Electronic Journal of Structural Engineering, 6 (2006)

Seismic performance of pile-to-pile cap connections: An investigation of

design issues
M. Teguh
University of Melbourne, Australia and Islamic University of Indonesia, Yogyakarta, Indonesia
Email: [email protected]

C.F. Duffield, P.A. Mendis and G.L. Hutchinson

University of Melbourne, Australia

ABSTRACT: Damage in recent earthquakes has resulted in the design of pile foundation systems becoming
more conservative, particularly pile-to-pile cap connections. However, the application of current international
design practice results in pile cap joint details having congested steel reinforcement in the pile cap and this is
extremely difficult to construct in accordance with the designers recommendations. The formation of plastic
hinges in the piles remains a serious risk. A review of critical design issues and former research investigations
into the soil-structure interaction of pile systems and the findings of a three-dimensional, nonlinear finite ele-
ment analysis of the system is reported. Significant gaps have been identified between current practice and
the performance of piling systems when subjected to seismic events. Preliminary findings indicate potential
for alternate connection details to improve performance under seismic action. The paper concludes with a
concise summary of current state-of-the-art design approaches and details further research requirements.

KEYWORDS: Pile-to-pile cap connections, nonlinear finite element analysis, seismic performance, interna-
tional design practice, state-of-the-art design approach

1 INTRODUCTION foundation systems and their response to earth-

quakes. Other studies pertain to concrete piles, yet
1.1 Background of problem again research work specific to performance of piles
Earthquakes that have occurred globally over the and their connection to pile cap is also very limited.
last two decades have resulted in an increased ex- Research work by Pam and Park (1990b) has pro-
pectation of acceptable performance and damage vided a starting point regarding the design and de-
control for different structures during seismic tailing of pile-to-pile cap connections. Specific in-
events. Catastrophic failures of piled foundation sys- vestigation into the damage of piled foundations
tems in the recent earthquakes of Loma Prieta, after earthquakes has seldom been conducted be-
Northridge (Mizuno 1987; Nogami 1987; Sheppard cause it requires excavation and is thus costly.
1983), and Kobe (Building Research Institute 1996) Past experience has shown poor connection detail
have led to considerable effort being directed to- of some reinforced concrete piles and hollow
wards safer civil infrastructure particularly in the prestressed concrete piles performing poor seismic
seismic zones. However, repair of damaged piles in performance of the pile foundation system particu-
high-rise building systems is impractical because of larly at the pile-to-pile cap connection (PPC). Lack
the expensive cost and difficulty associated with of careful detailing and poor confinement of core
ground excavation.
concrete appears to be the reason for the failure of
In the aftermath of many earthquakes, numerous
engineering inspections and investigations have most of the piles. Therefore, there are still many
been performed to assess the degree of structural questions left to ponder and be answered to assist
damage and to evaluate the performance of various the understanding of the pile-to-pile cap behaviour.
construction materials. Most of the reports, however, Potential inelastic damage has occurred at the in-
address only the upper structure of buildings or terface between pile heads and pile cap as evidenced
bridges; very little information is available on the in recent earthquakes. Figure 1 shows the formation
performance of under ground structures such as pile of typical potential plastic-hinges. It should be noted

8
 Electronic Journal of Structural Engineering, 6 (2006)

that because of the difficulties associated with the DIANA (2005), to obtain structural response that
repair of foundation damage occurring during severe can be used to validate the analytical design proce-
earthquakes, it is desirable to design the piles to re- dure and to determine the types of inelastic behav-
main undamaged. Therefore, damage due to recent iours of the system will experience under seismic
earthquakes has resulted in an increased conserva- loads.
tism in the design of piles and pile-to-pile cap con- In this paper, the particular case of a square
nections. In contrast, current recommendations pro- prestressed pile embedded in a cast-in-situ pile cap
duce connection detailing which result in high levels has been analysed. Linear and nonlinear finite ele-
of congestion of steel reinforcement and extreme ment and time history analyses have been conducted
difficulties in construction. to investigate the behaviour of pile-to-pile cap con-
nections under simulated seismic loads.

2 CURRENT PRACTICE
2.1 Design standard
In practice, the use of prestressed concrete piles and
their connection to cast-in-situ reinforced concrete
pile cap is inherent in pile foundations. The connec-
tion distributes shear, axial load, and moment to
each pile underneath the pile cap. All connections
should perform their function at all stages of loading
without any distress, and with an appropriate safety
factor against failure due to overload. From a previ-
ous investigation (Harries and Petrou, 2001) it has
been established that piles tend to develop their flex-
ural capacity without distress to the pile cap pro-
vided that a sufficient embedment length is fur-
nished.
Figure 1. Response and potential plastic hinges of pile group. In general, design criteria for the required seismic
performance of pile foundation system in a ductile
Ideally, the design concept should aim to dissi- earthquake resistant structure, are documented in
pate seismic energy by ductile yielding at plastic many standards, e.g. the American Concrete Institute
hinge regions in the structure above the foundations (ACI), the Australian/New Zealand Standards
or mechanical energy dissipating devices placed be- (AS/NZ), California Department of Transportation
tween the foundation and the structure. i.e., the System (CALTRANS), and Japanese Codes. How-
foundation should be provided with sufficient ever these Standards do not detail the performance
strength to ensure, as far as possible, that they re- of pile-to-pile cap connections. As a result, research
main in the elastic range while energy dissipation on the performance of pile-to-pile cap connections
occurs. Moreover, the embedment detail of the con- subjected to earthquake response is limited and the
nection between the precast prestressed piles and the behaviour of these connections not well understood
cast-in-situ reinforced concrete pile cap is currently due to its complexity. In fact, recent research relat-
investigated in the University of Melbourne to fully ing to the design and detailing of piles and pile-to-
comprehend the seismic behaviour of the overall pile cap connections has relied heavily on the NZS
structures. 3101, the CSA Code, and the ACI Code (Joen and
Park 1990).
Current design procedures for pile caps do not
1.2 Aim of study
provide engineers with a clear understanding of the
The aim of this study is to refine the design of pile- physical behaviour of these elements. The current
to-pile cap connections in conjunction with capabili- ACI Building Code procedure/guidance for the shear
ties of developing the moment demands on the piles design of pile caps does not predict adequately the
as a result of the seismic events. This aim is being actual behaviour because this procedure neglects
accomplished by subjecting analytical models of a certain important parameters, such as the amount of
typical connection detail of pile-to-pile cap connec- longitudinal reinforcement, and overemphasize other
tions. A three dimensional (3D) solid model of the parameters, such as the effective depth. In order to
pile and pile cap was generated with the finite ele- achieve efficient connection dimensions, new ad-
ment analysis (FEA) programs, SAP2000 and

9
 Electronic Journal of Structural Engineering, 6 (2006)

vanced materials and construction techniques need Figure 2a and 2b show the liquefacted soils during
to be considered. major earthquake affecting inelastic lateral deforma-
The structure and all its components must gener- tion of pile-to-pile cap interface and plastic hinges
ally be designed to resist all loads, deformations and and buckling along the pile length. Four years after,
environmental conditions likely to occur during con- the magnitude of 7.8 Off-Tokachi Earthquake and
struction and normal use, and have adequate durabil- magnitude of 7.4 aftershock caused serious damage
ity. Furthermore, design loads, actions and strengths to a region of Northern Japan, and in particular dam-
are intended to produce a very low probability of age the Anenuma Bridge. Tamura et al. (1973) has
failure during the design life. undertaken post-earthquake inspection revealing
cracks near the top of the piles, and over 2 ft of lat-
eral displacement and as much as 4 in of settlement
2.2 Observation of earthquake events
of the bridge.
Case histories of observed reinforced and The 1978 Off-Miyagi Prefecture Earthquake re-
prestressed concrete piles damage and failure during sulted in a number of cases of damage to prestressed
global major earthquakes are presented as a repre- concrete piles, which were principally caused by
sentative survey in the pile foundation problems. earthquake-induced vibration of the superstructure
The cases are described with extensive indications (Sugimura 1981).
of the pile-to-pile cap performance during strong The Loma Prieta Earthquake was occurred on
shaking and insight into modes of behaviour and 17th October 1989 with the magnitude of 7.0 causing
failure. the impressive failure of many pile-supported struc-
The Niigata Earthquake was the second major tures. It was observed that piles along an alignment
earthquake of 1964 with the magnitude of 7.3 result- that transitioned from stiff to soft foundation soils,
ing widespread soil liquefaction related to damage the site response and structural connection details
and numerous failures of pile supported structures. were the principal failure mechanism. Yashinsky
The 7.2 m long concrete piles of 0.18 m diameter (1998) provided a comprehensive summary of dam-
lost bearing capacity due to liquefaction, and the age to highway systems in the Loma Prieta Earth-
structure of Saiseikai Hostpital tilted and cracked quake such as at the Oakland Outer Harbor Pier 7,
(Kishida 1966). 16 in square prestressed concrete batter piles failed
at or near the connection to the pile cap. It has been
noted that soil liquefaction did not contribute to the
failure, as upper foundation soils comprising soft
clays and organic, with some alluvial sands present.
Similar to severe damage over a large area caused by
the Loma Prieta Earthquake is the 1991 Costa Rica
Earthquake with the magnitude of 7.5 that resulted
soil liquefaction relating to collapse of several pile-
supported bridges. As reported by Priestley et al.
(1991) two of the three spans on the Rio Viscaya
Bridge collapsed due to severe abutment rotation,
a. Inelastic lateral deformation at b. Plastic hinges and buckling
pile distress, and failure of an interior support, also
pile-to-pile cap interface occurred within soil layers
producing from extensive soil liquefaction.
The 1995 Hyogoken-Nanbu (Kobe) Earthquake
with the magnitude of 7.2 was the most destructive
earthquake over 60 years to strike Japan. It was di-
rectly hit on a major metropolitan area that damaged
too many pile foundation systems. The most devas-
tated structural failure during the Kobe Earthquake
was the collapse of an elevated section of the pile-
supported Hanshin Expressway. Mizuno et al.
(1996) has undertaken a survey over more than 30
cases of pile damage observed in precast concrete,
cast-in-place concrete, and steel pipe piles. Damage
patterns consisted of separation between piles and
c. Pile head released from pile d. Damage at pile-to-pile cap
pile caps, damage near the pile head, and damage at
cap connection
deeper portions of piles as shown in Figure 2c and
Figure 2. Pile foundation damage due to strong earthquakes 2d.
(Hamada 1991; Mizuno 1987).

10
 Electronic Journal of Structural Engineering, 6 (2006)

2.3 Theoretical design plated to verify prestressed concrete piles’ ability to

develop the required strength of the pile cap em-
Design of piles to resist earthquake forces includes bedment and ductility at the pile-to-pile cap inter-
the interaction of lateral loads and variations in the face (Silva et al. 2001).
axial loading characteristics of pile groups, as illus- An intensive study on laterally loaded piles has
trated in Figure 1. Under the imposed column axial also been conducted to review analytical concepts,
force, Pc, shear force, Vc, and moment, Mc, axial procedures and methods for evaluating the capacity
forces, Pp, shear forces, Vp, and moments, Mp de- of prestressed concrete piles. Further, a new ap-
velop in the piles to withstand the applied loads.. proach to analyse pile-to-pile cap connections sub-
In general, the pile foundation system is designed jected to seismic loadings is currently being formu-
as follows: (1) estimation of the applied forces from lated (Teguh et al. 2003b).
the column to the foundation system, (2) estimation Tests on solid prestressed concrete piles have
of soil properties to determine the soil-structure in- shown that properly confined members can with-
teraction required for analysis, (3) design the pile stand considerable displacement ductility factors
cap necessary to transmit effectively the applied without significant loss of their load carrying capac-
loads to the soil strata, and (4) selection of the pile ity. Furthermore, it has been demonstrated that extra
type required to transmit the applied loads to the soil longitudinal reinforcing steel improves the flexure
strata. strength but does not significantly improve the duc-
The fundamental basis of this seismic design tility of the prestressed concrete piles (Pam and Park
method relies on carefully selecting and detailing lo- 1990a).
cations of potential inelastic deformation. All other In early 1983 Sheppard has initially proposed
regions, defined to remain elastic during seismic ex- eight varieties of connection details where piles are
citation, are designed based on appropriate strength embedded in cast-in-place reinforced concrete pile
margins above the possible over strength in those caps (Harries and Petrou 2001b; Sheppard 1983).
regions with elastic response (Priestley et al. 1996). The first three types of connections is categorized as
In conjunction with capacity design principles, a simple pile foundation system used in non-
the ideal approach for the design of buildings or significant structures, for instance low-rise buildings
bridges would be for piles to remain elastic during a and short-span bridges. The other types of connec-
seismic event because of difficulties associated with tion details are applied to important structures re-
inspection and repair of subsurface foundation com- quiring higher levels of seismic resistance such as
ponents after an earthquake. However, under this high-rise buildings and long-span bridges. However,
moderate condition, the design approach may be im- Sheppard has not yet investigated all proposed con-
practical due to the potential inelastic deformations nection details.
that may form in the piles either in the connection A research project conducted by Pam (1988) in-
region or in subgrade regions. In addition, loads and vestigated the adequacy of different connection de-
deformations imposed on the piles of a pile group, as tails of precast prestressed concrete piles to the pile
a result of lateral translations and rotations of the cap. Four different types of connections were used in
pile cap, and due to seismic action, may led to un- the construction of six full scale tests. All full six
wanted inelastic response of piles at the connections scales tests were designed in accordance with the
to the pile cap and/or below ground level. Conse- New Zealand Standard (New Zealand Standard
quently, a simplified approach must be developed 1982; Standard 1982) and used 400 mm octagonal
which encompasses the main features of the prob- precast prestressed concrete piles.
lem, but which is analytically and numerically trac- Silva (1998) investigated three standard Caltrans
table. test units, which were tested under increasing cyclic
Current engineering practice designs pile-to-pile lateral load or deformation and fully reversed vary-
cap connections based on embedded longitudinal re- ing axial loads (compression and tension). The three
inforcement and confinement reinforcement rely on standard Caltrans pile test units were defined as test
careful selection and detailing around zones of po- unit STD1, STD2 and STD3. The pile of test units
tential inelastic deformation. Other regions are de- STD1 and STD2 was a Class 625 pile, and for test
signed on an appropriate strength method using elas- unit STD3 the pile was a Class 1780 pile. The test
tic response techniques. unit STD1 was a full scale of 305 mm square precast
prestressed concrete and the test unit STD2 was also
a full scale of 356 mm diameter steel encased un-
2.4 Past researches
reinforced concrete. The test unit STD3 was of 7/12
In view of recent proposed code changes restricting scale model of 356 mm diameter steel encased un-
the use of prestressed concrete piling in the United reinforced concrete.
States, several testing programs have been contem-

11
 Electronic Journal of Structural Engineering, 6 (2006)

Harries and Petrou (2001) undertook experimen- bedded in the cast-in-place pile cap. In the headed
tal tests at the University of South Caroline Struc- embedment (with treatment), pile strands confined
tures Laboratory. Two 18 in. (450 mm) square by 18 with round reinforcement are exposed and embed-
in (5.49 m) long piles were fabricated simultane- ded in the cast-in-place pile cap.
ously in a 40 ft (12.2 m) prestressing bed. In this ex- In this paper, prestressed concrete piles manufac-
periment, a detail connection recommended by tured in Indonesia (Teguh et al. 2003a) and two pile
Sheppard (1983) was adopted where the pile was units tested by Harries and Petrou (2001b), were re-
simply embedded in the pile cap without any treat- viewed and the validity of their hypotheses was rig-
ment so called plain embedment (i.e., no static bars). orously investigated through comparison of the ob-
The connection details used in the first three ex- served and predicted response from linear to
perimental test models of pile-to-pile cap connec- nonlinear behaviour.
tions were based on recommendations by the New
Zealand and Caltrans standards (Caltrans 1990;
4.1 Analytical approach
Standard 1982). The objectives of the studies were
to check the capabilities of pile-pile cap connections To determine the capacity of a plain pile-to-pile cap
to resist large inelastic deformations caused by cy- connection (Figure 3a), it is assumed that a rigid
clic lateral loading. body (pile) embedded in cast-in-place concrete
monolith (pile cap) and based on mobilization of an
internal moment arm between bearing forces Cf and
3 IDENTIFIED GAPS/WEAKNESSES IN Cb as shown in Figure 3b will be utilized. A para-
CURRENT APPROACH bolic distribution of bearing stresses is assumed for
A current state-of-practice design and analysis ap- Cb, and Cf is computed by a uniform stress equal to
plication of the soil-pile interaction considered in the 0.85fc’. The bearing stresses are distributed over the
pile-to-pile cap connections is a complex soil- width of the embedded pile, b. Based on these as-
structure interaction of the pile foundation system sumptions and calibrating the calculated stresses
and not well developed. The unavailability of stan- against experimental data, the required embedment
dardized and validated analysis techniques, and the length, Le, may be determined from Equation 1,
conservative perception, designers routinely ignore where a is the shear span of the pile (distance from
or greatly simplify the present of pile foundations in pile cap to assumed point of zero moment) and β1 is
their analyses. the concrete stress block factor defined in ACI 318-
A special challenge of soil-structure interaction 02, Section 10.2.7.3 (A.C.I. Committee 318 2002) as
problems is in dispute over two disciplines, geotech- presented in Equation 2. The shear span can be in-
nical and structural engineering, and the analysis is creased by an amount equal to the concrete cover, c,
frequently broken into two parts rather than ad- to account for possibility spalling of the soffit of the
dressed in a holistic manner. In fact, a geotechnical pile cap as shown in Figure 3b. The value of b’ is
engineer idealizes a complex multimode superstruc- given by Mattock and Gaafar as the width of the
ture as a single degree of freedom oscillator; while
the structural engineer often represents the poten-
tially nonlinear soil-pile interaction with a simple
linear spring. In this manner, nonlinear system inter-
action between the superstructure and substructure is
artificially prevented. The diverse and non-
standardized design approaches are basically in-
tended to illustrate the lack of professional consen-
sus and the gap between the current state-of-practice
and the current state-of-the-art. These approaches
are also considered as creative applications of lim-
ited tools to complex problems.

4 PRELIMINARY ANALYSIS
This section presents preliminary analysis of pile-to- Figure 3. Analytical approach for determining pile capacity
pile cap connections focusing on two types of con- plain embedment.
nections, i. e., plain embedment and headed embed-
ment, to improve the capacity of the pile-to-pile cap element into which, in this case, the pile is embed-
connections. In the plain embedment (without treat- ded (Mattock and Gaafar 1982).
ment), the prestressed concrete pile is simply em-

12
 Electronic Journal of Structural Engineering, 6 (2006)

Using a slightly different assumed stress distribu- 4.2 Linear finite element analysis
tion, Marcakis and Mitchell (1980) proposed Equa- The study continued with a review of modelling as-
tion 3 to determine the embedment length, Le. This sumptions to assess the effect of varying the degree
expression has also been calibrated against experi-
of structural complexity. The most comprehensive
mental data, where e is the eccentricity from the
form of analysis for reinforced concrete members
point of zero moment to the centre of effective em- uses the finite element (FE) method where individ-
bedment (Figure 3). They define b’ based on a strut- ual bars and elements are separately modelled. A fi-
and-tie approach as being effective width to assumed nite element method allows satisfactory simulation
“tie” steel, limited by a value of 2.5b (Marcakis and of test results, provided accurately defined constitu-
Mitchell 1980). tive models are used. A fixed base pile cap acting as
⎡ ⎤
0 . 66 ⎢ 0 . 58 − 0 . 22 β ⎥ a simple cantilever column was therefore employed
⎛ b' ⎞ in the model.
V u = 4 .5 f c' ⎜ ⎟ β 1 bL e ⎢ 1
⎥
⎝b ⎠ ⎢ 0 . 88 + a ⎥ (1)
A three dimensional (3D) finite element analysis
⎢ A e − c ⎥⎦
⎣ was performed to model the behaviour of the pile-to-
pile cap connection. The objectives of this modelling
For fc’ ≤ 4000 psi (27.6 MPa), β1 = 0.85 and were to determine elastic stress distribution and
for fc’ > 4000 psi (27.6 MPa), 0.65 ≤ β1 ≤ 0.85, load-displacement response of the pile-to-pile cap
where: connection due to axial and simulated seismic loads.
A linear static finite element analysis as initial stage
⎛ f c' − 4000 ⎞
β1 = 0.85 − 0.05⎜⎜ ⎟⎟ of elastic modelling was developed to simulate the
(2)
⎝ 1000 ⎠ seismic behaviour.
The two selected pile units (Harries and Petrou
In addition, the nominal moment capacity of the
2001b) presented in Table 1 to 3 respectively was
piles corresponding to an axial load is predicted us-
analysed using the structural analysis software
ing plane section analysis program RESPONSE- SAP2000 (Wilson 1996). Two element types were
2000 (Bentz 2001).
used for the finite element discretisation of the mod-
0.85 f c b' (A e − c ) els. The concrete pile-to-pile cap was meshed (mod-
Vu = elling surface delineation lines) using a special solid
3. 6 ⎛ A e − c ) ⎞ (3)
1+ ⎜ + c + a⎟ element type in order to easily perform the stress
A e − c) ⎝ 2 ⎠ concentration or failure of concrete in crack-
ing/crushing. Strands and reinforcing bars used in
Curves generated from Equations 1 and 3 are pre-
the pile and pile cap were meshed using truss ele-
sented in Figure 4 as a comparison of the shear and
ments with two different materials (i.e., strands and
moment capacities of the embedment for varying
reinforcing bars).
embedment length and having varying square pile
sizes manufactured in Indonesia (Figure 4). It should Table 1: Unit test variables of pile-to-pile cap connections
be noted that the results from Equations 1 and 3 are Pile unit number Unit P1 Unit P2
similar. It can be summarized that equation proposed Initial prestressing steel 1.394x109 1.394x109
by Marcakis and Mitchell tends to result is slightly (constant) N/m2 N/m2
more conservative embedment capacity values than Applied axial load (in-
the values suggested by the PCI design handbook. crement): assumed as 4.260x106 4.260x106
2
pressure load over the N/m N/m2
pile surface
Maximum lateral loads 1.112x105 N 9.600x104 N
(monotonic & cyclic)
Pile reinforcement:
1. Low relaxa- 12.5 mm 12.5 mm
tion strand (1/2”) (1/2”)
tendon 7 mm 7 mm
2. Wire plain (0.276”), R12 (0.276”), R12
spiral
Pile cap reinforcement: 22 mm 22 mm
1. Main (0.875”), D22 (0.875”), D22
2. Transverse D10 - 152 mm D10 - 152 mm
Pile dimension (B) 0.45 x 0.45 m 0.45 x 0.45 m
Figure 4. Embedment capacity predictions of varying pile sizes Pile cap dimension 2.14 x 0.92 x 2.14 x 0.92 x
manufactured in Indonesia (Teguh et al. 2003a). 2.14 m 2.14 m
Pile embedment length 0.61 m (1.3 B) 0.45 m (B)

13
 Electronic Journal of Structural Engineering, 6 (2006)

In addition to the experimental process, the mag-

Table 2. Mechanical properties of concrete nitude of the structural loading was incrementally
Pile Compressive Tensile Modulus of elas- increased in accordance with a certain predefined
unit strength (MPa) strength (MPa) ticity, 104 (MPa) pattern. During the investigation, the cyclic lateral
Pile Pile cap Pile Pile cap Pile Pile cap load was applied three times at each load or dis-
P1 46.2 34.5 4.62 3.45 3.215 2.778 placement level. With the increase in the magnitude
P2 46.2 20.7 4.62 2.07 3.215 2.152
of loading, weak links and failure modes of the
structure were found as shown in Figure 5b. The
Table 3. Mechanical properties of reinforcing bars
Reinforcement Yield stress Tensile Modulus of
loading carried out in the experiment was monotonic
(MPa) strength (MPa) elasticity (MPa) with the effects of the cyclic behaviour and load re-
Tendon 1791 1882 1.94 x105 versals being estimated by using a modified mono-
Wire 448 630 1.86 x105 tonic force-deformation criteria and with damping
D22 275 464 2.00 x105 approximations (Figure 6).
D12 275 464 2.00 x105
4.3 Time history analysis
Time history analysis incorporating a discrete model
where the inelastic action is effectively lumped or
concentrated over a short length known as a plastic
hinge is described in this section. The program
Ruaumoko (Carr 2001) incorporates the Giberson
one-component model, which is an elastic beam
with the potential to develop a plastic hinge at one or
both ends. The program has also a wide range of
hysteretic models for describing the inelastic behav-
a) FE deformed stress contour b) Experimental test iour of a member including the simple elasto-plastic
at maximum moment (Harries and Petrou 2001b) and bi-linear models and the more complex modified
Figure 5. Stress propagation and deformation of the structure. Takeda degrading stiffness rule.
In the time history analysis, a simple finite ele-
In the 3D linear static analysis, lateral and axial ment model of the pile-to-pile cap connection for
loads were predicted by the Response2000 program. test unit 1 was descretised into small elements and
This numerical analysis provided stress distribution established using the Giberson’s model with a plas-
at each element as shown in Figure 5a. tic hinge at the fixed end. The lumped mass proper-
ties and cyclic loading history applied to the free end
were identical to the values used for the experiment.
In the analysis, input data for ground acceleration
was based on the 1995 Kobe Earthquake record
(Figure 7). Using the Ruaumoko 3D program, the tip
displacement for unit 1 was predicted as shown in
Figure 8.

600
400
Input acceleration (g)

200
0
-200 1 21 41 61 81 101 121 141 161 181 201

-400
-600
Figure 6. A comparison of load-deformation response.
-800
As seen there are crack damage propagating along Time (second)
the tensile fibre as well as stress concentrations oc-
curred at the interface between the pile head and pile Figure 7 Ground acceleration input (Kobe Earthquake, 1995).
cap and these stress concentrations also occur at the
opposite face when the lateral load is reversed.
14
 Electronic Journal of Structural Engineering, 6 (2006)

method, which is less computationally demanding,

60 involves the use of embedded or smeared reinforce-
ment. The second method, more computationally
T ip dispacement (mm) .

40 expensive, involves separate discrete modelling of

20 the reinforcement. The second model allows for the
investigation of bond-slip behaviour of reinforce-
0 ment with respect to the surrounding concrete. This
0 2 4 6 8 10 12 technique becomes computationally expensive when
-20
carried out over the entire system.
-40

-60
Time (second)

Figure 8 Time-displacement response at cantilever tip.

4.4 Nonlinear cyclic analysis

To understand the nonlinear behaviour of pile-to-
pile cap connections subjected to seismic actions, re-
fined joint models for both pile units were developed
and are presented in this paper. These proposed
models were analysed using the commercial finite
element analysis program, DIANA 9.0 (de Witte and
Kikstra 2005). Each material displayed nonlinear
static behaviour in order to achieve the desired per-
formance, and material properties had to be accu-
rately modelled. Figure 9. Constitutive material models for concrete and rein-
Unit test variables from Harries and Petrou forcement.
(2001) are listed in Table 1 and the mechanical
properties of concrete and reinforcing bars are de- The constitutive relationships used for the FE
tailed in Tables 2 and 3, respectively. Each pile unit models are shown in Figure 9. The non-prestressed
was tested under a constant axial load of approxi- and prestressed reinforcements use Von Mises yield
mately 0.1 Agfc’ (890 kN) and a reverse cyclic load hardening criteria with constitutive models matching
of like magnitude. Effects of secondary moment on the behaviour determined from testing (Figure 9d).
the column were excluded. Pile embedment lengths Concrete is modelled using Von Mises yield criteria
for units P1 and P2 varied ranging 1.0-1.3 of pile for compression a tension cut-off from the concrete
width respectively. compressive strength, fc’, to the concrete tensile
A single prestressed concrete pile connected to a strength, ft’ (Figure 9a) for tension. The concrete
cast-in place pile cap was modelled with discrete 3- compressive behaviour models the behaviour ob-
D mesh solid brick elements to match the boundary tained from displacement controlled testing of con-
conditions and geometry of the tested pile units. The crete cylinders (Figure 9c). The Hordijk model
pile and pile cap were modelled by a twenty-node shown in Figure 9b was used for the concrete tensile
isoparametric solid brick element (HE20 CHX60) behaviour (de Witte and Kikstra 2002). It consists of
and the analysis based on quadratic interpolation and elastic response to the tensile capacity followed by a
Gauss integration. Longitudinal and transverse bars
in the pile cap, plain spiral reinforcement and tendon 140

in the pile were modelled by the discrete truss ele- 120

ment to include an interaction between reinforcing
Transverse Load (kN)

bar and concrete. This interaction was modelled by 100

introducing interface and/or linkage element(s) in 80

the interface between truss element and concrete Envelope of Experimental
element. All other reinforcing bars and tendons were 60
Response
modelled by an embedded element, assuming perfect Von Mises Model-P lain
40
E mbedment
bond. Von Mises Model-
20
In the FE method, reinforcement may generally H eaded E mbedment
be modelled by one of two methods. The first 0
0 0.05 0.1 0.15 0.2
15 Tip Displacement (m)
 Electronic Journal of Structural Engineering, 6 (2006)

nonlinear unloading branch. Cracking was modelled 5 FINDING AND DISCUSSION

using both multiple fixed cracks and rotating crack
formulations. The results presented in this paper are Seismic performance of piles is directly related to
limited to the fixed crack model. the earthquake response of pile-to-pile cap connec-
tions. The maximum loads at early stage of analysis
140 were computed from the member yield moment ca-
120
pacities generated in the structure at its state of near
collapse, but are limited to the values generated to
Transverse Load (kN)

100 the structure under elastic condition. The theoretical

80
method presented predicts shear and moment capaci-
Envelope of Experiment al
ties of the plain embedment of pile-to-pile cap con-
60
Response nection and performance of square prestressed con-
40
Von Mises Model - P lain crete piles. In addition, previous experimental tests
Embedment
Von Mises Model - Headed
of plain square pile-to-pile cap connections by Har-
20
Embedment ries and Petrou (2001a) show good agreement with
0 the analytical method.
0 0.05 0.1 0.15 0.2 This paper reports the preliminary results for a
Tip Displacement (m) pile-to-pile cap connection with both connection de-
a) Unit P1 tails; i.e., plain and headed embedments. The more
b) Unit P2 complex behaviour of pile-to-pile cap connections is
Figure 10. Comparative load-displacement response (nonlinear being investigated and will be reported in the near
FE analysis). future.
The response of the pile due to cyclic loading is
presented in Figure 6. The predicted pile moment
capacity of 285 ft-kips (386 kN-m) is shown as a
horizontal dotted line. It was observed that the ap-
plied load of 13.9 kips (62 kN) caused the first crack
at the pile-to-pile cap interface, while the predicted
load to cause cracking of the pile was 11.2 kips
(49.7 kN).
As seen from Figure 6, Response2000 (Bentz
2001) program gave a maximum load of 25.44 kips
a) Unit P1 b) Unit P2 (113.2 kN), with significant inelastic behaviour
Figure 11. Comparative load-displacement response (nonlinear close to this load. The maximum load corresponds to
FE analysis). a displacement of 3.13 in (79.5 mm) at the pile tip.
In addition, theoretical moment-curvature relation-
In this study, Modified Newton Raphson incre- ships were developed for the pile cross sections us-
mental-iterative method with tolerance for conver- ing the Modified Scott stress-strain models for spi-
gence of 0.0001 was used. The comparison of the rally confined concrete. The stress-strain constitutive
load-displacement response is shown in Figure 9 and model for concrete developed at the University of
the finite element results envelope the experimental Melbourne (Mendis 2001) took into account the in-
findings. As discussed earlier, the experimental in- crease in strength and ductility of the concrete re-
vestigations were performed cyclically while the fi- sulting from the confinement. This model was used
nite element analyses are static pushover (mono- to modify the existing unconfined stress-strain
tonic). The lateral loads were applied at the point of model in the Response2000. The force-displacement
zero moment or at the pile tip while static axial loads responses obtained from numerical analyses proce-
consist of static axial load of 0.1 fc’Ag (890 kN) and dures, finite-element analysis and Response2000 are
initial prestress of 1.394 kN/mm2. The pile embed- shown in Figure 6. The load-displacement response
ment length was varied from 1.0 to 1.5 of pile width. predicted using the linear static pushover method is
As seen from Figure 10a and 10b, a longer embed- conservative compared to the results of the experi-
ment length develops the flexural capacity of the mental investigation due to the linear material be-
pile without distress to the pile cap. haviour used during the analysis. However Re-
sponse2000 program gave a better prediction of the
load-displacement response.
The time history analysis using the Kobe Earth-
quake ground motion predicted a small value of
maximum cantilever tip displacement of 1.72 in.

16
 Electronic Journal of Structural Engineering, 6 (2006)

(43.7 mm) at 5.8 seconds (Figure 8 compared to the 6 CONCLUSIONS

experimental maximum of 3.10 in. (78.8 mm) (see
Figure 7. Therefore the response used in the time 6.1 Current state-of-the art
history analysis need to include other parameters Indicative conclusions have been drawn based on
such as soil pile interactions. the analysis of the experimental data and the com-
During the tests conducted by Harries and Petrou parison with simple numerical models.
(2001) it was noted that, at the end of the test, the The obtained numerical predictions for load-
compressive fibres at the pile head (close to the in- displacement response using the finite-element
terface between pile head and pile cap) spalled and method and Response2000 gave reasonable agree-
crushed, and subsequent cracking propagated else- ment compared to experimental results. It is evident
where near the connection (Figure 5b). Furthermore, that more work is required to understand the behav-
an interfering action resulted when moment was ap- iour of prestressed concrete pile-to-pile cap connec-
plied at the interface of the pile and the pile cap, tions between experimental results and numerical
thereby increasing the contact pressure at the junc- predictions will assist in confirming the indicative
tion of the pile face and the pile cap. The maximum findings. Consideration of refined numerical simula-
stresses predicted by the finite element model (Fig- tions may also be of significant help.
ure 5) reflected the observed crack propagation at Harries and Petrou (2001) conducted cyclic lat-
the pile-to-pile cap interface, thus clearly represent- eral loading tests on two full scale pile-to-pile cap
ing the pile damage after testing (Figure 5b). Either connections. The major test variable was pile em-
the linear static pushover analysis or the time history bedment length in the pile cap; however the pile was
analysis, however, did not accurately predict the pile embedded in the pile cap without treatment (plain
damage except the nonlinear analysis provides the embedment model). The effect of confined headed
damage locations as shown in Figure 11 where the reinforcement at the connection was not investigated
value of principal stress greater than 1.0. in the experimental tests.
When concrete systems are subjected to cyclic It appears that the overall joint performance
loading the entire system undergoes tension- would be improved if the shear strength of non-
compression reversals. As a result, cracks open and critical sections was reduced. By non-critical, it is
close leading to a greater rate of system stiffness meant that these elements are not required to restrain
degradation than that observed under monotonic transverse forces. Design of piles employing a con-
loading. Load reversal requires more sophisticated ditional ductility approach, will result in a more eco-
constitutive modelling and this issue is being inves- nomical pile foundation system. Also, a more flexi-
tigated to better represent pile-to-pile cap connec- ble foundation system may reduce the ductility
tions. demand on the columns, potentially reducing dam-
A comparison of experimental and analytical re- age to the building or bridge under seismic events.
sults shows that the multi-directional fixed crack This work is continuing at the University of Mel-
formulation provides a good prediction of crack for- bourne.
mation in a three-dimensional model. The joint
cracking pattern observed in the experiment is cap-
6.2 Future work
tured by the analytical model. Furthermore, the
crack strain is largest along the pile embedment Current international design standards for pile-to-
crack, as observed in the experimental investigation. pile cap connections do not provide adequate guid-
It is observed that regions of high tension occur ance for reinforcement detailing at joints, however it
close to the pile cap longitudinal faces. These re- is common sense that the resulting joint design
gions of high transverse tension indicate potential should have enough ductility and serviceability in
locations for plastic hinges and principle stress con- resisting seismic loads. The results show that the
centrations occurred in the joint region of the pile- use of headed embedment as opposed to plain em-
to-pile cap connection (Figure 11); this behaviour bedment provides effective confinement of the joint
was observed in both pile units. Figure 11 shows the region, allowing for less congestion with compara-
principal stress distribution along pile height at ini- ble levels of performance. It is observed that using a
tial, middle, and final steps of monotonically in- longer headed embedment (1.2 – 1.5 of pile width)
creasing transverse loading (load case 3) indicated produces a strong connection and reduces stress con-
as LC3 16, LC3 71, LC3 121 for unit P1 and LC3 centration and crack damage at the interface and in-
16, LC3 41, LC3 71 for unit P2 respectively. From ner pile-to-pile cap connections. This study has also
these results, it is inferred that 3-D modelling is an shown that spiral confinement at the joint and along
important tool for further understanding and accu- the pile may be better achieved through the use of
rate modelling of the pile-to-pile cap connections. headed reinforcement than continuation of the pile
into the pile cap (plain embedment).

17
 Electronic Journal of Structural Engineering, 6 (2006)

The use of finite element models is effective in New Zealand Standard. (1982). "Code of Practice for the De-
capturing global and local behaviour of pile to pile sign of Concrete Structures, NZS 3101, Part 1: 1982."
Standard Association of New Zealand, Wellington.
cap connections Additional investigation into the in- Nogami, T., Ed. (1987). Dynamic Response of Pile Founda-
fluence of bond-slip under cyclic loading is required. tions--Experiment, Analysis, and Observation, Proceedings
Ideally, effective tools and guidelines should be de- of a session of the Geotechnical Engineering Division of
veloped to assist in the practical and efficient design the American Society of Civil Engineers in conjunction
of pile-to-pile cap connections. with the ASCE Convention in Atlantic City, ASCE, New
Jersey,185 p.
Pam, H. J. (1988). "Seismic Performance of Prestressed Con-
crete Piles and Pile-Pile Cap Connections," Ph.D Thesis,
7 REFFERENCES University of Canterbury, Christchurch, New Zealand.
A.C.I. Committee 318. (2002). Building Code Requirements Pam, H. J., and Park, R. (1990a). "Flexural Strength and Duc-
for Structural Concrete (ACI 318-02) and Commentary tility Analysis of Spirally Reinforced Prestressed Concrete
(ACI 318R-02), Piles." Prestressed Concrete Institute Journal, Vol. 35, No.
Bentz, E. (2001). "User Manual: Response-2000, Shell-2000, 4, July/August 1990, pp. 64-83.
Triax-2000, and Membrane-2000." Version 1.1. Pam, H. J., and Park, R. (1990b). "Simulated Seismic Load
Building Research Institute. (1996). "A Survey Report for Tests on Prestressed Concrete Piles and Pile-Pile Cap Con-
Building Damages due to The 1995 Hyogo-ken Nanbu nections." Prestressed Concrete Institute Journal, Vol. 35,
Earthquake." Ministry of Construction Japan. No. 6, November/December 1990, pp. 42-61.
Caltrans. (1990). Bridge Design Specifications, California De- Priestley, M. J. N., Seible, F., and Calvi, G. M. (1996). Seismic
partment of Transportation, Sacramento, California Design and Retrofit of Bridges, Wiley, New York
Carr, A. J. (2001). "RUAUMOKO Program: Inelastic Dynamic Priestley, M. J. N., Singh, J., Youd, T., and Rollins, K. (1991).
Analysis (3D Version)." 3-Dimensional Version(University "Costa Rica Earthquake of April 22, 1991 Reconnaissance
of Canterbury, New Zealand). Report." Publications 91-02, Earthquake Engineering Re-
de Witte, F. C., and Kikstra, W. P., Eds. (2002). DIANA Finite search Institute, pp. 59-91.
Element Analysis, User's Manual: Release 8.1, TNO Build- Sheppard, D. A. (1983). "Seismic Design of Prestressed Con-
ing and Construction Research, Delft, The Netherlands crete Piling." Prestressed Concrete Institute Journal, V. 28,
de Witte, F. C., and Kikstra, W. P., Eds. (2005). DIANA Finite No. 2, March-April 1983, pp. 20-51.
Element Analysis, User's Manual: Release 9.0, TNO Build- Silva, P. F. (1998). "Experimental and Analytical Models to
ing and Construction Research, Delft, The Netherlands Predict the Response of Pile to Pile Cap Connections Under
Hamada, M. (1991). "Damage of Piles by Liquefaction- Simulated Seismic Loads," Ph.D dissertation, University of
Induced Ground Displacements." The Third US Conference California, San Diego, California.
Lifeline Earthquake Engineering, ASCE, Los Angeles, US, Silva, P. F., Seible, F., and Priestley, M. J. N. (2001). "Influ-
pp. 1172-1181. ence of Strand Development Length in Formation of Plastic
Harries, K. A., and Petrou, M. F. (2001a). "Behaviour of Pre- Hinges in Prestressed Piles." PCI Journal, V. 46 No.
cast, Prestressed Concrete Pile to Cast-in-Place Pile Cap 3(May-June), pp. 76-89.
Connections." PCI Journal( July-August), pp. 82-92. Standard, N. Z. (1982). Code of Practice for The Design of
Harries, K. A., and Petrou, M. F. (2001b). "Behaviour of Pre- Concrete Structures, NZS 3101, Part 1: 1982, Standards
cast, Prestressed Concrete Pile to Cast-in-Place Pile Cap Association of New Zealand, Wellington
Connections." PCI Journal(July-August), pp. 82-92. Sugimura, Y. (1981). "Eathquake Damage and Design Method
Joen, P. H., and Park, R. (1990). "Simulated Seismic Load of Piles." Stockholm, pp. 865-868.
Tests on Prestressed Concrete Piles and Pile-Pile Cap Con- Tamura, Fujiwara, T., and Morishige, R. (1973). "Damage of
nections." Prestressed Concrete Institute Journal, V. 35, pp. Rigid Frame Bridges in the Off Tokachi Earthquake 1968."
42-61. Japan Earthquake Engineering Symposium 1973, Tokyo,
Kishida, H. (1966). "Damage of Reinforced Concrete Build- Japan, pp. 147-154.
ings in Niigata City with Special Reference to Foundation Teguh, M., Duffield, C. F., and Mendis, P. A. (2003a). "Seis-
Engineering." Soils and Foundations, JSSMFE, Vol. 6, No. mic Performance of Piles Designed and Constructed in In-
1, pp. 71-88. donesia." The Ninth East Asia-Pacific Conference on Struc-
Marcakis, K., and Mitchell, D. (1980). "Precast Concrete Con- tural Engineering and Construction, Bali, Indonesia, 16-18
nections With Embedded Steel Members." PCI Journal, V. December 2003, pp. GPE-17-22.
25, No. 4(July-August 1980), pp. 88-116. Teguh, M., Hutchinson, G. L., Mendis, P. A., and Duffield, C.
Mattock, A. H., and Gaafar, G. H. (1982). "Strength of Em- F. (2003b). "Introduction to Seismic Performance of
bedded Steel Sections as Brackets." ACI Journal, V. 79, Prestressed Concrete Pile-to-pile Cap Connections." De-
No. 2, pp. 83-93. partment of Civil and Environmental Engineering, Univer-
Mendis, P. A. (2001). Design of High-Strength Concrete Mem- sity of Melbourne.
bers, Engineers Australia Pty Limited, Melbourne, Austra- Wilson, E. L. (1996). "SAP2000 Integrated Finite Element
lia Analysis and Design Structures." CSI Computers & Struc-
Mizuno, H. (1987). "Pile Damage During Earthquake in Japan tures Inc., Berkeley, California.
(1923-1983)." Proceedings of a session of the Geotechnical Yashinsky, M. (1998). "The Loma Prieta, California, Earth-
Engineering Division of the American Society of Civil En- quake of October 17, 1989-Highway Systems." Profes-
gineers in conjunction with the ASCE Convention in Atlan- sional Paper 1552-B, USGS, Washington.
tic City, New Jersey, pp. 53-78.
Mizuno, H., Iiba, M., and Hirade, T. (1996). "Pile Damage
During the 1995 Hyogoken-Nanbu Earthquake in Japan."
The 11th World Conference on Earthquake Engineering
(WCEE), Acapulco, Paper No. 977.

18