Вы находитесь на странице: 1из 13

Soil Dynamics and Earthquake Engineering 128 (2020) 105896

Contents lists available at ScienceDirect

Soil Dynamics and Earthquake Engineering


journal homepage: http://www.elsevier.com/locate/soildyn

Shaking table tests on shallow foundations over geocomposite and


geogrid-reinforced liquefiable soils
Hadi Bahadori a, *, Hooman Motamedi a, Araz Hasheminezhad a, Ramin Motamed b
a
Department of Civil Engineering, Urmia University, Urmia, Iran
b
Department of Civil and Environmental Engineering, University of Nevada, Reno, USA

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

Keywords: The present study presents the results of a series of 1 g shaking table tests modeling a 4 story building foundation
Geocomposite overlying liquefiable soil reinforced by two different types of geosynthetics as liquefaction mitigation counter­
Geogrid measures. Models of geosynthetics including geogrid and geocomposite were used as reinforcement elements.
Shaking table
Failure mechanisms; foundation and geosynthetic settlement; and pore water pressure built-up trends were
Liquefaction
Failure mechanism
considered as the indicators of geosynthetics performances. The failure mechanism was observed through a
Settlement digital camera and analyzed based on a mesh of colored sand. The obtained results indicated the negligible effect
of geogrids on the reduction of settlement as the failure mode was controlled by the pore water pressure and it
was not influenced by geogrids. However, based on the experimental results the effectiveness of geocomposits in
reducing the settlement due to impervious characteristics was evident. In addition, the a remarkable mitigation
occurred in the case of settlement of soils between two geocomposite sheets, but the general performance of
geocomposites was not satisfactory at all.

1. Introduction tendency for densification [10]. Consequently, there may be a situation


in which effective stress reaches zero and soil shows a fluid like behavior
Shallow foundations, when built on weak subgrade soils, have low or liquefaction, which has caused serious damages during past decades.
bearing capacity and undergo large settlements. To improve this situa­ Therefore, finding effective solutions to decrease damage due to lique­
tion they can be built over a compacted granular fill of sufficient depth faction in different conditions is one of the most important challenges in
placed on top of the weak subgrade. The granular fill allows the load to the geotechnical earthquake engineering.
be distributed to such an extent that the weak subgrade is no longer It should be noted that, the involved foundation in liquefaction
overstressed. A further improvement can be achieved by placing a pattern differs than free field. In free field, primary liquefaction occurs at
geosynthetic layer at the interface of the weak subgrade and the com­ shallow depth [11,12], but this pattern changes in the case of an
pacted granular fill [1]. Previous studies evaluating the bearing capacity involved foundation. Liu and Dobry during their study conducted eight
of shallow foundations supported by soils with a geosynthetic rein­ centrifuge model experiments to examine the mechanism of
forcement layer have shown the feasibility of this reinforced system. liquefaction-induced settlement. They found that in the case of
They all concluded that the soil reinforcement by geosynthetics in­ free-field, at the initial point liquefaction occurs at shallow depths and
creases the static bearing capacity of foundations resting on sandy soils then propagates downward but pore water pressure develops under the
(especially dry) and several recommendations have been proposed footing that significantly is less than the corresponding initial effective
regarding their best applications as the soil reinforcement [1–9]. How­ vertical pressure [13]. Moreover, Adalier et al. conducted a series of
ever, granular saturated soil mass has tendency to be densified when highly instrumented dynamic centrifuge model tests, through these tests
subjected to earthquake loading. Rapid earthquake loading can occur they stated that without surcharge, liquefaction propagates top to bot­
under undrained condition in the case of saturation of cohesion-less tom but in a model ground with surcharge, Ru (Excess pore water
soils, so both excess pore water pressure and effective stresses can pressure ratio) values lowers immediately below the foundation
experience increase and decrease respectively in their levels due to the centerline [14]. Furthermore, Karamitros et al. presented a simplified

* Corresponding author.
E-mail addresses: h.bahadori@urmia.ac.ir (H. Bahadori), hooman.motamedi@gmail.com (H. Motamedi), a.hasheminezhad@urmia.ac.ir (A. Hasheminezhad),
motamed@unr.edu (R. Motamed).

https://doi.org/10.1016/j.soildyn.2019.105896
Received 30 November 2018; Received in revised form 24 July 2019; Accepted 2 October 2019
Available online 9 October 2019
0267-7261/© 2019 Elsevier Ltd. All rights reserved.
H. Bahadori et al. Soil Dynamics and Earthquake Engineering 128 (2020) 105896

bottom of the model, but rather at some intermediate depth [11]. During
this study it was revealed that value of excess pore water pressure ratio is
low in shallow depths under foundation [11]. have also indicated that
maximum excess pore water pressure at diverse depths does not occur at
the same time which was observed in this series of experiments.
It has shown that reinforcing sand has a major outcome in increasing
soil’s resistance to liquefaction especially in low densities [17–22].
Moreover, Maheshwari, et al. (2012) investigated the effect of rein­
forcement on the liquefaction resistance of Solani sand (at 25% relative
density) by carrying out shaking table tests. They used coir fiber, geogrid
sheet, and synthetic fiber as reinforcement. They concluded that rein­
forced soil was more resistant to liquefaction at a lower acceleration (e.
g., at 0.1 g) and although geogrid sheets increased the liquefaction
resistance of sand, the improvement of loose Solani sand reinforced by
coir or synthetic fiber was much more significant. In another research
[23], investigated the effectiveness of geogrid layers on liquefaction
resistance at 40% relative density. They also asserted the effect of geo­
grid reinforcement on increasing the liquefaction resistance of sand. In
addition, during past years, the failure mechanism of shallow founda­
tions under seismic loading has been investigated by many analytical
studies (Paolucci and Pecker, 1997; [12,24,25]. One method which is
recently used in this field is measuring the soil deformation using digital
images and PIV (Particle Image Velocimetry) analysis in which variation
of brightness is followed in a series of images of soil [26,27]. Durga [28]
investigated the load-settlement response and proposed optimal depth
of reinforcement for aggregate layer overlying sand layer for various
cases [29]. carried out laboratory modeling tests and numerical studies
on circular and square footings resting on geosynthetic-reinforced sand
bed. According to their findings, the effect of a geosynthetic inclusion
was mainly dependent on the shape of footings. In addition, an increase
in the number of reinforcement layers decreased the tilt of circular
footing more than square footing and based on the settlement reduction
of reinforced condition at ultimate bearing capacity, the settlement was
remarkably dependent on load eccentricity which was not different from
the unreinforced one [30]. simulated a geocell-reinforced foundation
sandy and clay beds using FLAC3D and found that the optimal height of
the geocell layer in a clay basement was 30–50% while the optimal
range in sand was 30–40% [31]. used laboratory scaled plate load tests
to study the performance of a square footing resting on the
three-dimensional geogrid reinforced sand bed. They used two types of
3D geogrids with triangular and rectangular pattern. Based on their
obtained results, the performance of 3D geogrids is better than con­
ventional geogrids in surface deformation reduction [32]. in an exper­
imental study, investigated the bearing capacity of strip footing with
geogrid on fine loose sand. The increase of the geogrid reinforcing layers
caused an increase in the bearing capacity ratio.
A comprehensive review of the available literature indicates that a
few studies have reported the liquefaction resistance of saturated sand
Fig. 1. Test apparatus.
Table 2
methodology to estimate the post-shaking bearing capacity of shallow Summary of some of the scaling factors for modeling of geosynthetic ma­
foundations, and present a typical analysis for a strip foundation. They terials in 1 g model [38]].
predicted that excess pore pressure ratios that are available in the area Parameters Scale factors
beneath the footing remain lower than 1 in the free field [15]. Previous Geosynthetic strain (%) 1
studies propose that the decreased excess pore pressure underneath the Length (m) 1/N
foundation can be attributed to the shear-induced dilative behavior of Displacement (mm) 1/N
Tensile Strength (kN/m) 1/N2
the subsoil. This is due to the deformation of saturated soil and the
Secant modulus (kN/m) 1/N2
surcharge loads can also reduce the potential of liquefaction [13–16]. Pull-out force 1/N3
Yet another study has also indicated that the value of maximum excess Soil-geosynthetic friction angle (Deg) 1
pore pressure ratio does not occur beneath the foundation or at the

Table 1
Properties of Firoozkuh sand NO.161.
Firoozkuh sand No.161 Gs emax emin K (cm/s) cc cu D50 (mm) %FC
2.658 0.858 0.548 0.0125 0.97 2.58 0.3 1

2
H. Bahadori et al. Soil Dynamics and Earthquake Engineering 128 (2020) 105896

Fig. 2. Geocomposite and Geogrids used in this research.

Table 3
Properties of model reinforcements.
Geogrid Geocomposite

Mass/unit area (g/m2) 300 64


Aperture size (mm) 2 –
Thickness (mm) 1 0.13
Ultimate tensile strength (N/m) 1300 1582
Tensile strength (N/m) @ 2% strain 192
5% strain 445
Tensile stiffness (N/m) @ 2% strain 9600
5% strain 8900

reinforced with geosynthetics by adopting element tests on sands with


different densities and reinforcing elements, therefore we are going to
cover this gap during the present study. In this research, six 1 g shaking
table tests were conducted in which detailed instrumentation was per­
formed using pore water pressure, acceleration and displacement
transducers in order to measure and record the data. Although PIV was
not directly used in this study, but the idea of utilizing pictures to find
failure mechanism was adopted. A digital camera was used to record
videos and convert them into pictures to examine the failure mechanism
of foundation on sandy bed reinforced with geosynthetics during
liquefaction. The effect of geogrid and geocomposite on the settlement
reduction in the saturated sandy soils under dynamic excitation was
studied in detail. The main objective of this paper was to evaluate the
effectiveness of two different types of geosynthetics including geogrid
and geocomposite on the settlement mitigation due to liquefaction.

2. 1 g shaking table model tests

Fig. 1 displays the test layout and dimensions of the shaking table
and its instrumentation. The shaking table had two 36 mm-thick
200 � 60 cm wood blocks, where upper wood block served as a dock on
which a container was placed and the lower wood block served as a fixed
support; three 1.2 mm-thick steel plates which each of them were
enclosed by four 50 � 50 � 5 mm angle beams. The steel plates had
resilience property and were able to provide a harmonic movement in a
direction perpendicular to their facings. The container was a rigid and
transparent box made of plexiglass with inner dimensions of
200 � 50 � 70 cm. At the bottom of container a void chamber was made
for drainage purposes. The shaking table container was rigid; therefore
boundary conditions could affect the test results. Thus, a 2-cm thick
absorbing layer of foam was employed in shaking direction of the
container. In order to reduce this effect, band filtering data (acceleration
raw data) was used to eliminate part of erroneous data. During satura­
tion process, first chamber was filled with water; therefore a constant
and gradual flow of water was provided from the bottom of the
container upwards into the soil. Firoozkuh sand No. 161 was used as the
soil material. Table 1 shows the properties of this soil [33]. Wet tamping Fig. 3. Conducted shaking table test models.

3
H. Bahadori et al. Soil Dynamics and Earthquake Engineering 128 (2020) 105896

Fig. 4. Time histories of base acceleration transducer (A4) for different models.

method was used to prepare the model which has been also used by Refs. 1 2
to (25 ) of engineering geosynthetic. High strength engineering geogrids
[34,35]. In this method sand was mixed with 5% water and then have ultimate tensile strength of around 200 kN/m. Considering the
distributed in the container homogeneously in order to construct uni­ same scale factor of 1:25, ultimate tensile strength of model geo­
form soil layers. Then every layer is compacted till the weight per vol­ synthetics should be around 320, but geosynthetics with much more
ume of soil reaches to 14.72 kN/m3 (equivalent value of maximum void ultimate tensile strength (1300 and 1582 N/m for geogrid and geo­
ratio with relative density of 0%). Also a grid of colored sand was placed composite geosynthetics, respectively) were used in the models in order
on the front part of the container to observe the behavior of sand during to omit the possible effect of low tensile strength on liquefaction resis­
liquefaction. Meanwhile, carbon dioxide gas was inserted into the model tance and investigate the effectiveness of geosynthetics in general.
ground to accelerate the saturation process by replacing air trapped in Therefore, standard light Meshiran fence as geogrid and plastic sheet as
voids. Finally, a steady flow of low discharged water from beneath of the geogrid-geomembrane geocomposite were used as reinforcement in the
container saturated the soil. Vibration was applied to the container at models (Fig. 2). The plastic sheet demonstrates appropriate tensile
the shaking table’s resonant frequency which was 2.8 Hz. In order to strength and is impermeable, therefore it can simultaneously model both
record the time history of several parameters throughout shaking the geogrid and geomembrane properties. In the reinforced models, geo­
model ground and foundation were instrumented by acceleration, pore synthetics with dimensions of 190 � 45 cm were placed beneath the
water pressure transducers and LVDT. foundation in different depths. Properties of model reinforcements are
Foundation used in the models weighs 47 kg which can apply a summarized in Table 3. Fig. 3 shows the schematic illustrations from
4.26 kPa surcharge to the sand [36]. recommended scaling factors of tests without and with mitigation measures. In this study, results of six
different soil parameters and they were evolved by Ref. [37]. According different models were compared.
to Ref. [37] in 1 g models, scale factors for force and acceleration are 1=
n3 and 1 respectively in which n represents scale factor for the length, 3. Results and discussion
therefore 1=n3 is the scale factor for mass. Considering the weight of a 4
story building with dimensions of 6.12 � 11.25 m (about 750 tons), a 3.1. Time history of acceleration
47 kg 45 � 24.5 cm foundation can model a 4 story building with scale
factor of 1:25. Relationships required for modeling geogrids were Time histories of input acceleration that were recorded at the base of
introduced by Viswanadham and Ko €nig [38]. Table 2 summarizes the container (A4), as well as the ones that were recorded at the depth of
scaling factors for modeling of geosynthetic materials in 1 g model. 35 cm (A3) for different models, are shown in Fig. 4 and 5, respectively.
Therefore, model is in need of a material with tensile strength equivalent The main focus in this paper is to study the deformation pattern of loose

4
H. Bahadori et al. Soil Dynamics and Earthquake Engineering 128 (2020) 105896

Fig. 5. Time histories of acceleration at depth of 35 cm (A3) for different models.

saturated soil underlying a surcharge in the case of seismic force constant and in the third and final part, it decreased over time. Ac­
application. In another words, attention has been given to inspecting cording to Fig. 6 the slope of first part of Ru1 and Ru3 curves are lower
where the deformation of underlying soil started and how it propagated than Ru2 and Ru4 curves, respectively.
throughout the soil. Different input acceleration alters the extent of This result is more evident in the models reinforced with geo­
deformed soil. Therefore, through keeping the acceleration amplitude composites except in the model “d" in which there is no reinforcing
similar for all tests (~0.3 g), input acceleration was applied in all models element between higher and lower pore water pressure transducers. In
until the foundation showed a major settlement. As a result, while this case, below part of the geocomposite layer can built pore water
comparing the results, weak input motions in the models “b" and “c" and pressure more quickly than upper part of the geocomposite layer.
the strong motion in the model “e" should be considered. Therefore, the geocomposite layers can delay building up pore water
pressure due to the ground movement. Comparison of the results of test
3.2. Time history of excess pore water pressure ratio “e" in Fig. 6 and Fig. 7 with other models, makes it clear that despite
large input acceleration, the values of excess pore water pressure at the
Fig. 6 illustrates the time histories of excess pore water pressure ra­ shallow depths are not that much high. This can prove the effectiveness
tios Ru1 to Ru4 for transducers P1 to P4 respectively in the models “a" to of geocomposites in reducing building up of excess pore water pressure
“f". Fig. 7 shows the maximum values of excess pore water pressure especially in the shallow depth. Comparing results of test “e" with test “f"
ratios for various pore water pressure transducers in the different also indicates that despite larger input motion in the test “e", reinforcing
models. Ru is derived by dividing the excess pore water pressure to the with geocomposite is more effective than reinforcing with geogrid
initial effective stress at any point. As it can be observed from the Fig. 6 sheets.
and later at deformation patterns, this study also confirms that lique­
faction does not occur at shallow depths, and surface settlement is due to 3.3. Geosynthetic profile
liquefaction at profounder depths [11]. have also indicated that
maximum excess pore water pressure at diverse depths does not occur at Fig. 8 shows the geosynthetic profiles before and after strong shaking
the same time which also was observed in this series of experiments in the reinforced models (b-f) and maximum displacement values are
(Fig. 6). As depicted in Fig. 6 and presented in Fig. 7, no full liquefaction presented in Fig. 9. In all models, the values of settlement of first and
state was observed at shallow depth (transducers P1 and P3 at 14 cm second geosynthetic reinforcement layers were similar. Therefore, it can
depth) and the foundation settlement was due to liquefaction in be concluded that reinforced soil layers in the upper 25 cm of models
profounder depths. Each Ru curve had three parts. In the first part, the were not liquefied and liquefaction of soil lower than this depth caused
excess pore water pressure raised from 0 to its peak. In the second surface settlement. In Fig. 9, a comparison between the different models
portion of the figure, the excess pore water pressure was relatively indicates that upper reinforcement layers in model “e" despite

5
H. Bahadori et al. Soil Dynamics and Earthquake Engineering 128 (2020) 105896

Fig. 6. Time histories of excess pore water pressure ratios.

experiencing much more input acceleration didn’t show high settlement 3.4. Failure mechanism
values as other models. Therefore, liquefaction can improve a heavily
reinforced model with geocomposites resistance of surface soil to set­ During preparation of each model, a grid of colored sand was placed
tlement. This fact is not true when model is reinforced with geogrid in the soil at the facing of transparent container in order to directly
layers, which is another reason for the effectiveness of geocomposite observe the failure mechanism of liquefied soil. For this purpose, a
layers than geogrids in the liquefiable soils. digital camera was used to record videos in all tests and reproduction of
pictures demonstrating the failure mechanism of model foundation.
Figures 10–15 show the deformation patterns of different models after
initiation of acceleration. As can be seen in all models, liquefaction was

6
H. Bahadori et al. Soil Dynamics and Earthquake Engineering 128 (2020) 105896

Fig. 7. Maximum values of excess pore water pressure ratios.

initiated at lower half of the sandy model ground. Fig. 10 shows the (depth of 30–40 cm) and limitedly 3rd soil layer (depth of 20–30 cm)
deformation of soil under the foundation in the benchmark model. began to move laterally. Therefore, because of the reinforcement layer,
Initially at 40 cm depth beneath the foundation, the soil began to move two liquefied layers performed separately. Similar to the previous
laterally then the deformation expanded to the 20 cm layer (depth models, because of the movement of soil in the depth, the settlement was
30–50 cm) and at the same time, the soil beneath the foundation began toward the sides and due to sinking of foundation into the soil (loss of
to move downward. At this point, the settlement is due to the lateral bearing capacity); it was accompanied by the settlement.
movement of soil at the depth of 30–50 cm and it is not due to the Fig. 13 shows the deformation of soil under the foundation in the
reduction of bearing capacity of the foundation. Along with the model “d" during the shaking. In this model, due to the non-existence of
continuation of shaking, soil liquefaction expanded in arc form toward reinforcement layer in the depth of 30–50 cm, there was no significant
sides and upward. Therefore, the soil beneath the foundation deformed change in the pattern of soil movement. However, in comparison to the
laterally and the soil at both sides of the foundation moved upward benchmark model, the deformation initiated with delay. in this case the
which caused settlement in the middle and surface heave on both sides. installation of reinforcement layer and the other reinforcement layer at
With more strong movement, the arc like movement of soil beneath the the depth of 30 cm caused a relative movement of soil layer in the depth
foundation continued and settlement accumulated due to lack of bearing of 30–50 cm and 50–60 cm. Finally, upper reinforcement layer which
capacity of foundation on the soil (sinking of foundation into the soil) was installed at the depth of 10 cm caused a slight and straight heave on
and liquefaction in depth. Upper 30 cm of soil beneath the foundation the sides of foundation.
did not show much deformation and mainly deformation was caused by Fig. 14 shows the deformation of soil under the foundation in the
the movement of soil in the lower 30 cm. model “e" during the shaking. This model was reinforced with five
Fig. 11 shows the deformation of soil under the foundation in the geocomposite sheets at the depths of 10, 20, 30, 40 and 50 cm which
model “b" during the shaking. In this model, because of the placement of resisted the ground motion for 2 s. First signs of liquefaction initiation
reinforcement layers at shallow depths, they were not able to affect the were observed after 2.07 s after the initiation of strong motion in this
liquefied layer, therefore no significant change in the failure mechanism model. Movement of the soil from center to sides started at 45 cm depth,
was observed. Only due to the two reinforcement layers instead of arc and soil in the 4th (depth of 30–40 cm) and 5th (depth of 40–50 cm) soil
heave (as can be seen in the benchmark model), on sides of foundation, a layers moved in an arc like pattern as the shaking continued. At the last
straight heave was observed in this model. Also due to the low input seconds of the test, liquefaction was also observed in the 3rd soil layer
motion, settlement did not happen in this model. (depth 20–30 cm). Despite arc like movement of soil, due to heavily
Fig. 12 shows the deformation of soil under the foundation in the reinforcing soil with geocomposite layers, soil in profounder depths was
model “c" during the shaking. As can be observed in this model, lique­ not able to push soil in the shallower depths upward at sides of the
faction initiated at a profounder depth of 45 cm and didn’t expand to a foundation. In this model, liquefaction in deep depths did not cause
20 cm layer. Also, the fifth soil layer (depth 40–50 cm) moved to the serious surface settlement. Main reason for settlement in this model was
sides. In this model, due to the reinforcement layers, fewer arc like sinking of foundation into the soil which started about 6 s after the
movements of the soil were observed in this case. Soil in this layer was initiation of shaking. Consequently, reinforcing soil with 5 layers of
laterally moved. Therefore, colored sand at 30 and 40 cm depth didn’t geocomposites considerably decreased settlement due to liquefaction of
have a large upward movement. With more strong motion, 4th soil layer soil. Previous studies have attributed the increase in liquefaction

7
H. Bahadori et al. Soil Dynamics and Earthquake Engineering 128 (2020) 105896

Fig. 8. Geosynthetic profile in the different models.

resistance of reinforced sand to various parameters. For instance layers with geotextiles were considered as the evident reasons for the
Ref. [17], mainly attributed it to higher interface friction and stiffness of effectiveness of geotextile reinforcement on improvement of liquefac­
reinforcement element. Later, they suggested that the ability of rein­ tion resistance of sand [20].
forcement in increasing the liquefaction resistance of sand is due to the In this study due to impermeability property, geocomposite layers
increase of the effective confining pressure between the reinforcement may further reduce interstitial pressure distribution. On the other hand,
layers of soil, which is a function of the shear stress mobilized along the by separation of sand layers in the liquefiable zone, they may impose
soil-reinforcement interface [18]. They observed that angle of interface limitation on soil movement due to liquefaction.
friction, bending stiffness and compressibility plays a dominant role in Fig. 15 shows the deformation of soil under the foundation in the
mobilization of shear stress along the soil-fabric interface [18]. Another model “f" during the test. In this model, 4th (depth of 30–40 cm) and 5th
study performed by the use of cyclic triaxial instrument, indicated that (depth of 40–50 cm) soil layers started to move to the sides discretely. So
an increase in liquefaction resistance is either due to the deformability of it can be assumed that any reinforcement material used at any depth in
the inclusions or due to the soil-geosynthetic friction phenomenon based the case of the occurrence of liquefaction can separate soil deformation.
on the value of stress ratio [19]. Furthermore, the authors explained that In this model, no serious delay between initiation of strong motion and
compressible geotextile sheets can delay liquefaction. This process is initiation of deformation was observed, and surface settlement was
done through reducing the interstitial pressure in the sand because of considerable. Table 4 shows the delay between initiation of input mo­
expansion during successive extensions of the sample. Therefore, tion and location in which deformations were initiated.
liquefaction occurs when interstitial pressure gradient in the sample
establishes and geotextiles reaches a maximum expansion [19]. It was 3.5. Time history of LVDT
also proposed that reduction in interstitial pressure distribution may
increase the liquefaction resistance of sand reinforced by geogrid sheets. In all models, the settlement of foundation exceeded the recordable
Furthermore, reinforcing sand with geogrid sheets makes it a composite data range of LVDT which was about 7 cm but lower settlements of
material with higher strength and stiffness compared with the sand foundation was obvious in the models “c" to “f". In this paper due to
alone [39]. On the other hand, improvement in the frictional and tensile variation of input acceleration in different tests, a new parameter as
properties of sand with geotextile reinforcement and separation of sand cumulative input energy (CIE) was proposed. In order to calculate CIE,

8
H. Bahadori et al. Soil Dynamics and Earthquake Engineering 128 (2020) 105896

Fig. 9. Maximum settlement of different geosynthetic layers.

Fig. 10. Deformation pattern of benchmark model.

input force and velocity of the container should be calculated. Input of soil mass at t ¼ t1 where t1 varies from 0 to the end of energy appli­
force can be calculated by equation (1) and velocity of the container can cation to the container.
be calculated by equation (2). As can be seen in Fig. 16, no clear pattern can be observed before
settling 5 mm but afterwards models “b" to “e" dissipated more energy.
FðtÞ ¼ maðtÞ (1)
Consequently, also models “d" and “e" dissipated much more energy to
Z settle more than 10 mm. The reason can be attributed to the ability of
vðtÞ ¼ aðtÞdðtÞ (2) geocomposite to dissipate the waves induced by the input acceleration.
Therefore, much more energy is required to liquefy subsoil and settle the
where m is the mass of soil in the container and a(t) is the input accel­ foundation. Also, despite implementing less reinforcement layers in the
eration or base acceleration. Finally CIE can be calculated by equation model “d" compared with model “c”, due to implementing a reinforce­
(3): ment layer in a profounder depth, slope of settlement curve in this model
Z t¼t1 is less than the model “c". Therefore, it can be concluded that the
CIEðtÞ ¼ jðFðtÞ:vðtÞÞjdt (3) behavior of liquefiable sand can be improved by implementing rein­
t¼0 forcement layers in the profounder depths. In Fig. 16, the model “f"
Fig. 16 shows the foundation settlement versus CIE per unit volume surprisingly showed the minimum resistance to input acceleration

9
H. Bahadori et al. Soil Dynamics and Earthquake Engineering 128 (2020) 105896

Fig. 11. Deformation pattern of model “b”.

Fig. 12. Deformation pattern of model “c”.

Fig. 13. Deformation pattern of model “d”.

10
H. Bahadori et al. Soil Dynamics and Earthquake Engineering 128 (2020) 105896

Fig. 14. Deformation pattern of model “e”.

Fig. 15. Deformation pattern of model “f”.

compared with other models, so it can be said that geocomposite is a


Table 4 better reinforcement material than geogird.
Delay between initiation of input motion and deformation and location of
initiation of deformation. 4. Conclusions
Delay Depth

Test a 0.88 s 40 cm Settlement of a shallow foundation on saturated sandy soil layers


Test b 1.22 s 55 cm during a strong motion has two phases. One is the settlement due to the
Test c 0.98 s 45 cm movement of soil in the profound depths under the foundation toward
Test d 1.01 s 40 cm
sides caused by liquefaction. Another one is the sinking of foundation
Test e 2.07 s 35 and 45 cm
Test f 1s 40 cm into the soil due to the reduced bearing capacity of surface soil. Ac­
cording to the obtained results in this study:

� Liquefaction happened initially at some intermediate depth under


the foundation. No soil liquefaction occurred under the foundation

11
H. Bahadori et al. Soil Dynamics and Earthquake Engineering 128 (2020) 105896

Fig. 16. CIE/Unit Volume of soil versus settlement curves for different models.

and at the bottom of the model. Reinforcing sand with geocomposite Appendix A. Supplementary data
can cause lower excess pore water pressure ratios at shallow depths
under foundation.Reinforcing shallow depths under the foundation Supplementary data to this article can be found online at https://doi.
does not change the deformation pattern. It can only reduce or org/10.1016/j.soildyn.2019.105896.
eliminate surface heave at the sides of foundation. Reinforcing sand
with geosynthetics at liquefaction zone can discrete and confine its References
movement. Reinforcing sand with geocomposites at liquefaction
zone can increase soil resistance to liquefaction and consequently [1] Pinto MIM. Applications of geosynthetics for soil reinforcement. Proceedings of the
delay and decrease surface settlement. Reinforcing sand with geo­ Institution of Civil Engineers-Ground Improvement 2003;7(2):61–72.
[2] Binquet J, Lee KL. Bearing capacity tests on reinforced earth slabs. J Geotech
composite at the depth below the liquefaction zone can increase soil Geoenviron Eng 1975;101. ASCE #11792 Proceeding.
resistance to liquefaction and consequently delay and decrease sur­ [3] Huang CC, Tatsuoka F. Bearing capacity of reinforced horizontal sandy ground.
face settlement. So, geocomposite is a better reinforcement material Geotext Geomembranes 1990;9(1):51–82.
[4] Yetimoglu T, Wu JT, Saglamer A. Bearing capacity of rectangular footings on
than geogrid in the liquefiable soils. CIE as a new parameter was geogrid-reinforced sand. J Geotechnical Eng 1994;120(12):2083–99.
proposed in this paper through which the calculation of the amount [5] Dash SK, Krishnaswamy NR, Rajagopal K. Bearing capacity of strip footings
of energy applied to the container during each test was possible. supported on geocell-reinforced sand. Geotext Geomembranes 2001;19(4):235–56.
[6] Alawaji HA. Settlement and bearing capacity of geogrid-reinforced sand over
Results displayed that via reinforcing saturated sand with geo­ collapsible soil. Geotext Geomembranes 2001;19(2):75–88.
composites, in comparison to the virgin model and the reinforced [7] Patra CR, Das BM, Atalar C. Bearing capacity of embedded strip foundation on
model with geogrid elements, more energy is required for foundation geogrid-reinforced sand. Geotext Geomembranes 2005;23(5):454–62.
[8] Latha GM, Somwanshi A. Bearing capacity of square footings on geosynthetic
to settle.
reinforced sand. Geotext Geomembranes 2009;27(4):281–94.
� Heavily reinforcing soil with geocomposite, due to liquefaction, can [9] Latha GM, Somwanshi A. Effect of reinforcement form on the bearing capacity of
considerably reduce settlement in profounder depths but is not able square footings on sand. Geotext Geomembranes 2009;27(6):409–22.
to increase the bearing capacity of surface soil. Therefore, it can’t [10] Kramer SL. Geotechnical earthquake engineering. Upper Saddle River, New Jersey:
Prentice Hall; 1996.
decrease settlement of foundation due to its sinking into the soil. [11] Shahir H, Pak A. Estimating liquefaction-induced settlement of shallow
Reinforcing soil with geogrid will have insignificant effect on foundations by numerical approach. Comput Geotech 2010;37(3):267–79.
reducing the settlement which is due to the liquefaction especially in [12] Hasheminezhad A, Bahadori H. Seismic response of shallow foundations over
liquefiable soils improved by deep soil mixing columns. Comput Geotech 2019;
profounder depths. This is due to the failure mode and the fact that 110:251–73.
the phenomenon is controlled by the pore water pressure built-up at [13] Liu L, Dobry R. Seismic response of shallow foundation on liquefiable sand.
profounder depths and not influenced by geogrids. The results J Geotech Geoenviron Eng 1997;123(6):557–67.
[14] Adalier K, Elgamal A, Meneses J, Baez JI. Stone columns as liquefaction
approved that due to liquefaction, geosynthetics cannot properly countermeasure in non-plastic silty soils. Soil Dyn Earthq Eng 2003;23(7):571–84.
control the settlement only by itself. [15] Karamitros DK, Bouckovalas GD, Chaloulos YK, Andrianopoulos KI. Numerical
� This study argued that the used method in this investigation can alter analysis of liquefaction-induced bearing capacity degradation of shallow
foundations on a two-layered soil profile. Soil Dyn Earthq Eng 2013;44:90–101.
the soil movement pattern in deep depths and hence limit the set­
[16] Bahadori H, Farzalizadeh R, Barghi A, Hasheminezhad A. A comparative study
tlement but it can’t change the soil softening behavior of sand due to between gravel and rubber drainage columns for mitigation of liquefaction
cyclic motion. In this case, the foundation settles due to loss of hazards. Journal of Rock Mechanics and Geotechnical Engineering 2018;10(5):
924–34.
bearing capacity of shallow soil instead of the deep failure of lique­
[17] Krishnaswamy NR, Isaac NT. Liquefaction potential of reinforced sand. Geotext
fiable soil. Hence, this method can be much more effective if a Geomembranes 1994;13(1):23–41.
foundation is designed based on post-earthquake shear strength [18] Krishnaswamy NR, Thomas Isaac N. Liquefaction analysis of saturated reinforced
parameters. granular soils. J Geotechnical Eng 1995;121(9):645–51.
[19] Vercueil D, Billet P, Cordary D. Study of the liquefaction resistance of a saturated
sand reinforced with geosynthetics. Soil Dyn Earthq Eng 1997;16(7–8):417–25.

12
H. Bahadori et al. Soil Dynamics and Earthquake Engineering 128 (2020) 105896

[20] Altun S, G€oktepe AB, Lav MA. Liquefaction resistance of sand reinforced with [30] Oliaei M, Kouzegaran S. Efficiency of cellular geosynthetics for foundation
geosynthetics. Geosynth Int 2008;15(5):322–32. reinforcement. Geotext Geomembranes 2017;45(2):11–22.
[21] Bahadori H, Farzalizadeh R. Dynamic properties of saturated sands mixed with tyre [31] Makkar FM, Chandrakaran S, Sankar N. Behaviour of model square footing resting
powders and tyre shreds. Int J Civ Eng 2018;16(4):395–408. on sand reinforced with three-dimensional geogrid. Int J Geosynthetic Ground Eng
[22] Xu R, Fatahi B. Novel application of geosynthetics to reduce residual drifts of mid- 2017;3(1):3.
rise buildings after earthquakes. Soil Dyn Earthq Eng 2019;116:331–44. [32] El-Soud SA, Belal AM. Bearing capacity of rigid shallow footing on geogrid-
[23] Senapati S, Maheshwari BK. Effects of geogrid on dynamic strength characteristics reinforced fine sand—experimental modeling. Arabian Journal of Geosciences
of solani sand. Indian Geotech J 2012;42(4):287–93. 2018;11(11):247–57.
[24] Richards Jr R, Elms DG, Budhu M. Seismic bearing capacity and settlements of [33] Bahadori H, Ghalandarzadeh A, Towhata I. Effect of non plastic silt on the
foundations. J Geotechnical Eng 1993;119(4):662–74. anisotropic behavior of sand. Soils Found 2008;48(4):531–45.
[25] Soubra AH. Upper-bound solutions for bearing capacity of foundations. J Geotech [34] Manafi Khajeh Pasha S, Hazarika H, Bahadori H, Chaudhary B. Dynamic behaviour
Geoenviron Eng 1999;125(1):59–68. of saturated sandy soil reinforced with non-woven polypropylene fibre. Int J
[26] White DJ, Take WA, Bolton MD. Measuring soil deformation in geotechnical Geotech Eng 2018;12(1):89–100.
models using digital images and PIV analysis. In: 10th international conference on [35] Bahadori H, Manafi S. Effect of tyre chips on dynamic properties of saturated sands.
computer methods and advances in geomechanics, vol. 1; 2001. p. 997–1002. 1. Int J Phys Model Geotech 2015;15(3):116–28.
[27] Knappett JA, Haigh SK, Madabhushi SG. Mechanisms of failure for shallow [36] Iai S. Similitude for shaking table tests on soil-structure-fluid model in 1g
foundations under earthquake loading. Soil Dyn Earthq Eng 2006;26(2–4):91–102. gravitational field. Soils Found 1989;29(1):105–18.
[28] Prasad BD, Hariprasad C, Umashankar B. Load-settlement response of square [37] Wood DM. Geotechnical modeling. Spon Press; 2004. ISBN 0-415-34304-6 (hbk),
footing on geogrid reinforced layered granular beds. Int J Geosynthetic Ground ISBN 0-419-23730-5 (pbk).
Eng 2016;2(4):36. [38] Viswanadham BVS, K€ onig D. Studies on scaling and instrumentation of a geogrid.
[29] Badakhshan E, Noorzad A. Effect of footing shape and load eccentricity on Geotext Geomembranes 2004;22(5):307–28.
behavior of geosynthetic reinforced sand bed. Geotext Geomembranes 2017;45(2): [39] Maheshwari BK, Singh HP, Saran S. Effects of reinforcement on liquefaction
58–67. resistance of Solani sand. J Geotech Geoenviron Eng 2012;138(7):831–40.

13

Вам также может понравиться