See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/269084042

Study on Bearing Capacity of Pile in Liquefiable and Unliquefiable Soil Layers

Conference Paper · March 2011

DOI: 10.1061/41165(397)380

CITATIONS READS
2 1,110

4 authors, including:

Issa Shooshpasha Maryam Afzali

Babol Noshirvani University of Technology Islamic Azad University Karaj Branch
81 PUBLICATIONS 1,027 CITATIONS 4 PUBLICATIONS 23 CITATIONS

SEE PROFILE SEE PROFILE

Mohammad Ghavami
Babol Noshirvani University of Technology
18 PUBLICATIONS 403 CITATIONS

SEE PROFILE

All content following this page was uploaded by Mohammad Ghavami on 06 December 2015.

The user has requested enhancement of the downloaded file.

Geo-Frontiers 2011 © ASCE 2011 3716

Study on Bearing Capacity of Pile in Liquefiable and

Unliquefiable Soil Layers

I. Shooshpasha 1; M. Afzali Rad 2; M. Ghavami3; H. Kamalijoo3

1
Assist. Prof., Civil Eng. Faculty, Mazandran University (UM), P.O. Box 47144,
Babol, Iran, Tel: 981113231707, Email: [email protected].
2
Faculty Member of the Islamic Azad University, Ghaemshahr branch, Iran,
email: [email protected]
3
B.S. of Civil Engineering. Islamic Azad University, Ghaemshahr branch, email:
[email protected]

ABSTRACT

Increasing usage of piles in harbors on sandy coasts and liquefaction after earthquakes
motivate the study of bearing capacity of pile foundation. Due to high costs of test
performance on piles, computer simulation of piles displacement under various loads are
exercised. Hence, a 2D model of pile-soil has been analyzed by finite difference method
in FLAC software and after controlling the validity of software, the liquefaction potential
of liquefiable soil layer between two unliquefiable layers due to earthquake loading is
checked. Afterwards the pile behavior is analyzed in liquefiable and unliquefiable soil
layers and total skin friction and end bearing capacity of the pile are calculated. For
calculation of pile bearing capacity, force-settlement behavior of pile is noticeable.
Regarding axial force distribution along the pile, we have used axial force reduction
mechanism which occurs by soil shear strength action. This action happens in result of
pile displacement rather than the soil and force transmission from the pile to the soil.

INTRODUCTION

Using deep foundations (piles) during recent decades has significantly

increased. In addition to implementing ordinary and common structure projects
like bridge construction in which deep foundations were used in far past as well,
the development of harbors of the country, offshore structures and also substantial
investments in oil and petrochemical resource section in the southern coasts of the
country has considerably increased the need to use piles. According to the Iran
seismic code 2800, these areas are regarded as regions in which there is a high or
very high risk of earthquake and in which grain orientation mostly includes loose
sand underground water.
On the other hand, stresses doe to earthquakes causes the increase in the
pore water pressure and losing soil strength in areas with saturated earth grain
structure and finally soil gets liquefiable. The seismic design of pile foundation in
liquefiable soil creates complicated issues and problems in design and analysis.
Depending on the liquefaction event, pile foundations are disposed to any damage
due to disregarding liquefiable soil layer in designing. Although piles are used
since long ago in the world and different methods are suggested by researchers for
designing pile, the investigation of bearing capacity and optimum designing with
regard to liquefiable soil layers has been still considered as an important issue.
Geo-Frontiers 2011 © ASCE 2011 3717

Designing pile against axial load is usually the most important factor in
pile design in terms of bearing capacity. In other words, the axial force on pile is
the main deciding factor in determining the bearing capacity of piles in most
applicable models. In this study, the behavior of pile in liquefiable layer between
two unliquefiable layers was studies through modeling and numerical analysis.
Hence, 2D pile-soil models were analyzed in FLAC using finite difference
method. Besides, the effect of such factors as the thickness of the liquefiable layer
on the bearing capacity of pile was studied.

The destruction mechanism of deep foundations due to liquefaction

The effects of liquefaction on deep foundation are very destructive and

costly. Providing that an unliquefiable soil layer moves on a liquefiable soil, such
a displacement is very destructive for pile foundations. Permanent or lateral
displacements of earth or lateral extension whether separately or in combination
with superstructure inertia forces and created torques during the earthquake on the
soil which has been already weaken with the increase in the pore water pressure is
one the important sources of danger and worry. Cracks and fractures of pile in
surface and deep heights, the fracture of pile couplings, the vertical, lateral and
rotational movements of pile head and pile cap are some models observed in
previous earthquakes. This issue influences buildings, bridges, harbor facilities
and equipments and other building in Japan and America during Kobe earthquake
in Japan (1995) and Loma Perita earthquake in America (1989). Most pile
destructions during Nigata earthquake (1964) and Hyogoken Nambu earthquake
(1995) were found in liquefiable lands
The significance of upper and lower liquefiable ranges of liquefiable
layers is clearly illustrated in Figure (1) in an example of pile destruction in
Nigata earthquake. This Figure shows a building which underwent a displacement
of about one meter. The pile(1) is a frictional one bent at the depth of 2 meters and
cracked which was located near ground water level and in the upper range of
liquefiable soil. The pile(2) is a support one cracked at the depth of 8 meters and
under liquefiable layer.

Figure 1. The destruction of frictional and support piles and standard

penetration number in Nigata earthquake (1964).
Geo-Frontiers 2011 © ASCE 2011 3718

The analysis of finite difference and development model of pore water

pressure

2D FLAC software was used for the numerical modeling of pile and soil
liquefaction in this research. This program can appropriately model the behavior
of soil and rock under different loads. Furthermore, since this program is written
in finite difference method, different regions can be created by defining points and
nods. The displacement of regions is defined with the displacement of nods. The
behavior of each element in the region follows the coordination number defined
for it. It means that a suitable model can be specified to different materials
through the knowledge of the behavior of different materials towards stress and
discontinuity and flowing can be observed in different models with regard to the
related mode.
During an earthquake, the application of cyclic shear stresses induced by
the propagation of shear waves causes the loose sand to contract, resulting in an
increase in pore water pressure. Volume change for saturated sand is only due to
drainage. When drainage is not possible, volume reduction leads to rise in the
pore water pressure which influences the behavior of soil through changing
effective stress. In fact, the created pore pressure is somehow a created secondary
effect in model while it is assumed that this issue is the main response to dynamic
loading. The main effect of this loading includes an irrevocable volume
contraction which takes place in the network of soil grains. Martin et al. presented
the following experimental equation which indicates incremental volume strain:
§ CH 2 ·
'H vd C1 (J  C 2 H vd )  ¨ 3 vd ¸ (1)
¨J C H ¸
© 4 vd ¹

In which H vd is plastic volume strain, 'H vd is the plastic volume strain

changes, J the shear strain range of the period and C1 , C2 , C3 , C4 are fixed
parameters which depend on the relative density of sand Byrne et al. changed
equation (1) and presented the following simplified one:
§  C 2H vd · (2)
'H vd JC1 exp¨¨ ¸¸
© J ¹

The quantity of Byrne fixed numbers i.e. C1 , C2 are presented as a relative

density function and SPT number:
C1 7600( Dr ) 2.5 (3)

Dr 15( N 1 )160/ 2 (4)

0 .4
C2 (5)
C1

SOIL AND PILE MODELING

The soil sample was of grain type in three layers and a depth of 15 meters
that the elasticity module, specific weight and other parameters needed by soil for
Moher-Colomb model achieved through geotechnical studies are presented in
Table (1). The parameters of shear resistance between soil and pile are considered
Geo-Frontiers 2011 © ASCE 2011 3719

to be equal to 1/3 of the resistance parameters of soil using Moher-Colomb model.

Pile parameters for elastic behavioral model are also presented in Table (2).
It is necessary to model pile for axial static load in finite difference
software before preparing the model for pile analysis in liquefiable earthquake
during earthquake and the results of the program were checked with the results
obtained from theoretical relations of determining the bearing capacity of pile. In
order to do so, a pile was modeled under axial load and the force-settlement
behavior of the pile was taken into account. Due to geometrical and loading
symmetry, modeling was conducted as axial symmetry and only half of the pile
and the soil around it were used in model.

Table 1. Parameters for different soil layers in Moher-Colomb model

Permeabilit Angle of Shear Bulk Elasticity
Cohecen Poisson' Dry Density depth
y coefficient internal modules moulds module
kPa s ratio kN/m3 (m)
m/s friction kPa kPa (Kpa)

1 u 107 5 30 11923 25833 0.30 31000 14.0 0-2

1u10 6 0 35 12000 20000 0.25 30000 12.5 2-6

8
1u10 5 30 12308 26667 0.30 32000 15.0 6 - 15

Table 2. Parameters for pile

Diameter Length Shear Bulk Poisson' s Elasticity Unite
(m) (m) Module Module Ratio Module Weight
(Kpa) (Kpa) (Kpa) (KN/m3)
1.333u 107
1.0 10.0 1u 10 7 0.20 2.4 u 10 7 24

Axial force and frictional bearing along the pile

The behavior of pile under static axial load can be seen in Figure (2).
Under static axial force in different stages, the displacement of pile to nearby soil
increases which itself decreases the axial force exercised on pile with the increase
in depth. Frictional resistance changes are also shown in Figure (3). The body
resistance of pile is maximized at first and in the displacement of lower pile body
and, the end resistance of the pile is activated with the increase in end pile
displacement. Axial force changes to pile settlement are also presented in Figure
(4).
Geo-Frontiers 2011 © ASCE 2011 3720

Figure 2. Axial force changes along the pile for 2000, 1500 and 500 kilo
Newton loads

Figure 3. Body resistance changes

Figure 4. Axial force changes for
along the pile for 500 kilo Newton
pile settlement
load level

Comparison between numerical results and load bearing capacity relations

In order to compare results with present relations for determining the

bearing capacity of pile, we can use Mayerhof, Vesic and Janbu relations for the
end bearing of pile and also use D and O relations in order for determining the
frictional bearing of the pile as a summary of computations is given below:
Using Meyerhof relation, we will have: Qt At (CN c*  q N q* ) 4172kN

Using Vesic relation, we will have: Qt At (CN c*  V 0c N V* ) 2719kN

Using Janbu relation, we will have: Qt At (CN c*  q N q* ) 2042kN

Using O relation, we will have: f s O (V vc  2Cu ) 7.85 o Qs ¦ A 'Lf

s s 247kN

Using D relation, we will have:

Geo-Frontiers 2011 © ASCE 2011 3721

rs DCu  V vc K tan G 9.47 o Qs ¦ r A 'L

s s 298kN

Using above calculations, we will obtain the bearing capacity of pile:

4172  2719  2042 247  298
Qult Qs  Qt  3250kN
3 2

The results in Figure (4) indicates that the numerical model has reached a
capacity of 3000 KN but it still shows a rising trend and almost conforms with
conducted calculations. However, it should be also mentioned that all capacities
calculated for pile in this research without factor of safety.

THE DYNAMIC MODEL

Dynamic loading is exercised as acceleration-time history input to the
model. Given liquefying Roadbar and Manjil regions in Gilan Province during the
earthquake, Manjil earthquake is scaled with following parameters and is
exercised on the model. The properties of Manjil earthquake exercised on the
model are presented in Table 3.
Table 3. Properties of Manjil earthquake
Distance Peak Ground year magnitude location earthquake
from fault Acceleration (PGA)
12 Km 0.55 1369 7.1 Iran Manjil

Assessing liquefaction potential of the model

As it was mentioned earlier, Bayern experimental relation which states
incremental volume strain is used for the dynamic analysis of soil in liquefaction
model and increased of execs pore water pressure. This incremental volume strain
appears due to a cycle of simple shearing load in dry sand. The soil parameters for
liquefying analysis are presented in Table (4).

Table 4. Soil parameters for Finn model

Depth(m) N 60 C1 C2 C3

0-2 20 0.206 1.946 0

2-6 10 0.489 0.818 0
6 - 15 30 0.124 3.228 0

Figure (5) depicts a graph which indicates the pore water pressure to total
stress against soil depth (pile height). As it can be seen, soil is almost liquefiable
at a depth of 2 to 6 meters.
Geo-Frontiers 2011 © ASCE 2011 3722

Figure 5. Variation of pore water pressure to total stress against soil depth
due to earthquake

Earthquake and liquefaction-caused displacement under the pile

As it was mentioned, FLAC was used for numerical modeling of pile and
soil liquefaction. After checking the numerical model for static loading with
current relations and studying the liquefaction of modeled soil between two
unliquefiable soil layers, we will now study the axial bearing capacity of pile in
liquefaction model of soil.
As we expected, soil and pile settlement is higher than in static sate in the
same load level due to soil compaction as a result of earthquake and soil
liquefaction. A comparison of displacement scales under the pile in dynamic and
static models is presented in Table (5). As it can be seen, pile head displacement
in dynamic model is almost 10 times more than pile head displacement in static
model.

Table 5. Comparison of displacement in static and dynamic model for 500,

1000 and 1500 KN load levels

Load level Displacement in dynamic Displacement in static

(kN) model(cm) model (cm)
500 1.907 0.093
1000 8.478 0.858
1500 22.94 3.31

Axial force and body resistance distribution along the pile in soil liquefaction
model and a comparison with static mode

Variation of Axial load along the pile in liquefaction soil model is

illustrated in Figure (6) and also is compared to the static model of introduced
load levels. Since shear resistance is inclined to zero in liquefiable soil and soil
gets liquidized, pile will also have no frictional resistance which can be seen in
Figure (7). Given that soil is liquefied at a height of 2 to 6 meter, frictional
Geo-Frontiers 2011 © ASCE 2011 3723

resistance decreases and the remained bearing capacity goes to end bearing
capacity. In Figure (7), there is a comparison between changes in the bearing
capacity of body along the pile for 500 KN loading level in static and dynamic
models. As it can be seen in this Figure, frictional resistance decreased in
proportion to static model.

Figure 6. Comparison of axial force along the pile length for 500, 1000 and
1500 KN load levels in static and dynamic models

Figure 7. Comparison between friction resistance along the pile for 500 KN
load level in static and dynamic model

Load-settlement curve of the pile in soil liquefaction model and a comparison

with static model
Variation of Axial force along the pile versus pile settlement is presented
in Figure (8). The present results reveal that the numerical data in liquefaction
model has reached 1500 KN of the total capacity and the bearing capacity of soil
has almost 50% decreased versus the numerical data is static model which reached
about 3000 KN of the total capacity. Hence, axial force 50% decreases in a fixed
end pile displacement.

Load-displacement curve of the pile in variable liquefiable soil layer

thickness model
Geo-Frontiers 2011 © ASCE 2011 3724

A comparison between axial force Variation along the pile versus pile
settlement in the liquefaction model of the soil between unliquefiable soil layers
(with the loading of Manjil earthquake) where the liquefiable soil layer started at a
height of 2 meters from ground level but the thickness of the liquefiable layer was
variable between 2 to 8 meters is presented in Figure (9). As it can be seen, the
bearing capacity of pile decreases with an increase in the thickness of the
liquefiable layer.

Figure 8. Comparison between axial force along the pile length versus
settlement in the static and dynamic models.

Figure 9. Comparison between axial force variations along the pile versus
settlement in soil liquefaction model between the two unliquefiable soil layers

CONCLUSION

According the conducted pile modeling in liquefiable and unliquefiable

soils by the finite difference software called FLAC; here is an outline of the
results:
1. As expected, soil and pile settlement was (up to 10 times) higher than
static mode in the same load level due to soil compaction and settlement as
a result of earthquake and soil liquefaction.
2. Axial force distribution along the pile has changed in proportion to static
model and soil has almost no shear resistance in liquefiable layer;
 Geo-Frontiers 2011 © ASCE 2011 3725

consequently, the frictional resistance of pile approaches zero.

Accordingly, no resistance should be assigned to pile in liquefiable layer.
3. The total bearing capacity of pile has 50% decreased in proportion to static
model. It can be concluded that reduction in bearing capacity from soil
liquefaction can be made up for through exercising factor of safety (2) in
static model in order for soil liquefaction not to cause pile destruction due
to reduction in bearing capacity. However, it should be mentioned that the
results of static bearing capacity exclude factor of safety and another
general factor of safety, apart from the factor of safety (2) from this
analysis, should be exercised.
4. Parameter studies were conducted concerning the thickness of the
liquefiable layer from ground level which revealed that the total bearing
capacity of pile decreases with an increase in the thickness of the
liquefiable layer.

REFERENCES

Abdoun Tarek, Dorby Ricardo, (2002), “Evaluation of the foundation response to

lateral
spreading.” Soil Dynamic and Earthquake Engineering, 22, pp. 1051-
1058.
Byrne PM, McIntyre J, (1994), “Deformation in granular soils due to cyclic
loading.” ASCE GSP, 101, pp. 1864-1896.
Editorial Committee for the Report on the Hanshin – Awaji Earthquake Disaster,
(1998), “Report on the Hanshin – Awaji earthquake disaster.” series
volume 4, Maruzen, p. 548.
Finn W.D.L, Fujita N, (2002), “Piles in liquefaction soils: seismic analysis and
design issues.” Soil Dynamic and Earthquake Engineering, 22, pp.
731-742.
Hamada M, Yasuda S, Isoyama R, Emoto K, (1986), “Study on liquefaction
induced permanent ground displacements.” Report for the development of
earthquake prediction, Tokyo Japan.
Itasca Consulting Group, (2002), “FLAC 2D ver5.00 manual.” Minnesota, USA.
Martin GR, Finn LWD, Seed HB, (1975), “Fundamentals of liquefaction under
cyclic loading.” ASCE, 101, pp. 423-438.
Pile Damage Investigation Committee, (1996), “Report of damage to building
foundation during Hyogoken - Nanbu earthquake.” Kinki Branch of AIJ,
p. 399.
Tokimatsu K, (1999), “Performance of pile foundations in laterally spreading
soils.” proceeding of the Second International Conference
Earthquake Geotechnical Engineering, Vol 3, pp. 957-964.
Youd, T. L., et al. (2001). “Liquefaction resistance of soils: summary report from
the1996 NCEER and 1998 NCEER/NSF workshops on evaluation of
liquefaction resistance of soils.” J. Geotech. Geoenviron. Eng. 127(10), pp.
817–833.
Fakharian k, Ealami A. (2005). “ Bearing Capacity of Pile Axial”. Tehran, Iran
Ministry of Roads and Transportation, Transportation Research Institute.
(In Persian).

View publication stats