Shaking Table Tests to Evaluate the Seismic Performance

of Soil Nailing Stabilized Embankments

Smrutirekha Sahoo1; Bappaditya Manna, M.ASCE2; and K. G. Sharma3
Downloaded from ascelibrary.org by Jaypee University Of Information Tech - Solan on 05/07/24. Copyright ASCE. For personal use only; all rights reserved.

Abstract: A series of shaking table tests will be carried out on 0.4 m high instrumented soil nailed embankments. The important parameters
considered in this study consist of a combination of three different slope angles (β = 30°, 45°, and 60°), three different nail inclinations (i = 0°,
15°, and 30°), and three nail lengths (l = 0.2, 0.3, and 0.4 m). This study revealed that the effect of slope angle on the seismic resistance of
steep nailed soil slopes was significant and the minimal magnitude of the facing displacement was obtained with nail inclinations of 0° and
15° for the nailed slopes without facing and with facing, respectively. A model nailed slope with a nail inclination of 15° provided better
reinforcement action than the horizontally and 30° inclined nails when stabilizing soil slopes. The maximum axial nail force was obtained
at approximately two-thirds of the nail length in the model slopes with facing and at the slope facing for model slopes without facing.
DOI: 10.1061/(ASCE)GM.1943-5622.0001981. © 2021 American Society of Civil Engineers.
Author keywords: Shaking table test; Seismic performance; Facing; Nailed soil slope; Nail force.

Introduction nail walls, and the development of nail forces have been evaluated
under static and seismic conditions. From the observations made on
Soil nailing is an important ground reinforcement technique that is various failure modes, the factor of safety values obtained conven-
used for the repair, stabilization, and reconstruction of existing re- tionally and from the numerical simulations for the static, pseudo-
taining structures for excavation support, support of hill cuts, static, and dynamic cases satisfy the corresponding minimum
bridge abutments, highways, railways, and tunnel portals. It is an recommended values according to various standard guidelines. In
effective and economical method that has been used successfully addition, pseudostatic analyses resulted in conservative estimates
in highway and railway cuts, and for end-slope removal under ex- of displacements and factor of safety values compared with those
isting bridge abutments during underpass widening. The installa- obtained from the time history analyses of the considered earth-
tion of soil nail walls is relatively simple and fast, uses fewer quake motions. It was concluded that soil nails could provide an ef-
construction materials, and easy adjustments can be made to nail ficient, feasible, and economical alternative to conventional
positions when obstructions, such as stones, cobbles, or under- retaining structures under seismic conditions, in particular, to sup-
ground utilities, are encountered during drilling without additional port vertical or near-vertical cuts that were made in the soil for var-
cost or time. Because of this, soil nailing is a viable and economical ious slope stability applications in geotechnical engineering.
option to support vertical cuts, particularly in locations where site The coherent and flexible nature of nailed soil structures means
constraints are more predominant and project duration is limited. that they are inherently capable of resisting a substantial amount of
In the literature, research has been conducted into the design and deformation under seismic loading conditions. This is why a large
construction methodologies (Fan and Luo 2008; Chu and Yin part of the literature on nailed soil structures emphasizes the design
2005; Sheahan and Carlton 2003), laboratory techniques (Li et al. and reinforcement mechanism of nailed structures under static load
2008; Yin and Lok-Man 2006), and numerical techniques (Patra only. However, limited studies are available on the evaluation of
and Basudhar 2005; Yuan et al. 2003) for nailed structures under the seismic response of nailed soil slopes that do not cover some
static condition. The performance of a soil nail wall that supports of the important aspects, for example, mobilized nail forces along
an 8 m high vertical cut was studied under both static and seismic their length, the effect of slope facia, and acceleration amplification
conditions (Babu and Singh 2008). The results for aspects such as under seismic excitation, which is essential to understand the seis-
maximum lateral displacements, important failure modes of soil mic behavior of steep nailed slopes and designs. The consideration
of bending and shear resistance of soil nails has been debated in soil
1 nailing analysis and design. In numerical simulations of soil nailed
Assistant Professor, Dept. of Civil Engineering, National Institute of
Technology Meghalaya, Shillong 793003, India. Email: smrutirekha walls (Fan and Luo 2008), bending stiffness of the soil nails are
[email protected] mostly considered. However, some researchers (Jewell and Pedley
2
Associate Professor, Dept. of Civil Engineering, Indian Institute of 1992; Liew and Khoo 2006) ignored the effects of bending and
Technology Delhi, Hauz Khas, New Delhi 110016, India (corresponding shear resistance of the soil nails during the analysis and design of
author). ORCID: https://orcid.org/0000-0003-0739-1035. Email: soil nail walls. In practice (FHWA 2003), this approach that ignores
[email protected] the effects of shear and bending resistances of soil nails is accepted
3
Emeritus Professor, Dept. of Civil Engineering, Indian Institute of when carrying out soil nailing analysis and design. The incorpora-
Technology Delhi, Hauz Khas, New Delhi 110016, India. Email:
tion of the bending stiffness of nails in the analysis and design of
[email protected]
Note. This manuscript was submitted on May 5, 2020; approved on
soil nailed slopes has a significant effect on the development of
November 18, 2020; published online on February 15, 2021. Discussion nail forces (Joshi 2003).
period open until July 15, 2021; separate discussions must be submitted It has been observed that (S. S. Liew, and C. H. Liong, “Two
for individual papers. This paper is part of the International Journal of Case Studies on Soil Nailed Slope Failures,” submitted, Int.
Geomechanics, © ASCE, ISSN 1532-3641. Conf.) after the failure of a nailed slope, most of the soil nails

© ASCE 04021036-1 Int. J. Geomech.

Int. J. Geomech., 2021, 21(4): 04021036

 remained intact at the failure surface whereas only the facing failed, 2. To develop the failure mechanism of the nailed soil slopes and
which aggravated the soil mass inbetween the nails to come out and delineate the effect of slope angle, nail inclination, nail length,
caused total slope failure. But, in practice, the provision of proper and slope facing on the behavior of model soil slopes under
facing is often ignored by only shotcreting the slope face or cover- the seismic loading condition.
ing the entire slope face by vegetation when designing nailed 3. To study the maximum lateral displacements at various heights
slopes, which substantially affects this nailed slope when it is sub- of the facing, maximum crest settlement, and acceleration am-
jected to seismic conditions. Although the role of bending stiffness plification at the crest for nailed soil slopes under seismic
of soil nails has been addressed in the literature (Juran et al. 1990; loading.
Fan and Luo 2008; Singh and Babu 2010), the effect of facing and 4. To examine the development of maximum nail forces and the
its connection with the nails under static and seismic conditions has variation of the nail forces along the nail length with time at var-
Downloaded from ascelibrary.org by Jaypee University Of Information Tech - Solan on 05/07/24. Copyright ASCE. For personal use only; all rights reserved.

not received significant attention from researchers. ious stages of seismic loading.
Large-scale field tests on virtually vertical nailed cuts, or slopes,
or both in cohesionless soil have been conducted (Stocker et al.
1979; Gassier and Gudehus 1981). Small-scale model tests on rein- Experimental Program
forced slopes have been conducted (Kitamura et al. 1988; Gutierrez
and Tatsuoka 1988) in which the tensile reinforcement forces and
developed strain were measured. The small-scale models were Uniaxial Shaking Table
used (Juran et al. 1988) to study the effect of the construction tech- A total of 30 shaking table tests were conducted on 0.4 m high
nique on the behavior of nailed soil structures. The behavior of steel model slopes to investigate the effects of parameters, such as
reinforced slope was investigated (Davis et al. 1993) by a series of slope angle, nail length, and nail inclination, and presence or ab-
model tests conducted in clayey sand. Model testing that uses a sence of a facing plate on the seismic behavior of nailed slopes.
shaking table is one of the limited methods to study the behavior An MTS uniaxial shaking table (MTS SYSTEMS, Eden Prairie,
of reinforced slopes under seismic conditions in the laboratory. A MN, USA) was used in this study. The shaking table had a horizon-
detailed report by Clough and Pirtz (1956) established shaking tal platform, one end was attached to the servohydraulic actuator
table studies for unreinforced slopes for the first time. Shaking that was vibrated in a single horizontal direction for harmonic as
table tests were carried out (Seed and Clough 1963; Goodman well as random vibration (artificial earthquake ground motion).
and Seed 1965; Arango and Seed 1974) to study the earthquake re- The shaking table had a high strength steel platform 0.9 × 0.9 m,
sistance of embankment slopes and dams. However, limited inves- a payload capacity of 9 kN, and a hydraulic actuator capacity of
tigations have been carried out using shaking table tests (Hong et al. 36 kN and 150 mm (±75 mm) stroke length. The shake table re-
2005; Giri and Sengupta 2009, 2010; Yazdandoust 2018, 2019) to sults are frequently the preferred data to understand the behavior
study the earthquake resistance of nailed soil slopes. In addition, of a structure under different ground motions. The shaking table
the details related to the development of nail forces under earth- used in this study had the facility to use displacement as input,
quake loading were limited, which is an important parameter to for example, the displacement history of the 1989 Loma Prieta
understand the seismic behavior of nailed soil slopes. However, earthquake. Therefore, the necessary mathematical steps were in-
the failure mechanism and the seismic resistance of nailed soil troduced to convert the acceleration–time history to displace-
slopes during a seismic event are not distinctly understood, and ment–time history. The MTS Test Star II M (digital controller
therefore, a comprehensive investigation is required. The perfor- model 493.10) was used to control the hydraulic actuator that
mance of reinforced soil structures under seismic condition is im- was connected to the shaking table by station manager software.
portant, especially in earthquake-prone zones. Due to a Multipurpose test ware of the station manager software was used
deficiency of monitored data from correctly instrumented model to generate earthquake excitation. The software gave the best re-
tests that can yield reliable results, it is essential to conduct this sults for peak/valley compensators (PVC) mode especially at
type of test in a laboratory to gain insight into the seismic behavior high frequency, and therefore, the PVC compensator was used in
of nailed slopes. The failure or distress reported in the literature this study. A PVC compensator monitors cyclic command feed-
(Bathurst 1997) due to earthquakes demonstrated how vulnerable back for any amplitude roll-off or mean level divergence.
nailed soil structures were to ground motions. An extensive de-
scription of the soil nailing technology and design methods has
been reported in Hussin et al. (1997). However, limited studies Instrumentation
are available that evaluate the seismic responses of nailed soil Three one-dimensional microelectromechanical system (MEMS)
slopes and the corresponding failure mechanism of nailed soil based accelerometers were used throughout the experiment that
slopes. had a measuring range of ±2 g and frequency range ≤200 Hz.
Therefore, in this study, a series shaking table tests are con- All three accelerometers were fixed at various positions onto the
ducted to study the seismic resistance and failure mechanism of model slopes to record the acceleration response in the direction
nailed soil slopes. This study will attempt to understand the effect of excitation of the shaking table. One accelerometer was fixed at
of slope angles, nail length, nail inclination, and the presence of the base of the shaking table and this reading was taken as the
facing on the seismic behavior of model slopes. The results for ac- input acceleration history throughout this study. The readings mea-
celeration response, facing displacement, failure pattern, and nail sured at the base of the shaking table gave the actual acceleration
behavior are presented comprehensively for the failure assessment values that were induced to excite the model slope. The second ac-
of nailed soil slopes under seismic loading conditions. celerometer was fixed to the frame of the model test box to examine
This study aims to understand the seismic behavior of nailed soil its acceleration response, because of the excitation at the base of the
slopes with the help of shaking table tests. The following are the shaking table. The third accelerometer was placed inside the soil at
specific objectives of this study: a depth of 0.5 m from the crest. The various locations of the accel-
1. To carry out laboratory shaking table tests to study the seismic erometers in the model slope and the data acquisition system that
behavior of nailed soil slopes under a particular seismic loading received the signals from all the accelerometers are shown in
condition. Fig. 1. Four linear variable differential transformers (LVDTs)

© ASCE 04021036-2 Int. J. Geomech.

Int. J. Geomech., 2021, 21(4): 04021036

Downloaded from ascelibrary.org by Jaypee University Of Information Tech - Solan on 05/07/24. Copyright ASCE. For personal use only; all rights reserved.

Fig. 1. Schematic diagram of the whole instrumental arrangement with the model slope.

were used to measure the facing displacement values at different el-

evations of facing and crest settlement due to earthquake excitation
that was provided at the base of the shaking table. The three LVDTs
were fixed to a heavy frame horizontally, and one LVDT was fixed
vertically using some metallic wires that were attached to them to
measure horizontal displacement and vertical displacement output,
respectively. The moving needle of the LVDTs was then attached
to the predefined locations of the model slope on the shaking table
and the cables from them were connected to the data acquisition
system. The LVDT used in this study was a plunger-type displace-
ment LVDT stroke length ±50 mm. Fig. 1 shows the various loca-
tions of the LVDTs on the model slope and the data acquisition
system that received the signals from the LVDTs.
The bonded metallic strain gauges that had full-bridge circuits
were used to evaluate the axial nail force in each nail. Fig. 2
shows the isometric view of the model slope with all the strain
gauges attached to the respective nails. The separate modules of
the accelerometers, strain gauges and LVDTs were connected to
the data logger for each respective output. The instrumentation
along with the details of the model slope are shown in Fig. 1.
The pictorial form of the complete shaking table test setup in the
model slope and data acquisition system that received the signals
is shown in Fig. 3. Fig. 2. Isometric view of the model slope with a neat detailing of the
strain gauges attached to nails.

Scaling of the Earthquake Time History

The earthquake motion considered for this study was the N00E The reports (Felio et al. 1990) on failure or distress in the
component of the1989 Loma Prieta earthquake that was recorded nailed soil slopes during the 1989 Loma Prieta earthquake con-
at Los Gatos presentation center (US Geological Survey 2002). firmed the vulnerability of nailed soil structures to ground mo-
The peak ground acceleration of the Loma Prieta earthquake mo- tions. Therefore, research is required on the nailed soil
tion was 0.57 g. This seismic time history was chosen for this structures under this type of seismic conditions to develop an
study, because as it is a well-known and standard seismic history understanding of their seismic behavior and to develop design
that was recorded in the past and was frequently a preferred seismic guidelines to make this type of structure more resilient to earth-
history for research purposes (Felio et al. 1990; Keefer 2000; quakes. Fig. 4(a) shows the acceleration–time history of the
Khazai and Sitar 2004; Wasowski et al. 2011). Loma Prieta Earthquake. The previous acceleration history

© ASCE 04021036-3 Int. J. Geomech.

Int. J. Geomech., 2021, 21(4): 04021036

Downloaded from ascelibrary.org by Jaypee University Of Information Tech - Solan on 05/07/24. Copyright ASCE. For personal use only; all rights reserved.

Fig. 3. Complete shaking table test setup.

(a)

(b)

(c)

Fig. 4. Scaling of the earthquake time histories of the 1989 Loma Prieta earthquake: (a) acceleration time history; (b) displacement time history; and
(c) Fourier transformation of acceleration history.

© ASCE 04021036-4 Int. J. Geomech.

Int. J. Geomech., 2021, 21(4): 04021036

 Table 1. Similitude law for shaking table tests of slopes
Parameter Scaling law Model slope Prototype slope
Slope height λ 0.4 m 8m
Unit weight of soil Field density could not be achieved in model testing 15.45 kN/m3 17 kN/m3
Nail length λ 0.2, 0.3, 0.4 m 4, 6, 8 m
Nail inclination Equal 0°, 15°, 30° 0°, 15°, 30°
Nail spacing Horizontal λ 0.225 m 4.5 m
Vertical 0.130 m 2.6 m
Nail/facing thickness λ 2 mm 40 mm
Area of cross section of nail λ2 50.265 mm2 20,106 mm2
Downloaded from ascelibrary.org by Jaypee University Of Information Tech - Solan on 05/07/24. Copyright ASCE. For personal use only; all rights reserved.

(Anail)
Nail material Practical testing considerations Aluminum pipe Steel bar
Facing material Practical testing considerations Aluminum plate Reinforced concrete
Elastic modulus of nail (Enail) λ 0.5 10.5 GPa 46.96 GPa

was numerically double integrated to produce the displacement

record. Fig. 4(b) shows the displacement–time history. Then,
the displacement data were filtered and applied to the shaking
table that was used to conduct the model tests. The aim of fil-
tering the displacement data was to obtain the displacement
maximums within the maximum stroke of the shake table actu-
ator without substantially changing the overall acceleration time
history before it was applied to the shaking table tests. The
moving average method (Elvin 2009) was selected for this pur-
pose, because it is a high-pass filtering method that filters real-
istic earthquake loading earthquake signals; therefore, they can
be applied by servo hydraulic test machines with limited dis-
placement capabilities. After application of the moving average
filter method for scaling, the displacement–time history was re- Fig. 5. Grain size distribution of Yamuna sand.
duced to the available actuator stroke length of 150 mm.
Fig. 4(b) shows the comparison between the original and the
scaled displacement time history of the earthquake. To estimate Material Properties
the acceleration–time history that corresponded to the scaled-
Soil Material
down displacement–time history shown in Fig. 4(b), the shak-
Yamuna sand was used as the earth fill material. The engineering
ing table ran without fixing the model test tank onto it. After
properties of the sand, such as grain size distribution, maximum
applying the scaled displacement–time history as input to the and minimum dry unit weight, and shear parameters were deter-
shaking table, the acceleration–time history was recorded at mined from laboratory tests. Fig. 5 shows the grain size distribution
its base and was plotted, as shown in Fig. 4(a). Therefore, it of Yamuna sand that was classified as poorly graded sand (SP) ac-
has was confirmed that the double derivative (i.e., the acceler- cording to the Unified Soil Classification System. The sand was
ation–time history) of the supplied input displacement–time fine and was composed of quartz and mica flakes. The properties
history was not lost. Fig. 4(c) shows the comparison with the of the soil used for the preparation of model slope are presented
Fourier transformation between the original and the scaled- in Table 2. During the shaking table tests on model slopes, the
down acceleration history of the earthquake. The Fourier trans- bulk unit weight and relative density (Dr) of the sand were main-
formation of an earthquake for frequency–time information is tained at 15.45 kN/m3 and 65%, respectively using a controlled
useful to understand the respective earthquake motion. The volume compaction method.
maximum values of the predominant frequencies of the original
and scaled-down earthquake time–history were 1.56 and Nail and Facing Plate
1.57 Hz, respectively. Hollow aluminum pipe was used as the nail material in this study.
The outer diameter and thickness of the soil nails were 10 and
2 mm, respectively. Aluminum plate 2 mm thick was used as the
Similitude Law of Model Tests
facing plate. A slope facing usually provides slope surface protec-
A series of large-scale 1 g model tests were conducted to assess tion, minimizes erosion, and other adverse effects on the slope. A
the effects of various nail and slope combinations on the seis- slope facing might be soft, flexible, hard, or a combination of all
mic response of nailed soil slopes. The model nailed slopes three (CIRIA 2005). Therefore, an aluminum plate 2 mm thick
were scaled-down appropriately using scaling law (Iai 1989) used as a facing plate could represent a flexible and hard combina-
to predict the behavior of the prototype nailed soil slopes tion type slope facing. The elastic modulus and ultimate yield
from the measured responses in the model testing. The scaling strength of a nail specimen were determined from uniaxial tests,
factors are listed in Table 1 for the model and prototype that and 0.3 m long aluminum nail specimens were tested in the labora-
had the same soil density. For example, the ratio of the proto- tory and the properties are tabulated in Table 2.
type slope height to the model slope height (λ) was 20 for the
0.4 m high model slope to simulate an actual slope of 8.0 m. Soil–Nail Interface
The characteristics of the prototype slope compared with the The properties of the soil and nail interface were found by conduct-
model slope is presented in Table 1. ing direct shear tests. A 2 mm thick aluminum plate, which had the

© ASCE 04021036-5 Int. J. Geomech.

Int. J. Geomech., 2021, 21(4): 04021036

 Table 2. Material properties
Properties Value
Soil properties
Material and grain size distribution Yamuna sand and SP
Maximum and minimum unit weight (γmax and γmin) 17.5 and 13.1 kN/m3
ϕn 33°
Cohesion (c) 0.5 kPa
Young’s modulus of elasticity (E) 9.5 MPa
Poisson’s ratio (ν) 0.3
Downloaded from ascelibrary.org by Jaypee University Of Information Tech - Solan on 05/07/24. Copyright ASCE. For personal use only; all rights reserved.

Nail (reinforcement) properties

Material Aluminum tube
Unit weight (γ) 21.03 kN/m3
Elastic modulus [E (at 2% strain)] and Poisson’s ratio 10.5 GPa and 0.24
Yield strength (Ty) 222.4 MPa
Interface properties between soil and nail
Ks 90,000 kN/m2/m
Kn 4,000,000 kN/m2/m
ϕn between soil and nail 30°

properties of the nail tubes given in Table 2, with sand particles

glued to one side of the plate surface was placed onto the lower
half of the direct shear box. This was carried out because, during
actual model testing, the sand particles were glued to the nails.
The upper half of the direct shear test box was filled with Yamuna
sand that had a Dr of 65%. Before conducting the direct shear test it
was ensured that the rough surface of the aluminum plate glued
with sand particles met the soil in the upper half of the box at the
junction of the two halves of the shear box. The interface properties
for shear stiffness modulus (Ks), normal stiffness modulus (Kn), and
angle of internal friction (ϕn) are tabulated in Table 1. The interface
adhesion (ca) was zero.

Pull-Out Capacity of Nails

The pull-out capacity of the soil–nail or the interface shear strength
behavior of the soil–nail (i.e., aluminum tube) was determined from
laboratory pull-out tests. An important parameter that governs the
safety assessment of soil nailing is the ultimate shear strength at
the interface between the soil nails and the surrounding soil. Nail Fig. 6. Pull-out force versus nail displacement curves.
specimens with an outer diameter of 10 mm and length 0.3 m
were used for the pull-out test. The Dr of 65% was maintained in-
side the pull-out test boxes by filling them using a controlled vol-
seismic behavior of the reduced scale model. The presence of an
ume compaction method. An aluminum nail was placed in the
artificial boundary on the nailed soil slopes is an important consid-
compacted sand from a hole on the side of a container box. A pull-
eration for the seismic studies. The boundary was rigid in this
out load was then applied through a pulley arrangement at the other
end of the nail. The pull-out capacity of the nail was interpreted study, because as the container was made from Perspex sheets,
from the pull-out force versus displacement curve and was 0.010 steel angles, and bars. The rigidity of the container walls could
and 0.013 kN for a 0.3 m long nail at the overburden pressures cause reflection of seismic waves, and therefore, could be respon-
of 2.2 and 4.2 kPa, respectively as shown in Fig. 6. The nails sible for many errors in the output information because of the de-
were subjected to these overburden pressures at the level of the velopment of unreal intensification of the whole system. From a
slope where the nails were inserted. numerical investigation (Bathurst and Hatami 1998) the far-field
boundary conditions could have a significant effect on the response
of a reinforced soil system. From the literature (Bhattacharya et al.
Model Testing Program and Procedure 2011), wave reflections could be averted by lining the container
A number of researchers (Tufenkjian and Vucetic 2000; Hong et al. walls with an appropriate absorptive material. To reduce the reflec-
2005; Giri and Sengupta 2009, 2010; Wang et al. 2010; Yazdandoust tion of waves from the rigid boundaries, a 50 mm thick conven-
2018, 2019) considered various slope heights to prepare physical tional foam sheet was provided at the bottom of the container
models when conducting model tests on nailed soil slopes. It is and between the back of the models and the container wall in the
well-known that the height of a traditional soil–nail wall is from direction of seismic excitation, as shown in Figs. 1 and 3.
3.0 to 14.0 m, with an average height of 8.0 m. The height of the A total of 30 tests on 0.4 m high model slopes that used a shak-
model slopes and foundation soil were 0.4 and 0.2 m, respectively ing table were performed on nailed soil slopes to investigate the ef-
and considered the payload capacity of the shaking table. As given fects of various parameters on their performance under seismic
in Table 1, the λ was 0.4 m, which manifests the prototype slope conditions. The reinforced model soil slopes were divided into
height of 8.0 m that has a scale factor of 20, to reflect realistic two types: (1) a set of reinforced model soil slopes provided

© ASCE 04021036-6 Int. J. Geomech.

Int. J. Geomech., 2021, 21(4): 04021036

Downloaded from ascelibrary.org by Jaypee University Of Information Tech - Solan on 05/07/24. Copyright ASCE. For personal use only; all rights reserved.

(a) (b)

Fig. 7. Front view of a prepared reinforced model soil slope: (a) without facing plate; and (b) with facing plate.

without facing plate that are depicted as model nailed soil slopes nails) without and with facing were tested for three different
without facing as shown in Fig. 7(a); and (2) a set of reinforced slope angles, such as 30°, 45°, and 60°.
model soil slopes with a single 0.002 m thick aluminum sheet fac- The parameters considered for this study for both the nailed soil
ing plate, which is depicted as model nailed soil slopes with facing, slopes (i.e., without and with facing) consisted of a combination of
as shown in Fig. 7(b). The model slopes without facing had six nails three different slope angles (i.e., β = 30°, 45°, and 60°), which were
that were attached to six smaller isolated head plates at the slope used for three different nail inclinations (i.e., i = 0°, 15°, and 30°),
face. The model slopes with facing had six nails that were attached and three different nail lengths (i.e., l = 0.2, 0.3, and 0.4 m). The
to a continuous facing plate that covered the entire slope face. shaking table test results were obtained for: (1) lateral displace-
A summary of the sequence of soil nailing technique and asso- ments of slopes at different elevations of slope facing; (2) acceler-
ciated parameters follows: ation amplification at the slope crest due to the excitation provided
First, the steep soil slope of appropriate (predefined) slope angle at the base of the model slopes; and (3) crest settlement for all
was constructed in the rigid box using a controlled volume compac- model slopes. The axial forces that developed throughout the nail
tion method to maintain a Dr of approximately 65%. Then, the length, as well as the variation of nail forces developed with the
model soil slopes were prepared by inserting aluminum nails into progress of earthquake load, were determined from the shaking
the predefined locations of the prepared soil slopes. The nail align- table tests on the model slopes. The seismic resistance and failure
ments for the slope facing for each of the model slopes were mechanism of the model soil slopes without and with facing
achieved using a protractor and measuring scale. The smaller were studied in detail. A summary of test results for all slope mod-
sized square (0.04 × 0.04 × 0.002 m) aluminum plates at the slope els for acceleration amplification at the crest, maximum crest settle-
ment, and average facing displacement is reported in Table 3.
face that acted as nail heads termed isolated head plates were
The acceleration amplification factor is the ratio of measured
used for a nailed slope without facing. The nailed slope without fac-
peak response acceleration at the crest of the model slopes to the
ing was provided with six nails that were attached to six smaller
peak input acceleration (measured from the accelerometer attached
isolated head plates at the slope face. The front ends of the nails
at the base). The maximum acceleration amplification at the crest
were attached to the isolated head plates using bolts whose external
was 1.369 for Model 29 (β = 60°, l = 0.3 m, and i = 15° without fac-
threads were fastened to the internal threads that were at the inside ing). The minimum acceleration amplification at the crest was
part of the aluminum tubes (i.e., nails). Therefore, this ensured 1.077 for Model 16 (β = 30°, l = 0.2 m, and i = 0° without facing).
that no failure was allowed between the front end of the nail and The measured facing displacement of all model slopes was inves-
the nail head. tigated and reported in Table 3 for the ratio of average facing dis-
Then, the whole instrumental arrangement was applied to the placement to slope height. The maximum value of the ratio of
model slope before exciting it on the shaking table. The model average facing displacement to slope height was 0.307 for Model
slopes were built in a rigid box length 0.9 m, width 0.9 m, and 11 (β = 60°, l = 0.2 m, and i = 0° with facing) and the minimum
height 0.9 m that was made out of Perspex sheets and steel angles value was 0.207 for Model 23 (β = 45°, l = 0.4 m, and i = 0° with-
and bars. The base of the rigid box was made from a steel plate that out facing). The measured maximum crest settlement for all model
was provided with a nut and bolt facility to ensure that the box slopes is given in Table 3. The maximum and minimum value of
could be perfectly connected to the base of the shaking table. crest settlement were 0.120 and 0.100 m for Model 11 (β = 60°,
Yamuna sand was used as the backfill material filled in the rigid l = 0.2 m, and i = 0° with facing) and Model 28 (β = 60°,
box that was fixed to the base of the shaking table. The sand was l = 0.4 m, i = 0° without facing), respectively.
poured through a screen hopper from a fixed falling height of
1 m to achieve a uniform distribution of sand. The controlled vol-
ume compaction method was used to achieve a Dr of approximately Results and Discussion
65%. Therefore, the steep soil slope of an appropriate (predefined)
slope angle was constructed in the rigid box using the controlled The seismic performance of all the model slopes given in Table 3
volume compaction method. To measure the development of the was studied from the model testing results for acceleration re-
slope failure surface during the shaking table tests, thin horizontal sponse, facing displacement at various heights of the model
layers of Yamuna sand that was dyed pink were placed at regular slope, failure pattern, and nail behavior. The effect of various pa-
vertical intervals. Both sets of reinforced soil slopes (that used rameters, such as the effect of reinforcement, slope angle, nail

© ASCE 04021036-7 Int. J. Geomech.

Int. J. Geomech., 2021, 21(4): 04021036

 Table 3. Summary of model test results
Model Acceleration Ratio of average facing Maximum crest
number Model name β (°) l (m) i (°) f amplification at crest displacement to slope height settlement (m)
1 [β1l1i1f1] 30 0.2 0 with 1.113 0.276 0.112
2 [β1l2i1f1] 30 0.3 0 with 1.111 0.258 0.109
3 [β1l3i1f1] 30 0.4 0 with 1.107 0.256 0.108
4 [β1l2i2f1] 30 0.3 15 with 1.113 0.257 0.110
5 [β1l2i3f1] 30 0.3 30 with 1.115 0.247 0.112
6 [β2l1i1f1] 45 0.2 0 with 1.114 0.283 0.114
7 [β2l2i1f1] 45 0.3 0 with 1.110 0.260 0.110
Downloaded from ascelibrary.org by Jaypee University Of Information Tech - Solan on 05/07/24. Copyright ASCE. For personal use only; all rights reserved.

8 [β2l3i1f1] 45 0.4 0 with 1.106 0.247 0.108

9 [β2l2i2f1] 45 0.3 15 with 1.112 0.257 0.109
10 [β2l2i3f1] 45 0.3 30 with 1.114 0.250 0.113
11 [β3l1i1f1] 60 0.2 0 with 1.115 0.307 0.120
12 [β3l2i1f1] 60 0.3 0 with 1.106 0.277 0.110
13 [β3l3i1f1] 60 0.4 0 with 1.110 0.257 0.109
14 [β3l2i2f1] 60 0.3 15 with 1.108 0.237 0.111
15 [β3l2i3f1] 60 0.3 30 with 1.109 0.283 0.119
16 [β1l1i1f2] 30 0.2 0 without 1.077 0.231 0.110
17 [β1l2i1f2] 30 0.3 0 without 1.078 0.230 0.106
18 [β1l3i1f2] 30 0.4 0 without 1.079 0.227 0.105
19 [β1l2i2f2] 30 0.3 15 without 1.078 0.215 0.109
20 [β1l2i3f2] 30 0.3 30 without 1.079 0.218 0.103
21 [β2l1i1f2] 45 0.2 0 without 1.089 0.237 0.113
22 [β2l2i1f2] 45 0.3 0 without 1.090 0.227 0.107
23 [β2l3i1f2] 45 0.4 0 without 1.086 0.207 0.104
24 [β2l2i2f2] 45 0.3 15 without 1.090 0.220 0.108
25 [β2l2i3f2] 45 0.3 30 without 1.089 0.213 0.106
26 [β3l1i1f2] 60 0.2 0 without 1.098 0.230 0.105
27 [β3l2i1f2] 60 0.3 0 without 1.109 0.217 0.104
28 [β3l3i1f2] 60 0.4 0 without 1.310 0.210 0.100
29 [β3l2i2f2] 60 0.3 15 without 1.369 0.270 0.110
30 [β3l2i3f2] 60 0.3 30 without 1.105 0.217 0.102
Notes: β = slope angle (β1 = 30°, β2 = 45°, and β3 = 60°); f = facing plate (f1 = with facing; f2 = without facing); i = nail inclination (i1 = 0°, i2 = 15°, and i3 =
30°); and l = nail length (l1 = 0.2 m, l2 = 0.3 m, and l3 = 0.4 m).
The bold values represents the maximum and minimum values of result parameters.

length, nail inclination, and slope facing was investigated. Research is, [β3l2i1f1] respectively. It can be seen in Figs. 8(a and b) that the
was undertaken by Sahoo et al. (2016) on shaking table tests on un- maximum acceleration values at crest were higher than the input ac-
reinforced slopes and nailed soil slopes both had a slope angle of celeration history measured at the base of the shaking table. This
45° and slope height of 0.4 m and used the same setup. The nail was evident from Table 3, because the acceleration amplification
length and nail inclination of the reinforced slope were 0.3 m and factor was >1 for all the model slopes. The peak acceleration at
0°, respectively and the remaining parameters along with the load- the crest of Model 27 and Model 12 was obtained straight
ing condition were the same as in this study. The acceleration am- after the occurrence of peak input acceleration. As given in Table 3,
plification at the crest of the unreinforced slope was 1.214 and for the maximum acceleration and minimum acceleration at crest val-
the nailed slope was 1.110. The crest settlement was 23.1 and ues were found for Model 29 (β = 60°, l = 0.3 m, and i = 15° with-
15.8 mm for unreinforced and nailed soil slopes, respectively at out facing) and Model 16 (β = 30°, l = 0.2 m, and i = 0° without
the end of the seismic sequence. The ratio of the average facing dis- facing), respectively. The acceleration values at the crest were
placement to slope height was 0.253 and 0.26 mm for unreinforced higher for slopes with facing compared with the same slope without
and nailed soil slopes, respectively. It was concluded that the soil facing. The effect of other parameters, such as slope angle, nail
nails could effectively improve the seismic stability of a 45° length, nail inclination, and facing on acceleration amplification
slope. A combination of translation with rocking was found in at the crest is given in Table 3.
nailed slopes and higher translational movement was found in un-
reinforced slopes. In addition, soil nails reduce the acceleration am-
plification at the crest of the slope. Facing Displacement
The typical normalized accumulated displacements of the slope
facing with elevation are shown in Figs. 9(a–d) to study the effect
Acceleration Response
of slope angle, nail length, nail inclination, and facing, respectively.
To compare the experimental and predicted acceleration response The ratio of facing displacement to slope height accumulated at the
in 30 model slopes, 1 model slope was considered from each of end of the seismic excitation of the model slopes was defined as the
the sets, for example, Model 27 from the model slopes without fac- normalized accumulated displacement of the facing and is shown in
ing and Model 12 from the model slopes with facing. The typical these Figs. 9(a–d) obtained from the model tests. The facing dis-
input acceleration at the base (measured from an accelerometer at- placement of the 30 model slopes obtained from model testing is
tached at the base) and the acceleration history at crest are shown in reported in Table 3 for the ratio of average facing displacement
Figs. 8(a and b) for Model 27, that is, [β3l2i1f2] and Model 12, that to slope height.

© ASCE 04021036-8 Int. J. Geomech.

Int. J. Geomech., 2021, 21(4): 04021036

Downloaded from ascelibrary.org by Jaypee University Of Information Tech - Solan on 05/07/24. Copyright ASCE. For personal use only; all rights reserved.

(a) (b)

Fig. 8. Response acceleration history in the soil mass near crest: (a) Model 27; and (b) Model 12.

(a) (b)

(c) (d)

Fig. 9. Normalized accumulated displacements of the slope facing to study the effect of (a) slope angle; (b) nail length; (c) nail inclination; and
(d) facing.

© ASCE 04021036-9 Int. J. Geomech.

Int. J. Geomech., 2021, 21(4): 04021036

 Effect of Slope Angle than the horizontal and 30° inclined nails when stabilizing steep
The effect of slope angle are shown in Fig. 9(a) by considering soil slopes under seismic conditions, which confirmed the findings
Model 2 (β = 30°, l = 0.3 m, and i = 0°), Model 7 (β = 45°, l = in the literature (Lin et al. 2013). Therefore, it could be beneficial to
0.3 m, and i = 0°), and Model 12 (β = 60°, l = 0.3 m, and i = 0°) provide a small inclination to the nails when reinforcing a steep soil
where the facing displacement with slope height is plotted from slope to minimize the seismic instability of the slope.
the test results. From the test results, the predominant displacement
mechanism in Model 2 and Model 7 was translation. The outward Effect of Facing
convex tendency found in Model 12 was evidence of a combination The effect of facings was examined by considering Model 27 (β =
of translational and a small rocking movement. It is noted that the 60°, l = 0.3 m, and i = 0° without a facing plate) and Model 12 (β =
magnitude of facing displacement at all the measured points in- 60°, l = 0.3 m, and i = 0° with facing plate) as shown in Fig. 9(d),
Downloaded from ascelibrary.org by Jaypee University Of Information Tech - Solan on 05/07/24. Copyright ASCE. For personal use only; all rights reserved.

creased with the increase in slope angle for all the reinforced slopes. where the facing displacement with slope height is plotted. From
However, a small variation in the average magnitude and pattern of the test results, the predominant displacement mechanism in both
facing displacement was observed for the model nailed slope with a model slopes was translation with similar movement pattern. It
slope angle of 45° and 60°. From the results given in Table 3, it was was observed that the displacement mechanism in all the model
concluded that the seismic instability of slopes increased with an slopes was a combination of a translation and rocking movement,
increase in slope angle. This observation was true for both the because of the outward convex tendency that was found in both.
sets of reinforced slopes, for example, without and with facing In addition, from the plot, it was observed that the magnitude of
conditions. the facing displacement at all the measured points for Model 12
(with facing) was higher than Model 27 (without facing). The fac-
Effect of Nail Length ing displacements were high in all the model slopes with facing,
In this study, the nail lengths were from 0.2 m (i.e., 50% of the because the facing plate detached with the nail in most of the
slope height) to 0.4 m (i.e., 100% of the slope height). The effect cases due to its weight and seismic excitation. However, this be-
of nail length was investigated by considering Model 11 (β = havior was not found for model slopes without facing. The mag-
60°, l = 0.2 m, and i = 0°), Model 12 (β = 60°, l = 0.3 m, and i = nitude of facing displacement at all the measured points from the
0°), and Model 13 (β = 60°, l = 0.4 m, and i = 0°) as shown in test results was obtained from the readings of the LVDTs whose
Fig. 9(b) where the facing displacement with slope height are plot- tips were attached to the facing plate. Therefore, the detachment
ted. All the slope models that were considered were provided with a of the facing plate from the soil mass compelled a higher magni-
facing plate. From the test results, it was observed that the predom- tude of facing displacement at all the measured points, although
inant displacement mechanism in all the model slopes was transla- the movement of actual soil mass of the model slope (Model
tion. The displacement mechanism was a combination of 12) was much less than the movement of the facing plate. There-
translation and rocking movement, because of the outward con- fore, the higher facing displacement of model slopes with facing
vex tendency in all model slopes. In addition, it was observed plates did not reflect the true representation of displacement of
that the magnitude of facing displacement at all the measured slopes under practical condition. As an excavation proceeds, the
points decreased with an increase in nail length. The reason shotcrete, concrete, or other grouting materials are applied onto
for these phenomena could be a larger embedment length of the excavation face to grout the nails to improve the seismic
the reinforcement in the soil for longer nails that resist the soil stability for steep soil slopes.
movement effectively under seismic loading conditions. How-
ever, no significant difference in magnitude of the facing dis-
placement was found in the lower rows of nails in Model 12 Failure Pattern
(l = 0.3 m) and Model 13 (l = 0.4 m). Therefore, it was con- The development of the sliding surface and the crest settlement
cluded that the seismic stability of nailed slopes could be were monitored during the tests. A digital camera was utilized to
achieved to its maximum when the reinforcement length was ap- record the deformations of the slopes during the shaking table
proximately 75% of the slope height (i.e., l = 0.3 m). The previ- tests. At the end of each test, the development of cracks and defor-
ous observations were valid in all the model nailed soil slopes mation of the slopes were surveyed accurately. The failure pattern
that were considered in this study. of two model slopes, for example, Model 27 (β = 60°, l = 0.3 m,
and i = 0° without facing) and Model 12 (β = 60°, l = 0.3 m, and
Effect of Nail Inclination i = 0° with facing) are presented that typically represent the failure
Fig. 9(c) shows the plot of the facing displacement with slope modes of all the model nailed soil slopes without facing and with
height that was obtained for Model 12 (β = 60°, l = 0.3 m, and facing, respectively. The side and top views of Model 27 [β2l2i1f2]
i = 0°), Model 14 (β = 60°, l = 0.3 m, and i = 15°), and Model 15 and Model 12 [β2l2i1f1] after failure are shown in Figs. 10 and 11,
(β = 60°, l = 0.3 m, and i = 30°) to assess the effect of nail inclina- respectively. The solid lines and the dashed lines in the side views
tion. From the test results, the predominant displacement mecha- of these figures represent the undeformed shape and the potential
nism in Model 12 was translation; however, the outward concave failure surface, respectively for the chosen model slopes. The dot-
tendency found in Models 14 and 15 indicated a translational and ted lines in the side views represent the deformed shape and the dot-
rocking movement. It was observed that the displacement mecha- ted lines in the top views represent the hairline cracks of the chosen
nism in all the model slopes was a combination of a translation model slopes at the end of the seismic excitation. It can be seen
and rocking movement, because of the outward convex tendency from the side views of these models (Fig. 10) that bulging and
found in all the model’s slopes. In addition, it was observed that the translational movements at the toe in the outward direction (di-
the magnitude of facing displacement at all the measured points rection away from the slope) were visible. The top views of failed
of Model 14 (i = 15°) was less than the face displacement of Mod- slopes with visible cracks of these models are shown in Fig. 11. It is
els 12 (i = 0°) and 15 (i = 30°). The facing displacement of slopes noted that from the top views of these models the sliding mass of
that had nail inclinations of 0° and 30° (Models 12 and 15) were soil was prominently visible for the reinforced model slope without
approximately the same. Therefore, it was concluded that a small facing with a rotational failure type. However, for the reinforced
nail inclination of 15° had a more efficient reinforcement action model slope with facing, although the potential failure curve was

© ASCE 04021036-10 Int. J. Geomech.

Int. J. Geomech., 2021, 21(4): 04021036

Downloaded from ascelibrary.org by Jaypee University Of Information Tech - Solan on 05/07/24. Copyright ASCE. For personal use only; all rights reserved.

(a) (b)

Fig. 10. Side view of the failure surface: (a) Model 27; and (b) Model 12.

(a) (b)

Fig. 11. Top view of the failure surface: (a) Model 27; and (b) Model 12.

identified no such sliding soil mass was found. The failure surfaces Soil Nails Behavior
for all the cases were shallow and of a rotational type similar to the
Development of Axial Nail Force Along Nail Length
reported literature (NCREE 1999). The development of the sliding
Results for the variation of maximum axial nail force developed
surface and separation of the soil body near to the crest was very along the length of each nail for Model 27 (β = 60°, l = 0.3 m,
prominent in both cases. The failure surfaces were circular in the and i = 0° without facing plate) and Model 12 (β = 60°, l = 0.3 m,
upper part of the slopes. At the toe and lower portion of the slopes, and i = 0° with facing plate) are shown in Figs. 12(a and b), respec-
bulging and translational movements were evident in both cases. tively. The development of maximum axial nail force for the model
This was consistent with the slope failures reported in previous slope without facing was found at the slope facing itself and dimin-
studies (Khazai and Sitar 2004; Wartman et al. 2005). It was ob- ished steadily to null toward the end. The development of maxi-
served that the deformations were not uniform across the lateral mum axial nail force for the model slope with facing was
width of the model slopes. In all cases, a slightly concave deforma- obtained at approximately two-thirds of the nail length that started
tion profile was observed at the crest. In addition, several hairline from the slope facing and had a parabolic distribution pattern of nail
tension cracks at the crest were observed. From these failure forces along each of the nails. In addition, it was observed that the
views, the reinforced slope with facing resisted the possible slope maximum axial nail force developed in the model slope without
failure, or any possible substantial deformation of slope facing, facing was approximately eight times more than that the model
or both although there were some hairline cracks at its crest that slope with facing. For the reinforced slope with a facing plate,
did not cause any seismic hazard. The results indicate that the the axial forces that developed in the nails were transferred to the
development of the sliding surface and separation of the soil facing plate that was attached to them, and therefore, the developed
body near the crest was more prominent than in the other parts of maximum axial nail force was lower. This observation was due to
the slopes in both cases. The failure surface was developed near the bending stiffness of nails and the facing plate in combination. It
the midwidth of the slope and appeared to be deeper for steep is noted that the nail alignments were in the same direction of seis-
slopes, which is consistent with the field observations of mic excitation. As shown in Fig. 12(a), near the facing, that is, at
earthquake-induced landslides (Khazai and Sitar 2004). the nail head, the nail forces exceeded significantly the pull-out

© ASCE 04021036-11 Int. J. Geomech.

Int. J. Geomech., 2021, 21(4): 04021036

Downloaded from ascelibrary.org by Jaypee University Of Information Tech - Solan on 05/07/24. Copyright ASCE. For personal use only; all rights reserved.

(a) (b)

Fig. 12. Variation of axial nail force that developed along the nail: (a) Model 27; and (b) Model 12.

(a) (b)

Fig. 13. Variation of axial nail force with time (s) for (a) N2 nail; and (b) N5 nail at both their ends for Model 27.

(a)
(b)

Fig. 14. Variation of axial nail force with time (s) for (a) N2 nail; and (b) N5 nail at both their ends for Model 12.

capacity of the nails in the upper and lower rows of nails. Hence, it connection for the nails with facing the seismic performance of
was concluded that pull-out failure happened for the nailed slopes nailed slopes could be significantly improved.
without facing under seismic conditions. But, from Fig. 12(b), it
can be seen that the nail forces mobilized near the slope face Development of Axial Nail Force with Earthquake Duration
were within and much lower than the pull-out capacity of the The variation in axial nail force with time for Model 27 (β = 60°, l =
same nails, although the maximum mobilized nail forces found at 0.3 m, and i = 0° without facing plate) and Model 12 (β = 60°, l =
approximately two-thirds of the nail length that started from the 0.3 m, and i = 0° with facing plate) from the test results are
slope facing were approximately equal to their pull-out capacity. shown in Figs. 13 and 14, respectively. In these figures, the typical
It was concluded that by providing a proper facing and a good plots at two strain pick-up points (i.e., one near the slope face and

© ASCE 04021036-12 Int. J. Geomech.

Int. J. Geomech., 2021, 21(4): 04021036

 the other at the further end of the nail) of N1 and N5 nails in Fig. 2 facing displacement were not significantly different for the
are shown. It can be seen from Figs. 13 and 14 that the nail force 45° and 60° nailed slopes.
versus time curves obtained from the test results followed a trend 4. The magnitude of facing displacement decreased with an in-
similar to that of the input acceleration–time history. For the crease in nail length and the seismic stability of nailed slopes
model slopes without facing (Model 27), the higher magnitude of could be achieved to its maximum when the reinforcement
axial nail force was concentrated near the slope facing itself and length is approximately75% of the slope height.
it diminished toward the further end of the nail from the slope 5. A nail inclination of 15° provided more efficient reinforcement
face as shown in Fig. 13. This was because the nails were con- action than 0° and 30° inclined nails to stabilize soil slopes
nected to the isolated head plates and there was no possibility of under seismic conditions.
nail force dispersion as in the model slopes with facing. Therefore, 6. The reinforced model slopes with facing provided better protec-
Downloaded from ascelibrary.org by Jaypee University Of Information Tech - Solan on 05/07/24. Copyright ASCE. For personal use only; all rights reserved.

more force was recorded near the front end of the nail. However, tion against seismic instability than the reinforced model slopes
for model slopes with facing (Model 12), the higher magnitude without facing, because no such collapsible sliding soil mass
of axial nail force was obtained at the further end of the nails. was found in the case with facing compared with the case with-
This behavior of nails could be due to the dispersion of the devel- out facing. The facing plate for the reinforced model slopes with
oped nail forces throughout the entire facing plate that the nails facing provided a containment effect that limited the deforma-
were attached to, as shown Fig. 12(b). Therefore, lower nail tion near the slope surface and therefore could be beneficial in
force was observed near the front end of the nail for the model providing stability.
slope with facing. It can be seen from Fig. 13 that the development 7. The development of maximum axial nail force for each of the
of maximum axial nail force in the nails of model slopes without nails in the model slopes with facing was obtained at approxi-
facing obtained for the lower rows was approximately 1.5 times mately two-thirds of the nail length from the slope face and
higher in magnitude than the upper rows of nails for the entire du- had a parabolic distribution pattern of nail forces along each
ration of earthquake time history. This phenomenon was attributed of the nails; however, for the nails in model slopes without fac-
to the lower soil–nail interface frictional force that was due to lower ing, this was obtained at the slope facing itself and diminished
overburden pressure on the upper rows of the nails installed into the steadily toward the end irrespective of any slope and nail
slope than that of the lower rows. However, there was no such dif- arrangements.
ference in the magnitude of nail force with depth that was obtained
for the model slope with facing, as shown in Fig. 14. The presence
of facing would have resulted in the same nail force in the lower
Data Availability Statement
and upper rows of nails in the model slopes.
All data, models, and codes generated or used during the study ap-
Discussion of Nail Force Distribution pear in the published article.
It was observed for nailed soil slopes that, if a proper facing and a
good connections/linkage of nails to it was not provided, soil nails
could not mobilize its bending stiffness under seismic condition.
Hence, pull-out failure combined with facing failure was responsi- Acknowledgments
ble for the failure of the entire slope in this type of situations, be-
cause the actual purpose of providing soil nails to strengthen the This study was funded by the Department of Science & Technol-
slope against seismic instability was not achieved. From this ogy (DST), Ministry of Science & Technology, Government of
study, the continuous redistribution of seismic forces between India under DST Sanction No.: SR/FTP/ETA-0102/2011.
soil nails and the slope facing through the nail-facing connection
occurred during the seismic event. Therefore, the provision of a
flexible structural slope facing could enhance the seismic stability References
of a soil nailed system. Structural beams, such as nails and grillages
combined with a steel wire mesh (to provide containment effect) as Arango, I., and H. B. Seed. 1974. “Seismic stability and deformation of
the facing might be constructed on the slope surface to connect the clay slopes.” J. Geotech. Eng. Div. 100 (2): 139–156.
soil nail heads to promote the integral action of the soil nailed sys- Babu, G. L. S., and V. P. Singh. 2008. “Numerical analysis of performance
tem under seismic conditions. of soil nail walls in seismic conditions.” ISET J. Earthquake Technol.
45 (1–2): 31–40.
Bathurst, R. J. 1997. “Review of seismic design, analysis and performance
of geosynthetic reinforced walls, slopes and embankments.” In Earth
Conclusions reinforcement, 887–918. Rotterdam, The Netherlands: A.A. Balkema.
Bathurst, R. J., and K. Hatami. 1998. “Seismic response analysis of a
Some conclusions based on the results obtained from this study are: geosynthetic-reinforced soil retaining wall.” Geosynth. Int. 5 (1–2):
1. Translation and rocking movement combined were found in the 127–166. https://doi.org/10.1680/gein.5.0117.
model slopes with facing and a combination with predominant Bhattacharya, S., D. Lombardi, L. Dihoru, M. Dietz, A. J. Crewe, and C. A.
translation and mild rocking type movement was found in Taylor. 2011. “Model container design for soil-structure interaction
model slopes without facing. studies.” In Vol. 22 of Role of seismic testing facilities in performance-
2. The maximum acceleration values at the crest were higher than based earthquake engineering. Geotechnical, Geological and
the input acceleration at the base for all model slopes. The max- Earthquake Engineering book series, edited by M. N. Fardis, and
Z. T. Rakicevic, 135–158. Dordrecht, Netherlands: Springer.
imum and minimum acceleration amplification at the crest were
Chu, L. M., and J. H. Yin. 2005. “Comparison of interface shear strength of
1.369 and 1.077 for Model 29 (β = 60°, l = 0.3 m, and i = 15° soil nails measured by both direct shear box and pullout tests.”
without facing) and Model 16 (β = 30°, l = 0.2 m, and i = 0° J. Geotech. Geoenviron. Eng. 131 (9): 1097–1107. https://doi.org/10
without facing), respectively. .1061/(ASCE)1090-0241(2005)131:9(1097).
3. The seismic instability of slopes increased with an increase in CIRIA (Construction Industry Research & Information Association). 2005.
slope angle. However, the average magnitude and pattern of Soil nailing—Best practice guidance. Rep. No. C637. London: CIRIA.

© ASCE 04021036-13 Int. J. Geomech.

Int. J. Geomech., 2021, 21(4): 04021036

 Clough, R. W., and D. Pirtz. 1956. “Earthquake resistance of rock-fill Kitamura, T., A. Nagao, and S. Uehara. 1988. “Model loading tests of re-
dams.” J. Soil Mech. Found. Div. 82 (2): 1–26. inforced slope with steel bars.” In Int. Geotechnical Symp. on Theory
Davis, M. C. R., C. D. Jacobs, and R. J. Bridle. 1993. An experimental in- and Practice of Earth Reinforcements, 311–316. Rotterdam, The
vestigation of soil nailing, retaining structures, 587–598. London: Netherlands: A.A. Balkema.
Thomas Telford. Li, J., L. G. Tham, S. M. Junaideen, Z. Q. Yue, and C. F. Lee. 2008. “Loose
Elvin, A. 2009. “Experimentally applied earthquakes and associated load- fill slope stabilization with soil nails: Full-scale test.” J. Geotech.
ing on a full- scale dry-stacked masonry structure.” J. South Afr. Inst. Geoenviron. Eng. 134 (3): 277–288. https://doi.org/10.1061/(ASCE)
Civ. Eng. 51 (1): 15–25. 1090-0241(2008)134:3(277).
Fan, C. C., and J. H. Luo. 2008. “Numerical study on the optimum layout of Liew, S. S., and C. M. Khoo. 2006. “Soil nail stabilization for a 14.5 m
soil–nailed slopes.” Comput. Geotech. 35 (4): 585–599. https://doi.org deep excavation at uncontrolled fill ground.” In Proc., 10th Int. Conf.
/10.1016/j.compgeo.2007.09.002. on Piling and Deep Foundations. Hawthorne, NJ: Deep Foundations
Downloaded from ascelibrary.org by Jaypee University Of Information Tech - Solan on 05/07/24. Copyright ASCE. For personal use only; all rights reserved.

Felio, G. Y., M. Vucetic, M. Hudson, P. Barar, and R. Chapman. 1990. Institute.

“Performance of soil nailed walls during the October 17, 1989 Loma Lin, H., X. Wei, and C. Ping. 2013. “Stability of soil nailed slope
Prieta Earthquake.” In Vol. 1 of Proc., 43rd Canadian Geotechnical using strength reduction method.” Eur. J. Environ. Civ. Eng. 17 (9):
Conf., 165–173. Quebec, Canada: Np. 872–885. https://doi.org/10.1080/19648189.2013.828658.
FHWA (Federal Highway Administration). 2003. Geotechnical engineer- NCREE (National Research Center on Earthquake Engineering). 1999.
ing circular No 7- soil nail walls. Rep. No. FHWA0-IF-03-017. Geotechnical engineering Failure Investigation after Chi-Chi earth-
Washington, DC: FHWA. quake. Reconnaissance Report of Chi-Chi Earthquake. Taiwan,
Gassier, G., and G. Gudehus. 1981. “Soil-nailing- some aspects of a new China: NCREE.
technique.” In Proc., 10th ICSMFE, 665–670. Rotterdam, The Patra, C. R., and P. K. Basudhar. 2005. “Optimum design of nailed soil
Netherlands: A.A. Balkema. slopes.” Geotech. Geol. Eng. 23 (3): 273–296. https://doi.org/10.1007
Giri, D., and A. Sengupta. 2009. “Dynamic behavior of small-scale nailed /s10706-004-2146-7.
soil slopes.” Geotech. Geol. Eng. 27 (6): 687–698. https://doi.org/10 Sahoo, S., B. Manna, and K. Sharma. 2016. “Seismic stability analysis of
.1007/s10706-009-9268-x. Un-reinforced and reinforced soil slopes.” In Geo-China 2016:
Giri, D., and A. Sengupta. 2010. “Dynamic behavior of small-scale model Advances in Numerical and Experimental Analysis of Transportation
of nailed steep slopes.” Geomech. Geoeng. 5 (2): 99–108. https://doi.org Geomaterials and Geosystems for Sustainable Infrastructure,
/10.1080/17486020903497415. Geotechnical Special Publication 257, edited by R. Bulut, X. Yu, and
Goodman, R. E., and H. B. Seed. 1965. Displacement of slopes in cohe- S.-R. Yang, 74–81. Reston, VA: ASCE.
sionless materials during earthquakes. Berkeley, CA: Dept. of Civil Seed, H. B., and R. W. Clough. 1963. “Earthquake resistance of sloping
Engineering, Institute of Transportation and Traffic Engineering core dams.” J. Soil Mech. Found. Div. 89 (1): 209–242.
Sheahan, T. C., and L. H. Carlton. 2003. “Simplified trial wedge method
Studies, Univ. of California.
for soil nailed wall analysis.” J. Geotech. Geoenviron. Eng. 129 (2):
Gutierrez, V., and A. F. Tatsuoka. 1988. “Role of facing in reinforcing co-
117–124. https://doi.org/10.1061/(ASCE)1090-0241(2003)129:2(117).
hesionless soil slopes by means of metal strips.” In Int. Geotechnical
Singh, V. P., and G. L. S. Babu. 2010. “2D numerical simulations of soil
Symp. on Theory and Practice of Earth Reinforcement, 289–294.
nail walls.” Geotech. Geol. Eng. 28 (4): 299–309. https://doi.org/10
Rotterdam, The Netherlands: A.A. Balkema.
.1007/s10706-009-9292-x.
Hong, Y. S., R. H. Chen, C. S. Wu, and J. R. Chen. 2005. “Shaking table
Stocker, M. F., G. W. Korber, G. Gassler, and G. Gudehus. 1979. “Soil
tests and stability analysis of steep nailed slopes.” Can. Geotech. J.
nailing.” In Vol. 2 of Int. Conf. on Soil Reinforcement, 469–474.
42 (5): 1264–1279. https://doi.org/10.1139/t05-055.
Paris, France: Np.
Hussin, J. D., L. W. Abramson, M. B. Addison, J. Baez, D. Bruce, and
Tufenkjian, M. R., and M. Vucetic. 2000. “Dynamic failure mechanism of
G. K. Burke. 1997. “Ground improvement, ground reinforcement,
soil nailed excavation models in centrifuge.” J. Geotech. Geo-Environ.
and ground treatment: Developments 1987–1997.” In Soil Eng. 126 (3): 227–235. https://doi.org/10.1061/(ASCE)1090
Improvement and Geosynthetics, Geotechnical Special Publication 69, -0241(2000)126:3(227).
edited by V. R. Schaefer, L. W. Abramson, J. C. Drumheller, J. D. US Geological Survey 2002. “1989 Loma Prieta, California Earthquake
Hussin, and K. D. Sharp, 72–305. Reston, VA: ASCE. Data.” Accessed February 9, 2021. https://escweb.wr.usgs.gov/nsmp
Iai, S. 1989. “Similitude for shaking table tests on soil-structure fluid model -data/data_sets/891018_1.html#Downloads.
in 1 g gravitational field.” Soils Found. 29 (1): 105–118. https://doi.org Wang, L. P., G. Zhang, and J. M. Zhang. 2010. “Nail reinforcement mech-
/10.3208/sandf1972.29.105. anism of cohesive soil slopes under earthquake conditions.” Soils
Jewell, R. A., and M. J. Pedley. 1992. “Analysis for soil reinforcement with Found. 50 (4): 459–469. https://doi.org/10.3208/sandf.50.459.
bending stiffness.” J. Geotech. Eng. 118 (10): 1505–1528. https://doi Wartman, J., R. B. Seed, and J. D. Bray. 2005. “Shaking table modeling of
.org/10.1061/(ASCE)0733-9410(1992)118:10(1505). seismically induced deformations in slopes.” J. Geotech. Geoenviron.
Joshi, B. 2003. “Behaviour of calculated nail head strength in soil-nailed Eng. 131 (5): 610–622. https://doi.org/10.1061/(ASCE)1090
structures.” J. Geotech. Geo-Environ. Eng. 129 (9): 819–828. https:// -0241(2005)131:5(610).
doi.org/10.1061/(ASCE)1090-0241(2003)129:9(819). Wasowski, J., C. T. Lee, and D. K. Keefer. 2011. “Toward the next gener-
Juran, I., G. Baudrand, K. Farrag, and V. Elias. 1990. “Kinematical limit ation of research on earthquake-induced landslides: Current issues and
analysis for design of soil-nailed structures.” J. Geotech. Eng. future challenges.” Eng. Geol. 122 (1–2): 1–8. https://doi.org/10.1016/j
116 (1): 54–72. https://doi.org/10.1061/(ASCE)0733-9410(1990) .enggeo.2011.06.001.
116:1(54). Yazdandoust, M. 2018. “Seismic performance of soil-nailed walls using a
Juran, I., A. Guermazi, C. L. Chen, and M. H. Ider. 1988. “Modeling and 1 g shaking table.” Can. Geotech. J. 55 (1): 1–18. https://doi.org/10
simulation of load transfer in reinforced soil: Part 1.” Int. J. Numer. .1139/cgj-2016-0358.
Anal. Methods Geomech. 12 (2): 141–155. https://doi.org/10.1002/nag Yazdandoust, M. 2019. “Shaking table modeling of MSE/soil nail hybrid
.1610120203. retaining walls.” Soils Found. 59 (2): 241–252. https://doi.org/10
Keefer, D. K. 2000. “Statistical analysis of an earthquake-induced landslide .1016/j.sandf.2018.05.013.
distribution—The 1989 Loma Prieta, California event.” Eng. Geol. Yin, J. H., and C. Lok-Man. 2006. “A laboratory device to test the pull-Out
58 (3–4): 231–249. https://doi.org/10.1016/S0013-7952(00)00037-5. behavior of soil nails.” Geotech. Test. J. 28 (5): 1–15.
Khazai, B., and N. Sitar. 2004. “Evaluation of factors controlling Yuan, J. X., Y. Yang, L. G. Tham, P. K. K. Lee, and T. Tsui. 2003. “New
earthquake-induced landslides caused by Chi-Chi earthquake and approach to limit equilibrium and reliability analysis of soil nailed
comparison with the Northridge and Loma Prieta events.” Eng. Geol. walls.” Int. J. Geomech. 3 (2): 145–151. https://doi.org/10.1061
71 (1–2): 79–95. https://doi.org/10.1016/S0013-7952(03)00127-3. /(ASCE)1532-3641(2003)3:2(145).

© ASCE 04021036-14 Int. J. Geomech.

Int. J. Geomech., 2021, 21(4): 04021036