Download as pdf or txt
Download as pdf or txt
You are on page 1of 9

Computers and Geotechnics 126 (2020) 103657

Contents lists available at ScienceDirect

Computers and Geotechnics


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

Research Paper

Effect of bedrock acceleration on dynamic and pseudo-static analyses T


of soil-pile systems
Hyunsung Lima, Sangseom Jeongb,

a
Korea Institute of Civil Engineering and Building Technology, 283 Goyang-daero, Ilsanseo-gu, Goyang-si, Gyeonggi-do 10223, Republic of Korea
b
Department of Civil and Environmental Engineering, Yonsei University, 50 Yonsei-ro, Seodaemun-gu, Seoul 03722, Republic of Korea

ARTICLE INFO ABSTRACT

Keywords: In this study, pseudo-static and dynamic analyses were performed for reasonable earthquake-resistant designs of
Pseudo-static analysis soil-pile systems. The emphasis was on evaluating the effect of bedrock acceleration on the earthquake-resistant
Dynamic analysis design of soil-pile systems in sand. Three-dimensional (3D) numerical analysis was performed at different
Finite-difference method bedrock accelerations and soil densities. Numerical solutions were verified against data from 1 g shaking table
Soil-pile systems
tests. In this study, it is found that as the peak bedrock acceleration increases, the peak superstructure accel­
Earthquake-resistant design
eration increases, whereas the peak ground surface acceleration tends to decrease due to the hysteretic damping
and nonlinear behaviors of the soil. It is also observed that compared with dynamic analysis, pseudo-static
analysis tends to more conservatively predict the soil-pile system behavior, particularly in the peak acceleration
interval of 0.13 g–0.3 g at the different soil densities.

1. Introduction Novak (Novak, 1974) formulated a simple approach to predict the


dynamic behavior of pile foundations by assuming a plane-strain con­
Recently, the number of large-scale earthquakes and tsunamis of dition at the pile-soil interface. Although one-dimensional Winkler
magnitude 6.0 or higher has increased worldwide, resulting in nu­ models have become popular for the seismic analysis of pile founda­
merous casualties and much property damage (Hyogoken-Nambu tions, most of them can be used only for the linear analysis of pile-soil
earthquake, 1995; Tohoku Earthquake, 2011; Maule Earthquake, 2010; interactions (Novak, 1974; Dobry et al., 1982; Kaynia and Kausel, 1982;
Sumatra-Andaman Islands Earthquake, 2004). Therefore, earthquake- Kavvadas and Gazetas, 1993). Winkler models that consider nonlinear
resistant design has become increasingly important to reduce the threat soil behavior have been developed (Penzien, 1970; Kagawa, 1980;
that earthquake-induced structural deformation and damage pose to Kagawa and Kraft, 1981; Norris, 1994; El Naggar and Novak, 1996;
communities and property. The pile foundation is an essential structure Nogami and Konagai, 1988; Tabesh and Poulos, 2000; Tabesh and
that is the basis of many important structures, such as bridges, high-rise Poulos, 2001). Extensive studies have been conducted to investigate the
buildings and nuclear facilities (Luo et al., 2016); and it is an important dynamic behavior of pile foundations under both pseudo-static and
part in earthquake-resistant design because it directly affects the sur­ dynamic loading conditions using advanced soil models compared to
rounding soil. the Winkler foundations (Phanikanth et al., 2013; Chatterjee et al.,
Pseudo-static analysis, which uses factored static loads to account 2015). These methods do not provide the variations in the bending
for the dynamic effects of pile foundations, is widely used for the moment, shear force and displacement of the pile along the depth over
earthquake-resistant design of pile foundations (Abghari et al., 1995; time, but they can be used to obtain the maximum pile bending mo­
Ishihara and Cubrinowski, 1998; El Naggar and Bentley, 2000; Tabesh ment, shear force and displacement during earthquake loading. Al­
and Poulos, 2001). In pseudo-static approaches, a static analysis is though these methods can accurately predict the dynamic behavior of
conducted to obtain the maximum pile bending moment, shear force the pile at very low frequencies, the response can vary significantly
and deflection developed in the pile due to earthquake loading during transient loading due to nonlinear behavior, damping and in­
(Liyanapathirana and Poulos, 2010). Several simplified approaches teractions (El Naggar and Bentley, 2000).
based on the Winkler hypothesis for the analysis of single piles have Performance prediction of pile foundations during earthquakes is a
also been developed that can be used with little computational effort. fundamental task for the seismic design of structures and bridges. Most


Corresponding author.
E-mail addresses: hslim85@kict.re.kr (H. Lim), soj9081@yonsei.ac.kr (S. Jeong).

https://doi.org/10.1016/j.compgeo.2020.103657
Received 3 March 2020; Received in revised form 12 May 2020; Accepted 12 May 2020
Available online 20 July 2020
0266-352X/ © 2020 Elsevier Ltd. All rights reserved.
H. Lim and S. Jeong Computers and Geotechnics 126 (2020) 103657

Table 1
Summary of material properties used in the analyses.
a
Properties Sand (Dr 85%) Sand (Dr 65%) Sand (Dr 35%) Steel pipe pile

a 3
γ (kN/m ) 15.7 15.2 14.4 77
a
Vs (m/s) 320 270 170 –
ν 0.35 0.35 0.35 0.2
a
K (Pa) 4.827 × 109 3.333 × 109 1.249 × 109 1.526 × 1011
a
G (Pa) 1.609 × 109 1.111 × 109 4.164 × 108 1.145 × 1011
c (Pa) 0 0 0 –
a
ϕ (°) 41 36.5 30 –

a
Note: γ is obtained from tests for Jumunjin sand (Lim and Jeong, 2018);
Vs = 65.6 N0.407 (Sun et al., 2006); G = ρVs2, K = E/(3(1-2ν)), ϕ=√15 N + 15
(Peck et al., 1974), Material properties of the steel pile pile is referred to the
literature (Hussien et al., 2014; Ko et al., 2016; Al-Baghdadi et al., 2015).

Fig. 2. Typical 3D finite-difference model. (a) Plane view (x-z direction); (b) 3D
view (x-y-z direction).

time consumption is very high. Compared to the more complex dy­


namic analysis methods, pseudo-static methods are attractive for design
engineers. In engineering practice, one-dimensional Winkler models
based on finite-element and finite-difference methods are widely used
(Tabesh and Poulos, 2000). Therefore, it is necessary to investigate the
applicability of pseudo-static analysis to the earthquake-resistant design
of pile foundations.
This study evaluates the effect of bedrock acceleration on the soil-
Fig. 1. Modulus reduction and damping ratio curves: (a) modulus reduction pile system under dynamic loading. To achieve the study objective,
curve; (b) damping ratio curve.
pseudo-static and dynamic analyses were performed of soil-pile systems
using numerical analysis. Based on the numerical analysis results, the
peak bedrock acceleration obtained in pseudo-static analysis was
modern seismic codes (e.g., ASCE 41-06, ASCE 7–10, Eurocode 8, evaluated under varying soil conditions.
Italian technical code NTC, 2008, Mexico City Building Code) re­
commend accounting for soil–structure interaction (SSI) effects in the
seismic design of both foundations and superstructures (Tombari et al., 2. Numerical modeling
2017). Many researchers have developed two- and three-dimensional
numerical models to predict the static and dynamic behavior of pile and A three-dimensional finite-difference method was implemented
piled-raft foundations considering SSI (Hamada et al., 1994; Sakajo using commercial FLAC 3D software version 5.01 (Itasca Consulting
et al., 1995; Zheng et al., 1996; Shahrour and Ousta, 1998; Finn et al., Group, 2013). FLAC 3D is a program that can simulate the behavior of
2001; Jeong and Cho, 2014; Roy et al., 2018; Bhaduri and Choudhury, soil, structures or other materials that may experience plastic flow
2020). Since these models take into account the complex dynamic in­ when the yield limit is reached. The explicit scheme and the mixed-
teractions of soil and pile, the dynamic behavior of the pile foundation discretization zoning technique enable accurate modeling of plastic
can be relatively accurately predicted. However, these models are dif­ collapse and flow of the materials (Rayhani and El Naggar, 2008). As
ficult to apply in engineering practice because it is difficult to obtain the such, in this study the dynamic behavior of soil-pile interaction was
complex material properties required for the analysis and the analysis investigated by using this code.

2
H. Lim and S. Jeong Computers and Geotechnics 126 (2020) 103657

Table 3
Pile properties of the test.
Pile Scaling factor Prototype Pile (D = 0.5 m)
(λ = 26.72)

Diameter (cm) λ 91.44 2.0


Thickness (cm) λ 1.4 0.2
Pile depth (cm) λ 1,710 64
Flexural rigidity (kg cm2) λ4.5 8,429.83 0.0032

velocity of soil was applied according to the correlation between the


shear wave velocity and N value reported by (Sun et al., 2006). The
shear modulus was calculated by using the density and shear wave
velocity, and the bulk modulus was calculated using the elastic modulus
and Poisson's ratio. The friction angle of soil was applied according to
the relation of the friction angle and N value reported by (Peck et al.,
1974). The material properties and soil conditions for the numerical
analyses are summarized in Table 1.
A strong earthquake can cause slippage and separation at the in­
terface between the soil and pile foundation. This behavior should be
accurately simulated during dynamic analysis. The interface model of a
linear spring-slider system was applied to simulate the behavior of the
interface between the soil and pile. The motion of the relative interface
was controlled by the normal and tangential interface stiffness. The
normal and tangential interface stiffness were determined by the fol­
lowing equation (Itasca Consulting Group, 2013):

K + (4/3) G
kn = ks = max
zmin (1)

where K and G are the bulk and shear moduli, respectively, of the
neighboring zone, and Δzmin is the smallest width of the adjoining zone
in the normal direction. The max [ ] notation indicates that the max­
imum value over all zones adjacent to the interface is to be used.
Fig. 3. Testing apparatus: (a) 1 g shaking table and soil box; (b) sectional view The available hysteretic damping algorithm in FLAC 3D is em­
of the test model. ployed. The numerical model adopts the hysteretic damping algorithm
representing the variations in shear modulus and damping ratio of the
soil with the cyclic shear strain, which captures the energy absorbing
Table 2
characteristics of the soil. For this purpose, the default model built-in
Physical properties of the test soil.
tangent-modulus function is adopted in this study.
Physical properties Symbol Properties (Kwon and Yoo, 2016) obtained a G/Gmax-γ (cyclic strain) curve of
Jumoonjin sand from triaxial and resonant column tests. Based on the
Specific gravity Gs 2.65
Maximum dry density γdmax 16.2 kN/m3 test results, they determined the input values of L1 and L2 necessary for
Minimum dry density γdmin 13.6 kN/m3 the hysteretic damping model. L1 and L2 are the coefficients that de­
Effective particle size D10 0.38 termine the decrease rate and decrease starting point, respectively, of
D30 particle size D30 0.49 the G/Gmax value in G/Gmax-γ curve. L1 and L2 were −3.65 and 0.5,
D60 particle size D60 0.61
Uniformity coefficient Cu 1.59
respectively. Fig. 1 shows the G/Gmax-γ (cyclic strain) curve obtained by
Coefficient of curvature Cc 0.99 applying the default model and the damping ratio curve calculated in
USCS – SP FLAC 3D.
Numerical distortion of the propagating wave can occur in dynamic
analysis. The numerical accuracy of wave transmission is influenced by
both the frequency content of the input wave and the wave speed
The pile foundation was modeled as a linearly elastic material, characteristics of the system. To precisely represent wave transmission
while the surrounding ground was idealized as an elasto-plastic mate­ through a model, (Kuhlemeyer and Lysmer, 1973) indicated that the
rial that deformed plastically according to the Mohr-Coulomb criteria. spatial element size, Δl, must be smaller than approximately one-tenth
Solid elements were adopted to simulate the pile foundation con­ to one-eighth of the wavelength associated with the highest frequency
sidering a 17 m height and a 0.914 m outer diameter of the single pile. component of the input wave. For instance:
The pile was assumed to be a typical steel pipe pile. The steel pipe pile
was approximated as an equivalent solid pile according to the flexural l
10 (2)
rigidity of the steel pipe pile. This approximation was used for simpli­
city and to avoid a complicated and expensive mesh. The surrounding where λ is the wavelength associated with the highest frequency
soil consisted of a single-layer system. The soil conditions are different component that contains appreciable energy. Therefore, the mesh size
according to the shear wave velocity of the soil. The shear wave of the 3D model was estimated by Eq. (2).

3
H. Lim and S. Jeong Computers and Geotechnics 126 (2020) 103657

length of 1,500 mm. The maximum sample weight, acceleration and


frequency are 2 tons, 1.1 g and 50 Hz, respectively. A soil box that was
1,200 × 600 × 800 mm in size was constructed from transparent
polycarbonate and aluminum plates. Sponge pads that were 50 mm
thick were placed on the sidewalls of the soil box to reduce the re­
flection waves in the shaking direction. Fig. 3 shows the 1 g shaking
table and soil box.
Jumoonjin sand, classified as poorly graded sand (SP, according to
the Unified Soil Classification System), with a specific gravity
Gs = 2.65, an effective grain size D10 = 0.38 and a uniformity coef­
ficient Cu = 1.59, was used for the 1g shaking tests. The properties of
the test soil are listed in Table 2. The hysteretic damping curve for the
Jumunjin sand proposed by (Kwon and Yoo, 2016), shown in Section 2,
was used.
The model pile was made considering the similitude law proposed
by (Iai, 1989). In this study, Type 2 of Iai’s similitude law was adopted
Fig. 4. Typical mesh for validation.
because the model ground consisting of dry sand was expected to show
cyclic mobility behavior (Choi et al., 2015). The prototype is assumed
to be a general steel pipe pile with a diameter of 0.9144 m, a thickness
of 0.014 m and a length of 17 m. The pile is a flexible and long pile. The
To accurately conduct dynamic analysis, the lateral boundary con­ model pile was made of aluminum alloy (6061-T6) and had a hollow
ditions of the numerical model must consider the free-field ground circular section with an outer diameter of 2.0 cm. The embedment
motion without a structure and pile foundation. To simulate the dis­ depth was 64 cm. The properties of the model pile are summarized in
sipation of seismic waves at the horizontal model interface, a free-field Table 3.
boundary of infinite elements was used. Fig. 2 shows the typical mesh Fig. 4 shows the typical mesh used in this study. The pile was em­
used in this work and the general configuration of the boundary con­ bedded in dense sand with a relative density of 80%. The pile was
ditions. The overall dimensions of the model boundaries comprise a modeled as a fixed pile similar to the 1 g shaking table test, in which the
width of 10 times the pile diameter (D) from the pile center. The results pile was fixed, and a rock-socketed pile was simulated. The material
obtained by (Remaud, 1999) demonstrated that there was no interac­ properties of the soil and pile are listed in Table 4.
tion effect between the piles when the pile spacing was larger than 10D. The input acceleration was 0.154 g, and dynamic loading was ap­
Therefore, the boundary between the near- and far-field regions was plied to the base as a sine wave. The superstructure was modeled by
determined to be located at 10D from the pile center. The outer varying the unit weight at the pile head. The loading frequency was the
boundary of the mesh is a free-field boundary to reduce wave reflection. resonant frequency with the largest dynamic response. The resonant
A bilinear Mohr-Coulomb element is employed to simulate the soil-pile frequency was calculated by applying the sine wave to each frequency
interface. The interface element is treated as a zone of virtual thickness. and generating the largest response to the superstructure. The resonant
The mass of the superstructure was modeled by adjusting the unit frequency of entire system combined with the soil, pile and super­
weight of the superstructure. structure needs to be analyzed in this study. Therefore, the analysis was
performed under various frequency conditions in order to obtain the
resonant frequency of the soil-pile system considering the soil-pile in­
3. Validation by 1 g shaking table tests teraction.
Fig. 5 shows the acceleration response of the superstructure to the
The validity of the three-dimensional finite-element model was loading frequencies. As shown in Fig. 5, the largest response was ob­
evaluated by comparing the results of the existing study to those of (Lim tained at a loading frequency of 10 Hz, so a resonant frequency of the
and Jeong, 2018). This study examines the effect of bedrock accelera­ soil-pile system of 10 Hz was adopted. Fig. 6 compares the horizontal
tion on full-scale soil-pile systems under dynamic loading. However, the displacement of the superstructure obtained by numerical analysis with
experimental results for full-scale soil-pile systems are insufficient. the experimental results. The maximum displacements of the super­
Therefore, small-scale model tests were used to verify the suitability of structure revealed that the present analysis results were in good
the numerical solutions. The paper presents a brief description of the agreement with previous results. Fig. 7 compares the ground accel­
1 g shaking table test results and provides the details of the constitutive eration obtained by numerical analysis with the experimental results.
model and parameters used for the component materials in the model. The acceleration amplitudes confirmed that the numerical analysis
The shaking table has a platform width of 1,500 mm and a platform

Table 4
Summary of material properties for validation.
Properties γ (kN/m3) Vs (m/s) ν K (Pa) G (Pa) c (Pa) ϕ (°)

Sand 15.6 310 0.35 4.471 × 109 1.490 × 109 0 40


(Dr: 80%)
Pile 27 – 0.2 2.263 × 1010 1.697 × 1010 – –
(D = 0.02 m)

4
H. Lim and S. Jeong Computers and Geotechnics 126 (2020) 103657

results are consistent with the model test results. However, acceleration
and displacement phase differences are observed in Figs. 6 and 7. The
reason is that the loading frequencies in the model test and numerical
analysis are 10 Hz and 10.86 Hz, respectively, to compare the dynamic
response when the resonant frequency of the soil-pile system is applied.

4. Pseudo-static and dynamic analyses

Pseudo-static methods were proposed by (Abghari et al., 1995) and


(Tabesh and Poulos, 2001) for nonliquefying soil, and the inertial force
acting at the pile head is represented by the product of the cap mass and
the spectral acceleration (Dowrick, 1977). In this study, dynamic ana­
lysis, in which the soil was modeled as fully nonlinear, was first per­
Fig. 5. Acceleration of the superstructure with the loading frequency. formed to compare the results given by the pseudo-static method with
the results given by the dynamic finite-difference method. The peak
ground acceleration was obtained through dynamic analysis, and the
equivalent static load was determined using the peak ground accel­
eration and mass of the superstructure. The equivalent static load was
applied to the pile head to carry out pseudo-static analysis. The soils
were sandy soils with relative densities of 35, 65 and 85%. The input
wave was a sine wave. The input accelerations ranged from
0.11 g to 0.4 g, and the loading frequency were the resonant fre­
quencies of the soil-pile systems that realized the greatest dynamic
response to the superstructures. A series of numerical analyses in this
study is summarized in Table 5. Fig. 8 shows the peak superstructure
Fig. 6. Displacement of the superstructure. acceleration according to loading frequencies ranging from 2 ~ 20 Hz
when the acceleration amplitude is 0.154 g in dense sand (Dr: 85%).
The greatest acceleration response is attained when the loading fre­
quency is 11 Hz. Thus, 11 Hz is used as the resonant frequency of the
soil-pile system under this condition. Fig. 9 shows the resonant fre­
quency of the soil-pile system according to the acceleration amplitude
in dense sand (Dr 85%). As the acceleration amplitude increases, the
resonant frequency of the soil-pile systems decreases. This is similar to
the results reported by (Lim and Jeong, 2018), who reported a decrease
in the resonant frequency for each acceleration amplitude due to the
reduced elastic modulus of the soil. In other words, the resonant fre­
quency of the soil-pile system varies depending on the soil conditions
and input acceleration. Therefore, a resonant frequency corresponding
to each input acceleration and soil condition was used as a loading
frequency.

5. Results and discussion

Figs. 10 and 11 show the results of the pseudo-static and dynamic


analyses in sandy soil at a relative density of 85%. Fig. 10 shows the
displacement distribution of the pile with the depth when the maximum
displacement appears in the superstructure. Pseudo-static analysis
provides a larger pile displacement than that of dynamic analysis when
Fig. 7. Ground acceleration with the depth: (a) surface; (b) depth of 60 cm. the input acceleration is 0.154 g. In contrast, when the input accel­
eration is 0.4 g, dynamic analysis provides a larger pile displacement

Table 5
Summary of numerical analyses conducted.
Method Soil condition Acceleration amplitude (g) Loading frequency (Hz)

a
- Dynamic analysis Loose sand (Dr 35%) 0.11–0.22 fr
- Pseudo-static analysis Medium sand (Dr 65%) 0.11–0.3
Dense sand (Dr 85%) 0.154–0.4

a
Note: fr = resonant frequency of the soil-pile system, fr depends on soil conditions and acceleration amplitude.

5
H. Lim and S. Jeong Computers and Geotechnics 126 (2020) 103657

Fig. 8. Peak superstructure acceleration (acceleration amplitude of 0.154 g, Dr


85%).

Fig. 9. Resonant frequencies of soil-pile systems according to the acceleration


amplitude (Dr 85%).

than that of pseudo-static analysis. Fig. 11 shows the distribution of the


bending moment of the pile, which is dependent on the depth at which
the superstructure produced the maximum dynamic response. Pseudo-
static analysis provides a larger pile bending moment than that of dy­
namic analysis when the input acceleration is 0.154 g. However,
pseudo-static analysis provides a smaller pile bending moment than
that of dynamic analysis when the input acceleration is 0.4 g. In other
words, if the input acceleration is 0.154 g, the earthquake-resistant Fig. 10. Comparison of the pile displacement in dense sand (Dr 85%): (a) ac­
design of the pile using pseudo-static analysis is considered a stable celeration amplitude of 0.154 g; (b) acceleration amplitude of 0.4 g.
design. However, if the input acceleration is 0.4 g, pseudo-static ana­
lysis underestimates the pile displacement, which results in an unstable the peak ground acceleration. In other words, the point where the two
pile. curves intersect (0.3 g) indicates the peak bedrock acceleration, where
Fig. 12 shows a comparison of the p-y curves. When the input ac­ pseudo-static analysis is appropriate from the viewpoint of a con­
celeration is 0.154 g, dynamic analysis provides a stiffer p-y curve than servative design.
that of pseudo-static analysis. However, when the input acceleration is Based on the results, the peak earthquake acceleration up to which
0.4 g, the p-y curve of pseudo-static analysis appears stiff. pseudo-static analysis can be applied is determined. Fig. 14 reveals that
To determine the cause, the peak superstructure and surface ac­ an input acceleration occurs that is higher than the surface acceleration
celerations were compared according to the acceleration amplitude. as a result of the pile head acceleration being amplified by the ground
Fig. 13 shows that when the peak bedrock acceleration increases, the condition. Therefore, the applicability of pseudo-static analysis is
peak superstructure acceleration increases, but the peak ground accel­ quantified using a trend line. As a result, pseudo-static analysis can be
eration decreases. This phenomenon seems to occur due to the occur­ applied when the bedrock acceleration is lower than 0.3 g in sand at a
rence of hysteretic damping in the soil. Because the shear modulus of relative density of 85%. For sand at a relative density of 65%, pseudo-
soil is a function of the shear strain level, with the increase in ampli­ static analysis is stable below 0.18 g, and for sand at a relative density
tude, the shear strain in the soil mass increases, which causes a decrease of 35%, stable pseudo-static analysis is expected to be obtained at
in the shear modulus and consequently an increase in damping. When 0.13 g or lower.
the peak bedrock acceleration is 0.3 g, the load on the superstructure
caused by the earthquake is larger than the equivalent load due to the 6. Conclusions
maximum ground surface acceleration. Conversely, when the maximum
bedrock acceleration is 0.3 g or lower, the load on the superstructure This paper investigates the dynamic response of soil-pile systems by
affected by the earthquake is smaller than the equivalent load due to comparing dynamic and pseudo-static analysis results. The focus of this

6
H. Lim and S. Jeong Computers and Geotechnics 126 (2020) 103657

1000

600

Soil resistance (kN/m)


200

-200

-600
Psudo-static analysis

Dynamic analysis
-1000
-2 -1 0 1 2
(a) Pile deflection (mm)

1200

800

Soil resistance (kN/m)


400

-400

Dynamic analysis
-800
Psudo-static analysis
-1200
-6 -1 4
(b) Pile deflection (mm)

Fig. 12. Comparison of the p-y curve in dense sand (depth 1.0 m, Dr 85%): (a)
Fig. 11. Comparison of the bending moment in dense sand (Dr 85%): (a) ac­ acceleration amplitude of 0.154 g; (b) acceleration amplitude of 0.4 g.
celeration amplitude of 0.154 g; (b) acceleration amplitude of 0.4 g.

system compared to dynamic analysis at peak bedrock accelerations


study is the effect of the peak bedrock acceleration on the earthquake- lower than 0.3 g, 0.22 g and 0.13 g for dense, medium and loose
resistant design of soil-pile systems. From the findings of this study, the sands, respectively.
following conclusions can be drawn:

1. Numerical analysis was conducted on full-scale soil-pile systems. In CRediT authorship contribution statement
numerical analysis, the interface of the soil-pile system, free-field
boundary, hysteretic damping and dynamic properties of the soil Hyunsung Lim: Methodology, Software, Validation, Formal ana­
were appropriately considered. The comparison with 1 g shaking lysis, Data curation, Investigation, Writing - original draft. Sangseom
table test results showed that the numerical model was consistent Jeong: Conceptualization, Methodology, Visualization, Supervision,
with the general trend observed in the model tests. Writing - review & editing.
2. As the peak bedrock acceleration increases, the peak superstructure
acceleration increases, while the peak ground surface acceleration
decreases due to the hysteretic damping and nonlinear behaviors of Declaration of Competing Interest
the soil.
3. The dynamic behavior of the soil-pile systems is highly influenced The authors declare that they have no known competing financial
by the bedrock acceleration. Based on the comparison of the dy­ interests or personal relationships that could have appeared to influ­
namic and pseudo-static analysis results, pseudo-static analysis ence the work reported in this paper.
more conservatively predicts the dynamic behavior of the soil-pile

7
H. Lim and S. Jeong Computers and Geotechnics 126 (2020) 103657

1 Acknowledgments
0.9
Response Acceleration (g)

0.8 This work was supported by the Basic Science Research Program
0.7 through the National Research Fund of Korea (NRF) (No.
0.6 2018R1A6A1A08025348) and the Ministry of Land, Infrastructure and
0.5
Transport (Grant No. 20SCIP-B119947-05) of the Republic of Korea,
0.4
and we express our gratitude to them.
0.3
0.2 Superstructure_acc References
0.1 Surface_acc
0
0 0.1 0.2 0.3 0.4 0.5 Abghari, A., Chai, J., 1995. Modelling of soil–pile–superstructure interaction for bridge
foundations. In: Turner, J.P. (Ed.), Proceedings of Performance of Deep Foundations
Acceleration Amplitude (g) under Seismic Loading. ASCE, New York, pp. 45–49.
Al-Baghdadi, T.A., Brown, M.J., Knappett, J.A., Ishikura, R., 2015. Modelling of laterally
Fig. 13. Comparison of the superstructure and surface accelerations in dense loaded screw piles with large helical plates in sand. Proceedings of the 3rd
sand (Dr 85%). International Symposium on Frontiers in Offshore Geotechnics. Oslo Norway.
Bhaduri, A., Choudhury, D., 2020. Serviceability-based finite-element approach on ana­
lyzing combined pile-raft foundation. Int. J. Geomech. 20 (2), 04019178.
Chatterjee, K., Choudhury, D., Poulos, H.G., 2015. Seismic analysis of laterally loaded pile
under influence of vertical loading using finite element method. Comput. Geotech.
67, 172–186.
Choi, J.I., Yoo, M.T., Yang, E.K., Han, J.T., Kim, M.M., 2015. Evaluation of 1-G similitude
law in predicting behavior of a pile-soil model. Mar. Georesour. Geotechnol. 33 (3),
202–211.
Dobry, R., Vicente, E., O’Rourke, M., Roesset, J., 1982. Horizontal stiffness and damping
of single piles. J. Geotech. Eng. ASCE 108 (3), 439–458.
Dowrick, D.J., 1977. Earthquake Resistant Design: A Manual for Engineers and Architects.
Wiley, New York.
El Naggar, M.H., Bentley, K.J., 2000. Dynamic analysis for laterally loaded piles and
dynamic p-y curves. Can. Geotech. J. 37 (6), 1166–1183.
El Naggar, M.H., Novak, M., 1996. Non-linear analysis for dynamic lateral pile response.
J. Soil Dyn. Earthquake Eng. 15 (4), 233–244.
Finn, W.D.L., Thavaraj, T., Fujita, N., 2001. Piles in liquefiable soils: seismic analysis and
design issues. Proceedings of the 10th International Conference on Soil Dynamics and
Earthquake Engineering. Philadelphia Pa.
Hamada, M., Sato, H., Nakamura, T., 1994. An experimental and numerical study on
liquefaction-induced ground displacement. In: Proceedings of the 5th US National
Conference on Earthquake Engineering. Chicago USA, pp. 169–178.
Hussien, M.N., Tobita, T., Iai, S., Karray, M., 2014. On the influence of vertical loads on
the lateral response of pile foundation. Comput. Geotech. 55, 392–403.
Iai, S., 1989. Similitude for shaking table tests on soil-structure-fluid model in 1g grav­
itational field. Soils Found. 29 (1), 105–118.
Ishihara, K., Cubrinowski, M., 1998. Performance of large-diameter piles subjected to
lateral spreading of liquefied deposits. Thirteenth Southeast Asian Geotechnical
Conference, Taipei.
Itasca Consulting Group, 2013. FLAC3D (Fast Lagrangian Analysis of Continua in
3Dimensions) User's Guide. Minnesota USA.
Jeong, S., Cho, J., 2014. Proposed nonlinear 3-D analytical method for piled raft foun­
dations. Comput. Geotech. 59, 112–126.
Kagawa, T., 1980. Soil-pile-structure interaction of offshore structures during an earth­
quake. Offshore Technology Conference.
Kagawa, T., Kraft, L., 1981. Lateral pile response during earthquakes. J. Geotech. Eng.,
ASCE 107 (12), 1713–1731.
Kavvadas, M., Gazetas, G., 1993. Kinematic seismic response and bending of free-head
piles in layered soil. Geotechnique 43 (2), 207–222.
Kaynia, A.M., Kausel, E., 1982. Dynamic behavior of pile groups. 2nd Int. Conf. on Num.
Methods in Offshore Piling, Austin, TX, pp. 509–532.
Ko, J., Jeong, S., Lee, J., 2016. Large deformation FE analysis of driven steel pipe piles
with soil plugging. Comput. Geotech. 71, 82–97.
Kwon, S.Y., Yoo, M.T., 2016. Parametric study of dynamic soil-pile-structure interaction
in dry sand by 3D numerical model. J. Korean Geotech. Soc. 32 (9), 51–62 (in
Korean).
Kuhlemeyer, R.L., Lysmer, J., 1973. Finite element method accuracy for wave propaga­
tion problems. J. Soil Mech. Foundation. Div. ASCE 99 (SM5), 421–427.
Lim, H., Jeong, S., 2018. Simplified p-y curves under dynamic loading in dry sand. Soil
Dyn. Earthquake Eng. 113, 101–111.
Liyanapathirana, D.S., Poulos, H.G., 2010. Analysis of pile behaviour in liquefying sloping
ground. Comput. Geotech. 37 (1), 115–124.
Luo, C., Yang, X., Zhan, C., Jin, X., Ding, Z., 2016. Nonlinear 3D finite element analysis of
soil-pile-structure interaction system subjected to horizontal earthquake excitation.
Soil Dyn. Earthquake Eng. 84, 145–156.
Nogami, T., Konagai, K., 1988. Time domain flexural response of dynamically loaded
single Piles. J. Eng. Mech. ASCE 114 (9), 1512–1525.
Norris, G.M., 1994. Seismic bridge pile foundation behaviour. In: Proceedings of the
International Conference on Design and Construction of Deep Foundations, pp.
127–136.
Fig. 14. Comparison of the superstructure and surface accelerations under de
Novak, M., 1974. Dynamic stiffness and damping of piles. Can. Geotech. J. 11 (4),
different soil conditions: (a) Dr: 85% (Vs: 320 m/s); (b) Dr: 65% (Vs: 270 m/s); 574–598.
(c) Dr: 35% (Vs: 170 m/s). Peck, R.B., Hanson, W.E., Thornburn, T.H., 1974. Foundation Engineering, second ed.

8
H. Lim and S. Jeong Computers and Geotechnics 126 (2020) 103657

Wiley & Sons Inc., New York. Earthquake Engineering. Paris France.
Penzien, J., 1970. Soil-pile foundation interaction. In: Wiegel, R.L. (Ed.), Earthquake Sun, C.G., Kim, H.J., Jung, J.H., Jung, G., 2006. Synthetic application of seismic piezo-
engineering, chapter14. New York. Prentice-Hall. cone penetration test for evaluating shear wave velocity in Korean soil deposits.
Phanikanth, V.S., Choudhury, D., Reddy, G.R., 2013. Behavior of single pile in liquefied Mulli-Tamsa 9 (3), 207–224 (in Korean).
deposits during earthquakes. Int. J. Geomech. 13 (4), 454–462. Tabesh, A., Poulos, H.G., 2000. A simple method for the seismic analysis of piles and its
Rayhani, M., El Naggar, M., 2008. Numerical modeling of seismic response of rigid comparison with the results of centrifuge tests. In: Proceedings of the 12th World
foundation on soft soil. Int. J. Geomech. 8 (6), 336–346. conference on Earthquake Engineering, Auckland. New Zealand, pp. 1203.
Remaud, D., 1999. Piles under Lateral Forces: Experimental Study of Piles Group (Ph.D. Tabesh, A., Poulos, H.G., 2001. The effect of soil yielding on seismic response of single
Dissertation). University of Nantes France. piles. Soils Found. 41 (3), 1–16.
Roy, J., Kumar, A., Choudhury, D., 2018. Natural frequencies of piled raft foundation Tabesh, A., Poulos, H.G., 2001. Pseudostatic approach for seismic analysis of single piles.
including superstructure effect. Soil Dyn. Earthquake Eng. 112, 69–75. J. Geotech. Geoenviron. Eng. ASCE 127 (9), 757–765.
Sakajo, S., Chai, J.C., Nakajima, K., Maeda, M., 1995. Effect of group pile on liquefaction Tombari, A., El Naggar, M.H., Dezi, F., 2017. Impact of ground motion duration and soil
resistance of sandy ground. In: Proceedings of the First Conference on Earthquake non-linearity on the seismic performance of single piles. Soil Dyn. Earthquake Eng.
Geotechnical Engineering. Tokyo Japan, pp. 755–760. 100, 72–87.
Shahrour, I., Ousta, R., 1998. Numerical analysis of the behaviour of piles in saturated Zheng, J., Susuki, K., Ohbo, N., 1996. Evaluation of sheet pile-ring countermeasure
soils under seismic loading. Proceedings of the 11th European Conference on against liquefaction for oil tank site. Soil Dyn. Earthquake Eng. 15 (6), 369–379.

You might also like