Equivalent Plastic Hinge Length of Extended Pile-Shafts Embedded in Sand

Tengfei LIU
Ph.D. Student
Tongji University
Shanghai, China

Aijun YE
Professor
Tongji University
Shanghai, China

Xiaowei WANG
Ph.D. Student
Tongji University
Shanghai, China

15liutengfei@tongji.edu.cn

yeaijun@tongji.edu.cn

10_xiaoweiwang@tongji.edu.cn

Tengfei Liu, born 1990, received

his civil engineering degree from
Tongji University, China. He is
working for his Ph.D. degree in
Tongji University. His main area
of research is related to soil-pile
interaction.

Aijun Ye, born 1970, received his

civil engineering degree from Tongji
University, China. She is a professor
in Tongji University now. Her main
area of research is related to
earthquake resistance for bridges,
soil-pile interaction and seismic
isolation devices.

Xiaowei Wang, born 1987, received

his civil engineering degree from
Tongji University, China. He is
working for his Ph.D. degree in
Tongji University. His main area of
research is related to earthquake
resistance for bridges and soil-pile
interaction.

Summary
Under strong seismic excitation, the extended pile-shaft will inevitably form a plastic hinge in its
underground portion. The underground damage is difficult for post-earthquake restoration and may
cause large residual displacement at the top of the pile-shaft. In this regard, the ductile behavior,
especially the characterization of the underground plastic hinge needs to be investigated. In this
paper, a Beam on Nonlinear Winkler Foundation (BNWF)-based finite element (FE) model
embedded in a homogeneous sand layer was developed in the OpenSees platform and calibrated by
an in-situ full-scale test. This study focused on the effect of different structural and soil parameters
on the equivalent plastic hinge lengths (EPHLs) of extended pile-shafts. Results indicate that the
EPHLs of piles are affected by many parameters of pile-soil systems. A typical range of EPHLs of
the pile is also obtained from the parametric study.
Keywords: Extended Pile-Shafts, BNWF, Seismic Performance, Equivalent Plastic Hinge Length.

1.

Introduction

The current specifications for seismic design of bridges [1-2] follow the Philosophy of Capacity
Design, i.e. providing a differential strength between the ductile and nonductile components in
bridge structures to prevent potential brittle failures. Since the foundation is difficult to inspect and
expensive to repair after earthquakes, it may be strategically designed to remain elastic while the
pier portion of substructures is detailed carefully for inelastic deformation and energy dissipation.
However, sometimes plastic hinging cannot be avoided in the foundation system, for example, in
the extended pile-shaft. An extended pile-shaft is a kind of foundation where the column is
continued below the ground level as a pile shaft of approximately the same diameter. Extended pileshafts, without any expensive pile-caps, are cost-effective and widely used in urban viaducts. Under
seismic loading, the maximum bending moment will occur in the below ground portion of the pile,
which will definitely leads to plastic hinging as long as the earthquake is sufficiently severe. The
ductile seismic design for foundation is needed if necessary, particularly for the extended pile-shafts.
Actually, seismic design specifications in America and Japan [3-5] have had the rules that withhold
ductile damage from arising in foundation system during some extreme design situations.
The most important question in extended pile-shaft ductile design is how to determine the ductility
capacity with the aid of the ductility index, such as the lateral displacement at the top of the pile. A
common approach assumes that the extended pile-shaft can be replaced by an equivalent cantilever,
which is fully restrained against lateral translation and rotation at the base. Then a relationship
between the lateral displacement at the top of the pile and the EPHL is obtained. In order to limit
the yielding of the pile, the EPHL should be determined exactly.
Many similar studies have been conducted by previous researchers. Budek and Priestley (1996,
2000) [6] studied the effects of above-grade height and soil stiffness on the EPHL by numerical
simulation. Chai and Hutchinson (1999, 2002) [7-9] carried out four full-scale tests of the extended

pile-shaft to identify the parameters influencing the EPHL, which include above-ground height and
soil stiffness. However, these investigations either ignored the influence of the inelastic soil or drew
opposite conclusions, and few studies have focused on the effects of the axial force ratio
In this paper, nonlinear static pushover analyses of a broad range of soil-pile systems are conducted
to investigate different parameters influencing the EPHL of the extended pile-shaft at various
displacement levels. A finite element (FE) model of one of the four full-scale tests conducted by
Chai is developed in the OpenSees [10] (Open System for Earthquake Engineering Simulation)
platform to verify the correctness of FE model. It not only simulates soil nonlinearity, but also
directly models pile fiber cross sections to simulate its nonlinearity, instead of using reduced
sectional modulus to account for pile cracking in current engineering practices. Based on a practical
bridge, a numerical parametric study using this FE model is undertaken, considering the diameter of
piles, the aboveground height, the relative density of sand, the longitudinal reinforcement ratio, the
transverse volumetric reinforcement ratio and the axial force ratio. After that, a typical range of the
EPHLs of common extended pile-shafts in bridge engineering is obtained. Some suggestions are
also offered to contribute to the performance-based seismic design in future.

2.

Analytical model

2.1
BNWF model calibration
In 1997 and 1998, Chai et al. from the University of California at Davis conducted in-situ tests on
full-sized reinforced concrete piles to investigate the seismic behavior, especially the inelastic
behavior, of extended pile-shafts embedded in sand under lateral cyclic loading. In this section, a
new FE model of the first specimen in that test is developed in the OpenSees platform, and the
model calibration is conducted to verify the correctness of the model.
The experimental program consisted of reversed cyclic quasi static
testing of four full-scale reinforced concrete(RC) piles, each with a
diameter, D, of 0.406 m, and embedded in loose dry sand and dense
dry sand to a depth of 13.5D. The aboveground height was 2D for two
piles, and 6D for the other two. The axial force applied to the test
free-head piles was P=445kN, corresponding to a nominal axial force
ratio of 0.10. Each pile tip was fixed in the horizontal and vertical
degree-of-freedoms (DOFs), while the rotational DOF is free.
Based on BNWF framework, the pile was represented by 20
displacement-based beam-column elements and the soil was modelled
with nonlinear p-y elements connected to every other node on the pile.
The pile section was discretized into an array of fibers representing
the confined (72 fibers) and unconfined(24 fibers) concrete and the
longitudinal steel reinforcement(7 fibers).All details about the BNWF
model is illustrated in Fig.1.The uniaxial cyclic response of unconfined
and confined concrete was both simulated by the modified KentFig. 1: Details of the
Park[11] model(1971). Confinement effect on core concrete was
BNWF model
considered by using the Mander [12] et al. (1988) model. A parallel
material of Steel02, which was described by the Giuffre-Menegotto-Pinto model (1973), and
Hysteretic material was used to describe the cyclic stress-strain relationship of the longitudinal
reinforcement. Nonlinear zero length p-y springs were used to simulate the interaction between the
pile and cohesionless soil, and the backbone of p-y curve was based on API guidelines (1987) [13].
Since the rate of increase of modulus of horizontal subgrade reaction nh was overestimated in API
(1987) according to [9, 15], the value of nh recommended by ATC-32[14] (Fig.2) was used in this
paper. Scaling the API p-y curve by two multipliers of 0.85 and 4 applied to initial stiffness and
strength, respectively, was used to reach improved agreement with the load test.
P=445 kN

Pull

Push

2 Elements(2D)

Ground Line
Sand

D=Diameter of the Pile

12 Elements
(6D)

3 Elements
(3D)

Soil Nodes

Confined
Concrete
(72 fibers)

3 Elements
(4.5D)

Zero-Length Element

Unconfined
Concrete
(24 fibers)

Longitudinal
Reinforcement
Steel
(7 fibers)

Fig. 2: Value of nh (ATC-32)

Fig. 3: Measured and FE calculated hysteretic

response of Pile No.1 in full-scale tests

As shown in Fig.3, both the hysteresis loops and the skeleton curves obtained from the experiment
and nonlinear pushover analysis fits very well. Several key parameters describing the elastoplastic
process of Pile No.1 in the load test have rather good agreement.
2.2
FE model of the extended pile-shaft
Calibration of the BNWF model described in the last section has been completed. Based on the
developed FE model before, a range of FE models of soil-pile systems are built and used for the
nonlinear pushover analysis to investigate parameters influencing the EPHL of piles. Pushover
analyses are used because of its efficiency in study and extensive use in design practice. In this
paper, the first-yield limit state is defined as the material yielding of the longitudinal reinforcing
steel, and the ultimate limit state is assumed to be the instance when the displacement ductility is
D=3.0 ( ATC-32,1996) ,where D=D/Dy ,D=pile head displacement and Dy is the elastoplastic pile
head displacement.
2.2.1 The prototype
Based on a practical bridge, with little modification for better analysis, the prototype of FE models
of piles for parametric study is developed: the diameter of the extended pile-shaft, D, is 1.5m, with
a concrete cover of 80 mm; the pile is embedded in homogeneous dry sand, with a relative density,
Dr, of 50%, to a depth of 12D and the pile tip is fixed against all DOFs; the above ground height of
the pile is 4D;the free-in-all-DOFs pile head is subjected to an axial stress level of 0.05fcAg,where
fc=uniaxial compressive strength of the concrete ( 26.8 MPa) and Ag=the gross cross sectional area.
36 HRB400 reinforcement B25 bars (A=490.0 mm2) are used to provide a longitudinal steel ratio of
l=1.2%, and the pile is confined by circular HRB500 A24 bars at 150 mm pitch , corresponding to
a confined steel ratio of t=0.8%.
The prototype of the pile is represented by 20 displacement-based beam-column elements and the
soil was modelled with nonlinear p-y elements connected to every other node on the pile. The cross
section is discretized into an array of fibers representing the confined (432 fibers) and unconfined
(48 fibers) concrete and the longitudinal steel reinforcement (30 fibers)
2.2.2 All cases in parametric study
In this paper, up to 13 cases including the prototype are studied to investigate the effects of diameter
of piles, aboveground height, relative density of sand, longitudinal reinforcement ratio, transverse
volumetric reinforcement ratio and axial force ratio on the EPHL of piles. A summary of all 10
cases is presented in Table 1.

Table 1: Summary of all cases in parametric study

Case No.

3.

Diameter
of piles,
D (m)

Relative
density
of sand,

Aboveground
height,
La

Longitudinal
reinforcement
ratio,

l (%)

Dr (%)

Transverse
volumetric
reinforcement
ratio,

t (%)

Axial
force
ratio ,

S1
S2
S3
S4

1.0
1.5
2.0
1.5

4D
4D
4D
2D

50
50
50
50

1.2
1.2
1.2
1.2

0.8
0.8
0.8
0.8

0.05
0.05
0.05
0.05

S5
S6
S7
S8
S9
S10
S11
S12
S13

1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5

6D
4D
4D
4D
4D
4D
4D
4D
4D

50
30
80
50
50
50
50
50
50

1.2
1.2
1.2
0.8
1.5
1.2
1.2
1.2
1.2

0.8
0.8
0.8
0.8
0.8
1.0
1.2
0.8
0.8

0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.1
0.2

Results and discussion

In this section, the normalized EPHL Lp*=Lp/D is used to investigate the effects of different
parameters, because it can eliminate the effect of D (Diameter of piles) which is not the emphasis of
this paper. The method of calculating the EPHL from Chai(2002) is widely recognized ,but his
formula ,namely, Eq. (4) in this paper overestimates the lateral plastic displacement of pile head,
which will definitely lead to a higher value of the EPHL of the extend pile-shaft than the precious
one. Considering this, the strategy of calculating the EHPL is consistent with Zhang (2012) [16] in
this paper (Fig. 4).

Fig. 4: Strategy for calculating the EPHL

Fig. 5: Effect of D

Fig.5 illustrates the effects of diameter of piles, D, from 1.0m to 2.0m, on the normalized EPHLs of
piles. The Lp* of all cases fall quickly with the increase of the displacement ductility factor, D,
from 1.33 to 2.33, and then a small fluctuation happens when D=2.33-3.0. A bigger diameter of
piles leads to a higher values of Lp* when D>2.0, on the contrary, the differences of all three cases

can be neglected if D<2.0. The reason is obvious: bigger diameter of piles can heightens the
effective stiffness of cross sections of piles, however, it begins to work only when the displacement
ductility factor is enough to cause the plastic damage of core concrete to develop to a specific extent.
In Fig.6, the Lp* of all cases show a dramatic drop trend when increase from D 1.33 to 3.0. As is
expected, bigger aboveground height of extended shaft-piles leads to higher value of Lp*,but this is
only applied to D 2.0This is partly because higher value of aboveground height of piles leads to
a lower gradient of the increase of cross sectional curvature. When D>2.0, the difference is
complicated, and the reason may be that the combination of complex pile-soil interaction and the
degradation and redistribution of pile stiffness.

Fig. 6: Effect of La

Fig. 7: Effect of Dr

As shown in Fig.7, the normalized EPHLs of different cases with various relative densities of sands
drop dramatically for the displacement ductility factor, D=1.33- 3.0.This is partly because the
curvature in the plastic region of piles increases extremely faster than that in the elastic region. And
the result indicates that it may be more reasonable to
suggest different values of the EPHL in different design
levels .Besides, the pile embedded in sand with lower
relative density has a higher value of the EPHL when D
ranges from 1.33 to 2.0. The reason is obvious: a lower
relative density of sand will lead to a lower gradient of the
increase of cross sectional curvature; the length of plastic
region will increase because of a more gentle distribution of
curvature along the pile length. When D=2.0- 3.0, the
process of cave-in and recompression of sands is so
complicated that the EPHLs of piles in some cases fluctuate
gently along with the increasing D.
Fig.8 illustrates the effects of longitudinal reinforcement
Fig. 8: Effect of l
ratio, l, from 0.8% to 1.2%, on the normalized EPHLs of
piles. The Lp* of the stiff pile is smaller than that of the
flexible pile when D=1.67-2.33, as observed for piles with larger longitudinal reinforcing ratio
have shorter normalized EPHLs. However, the EPHLs have a complicated fluctuation afterwards.
The mechanism is still unknown, which needs the support of the experimental data in the future. A
speculation is that stiffer piles will have lower curvature in the plastic region and shorter plastic
region will occur at the same value of D.
In Fig.9, the normalized EPHLs of 3 cases all have a gradual drop when D ranges from 1.33 to
2.33, and a fluctuation with the same trend (raise then fall) occurs afterwards. All 3 cases have the
same values of the normalized EPHL when D 2.0, while the difference among different cases is
bigger, not big enough to be considered though, along with D increases .This phenomenon is
predictable: the normalized EPHL is effected by the ductile behavior of the pile, and higher
volumetric transverse ratio means better confinement effect on core concrete, which leads to a
better ductile behavior of the pile.

Fig. 9: Effect of t

Fig. 10: Effect of

The effect of axial force ratio, , on the normalized EPHL is

illustrated in Fig.10. When =0.05, the normalized EPHL
decreases from 2.10 to 1.26. However, the normalized
EPHL drops quickly first (D=1.33-2.0) and then remains
almost a constant of 0.4 (D=2.0-3.0) when =0.2. And the
value of the normalized EPHL when =0.1 is always
between the two values of the other cases except D=1.33. It
can be concluded that a higher axial force ratio means a
lower normalized EPHL except the first beginning of the
appearance of plastic hinging in piles. The reason is that
higher axial force ratio arouses a better lateral resistance of
piles, and the concentration of curvature along the pile
length increases accordingly, leading to a shorter normalized
EPHL.

Fig. 11: Range of Lp*

It can be seen that the EPHLs of piles at various displacement levels in all cases are illustrated in
Fig 11. Two cases, C12 and C13 are not included to avoid too big scatter in the values of Lp*. Based
on this, the maximum value curve, the minimum value curve and the average value curve of Lp*
with the increase of D are obtained. When D changes from 1.33 to 3.00, the maximum value
decreases from 2.56 to 1.42, reduced by 44.5%; the average value decreases from 2.22 to 1.22,
reduced by 45.0%; the minimum value decreases from 1.89 to 1.03, reduced by 45.5%. All the three
values(maximum, mean and minimum value) of Lp* almost remain constant when D>2.0,
corresponding to 1.50, 1.28 and 1.07, respectively.

4. Conclusion
In some extreme design situation, the plastic hinging of the extended pile-shaft is inevitable. For
performance-based design, it is fully important to evaluate the ductile behavior of piles with a
specific ductility index, such as the displacement ductility factor of pile head. And the displacement
of head can be easily obtained through the EPHL. So it is critical to get the exact value of the
EPHLs of piles. In this paper, after a calibration of in-situ quasi static test, a numerical parametric
study of RC extended pile-shafts embedded in cohesionless soil is conducted, considering some
critical parameters of pile-soil systems. Static nonlinear pushover analysis were conducted to
investigate the parameter influencing the EPHLs of piles. The main conclusion as follows:
(1) The normalized EPHL of piles changes with the displacement ductility factor, D, and it may
be more reasonable to adopt different recommended values of the EPHL of piles in various design.
(2) The effect of diameter of piles, D, on Lp* is obvious only when D > 2.0. For limited ductile
structures, the effect of D can be omitted.
(3) The aboveground height of piles has a big influence on the EPHL of piles when D < 2.0. The
longer La leads to higher value of Lp*. It must be considered when calculating the EPHLs of piles.

(4) The normalized EPHL is sensitive to the relative density of soil, Dr: when embedded in sands
with higher Dr, the normalized EPHL tends to be shorter. The ignorance of soil density can arouse a
rather big error in evaluating the EPHL of piles.
(5) The effect of longitudinal reinforcing ratio, l, is that: the normalized EPHL increases with the
reduction of l only when D=1.67-2.33; there is no clear relationships between l and Lp*. The
explanation needs the aid of experimental data.
(6) The difference of the normalized EPHL is very small when volumetric transverse ratio, t
changes. So its effect can be neglected.
(7) Axial force ratio, , has a big impact on the normalized EPHL of piles: if is small enough, the
value of normalized EPHL fluctuates with the change of D .At any displacement ductility level,
when increases, the normalized EPHL will have a big drop.
(8) The statistical results reveal that, there is a clear range of EPHLs of typical piles at various
displacement factor when is not too big. The values of EPHLs of piles show a relatively constant
trend when D >2.0.

5. Acknowledgements
This work was supported by the Ministry of Science and Technology of China, Grant No.
SLDRCE15-B-05.

6. References
[1]

[2]
[3]
[4]
[5]
[6]

[7]
[8]

[9]

[10]

[11]
[12]

Ministry of housing and urban rural development of the Peoples Republic of China, CJJ
166-2011, Code for seismic design of urban bridges, China Architecture & Building Press,
Beijing, 2011.
Ministry of transport of the Peoples Republic of China, JTG/TB 02-01-2008, Guidelines for
Seismic Design of Highway bridges, China Communications Press, Beijing, 2008.
American Association of State Highway and Transportation Officials, AASHTO Guide
Specifications for LRFD Seismic Bridge Design, 2nd Edition, Washington, DC, 2011
California Department of transportation (Caltrans), Seismic Design Criteria, Sacramento,
CA, 2011.
Institute of Japan Highway, Design Specification of Japan Highway Bridge (Explanation),
Maruzen Ltd ,Tokyo,2001.
Budek AM, Priestley MJN, Benzoni G., "Inelastic seismic response of bridge drilled-shaft
RC pile/columns", ASCE Journal of Structural Engineering, Vol. 126, No. 4, 2000, pp.510517.
Chai, Y. H., "Flexural Strength and Ductility of Extended Pile-Shafts. I: Analytical Model",
ASCE Journal of Structural Engineering, Vol. 128, No. 5, 2002, pp. 586-594.
Chai YH, Hutchinson TC, "Flexural strength and Ductility of extended pile-shafts-II:
experimental study", ASCE Journal of structural Engineering, Vol. 128, No. 5, 2002,
pp.595-602.
Hutchinson TC, Chai Y.H., "Boulanger RW. Simulation of full-scale cyclic lateral load tests
on piles", ASCE Journal of Geotechnical and Geoenvironmental Engineering, Vol. 131, No.
9, 2005, pp. 1172-1175.
Mazzoni S, McKenna F, Scott MH, Fenves GL, Open System for Engineering Simulation
user Command Language Manual, version 2.0, Pacific Earthquake Engineering Research
Center, University of California, Berkeley ,2010.
Kent DC, Park R., "Flexural members with confined concrete", Journal of Structural
Division, Vol. 97, No. 7, 1971, pp. 1964-1990.
Mander JB, Priestley MJN, Park R., "Theoretical stress-strain model for confined concrete .
" ,Journal of Structural Division , Vol. 114, No. 8, 1988, pp. 1804-1826.

[13]
[14]
[15]
[16]

American Petroleum Institute, Recommended practice for planning, designing, and

constructing fixed offshore platforms, American Petroleum Institute, Washington, DC, 1987.
Applied Technology Council (ATC), Improved seismic design criteria for California Bridges:
provisional recommendations, Report no. ATC-32, Redwood City, CA, 1996.
Allotey N, El Naggar M H., "A numerical study into lateral cyclic nonlinear soil-pile
response ", Canadian Geotechnical Journal, Vol. 45, No. 9, 2008, pp. 1268-1281.
Zhang J, Hutchinson TC, "Inelastic pile behavior with and without liquefaction effects ",
Soil Dynamics & Earthquake Engineering, Vol. 36, No. 3, 2012, pp.12-19.2012.