− 3 m).

995
Y.-X. Gao et al. Geotextiles and Geomembranes 50 (2022) 994–1003

consistent with that of the model box, but the length was slightly larger
to reserve the space for fixing the geogrids. An additional control group
was set up without geogrids. To obtain the strain distributions of geo­
grids during testing, fiber optic cables were laid in an S-shape and
adhered to the upper surface of the geogrid cross-rib with epoxy resin, as
shown in Fig. 2(c). The fiber optic cables allow distributed strain sensing
over a long distance, working like human nerves for feeling strain dis­
tribution along the fiber with markedly improved sensing point and
measurement frequency (Ye et al., 2022;Zhu et al., 2022; Li et al., 2022).
By attaching the fiber optic cables to geogrids, they can sense the loading
condition of geogrid-reinforced soils in real-time (Wang et al., 2009).
The embedment ratio of an anchor plate is defined as H/D, where H
= anchor embedment depth and D = anchor diameter. Referring to
previous studies (Rahimi et al., 2018; Liu et al., 2019), four embedment
ratios (i.e., 1.5, 2, 2.5, 3) were set to analyze the failure mechanisms of
anchor plates. It is necessary to analyze the influence of the bedding
layer thickness to evaluate the scale effect (Sakai and Tanaka, 2007). A
0.09-m thick sand bed was placed at the bottom of the model box in this
paper, eliminating the boundary effect at the bottom. Merifield et al.
(2006) proposed that when the distance between the sidewall of the
model box and the anchor plate is greater than the embedment depth, it
Fig. 3. Geogrid laying instructions diagram. is sufficient to get the accurate bearing capacity of the anchor plate.
Except for the sidewall in direct contact with the anchor plate, other
sidewalls in these tests are 0.215 m away from the edge of the anchor
resistance, and durability. Many scholars have studied the geogrid size
plate, which meets the requirements of the uplift tests of the anchor
in coarse-grained sand or gravel soil, and it is found that the ratio of
plate. Additionally, the soil displacement zones were contained well
geogrid aperture size and average particle size is mainly distributed in
within the boundary of the model box, which will be explained in Sec­
0.96–4 (Mehrjardi and Khazaei, 2017; Han et al., 2018; Sweta and
tion 3.4.
Hussaini, 2018). However, the average particle size of the test sand is
only 0.4 mm, so the ratio range obtained by predecessors is not appli­
2.3. Test program
cable. Tang et al. (2017) used the same sand as this paper, and they
found that when the contact area of the geogrid with the soil accounts
First of all, a semicircular anchor was installed on the sand bed and
for about 1/3 of the overall interface, the interface friction provided by
horizontally placed against the front window of the model box. Sec­
geogrids is the largest. As shown in Fig. 2(d), the geogrid aperture size is
ondly, geogrids were placed at the desired location and rigidly fixed to
0.025 m × 0.025 m, and the width of longitudinal and transverse ribs is
the model box using steel plates and bolts, as shown in Fig. 2(d). In the
4 × 10− 3 m. The contact area of the geogrid–soil interface is about 36%
studies of Hatami and Bathurst (2005) and Huang et al. (2009), geo­
of the overall interface. The physical and mechanical properties of the
synthetics were fixed at one end to reinforced soil segmental walls to
geogrids are given by the manufacturer. The breaking elongation and
provide larger resistance. Yang (1995) proposed a new retaining wall
strength of the geogrids are 3% and 5 × 104 N/m, respectively.
design combining anchor plates and geogrids, which required that the
geogrids on both sides of the retaining wall were anchored with con­
2.2. Experimental setup crete. In this study, the geogrids were fixed on the two sides of the model
box and their reinforcing mechanism on the soil was further discussed.
The tests were carried out in a model box. As shown in Fig. 2(b), this Then the soil was compacted to the predetermined density (1.75 × 103
box was 0.5 m long, 0.25 m wide, and 0.5 m high. It was made of an kg/m3) layer by layer, and the thickness of each layer was approxi­
aluminum plate at the bottom and transparent plates around it. A mately 0.035 m. After compaction, the soil surface shall be ground to
semicircular anchor plate with a diameter of 0.07 m and a thickness of 6 make the filling of the upper and lower layers closely and evenly fit. As
× 10− 3 m was used in the tests. The anchor was connected to the loading the soil compaction was completed, the anchor plate was pulled up at
beam of the universal testing machine through a fixture. The coordinate the rate of 8.3 × 10− 5 m/s to keep consistent with the actual working
system shown in Fig. 2(c) was established with the origin located in the conditions in the field (Tian et al., 2013). The tests were terminated
plane of the top of the anchor plate. Liu et al. (2012) found that the when the anchor plate was pulled up by 0.025 m.
interface friction makes the uplift resistance of the semicircular anchor A high-precision digital camera with 18 million pixels was placed at
plate higher than the actual value. However, the friction effect was approximately 0.5 m on the front side of the model box, and the camera
neglected in this study because the contact area of the anchor plate with parameters were adjusted to obtain the best image. The camera was set
the model box was small. This study focuses on the failure modes of to automatic shooting mode during the tests, and one photograph was
anchor plates in geogrid-reinforced soil during uplift, and the failure taken every 10 s for PIV analysis. At the same time, an optical frequency
displacements from tests on the semicircular anchor are comparable to domain reflectometry (OFDR) demodulator was used to obtain the strain
those on the full circular anchor (Liu et al., 2012). Hence, the soil measurements of the fiber optic cables in real-time. Before the anchor
displacement fields obtained in semicircular anchor tests can reveal the uplift tests, two calibration tests were carried out on fiber optic cables to
true failure mechanism. investigate their performance in distributed strain sensing. Prior to the
As shown in Fig. 3, in the uplift tests, there were three layout schemes calibration, the fiber optic cables were pasted on a uniform-strength
of geogrids: (1) one layer of geogrid was placed 0.07 m from the ground beam to apply uniform strains in stages. The strain measurements of
surface without contact with the anchor plate (i.e., Reinforced Layout the fiber optic cables were consistent with the theoretically calculated
1); (2) one layer of geogrid was in contact with the anchor plate (i.e., strains. These calibration results verify the high accuracy of the strain
Reinforced Layout 2); (3) two layers of geogrids were laid in soil, one in sensing cables.
contact with the anchor plate and the other placed 0.07 m from the
ground surface (i.e., Reinforced Layout 3). The width of the geogrids was

996
Y.-X. Gao et al. Geotextiles and Geomembranes 50 (2022) 994–1003

unnecessary in this study (Wu et al., 2020).

PIV is a velocity-measuring technique that was first applied to solve
problems in fluid mechanics. This technique has been successfully
applied to deformation monitoring in physical model tests (Adrian,
1991; White et al., 2003). As shown in Fig. 4(b), the initial photograph of
soil is divided into a grid of test patches. By tracking characteristics such
as texture or brightness within one searching patch in ‘‘Image 1′′ and
using specific algorithms to search this patch in ‘‘Image 2”, the
displacement of the test patch during the interval between images is
evaluated (Stanier et al., 2016). Planar deformation can be captured by
repeating this operation for the entire mesh of patches.

3. Analysis of results

3.1. Load versus displacement curves

The friction between the pulling rods and the surrounding soil may
have some influence on the uplift capacity of anchor plates. According to
the results of pulling rod uplift tests, the pullout resistance of the rod is
generally less than 3% of the total resistance. Hence, this effect is
negligible.
Fig. 5 shows the typical curves for anchor uplift resistance versus
displacement in unreinforced and reinforced soils. In this study, the
geogrids were rigidly fixed to the model box so that the anchor plate in
reinforced soil could provide larger uplift resistance. Although the uplift
forces of the anchor plate in reinforced soil are significantly greater than
Fig. 4. Working principles of (a) the OFDR technology and (b) the in the unreinforced case, the general trend of the uplift resistance with
PIV technique. uplift displacement is consistent. The anchor plate shows the hardening
behavior in reinforced soil, therefore, the ultimate bearing capacity is
2.4. Principle of the monitoring system defined as the uplift resistance corresponding to the point at which the
hardening stage begins. The ultimate bearing capacity and residual
In this experimental study, the Rayleigh-backscattering-based OFDR bearing capacity increase with the embedment ratio under reinforced
technique is used to obtain the strain profiles of geogrids during the and unreinforced conditions. Further, the uplift curves show significant
uplift of the anchor plate. Fig. 4(a) shows the demodulation principle of fluctuation at large displacements. This is because the soil flowed from
OFDR. The light source emits highly coherent continuous waves with a the upper part of the anchor plate to the bedding layer, reducing the
linear sweep optical frequency (Gomes et al., 2018). The linear light is uplift resistance. Then, with the increase in uplift displacement, the soil
divided into two parts by the coupler. One part is used as a local oscil­ above the anchor plate was squeezed, and the uplift resistance increased
lator to reflect the reference light, while the other part is reflected due to again (Murray and Geddes, 1987).
the random refractive index in an optical fiber (Palmieri and Schenato, For the Reinforced Layout 1 and unreinforced cases, the anchor plate
2013). Both the reference light and signal light are received by the uplift curves can be divided into three stages:
photodetector in a mixed state. The mixed signal, which is sent to the Stage 1) The uplift resistance rapidly increased with the anchor
spectrum analyzer, is separated into several signal windows according to displacement towards a peak value.
the spatial resolution of the interrogator and uses the cross-correlation Stage 2) After reaching the ultimate bearing capacity of the anchor
operation to calculate the spectral shift of each signal window (Wu plate, the dilatancy angle of the soil on failure surfaces rapidly decreased
et al., 2020). The refractive index of the optical fiber will change due to to 0 and the friction angle decreased to the residual value, resulting in
strain or temperature variations, resulting in a frequency shift. The the uplift resistance decreasing significantly as the uplifted displacement
frequency shift is linearly related to the change in strain and tempera­ continues to increase.
ture to which the fiber is subjected (Henault et al., 2012). In the anchor Stage 3) The soil above the anchor plate moved to the bedding layer,
uplift tests, the temperature variation was controlled within 1.5 ◦ C. resulting in the slow reduction of the uplift resistance with increasing
Therefore, temperature compensation for the strain data was displacement.

Fig. 5. Typical load versus displacement curves at (a) H/D = 2 and (b) H/D = 1.5.

997
Y.-X. Gao et al. Geotextiles and Geomembranes 50 (2022) 994–1003

As shown in Fig. 5 (a), the uplift curves in Reinforced Layouts 2 and 3

can be roughly divided into two stages. The uplift bearing capacity
increased rapidly to point A3 or A4 and then gradually reduced.
To quantify the contributions of geogrids to the overall uplift resis­
tance under different embedment ratios, two critical parameters at the
ultimate state are introduced as:
Qu
Nqu = (1)
γAH

Fig. 6. Variation of breakout factor and improvement factor with anchor

embedment ratio in unreinforced and reinforced soil. Fig. 8. Deformation patterns of the fiber optic cables.

Fig. 7. Vertical slip mechanical models in (a) unreinforced and (b) reinforced soil.

998
Y.-X. Gao et al. Geotextiles and Geomembranes 50 (2022) 994–1003

Fig. 9. Strain distributions of sensing cables at H/D = 3 under the reinforced condition of (a) Reinforced Layout 1, (b) Reinforced Layout 2, and (c) Reinforced
Layout 3.

(Nqu )rein. of Reinforced Layout 2 is similar to Reinforced Layout 3, which indicates

I= (2)
(Nqu )unrein. that the upper geogrid within the studied range plays an insignificant
role in the overall uplift resistance.
where Nqu is the breakout factor, I is the reinforcement improvement
factor, Qu is the anchor ultimate bearing capacity, γ is the unit weight of 3.2. Calculation of ultimate bearing capacity
the soil, H is the embedment depth of the anchor, and A is the cross-
sectional area of the anchor plate. The mechanical responses of unreinforced and reinforced soils are
As shown in Fig. 6, geogrids enhance the uplift resistance of the shown in Fig. 7. The influence of different factors on the anchor bearing
anchor plate, specifically in the form of an increase in the breakout capacity is quantified in this section.
factor. Increasing the anchor plate embedment ratio will greatly The ultimate bearing capacity (Qu) consists of two main parts in
improve the ultimate bearing capacity in unreinforced soil. At the same unreinforced soil. One is the weight (G) of the soil within the failure
time, altering the anchor plate embedment ratio also affects the surface, and the other is the combined force in the vertical direction of
improvement factor of geogrids. In addition, the anchor uplift capacity normal stress (σn) and shear stress (τ) applied to the failure surface.

999
Y.-X. Gao et al. Geotextiles and Geomembranes 50 (2022) 994–1003

can therefore be written as:

∫( [ ])
( ) 1 + k0 (1 − k0 )cos 2 θ
Qu 0 = γz tan φp cos θ − sin θ − dA + W (6)
2 2
A

where W is the weight of the soil within the failure surface, φp is the peak
friction angle, and θ is the angle of the shear band with the vertical and
its value is determined by the failure mechanism. For the cylinder sur­
face failure mode, θ = 0◦ .
Due to complex interactions between soil particles and geogrid ribs,
the geogrid reinforcement can improve the bearing capacity of the soil
and restrain the soil deformation (Teixeira et al., 2007). As presented in
Fig. 7(b), Segments a and d are subjected to interface friction and ver­
tical stress. In addition, the zone where geogrids exert their reinforcing
effects tended to be slightly larger than the displacement influence zone
(as shown in Segments b and e). The geogrids near the sidewalls of the
model box did not move at all and therefore they were unaffected by
geogrid-soil interface friction.
If the geogrid is subjected to interface friction, the reaction force
produces additional confining pressure on the soil, which is manifested
in the increase of soil cohesion (i.e. Δc). Whether the geogrid is fixed or
not, the mechanism of enhancing soil strength is the same. The Δc value
under the fixed condition is higher than that under the unfixed condi­
tion. From the geometry of these Mohr’s circles in Fig. 7(b), the ex­
pressions for the peak normal stress (σn1) and the peak shear stress (τp1)
of the reinforced soil can be obtained as:
1 [( 2 ) ( ) ]
σ n1 = k0 γz + k0 γz + Δσ h − k0 γz − k0 2 γz + Δσ h cos 2 θ (7)
2k0

tan φp [( 2 ) ( ) ]
τP1 = k0 γz + k0 γz + Δσh − k0 γz − k0 2 γz + Δσ h cos 2 θ + Δc
2k0
Fig. 10. Comparison of the improvement ratio between the upper and lower
geogrids in Reinforced Layout 3 at H/D = 3. (8)
The ultimate bearing capacity of the anchor plate in reinforced soil
can therefore be expressed as:
Table 1 ∫
Summary of uqu values. ( )
Qu 1 = τp1 cos θ − σ n1 sin θ dA + W (9)
3
H/D uqu ( × 10− m) A

Unrein. Rein. Layout 1 Rein. Layout 2 Rein. Layout 3

1.5 0.93 1.52 2.02 3.26 3.3. Strain analysis of geogrids

2 1.41 2.35 3.58 3.89
2.5 1.99 3.15 3.24 4.95
Ignoring the strain transfer loss due to the fiber optic sheath, soil
3 5.07 5.72 5.61 6.12
type, and other factors, the strains measured by fiber optic cables are
assumed to be equal to the true strains of the host materials (Zhang et al.,
There are three failure modes in unreinforced soil, including cylinder 2020). Hence, the fiber optic strain measurements are used to charac­
surface failure, truncated cone failure, and circular surface failure. As terize the deformation of geogrids at different positions in the following
shown in Fig. 7(a), White et al. (2008) proposed that the normal stress sections. As shown in Fig. 8, there may be three deformation patterns of
on the failure surface in unreinforced soil is constant, and the shear geogrids. When some part of the soil moves upward, the bending of
stress is determined by the Mohr circle of stress. geogrids leads to tensile strains of the fiber optic cables. The second
The horizontal stress (σ h0) and vertical stress (σz0) along the failure pattern indicates that the uniform vertical displacement of soil will
surface in the initial condition can be obtained by the following induce negligible strains of the cables. In the third pattern, the bending
equation: of geogrids will induce compressive strains of the cables.
As indicated in Fig. 9, the strain distributions of sensing cables in the
σ h0 = k0 σz0 = k0 γz (3)
horizontal plane were symmetrical. The closer the geogrid was to the
where k0 is the coefficient of earth pressure at rest. anchor plate, the larger the peak strain it received. In addition, the peak
From the geometry of the two Mohr’s circles in Fig. 7(a), in the strains of geogrids increased with the uplift displacement, and they were
unreinforced soil, the peak mobilized normal and shear stresses along closely related to the deformation of geogrids.
the failure surface can be expressed as: When the single-layer geogrid was not in contact with the anchor
[ ] plate, the sensing cable at y = 0.035 m showed significant compressive
1 + k0 (1 − k0 )cos2θ strain at the sides of the anchor plate, which indicated the boundary of
σ n0 = γz − (4)
2 2 the displacement influence zone. Prior to each uplift test, zero settings
[ ] were applied to the sensing cables. Therefore, the compressive strain
τp0 = γz tan φp
1 + k0 (1 − k0 )cos2θ
− (5) during uplift manifested the bending of the geogrids and relaxation of
2 2 the cables. When the anchor plate was uplifted, the downward soil
The ultimate bearing capacity of the anchor plate in unreinforced soil movement surrounding the anchor caused by the flow-around mecha­
nism was observed in the tests. Additionally, the boundary of the

1000
Y.-X. Gao et al. Geotextiles and Geomembranes 50 (2022) 994–1003

3 3
Fig. 11. Soil displacement field under unreinforced and reinforced conditions at H/D = 1.5 with different uplift displacements: (a) u = 1 × 10− m; (b)u = 2 × 10−
m; (c)u = 4 × 10− 3 m.

3
Fig. 12. Soil displacement field under unreinforced and reinforced conditions at H/D = 3 with u = 6 × 10− m.

influence zone formed the failure surface. The peak strain values for the geogrids in contact with the anchor plate were subjected to larger tensile
cables furthest from the anchor plate (i.e., y = 0.185 m) were less than stress and had a more significant reinforcing effect. To quantitatively
80 με, indicating that the soil displacement fields were limited to the evaluate the reinforcing effect of each layer in Reinforced Layout 3, a
range that did not reach the sidewalls. This phenomenon further non-dimensional parameter ŋ is defined as:
confirmed that there were almost no boundary and scale effects in this
(ε1 )max
study. η= (11)
(ε2 )max
As shown in Fig. 9(b), Reinforced Layout 2 could provide a larger
uplift resistance than other geogrid layouts. At y = 0.035 m, the
where ŋ is the improvement ratio, (ε1)max is the peak strain value of one-
maximum strain monitored by the sensing cable was 8604 με, which was
layer geogrid at different locations in Reinforced Layout 3, and (ε2)max is
much larger than the strain of the geogrid without contact with the
the sum of the peak strain values of two layers on the same vertical
anchor plate at the same vertical position. This phenomenon corre­
plane.
sponds to the reinforcing effect of the geogrids discussed in Section 3.1.
Comparing the improvement ratio of the upper and lower geogrids
Outside the deformation influence zone (i.e. y = 0.185 mm), the strain
for the case of H/D = 3, Fig. 10 shows that the lower layer plays a more
values measured by cables tended to be zero.
important role at y = 0.035 m. However, with the expansion of the
Note that the strains of the lower geogrid at y = 0.11 m and 0.185 m
displacement influence zone, the improvement ratio of the upper layer is
were not symmetrical in Fig. 9(c), while the strains of geogrids at other
greater than that of the lower layer at y = 0.11 m. In addition, the peak
locations possessed good symmetry. The reason for this phenomenon
strain values were much small at y = 0.185 m.
may be the non-uniformity of the compacted foundation.
Compared with the geogrids without contact with the anchor, the

1001
Y.-X. Gao et al. Geotextiles and Geomembranes 50 (2022) 994–1003

3.4. Soil failure mechanisms 4. For reinforced soil, the failure mechanism gradually changes from a
vertical slip surface mode to a bulb-shaped surface mode as the
The ultimate displacement (uqu) means the displacement at which embedment depth of the anchor plate increases. At H/D = 3, the
the ultimate uplift capacity is mobilized. The values of uqu for different displacement fields of unreinforced and Reinforced Layout 1 extend
reinforcement conditions are summarized in Table 1. Typically, the to the ground surface, while Reinforced Layouts 2 and 3 both show a
value of uqu increases with the embedment depth and is higher in rein­ deep failure mechanism that is mainly characterized by a bulb-
forced soil than in unreinforced soil at the same embedment ratio. A shaped failure surface.
similar trend has been found in previous studies (Rahimi et al., 2018).
As depicted in Fig. 11, for the case of H/D = 1.5, the failure mech­
Declaration of competing interest
anism in the unreinforced soil exhibited a vertical slip surface mode,
while the shapes of failure surfaces in three kinds of reinforced soils
None.
were curved. The displacement profile in reinforced soil was larger than
that in unreinforced soil, which was also reported in the literature
(Rahimi et al., 2018; Yunkul et al., 2021). The angles of failure surfaces Acknowledgments
in Reinforced Layouts 1 and 2 were equal to φ/3, but it is slightly larger
in Reinforced Layout 3. The reason for this phenomenon is that the The authors gratefully acknowledge the financial support provided
two-layer geogrids provide larger interface friction and a stronger by the National Natural Science Foundation of China (No. 42077235).
interlock between the soil and geogrids. In addition, the displacement Special thanks are given to De-Yang Wang for his help in revising the
influence zones extended outwards as the uplift displacement increased. manuscript.
Once reaching the ultimate displacement, the boundaries of these zones
cease to change (see Fig. 12). References
Ilamparuthi et al. (2002) investigated that the slip surface of the deep
Adrian, R.J., 1991. Particle-imaging techniques for experimental fluid mechanics. Annu.
failure mechanism eventually developed into a bulb shape. At H/D = 3,
Rev. Fluid Mech. 23 (1), 261–304.
the displacement fields in Reinforced Layout 1 and unreinforced soils Al-Naddaf, M., Han, J., Xu, C., Jawad, S., Abdulrasool, G., 2019. Experimental
extended to the ground surface. However, in Reinforced Layout 2 or 3, investigation of soil arching mobilization and degradation under localized surface
loading. J. Geotech. Geoenviron. Eng. ASCE 145 (12), 04019114.
the deformation field was mainly characterized by a bulb-shaped failure
ASTM D2487-17, 2017a. Standard Practice for Classification of Soils for Engineering
surface, which indicated that the failure surfaces gradually deflect in­ Purposes (Unified Soil Classification System). ASTM International, West
ward as geogrids gradually play their roles. There is little difference Conshohocken, PA, USA.
between the failure modes of Reinforced Layouts 2 and 3. Badakhshan ASTM D6528-17, 2017b. Standard Test Method for Consolidated Undrained Direct
Simple Shear Testing of Fine Grain Soils. ASTM International, West Conshohocken,
et al. (2018) pointed out that the second layer geogrid has little effect on PA, USA.
improving the ultimate uplift force. Within the studied range of this Baadiga, R., Saride, S., Balunaini, U., Madhira, R.M., 2021. Influence of tensile strength
paper, the best reinforcement method is to place one-layer geogrid in of geogrid and subgrade modulus on layer coefficients of granular bases. Transp.
Geotech. 29, 100557.
contact with the anchor plate (i.e. Reinforced Layout 2). Furthermore, Badakhshan, E., Noorzad, A., Zameni, S., 2018. An updating void ratio model for large
the bedding layer was not affected during the uplift tests, and the deformation simulation of geogrid-granular strip anchors plates. Comput. Geotech.
boundaries of the displacement influence zones did not extend to the 94, 134–149.
Bhattacherjee, D., Viswanadham, B.V.S., 2019. Centrifuge model studies on performance
sidewalls of the model box. Therefore, it can be concluded that there was of hybrid geosynthetic–reinforced slopes with poorly draining soil subjected to
no boundary effect in these tests. rainfall. J. Geotech. Geoenviron. Eng. ASCE 145 (12), 04019108.
Bhowmik, R., Jagdish, T.S., Datta, M., 2019. Experimental studies on inclined pullout
behaviour of geosynthetic sheet Vis-À-Vis geogrid-Effect of type of anchor and sand.
4. Conclusions Geotext. Geomembranes 47 (6), 767–779.
Boushehrian, A.H., Hataf, N., Ghahramani, A., 2011. Modeling of the cyclic behavior of
This paper presents an experimental investigation of the uplift shallow foundations resting on geomesh and grid-anchor reinforced sand. Geotext.
Geomembranes 29 (3), 242–248.
mechanism of anchor plates in geogrid-reinforced soil. A series of small- Choudhary, A.K., Pandit, B., Babu, G.L.S., 2018. Three-dimensional analysis of uplift
scale model tests have been performed to study the influence of anchor behaviour of square horizontal anchor plate in frictional soil. Int. J. Geosynth.
embedment depth, the number of geogrid layers, and the location of Groun. 4 (2), 14.
Choudhary, A.K., Pandit, B., Babu, G.L.S., 2019a. Uplift capacity of horizontal anchor
geogrids on the uplift behavior. Furthermore, theoretical equations are
plate in geocell reinforced sand. Geotext. Geomembranes 47 (2), 203–216.
derived for the calculation of the uplift resistance of anchor plates in Choudhary, A.K., Pandit, B., Babu, G.L.S., 2019b. Experimental and numerical study on
unreinforced and geogrid-reinforced soils. The following conclusions square anchor plate groups in geogrid reinforced sand. Geosynth. Int. 26 (6),
can be drawn from this experimental study: 657–671.
Das, B.M., Shukla, S.K., 2013. Earth Anchors, second ed. J. Ross Publishing, Inc, Florida.
Dickin, E.A., Leung, C.F., 1983. Centrifugal model tests on vertical anchor plates.
1. With the increase in embedment ratios, the uplift resistance of the J. Geotech. Eng. ASCE 109 (12), 1503–1525.
anchor plates significantly increases, and the anchor uplift resistance El Sawwaf, M., Nazir, A., 2006. The effect of soil reinforcement on pullout resistance of
an existing vertical anchor plate in sand. Comput. Geotech. 33 (3), 167–176.
is also affected by the layouts of geogrids. Moreover, the second layer Evans, T.M., Zhang, N., 2019. Three-dimensional simulations of plate anchor pullout in
of geogrid has an insignificant reinforcing effect, and the optimal granular materials. Int. J. Geomech. ASCE 19 (4), 04019004.
reinforcement method is to place one-layer geogrid in contact with Ganesh, R., Sahoo, J.P., 2017. Vertical uplift resistance of plate anchors with the
fluctuations of groundwater table. Int. J. Geosynth. Groun. 3 (1), 8.
the anchor plate. Ghosh, A., Bera, A.K., 2010. Effect of geotextile ties on uplift capacity of anchors
2. The geogrids can improve the soil strength and restrain the soil embedded in sand. Geotech. Geol. Eng. 28, 567–577.
deformation. The complex interactions between soil particles and Gomes, M., Cruz, S., Lopes, H., Arcipreste, B., Magalhães, R., Da Silva, A.F., Viana, J.C.,
2018. Distributed optical fiber sensors for PCB-strain analysis. IEEE Trans. Ind.
geogrid ribs produce additional confining pressure on the soil, which Electron. 66 (10), 8181–8188.
is manifested in the increase of soil cohesion. Han, B., Ling, J., Shu, X., Gong, H., Huang, B., 2018. Laboratory investigation of particle
3. Geogrids in contact with anchor plates are subjected to higher strains size effects on the shear behavior of aggregate-geogrid interface. Construct. Build.
Mater. 158, 1015–1025.
and provide more significant resistance than the geogrids without
Hatami, K., Bathurst, R.J., 2005. Development and verification of a numerical model for
contact with anchors. The strain distributions of geogrids in the the analysis of geosynthetic-reinforced soil segmental walls under working stress
horizontal direction are symmetrical. The closer geogrids are to the conditions. Can. Geotech. J. 42 (4), 1066–1085.
anchor plate, the larger the peak strain. Furthermore, the magnitude Henault, J.M., Quiertant, M., Delepine-Lesoille, S., Salin, J., Moreau, G., Taillade, F.,
Benzarti, K., 2012. Quantitative strain measurement and crack detection in RC
of the peak strain of geogrids can be used to characterize the rein­ structures using a truly distributed fiber optic sensing system. Construct. Build.
forcing effect. Mater. 37, 916–923.

1002
Y.-X. Gao et al. Geotextiles and Geomembranes 50 (2022) 994–1003

Huang, B., Bathurst, R.J., Hatami, K., 2009. Numerical study of reinforced soil segmental Riyad, A.S.M., Rokonuzzaman, M., Sakai, T., 2020. Progressive failure and scale effect of
walls using three different constitutive soil models. J. Geotech. Geoenviron. Eng. anchor foundations in sand. Ocean. Eng. 195, 106496.
ASCE 135 (10), 1486–1498. Sakai, T., Tanaka, T., 1998. Scale effect of a shallow circular anchor in dense sand. Soils
Ilamparuthi, K., Dickin, E.A., Muthukrisnaiah, K., 2002. Experimental investigation of Found. 1998 38 (2), 93–99.
the uplift behaviour of circular plate anchors embedded in sand. Can. Geotech. J. 39 Sakai, T., Tanaka, T., 2007. Experimental and numerical study of uplift behavior of
(3), 648–664. shallow circular anchor in two-layered Sand. J. Geotech. Geoenviron. Eng. ASCE 133
Ilamparuthi, K., Ravichandran, P.T., Toufeeq, M.M., 2008. Study on uplift behaviour of (4), 469–478.
plate anchor in geogrid reinforced sand bed. In: Proceedings of Geotechnical Sharma, A., Alzaylaie, M., Vandanapu, R., Khalaf, K., 2021. Numerical and analytical
Earthquake Engineering and Soil Dynamics Congress IV, pp. 1–10. Sacramento, studies of 3D effects on pullout capacity of anchor blocks in granular compacted fill.
California. Int. J. Geosynth. Groun. 7 (1), 1–8.
Kanitz, M., Hager, A., Grabe, J., Goniva, C., 2019. Numerical and experimental analysis Stanier, S.A., Blaber, J., Take, W.A., White, D.J., 2016. Improved image-based
of the extraction mechanism of an anchor plate embedded in saturated sand. deformation measurement for geotechnical applications. Can. Geotech. J. 53,
Comput. Geotech. 111, 191–201. 727–739.
Krishnaswamy, N.R., Parashar, S.P., 1994. Uplift behaviour of plate anchors with Sweta, K., Hussaini, S.K.K., 2018. Effect of shearing rate on the behavior of geogrid-
geosynthetics. Geotext. Geomembranes 13 (6), 67–89. reinforced railroad ballast under direct shear conditions. Geotext. Geomembranes 46
Latha, G.M., Murthy, V.S., 2007. Effects of reinforcement form on the behavior of (3), 251–256.
geosynthetic reinforced sand. Geotext. Geomembranes 25 (1), 23–32. Tang, X.S., Zheng, Y.R., Wang, Y.F., Feng, Y.S., 2017. Study on the reasonable size of geo-
Latha, G.M., Somwanshi, A., 2009. Effect of reinforcement form on the bearing capacity grid meshes. Rock Soil Mech. 38 (6), 1583–1588 (in Chinese).
of square footings on sand. Geotext. Geomembranes 27 (6), 409–422. Teixeira, S.H., Bueno, B.S., Zornberg, J.G., 2007. Pullout resistance of individual
Li, H., Zhu, H., Li, Y., Zhang, C., Shi, B., 2022. Experimental study on uplift mechanism of longitudinal and transverse geogrid ribs. J. Geotech. Geoenviron. Eng. ASCE 133 (1),
pipeline buried in sand using high-resolution fiber optic strain sensing nerves. 37–50.
J. Rock Mech. Geotech. Eng. 14 (4) (in press). Tian, Y., Gaudin, C., Cassidy, M.J., Randolph, M.F., 2013. Considerations on the design
Liu, C.N., Zornberg, J.G., Chen, T.C., Ho, Y.H., Lin, B.H., 2009. Behavior of geogrid-sand of keying flap of plate anchors. J. Geotech. Geoenviron. Eng. ASCE 139 (7),
interface in direct shear mode. J. Geotech. Geoenviron. Eng. ASCE 135 (12), 1156–1164.
1863–1871. Vahedifard, F., Shahrokhabadi, S., Leshchinsky, D., 2016. Geosynthetic-reinforced soil
Liu, J.Y., Liu, M.L., Zhu, Z.D., 2012. Sand deformation around an uplift plate anchor. structures with concave facing profile. Geotext. Geomembranes 44 (3), 358–365.
J. Geotech. Geoenviron. Eng. ASCE 138 (6), 728–737. Wang, B.J., Li, K., Shi, B., Wei, G.Q., 2009. Test on application of distributed fiber optic
Liu, F., Sun, H., Jung, J., Zhang, X., Ju, X., 2019. Experimental study of pullout capacity sensing technique into soil slope monitoring. Landslides 6 (1), 61–68.
of plate anchors shallowly embedded in hydrate bearing sediments. Ocean. Eng. 173, White, D.J., Take, W., Bolton, M., 2003. Soil deformation measurement using Particle
548–555. Image Velocimetry (PIV) and photogrammetry. Geotechnique 53 (7), 619–631.
Mehrjardi, G.T., Khazaei, M., 2017. Scale effect on the behaviour of geogrid-reinforced White, D.J., Cheuk, Y., Bollon, M.D., 2008. The uplift resistance of pipes and plate
soil under repeated loads. Geotext. Geomembranes 45 (6), 603–615. anchors buried in sand. Geotechnique 58 (10), 771–779.
Merifield, R.S., Lyamin, A.V., Sloan, S.W., 2006. Three-dimensional lower-bound Wu, H., Zhu, H.H., Zhang, C.C., Zhou, G.Y., Zhu, B., Zhang, W., Azarafza, M., 2020.
solutions for the stability of plate anchors in sand. Geotechnique 56 (2), 123–132. Strain integration-based soil shear displacement measurement using high-resolution
Mukherjee, S., Kumar, L., Choudhary, A.K., Babu, G.S., 2021. Pullout resistance of strain sensing technology. Measurement 166, 108210.
inclined anchors embedded in geogrid reinforced sand. Geotext. Geomembranes 49 Yang, H., 1995. Reinforced earth retaining wall with anchor plates. Building Structure 24
(5), 1368–1379. (2), 35–36 (in Chinese).
Murray, E.J., Geddes, J.D., 1987. Uplift of anchor plates in sand. J. Geotech. Eng. ASCE Ye, X., Zhu, H.H., Wang, J., Zhang, Q., Shi, B., Schenato, L., Pasuto, A., 2022. Subsurface
113 (3), 202–215. multi-physical monitoring of a reservoir landslide with the fiber-optic nerve system.
Niroumand, H.K., Kassim, A., Nazir, R., 2013. The influence of soil reinforcement on the Geophys. Res. Lett.
uplift response of symmetrical anchor plate embedded in sand. Measurement 46 (8), Yunkul, K., Gurbuz, A., 2020. Uplift behavior of shallow horizontal plate anchors
2608–2629. reinforced with geocells in cohesionless soil. Eur. J. Environ. Civ. Eng. 1–24.
Palmieri, L., Schenato, L., 2013. Distributed optical fiber sensing based on Rayleigh Yunkul, K., Usluogulları, O.F., Gurbuz, A., 2021. Numerical analysis of geocell reinforced
scattering. Open Opt. J. 7 (1), 104–127. square shallow horizontal plate anchor. Geotech. Geol. Eng. 39 (4), 3081–3099.
Perazzelli, P., Anagnostou, G., 2017. Uplift resistance of strip anchors in cohesive Zhang, J., Cao, W.Z., Zhou, Y.J., 2021. Mechanical behavior of triaxial geogrid used for
frictional mediums of limited tensile strength. Int. J. GeoMech. 17 (9), 04017042. reinforced soil structures. Adv. Civ. Eng. 2021, 5598987.
Perez, Z.A., Schiavon, J.A., Tsuha, C.D.H.C., Dias, D., Thorel, L., 2018. Numerical and Zhang, C.C., Shi, B., Zhu, H.H., Wang, B.J., Wei, G.Q., 2020. Toward distributed fiber-
experimental study on influence of installation effects on behaviour of helical optic sensing of subsurface deformation: A theoretical quantification of ground-
anchors in very dense sand. Can. Geotech. J. 55 (8), 1067–1080. borehole-cable interaction. J. Geophys. Res. Solid Earth 125 (3).
Qian, Y., Han, J., Pokharel, S.K., Parsons, R.L., 2013. Performance of triangular aperture Zhu, H.H., Mei, G.X., Xu, M., Liu, Y., Yin, J.H., 2014. Experimental and numerical
geogrid-reinforced base courses over weak subgrade under cyclic loading. J. Mater. investigation of uplift behavior of umbrella-shaped ground anchor. Geomech. Eng. 7
Civ. Eng. ASCE 25 (8), 1013–1021. (2), 165–181.
Rahimi, M., Tafreshi, S.M., Leshchinsky, B., Dawson, A.R., 2018. Experimental and Zhu, H.H., Wang, D.Y., Shi, B., Wang, X., Wei, G.Q., 2022. Performance monitoring of a
numerical investigation of the uplift capacity of plate anchors in geocell-reinforced curved shield tunnel during adjacent excavations using a fiber optic nervous sensing
sand. Geotext. Geomembranes 46 (6), 801–816. system. Tunn. Undergr. Space Technol. 124, 104483.
Rajesh, S., Viswanadham, B.V.S., 2012. Centrifuge modeling and instrumentation of
geogrid-reinforced soil barriers of landfill covers. J. Geotech. Geoenviron. Eng. ASCE
138 (1), 26–37.

1003