Sei sulla pagina 1di 18

Behavior of End Bearing Piles in Liquefiable Soils during

Earthquakes

Bahia S. EL REFAI and Sara A. FAYEK

Abstract
Piles have been the first choice for geotechnical engineers in adopting deep foundation mainly for bridges
design. In a review of bridges damage during major earthquake, bridges displacement and partial failure or
total failure has been noticed widely. So that bridges failure is mainly taking into consideration to
understand the behavior on deep foundation. The liquefaction effect is investigated on End bearing pile (fix
top and free top) using Plaxis 3D. The pile geometry will be conserved in both cases, in addition to the
surrounding soil and earthquake effect. The behavior of pile changes depending on it type, where the
optimum results and most economical type were found using an end bearing pile with fixed top.

Keywords: end bearing, fix top, free top, liquefaction, analysis, dynamic behavior, factor of safety

Introduction
An overview of piles loading, piles can be subjected to vertical and horizontal loads coming from the
substructure under static and dynamic effect. The liquefaction of soil during earthquake will contribute to
have excess stress in pile section leading to failure. The calculated parameters usually used to compare the
behavior of the piles under the effect of liquefaction where pore water pressure is generated during an
earthquake are bending moment, shearing force and settlement.

Failure of pile foundation in liquefiable soil was observed by settlement and lateral displacement (according
to Hamada). However, Damages usually occurs at three different locations: at the pile head (for fixed-head
piles), at a depth of 1-3m below the pile cap (for free-head piles) and at the interface soil layers (confirmed
by others such as Tachikawaet al., Shamoto et al., and Onishi et al). Moreover other factor like crushing of
pile shaft, pile buckling, pile settlement and bearing capacity failure are considered as major factor affecting
pile stability under earthquake.

The main purpose of this study is to understand the behavior of end-bearing pile by comparing the behavior
of two different piles condition at their top having the same circumstances. Our study is simulated with
Plaxis 3D, where the soil is modeled using finite elements analysis. Several comparisons will be made using
the bending moment, shearing force and settlement of the pile to find out the safest and most economical
solution.
The liquefiable soil-pile interaction decrease during earthquake and can cause major settlement under the
pile, pile buckling and the redistribution of horizontal load over the pile (confirmed by Subhamoy
Bhattacharya). Common generalizations that skin friction is lost during seismic action, the pile will be only
supported by its end bearing capacity and the importance of skin friction is only considered during static
load (the design of pile in liquefiable soil should ignore the skin friction resistance during calculation).
Objectives
The main objective of this research is to understand the behavior of pile under the effect of liquefaction and
to discuss the safest and more economical solution. Also this study focuses on understanding the new forces
generated by seismic action and estimate a new factor of safety.

Analysis
The study consists of evaluation models by creating loaded concrete pile driven into UBC sand criterion,
end-bearing piles are modulated in the computational program Plaxis 3D. This program is based on a finite
element method with different monotonic and cyclic stress paths. UBC criterion consists of a powerful
approach in order to model the onset of the liquefaction phenomenon.

Plaxis 3D offers realistic three dimensional analysis in the geotechnical engineering industry and in
researches. So that understanding the behavior stability and deformation in a model will be accessible. In
the geotechnical model used, the pile and loads can be activated and deactivated depending on the
conditions needed in every phase (four different phases). With advanced features including UBC Sand, it
is possible to induce the behavior of liquefiable soil. The UBC3D-PLM is a non-linear elasto-plastic
fully coupled constitutive model. The UBC3D uses the Mohr-Coulomb criterion in 3-D principal stress
space in defining the yield surface of the model (Plaxis 3D Manuel 2017).

The UBC3-PLM model involves two yield surfaces of the Mohr-Coulomb type; the primary surface
evolves according to an isotropic hardening law and the secondary yield surface is a simplified kinematic
hardening rule. Pile is modeled as an embedded beam with a 9 m length and in order to apply the
earthquake signal on the lateral boundary, a dynamic multiplier is created by the input signal. A table is
created using the data from the smc loaded file as acceleration points. The models developed consist of
three static phases where stiff sand and sand layers are modeled (8m of sand on the top of 4 m of stiff
sand) and of a dynamic phase where UBC sand and sand are modeled (8m of UBC sand on the top of 4
m of stiff sand).

The below tables show the properties of all materials used in the Plaxis 3D model.
TABLE 1: Pile properties

PILE-30
Material Type Elastic
E kN/m² 30.00E6
γ kN/m³ 25.00
Beam Type Predefined
Diameter m 0.3000
A m² 0.07069
I_2 m⁴ 0.3976E-3

TABLE 2: Sand Properties

Identification Number 2
Material Model Mohr-Coulomb
γ_unsatkN/m³ 19.70
γ_sat kN/m³ 21.80
e_init 0.5000
e_min 0.000
e_max 999.0
Rayleigh α 0.000
Rayleigh β 0.000
E kN/m² 18.57E3
ν 0.3000
G kN/m² 7143
E_oed kN/m² 25.00E3
c_ref kN/m² 0.000
φ(phi) ° 22.00
ψ(psi) ° 19.00
V_s m/s 59.64
V_p m/s 111.6
Skempton-B 0.9783

ν_u 0.4950
K_w,ref /n kN/m² 696.4E3
R_inter 0.6500
K_0,x 0.6254
K_0,y 0.6254
k_x m/day 7.128
k_y m/day 7.128
k_z m/day 7.128
TABLE 3: Identification stiff sand

Identification Number 3
Material Model Mohr-Coulomb
γ_unsatkN/m³ 20.00
γ_sat kN/m³ 17.00
e_init 0.5000
e_min 0.000
e_max 999.0
Rayleigh α 0.000
Rayleigh β 0.000
E kN/m² 75.00E3
ν 0.3000
G kN/m² 28.85E3
E_oed kN/m² 101.0E3
c_ref kN/m² 1.00
φ(phi) ° 31.00
ψ(psi) ° 0.00
V_s m/s 129.0
V_p m/s 241.4
Skempton-B 0.9783
ν_u 0.4950
K_w,ref /n kN/m² 2.812E6
R_inter 0.6500
K_0,x 0.6254
K_0,y 0.6254
k_x m/day 0.00
k_y m/day 0.00
k_z m/day 0.00
TABLE 4: Identification of UBC sand

Identification Number 4
Material UBC3D-PLM
ɣunsat(kn/m3) 19.7
ɣsat(kn/m3) 21.8
einit 0.74
E(kPa) 25000
μ 0.3
G(kPa) 7143
cref 0.0
φ (o) 22.0
φcv(o) 20
φp(o) 22
𝐾𝐺𝑒 854.6
𝑝
𝐾𝐺 250.0
𝐾𝐵𝑒 598.2
me 0.5
ne 0.5
np 0.5
Rf 0.811
PA(kPa) 100.0
σt (kPa) 0.0
fachard 1
(N1)60 7.65
facpost 1

The study includes four stages:

1- Initial phase (soil model – static)


2- Pile (soil model + pile – static)
3- Pile under static load (soil model +pile+ static load at the top – static)
4- Dynamic calculation (soil model +pile+ static load at the top – dynamic).

The types of piles used in the model are the concrete end bearing pile:

1- Free top concrete end bearing pile


2- Fixed top concrete end bearing pile (manifesting the pile cap and tie beam between piles).
Findings and Discussion
1. Concrete End Bearing pile -free top

FIGURE 1: Pile Displacement in Z direction in the Static Phase

During static load only vertical uz is observed in the pile, the value maximum value is occurred at the top
of the pile of 0.312 cm and 0.19 cm at it tip. This is very small value displacement wise.

FIGURE 2: Pile Displacement in X and Z direction in the Dynamic Phase

During dynamic load both horizontal and vertical displacement ux and uz were observed in the pile, the
value maximum value is occurred at the top of the pile of 9.2 cm due of the free conditions given to pile
and 3.7cm of its top. However the vertical displacement increased to 9.6 cm all over the pile.
FIGURE 3: Effective stress-zz with static load

The effective stress during the static phase reaches its maximum under the pile which is equal to 204
kN/m2. The resultant effective stress due to the load acting on the pile is 204-118.9=85.1 kN/m2. However
if there is no skin friction acting along the pile we will find a huge value of stress under the pile, thus
𝐹
𝜎′ = 𝐴 (1)

𝜎 ′ =5661.71kN/m2

Since 𝜎 ′ =85.1 kN/m2, the force on the pile toe is:

F=𝜎 ′ .A (2)

F=𝜎 ′ .A=6.1kN

FIGURE 4: Effective stress-zz with Dynamic load

As it shows that during earthquake that the loose soil is liquefied and the effective stress tends to zero.
The effective stress during the Dynamic phase reaches its maximum under the pile which is equal to
258.7kN/m2, in the stiff soil layer. The value of the effective stress increases in the dynamic phase due to
liquefaction where the pile loses it skin friction and the axial forces increases too.
FIGURE 5: Shear stress-zx with Dynamic load

The shear stress during dynamic phase 215.3 kN/m2, this value occurred between the two different layers
of the soil.

N [kN]
0.00
-10 -8 -6 -4 -2 0
-100.00
-200.00
-300.00
-400.00
-451.24 -500.00
Depth(m)

FIGURE 6: Axial force Due to Static and Dynamic Load


The above figure present the axial force due to static and dynamic load for a single end bearing
pile having high skin friction value, the axial force acting on the head pile is 400 kN, the study shows that
this value increase to 451.24 kN during seismic action.

The original value increased by almost 13%, during seismic action and due to liquefaction, the highest
value as the graph shows is at 3.4 m from the top of the soil.
Q_12 [kN]
400.00
200.00
0.00
-10 -8 -6 -4 -2 0
-32.87
-200.00
-400.00
-513.10 -600.00
Depth(m)

FIGURE 7: Shearing force due to Dynamic Load


The above figure illustrates the shearing force during seismic action, at zero level, the
earthquake provoke a value of 32.87KN, and the shearing forces reaches its maximum of
513.10kN at 6.19m from it top. However, the shearing force is zero in case of static load since
there is no horizontal load is acting on the pile.

M_3 [kN m]
100.00
79.22
50.00

0.00
-10 -8 -6 -4 -2 0
-50.00
-74.98
-100.00
Depth(m)

FIGURE 8: Bending moment due to Dynamic Load


The above figure illustrates the bending moment during seismic action; the maximum value of
79.22 kN.m is at 2.72m of the top level. However, the shearing force is zero in case of static load since
there is no horizontal load is acting on the pile.
2. Concrete End Bearing pile - fixed top

FIGURE 9: Pile Displacement in z direction in the Static Phase

During static load only vertical uz is observed in the pile, the value maximum value is occurred at the top
of the pile of 0.32 cm. This is very small value displacement wise.

FIGURE 10: Pile Displacement in X and Z direction in the Dynamic Phase

During dynamic load both horizontal and vertical displacement ux and uz were observed in the pile, the
value maximum value is occurred at the tip of the pile of 3 cm due of the fixity of the pile top. However
the vertical displacement increased to 10.33 cm all over the pile.
FIGURE 11: Effective stress-zz with static load

The effective stress during the static phase reaches its maximum under the pile which is equal to 204
kN/m2. The resultant effective stress due to the load acting on the pile is 204-118.9=85.1 kN/m2. However
if there is no skin friction acting along the pile we will find a huge value of stress under the pile, thus
𝜎 ′ =661.71kN/m2 …(1)

Since 𝜎 ′ =85.1 kN/m2, the force on the pile toe is , F= =6.1 kN…(2)

FIGURE 12: Effective stress-zz with Dynamic load

As it shows that during earthquake that the dense soil is liquefied and the effective stress tends to zero.
The effective stress during the Dynamic phase reaches its maximum under the pile which is equal to
252.1kN/m2, in the stiff soil layer. The value of the effective stress increases in the dynamic phase due to
liquefaction where the pile loses it skin friction and the axial forces increases too.
FIGURE 13: Shear stress-zx with Dynamic load

The shear stress during dynamic phase 225.5 kN/m2, this value occurred between the two different layers
of the soil.

N [kN]
0.00
-10 -8 -6 -4 -2 0
-100.00

-200.00

-300.00

-400.00
-456.53
-500.00
Depth(m)

FIGURE 14: Axial force Due to Static and Dynamic Load


The above figure present the axial force due to static and dynamic load for a single end bearing
pile having high skin friction value, the axial force acting on the head pile is 400 kN, the study shows that
this value increase to 456.53 kN during seismic action.

The original value increased by almost 14%, during seismic action and due to liquefaction, the highest
value as the graph shows is at 3.4 m from the top of the soil.
Q_12 [kN]
400.00

200.00
-8.58
0.00
-10 -8 -6 -4 -2 0
-200.00

-400.00
-489.69 -600.00
Depth(m)

FIGURE 15: Shearing force due to Dynamic Load


The above figure illustrates the shearing force during seismic action, at zero level, the earthquake
provoke a value of 8.58 KN, and the shearing forces reaches its maximum of 489.69kN at 6.19m from it
top. However, the shearing force is zero in case of static load since there is no horizontal load is acting on
the pile.

M_3 [kN m]
100.00

50.00

0.00
-10 -8 -6 -4 -2 0
-50.00

-92.05 -100.00
Depth(m)

FIGURE 16: Bending moment due to Dynamic Load


The above figure illustrates the bending moment during seismic action; the maximum value of 92.05
kN.m is at 2.72m of the top level. However, the shearing force is zero in case of static load since there is
no horizontal load is acting on the pile.
TABLE 5: Results of Displacement

End bearing free End bearing fixed


Uz(static)cm 0.32 0.32
Ux(dynamic)cm 9.2 3.4
Uz(dynamic)cm 9.7 10.37

In case of end bearing pile: The parametric study carried out that the pile having fixed head and free head
will behave in same way under static load and dynamic in the z direction. However the behavior of the
pile having a free top increases about 37% in comparing to the fixed top pile, during seismic action in the
x direction.

Therefore, the behavior of the piles differs from a case to another, where the end bearing pile fixed
laterally at top shows the best results displacement wise.

TABLE 6: Results for End Bearing

End bearing free End bearing fixed


Max-Axial(dynamic) kN 451.24 456.53
Max-Shear(dynamic) kN 513.10 489.69
Max-Moment(dynamic) kN.m 79.22 92.06

Liquefied soil can cause pile failure in shear and bending due to the unexpected large pressure generated
by the soil movement.

As shown in table 6, and as mentioned in the introduction that the pile damages are found in two different
places, the first is at 1 to 3 meters from it top and the second is at the interaction between two different
soils.

The first location of pile failure (1-3m) pile head is due to the generation of high bending moment at this
zone on pile, where the pile start to act as column beam element.

As for the second location of failure is due to the generation of high shearing force between the two layers
of soil, where the liquefied soil loses some of its properties.

Thus two scenarios are possible in that case, the first and most probable as seen in previous case studies
and papers is the pile failure by shearing force and the second is the soil end bearing failure at the toe.
Factor of safety

As noted before the forces acting on the different studies, the factor of safety will be calculated that
should be considered during pile design. For an elastic pile embedded subjected to an axial load 400 kN at
the head having the below concrete characteristics: Fc28=40Mpa and Fy=500 Mpa

Using the basic subgrade reaction theory the effective length of a pile subjected to buckling:

𝑑4 𝜌 𝑃𝑑 2 𝜌
EpIp(𝑑𝑧4 )+ 𝑑𝑧2 +𝑘ℎ 𝑑𝜌 (3)

d=30 cm, Ep=321600 kg/cm2,Kh=3kg/cm2

Where:

4 𝐸 𝐼
R=√ 𝐾𝑝 𝑑𝑝 (4)

EpIp= flexural stiffness of pile

Kh=modulus of subgrade

d=pile diameter

Thus, R=109.16cm

l’=πR (5)

l’ =342.77cm

Since the length of the pile L=900 cm


𝐿
= 2.655 using table proposed by (Francis &al 1965).
𝑙′

𝐿𝑒
𝑙′
=0.6, Le=205.6 cm

Calculating the slenderness ratio using EC8, for a cylinder shape column
4𝐿𝑒
λ= 𝑑
(6)

λ = 27.41

α= 0.87 (Ec2- preliminary design – simplified method).

For free head pile, by calculating the load capacity that the pile could resist, we assumed that the area
steel =3% are concrete = 19 cm2.

By using Archeffel, the pile could resist a force of 600 kN(axial load).
Thus the factor of safety F.O.S=1.5

For fix head pile, by calculating the load capacity that the pile could resist, we assumed that the area steel
=3%are concrete = 19 cm2 (EC2 authorize a maximum of 4% of area concrete)

The pile could resist a force of 1350 kN

Thus the factor of safety F.O.S= 3.37

During liquefaction and due to this high forces occurred to the pile, the diameter won’t resist this new
forces. And since the pile loses it lateral support, thus the critical length will the total height of the pile in
the liquefied soil wish is 8m. Furthermore by using the values in table 6, the pile diameter won’t be
sufficient to resist the new forces, shearing and bending moment actin on it during seismic action.

A new design was made, that has led to a new pile diameter of 130 cm and new area steel depending on
each pile type as shown in table 6

The pile new conditions will be:

d= 130cm, l=800 cm

Calculating the new slenderness ratio


4𝐿𝑒
λ= 𝑑
= 24.61… (6)

α= 0.89

Table 1- 1: Results of Area steel

End bearing free top End bearing fix top


2
AS(cm ) 321 285

As the study shows that an end bearing pile having a free top and with soft skin friction has the highest
value of area steel because it will act as a cantilever beam.

Concerning the concrete dimension, the pile diameter has increased 4.33 times.

And as for the steel area it has increased consequently depending on the nature of the pile, between 15
and 16.9 times.

In conclusion we have recommended the following calculation steps:

1-Start to design the pile dimension and anchor length under static loads

2-Apply a seismic action according to reading from a real earthquake

3-Calculate the new forces occurred (N,T,BM)

4-Redesign the pile with the new value


Conclusion

After modeling two types of piles, end bearing free top and end bearing fix top using Plaxis 3d finite
element, and using an appropriate type of sand modulus UBC3d-PLM that was made to create the
liquefaction behavior of soil.

The soil was used as normal loose sand in the static case of loading, and using the Mohr coulomb model
as for the dynamic phase same soil properties was induced but this time using the UBC-3DPLM model to
stimulate the sandy layer that leads to a liquefied soil during Sierra Madre earthquake (5.6Mw).

As per many studies the result were found similar in the case of end bearing pile, after seismic action a
horizontal load will be created on the pile, and while the sandy soil loses it properties and become
liquefied the pile loses it lateral support where four matters appears; buckling, loss of skin friction,
generation of high moments and shearing force on the pile.

The pile will suffer from those problems and it is able to fail due to the below circumstances:

1-Buckling failure: creation of excessive eccentricity

2-Failure due to instability: excessive displacement leading to damage in the superstructure

3-Base capacity failure: creation of penetrating cone

4-Failure due to the excess moment and shearing force in reinforced concrete

In conclusion the studies show that the loads arise during earthquake from both soil displacement and
inertial loads from the static load. The occurrence of liquefaction results in a degradation of soil stiffness
and an increase in soil displacement.

In general, the most severe damage will occur at the pile head and at interfaces between liquefied and
non-liquefied soil in case of end bearing pile.

It’s important that during pile design, the effect of liquefaction should be prevented by factors of safety
applied on the pile section the anchor length and the reinforcement bar.
References

1. Canadian Foundation Engineering Manual 4th EDITION, Canadian geotechnical society 2006.
2. S. Bhattacharya and M. Bolton, “Buckling of Piles during Earthquake Liquefaction.” 13 the
World conference on Earthquake Engineering, Vancouver, B.C., Canada, 2004
3. PLAXIS 3D Reference Manual 2017, PLAXIS
4. PLAXIS 3D Scientific Manual 2017, PLAXIS
5. S. Bhattacharya, “Pile Instability during Earthquake Liquefaction”, September 2013.
6. Design of Pile Foundations in Liquefiable Soils, G. Madabhushi, J. Knappett, S. Haigh, Imperial
College Press.
7. S. Bhattacharya, K. Tokimatsu, K. Goda, R. Sarkar, M.Shadlou, M. Rouholamin, “Collapse of
Showa Bridge during 1964 Niigata earthquake: A quantitative reappraisal on the failure
mechanisms.” Soil Dynamics and Earthquake Engineering 65 (2014) 55-71.
8. S. Bhattacharya, A. Kappos, “On the Collapse of Bridge Foundations in Liquefiable Soils during
Earthquakes.” Second European Conference on Earthquake Engineering and Seismology,
Istanbul AUG. 25-29, 2014.
9. Rahmani, A. Pak, “Dynamic Behavior of Pile Foundations under Cyclic Loading in Liquefiable
Soils.” Computers and Geotechnics 40 (2012) 114-126.
10. Advanced Foundation Engineering 2013, Prof. T.G. Sitharam, Indian Institute of Science,
Bangalore.
11. B. J. Meyer, L.C. Reese, “Analysis of Single Piles Under Lateral Loading.” Texas State
Department of Highways and Public Transportation; Transportation Planning Division P.O. Box
5051 Austin, Texas 78763.
12. S. Bhattacharya, S.P.G. Madabhushi, M. Bolton, “An Alternative Mechanism of Pile Failure in
Liquefibale Deposits During Earthquakes”, CUED/D-SOILS/TR324 (Oct 2002).
13. Deka, “Analysis of Pile Group under Lateral Load.” EJGE.
14. M. Achmus, K. Abdel-Rahman, K. Thieken, Behavior of Piles in Sand Subjected to Inclined
Loads, Institute of soil mechanics, Foundation Engineering and Waterpower Engineering,
Germany.
15. T.P.T. DAO, Delft University of Technology, Plaxis bv, Validation of PLAXIS Embedded Piles.
16. O. Aydan, M. Hamada, J. P. Bardet, R. Ulusay and A. Kanibir, “Liquefaction Induced Lateral
Spreading in the 1999 Kocaeli Earthquake, Turkey: Case Study Around The Hotel Sapanca”,
13th World conference on Earthquake Engineering, Vancouver, B.C., Canada, 2004.

Potrebbero piacerti anche