Professional Documents
Culture Documents
Assessment of Seismic Performance of Extended Pile-Shafts
Assessment of Seismic Performance of Extended Pile-Shafts
Correspondence to: Y. H. Chai, Department of Civil and Environmental Engineering, University of California,
Davis, Engineering Unit III, One Shields Avenue, Davis, CA 95616, U.S.A.
E-mail: yhchai@ucdavis.edu
Received 7 October 2002
Copyright
?
2003 John Wiley & Sons, Ltd. Accepted 17 January 2003
1938 Y. H. CHAI AND S. T. SONG
Figure 1. Transverse seismic response of extended pile-shafts.
the pile-shaft and column and the presence of a construction joint at the pile-shaft}column
interface.
Under a horizontal earthquake ground motion, the lateral force associated with the inertia
of the superstructure generates a bending moment distribution that varies with height but di-
minishes after attaining a maximum bending moment below the ground level, as shown in
Figure 1(b). A typical depth for maximum bending moment, and hence the location of the
plastic hinge, ranges from one to three or four diameters below the ground level, depend-
ing on the aboveground height and soil properties. Damage below the ground level due to
cracking or spalling of concrete, fracture of transverse reinforcement or buckling of longi-
tudinal reinforcement is generally dicult to assess after an earthquake. This, coupled with
the potential high-cost of repair, resulted in the current use of a design displacement duc-
tility factor that is smaller than that of columns in order to limit the amount of yielding in
the pile below the ground level. For example, ATC-32 [1] prescribes a displacement duc-
tility factor of 3 for extended pile-shafts compared to a displacement ductility factor of 4
for well-conned xed-base reinforced concrete columns. The reduction in design displace-
ment ductility factor led to the terming of extended pile-shafts as Limited Ductility Struc-
tures by ATC-32 [1]. The California Department of Transportation (Caltrans), on the other
hand, calls the extended pile-shaft type-I pile-shaft to dierentiate it from other types of pile
construction [2].
Since damage in extended pile-shafts may potentially occur below the ground level, an
assessment of the curvature ductility demand under dierent soil conditions becomes impor-
tant. The curvature ductility demand in extended pile-shafts depends on the exural strength
and rigidity of the pile, the lateral strength and stiness of the soil, as well as the plas-
tic hinge length of the pile. In this paper, the performance of extended pile-shafts is as-
sessed by determining the curvature ductility demand for a wide range of soil properties
and then is compared with the curvature ductility capacity available from current
design.
Copyright ? 2003 John Wiley & Sons, Ltd. Earthquake Engng Struct. Dyn. 2003; 32:19371954
ASSESSMENT OF SEISMIC PERFORMANCE OF EXTENDED PILE-SHAFTS 1939
CONFINEMENT OF CONCRETE
Current code requirements
Current seismic design of extended pile-shafts is similar to that of reinforced concrete in that
sucient conning steel must be provided in the yielding region to ensure a deformation
capacity that is commensurate with the level of ductility demand assumed in design. The
current approach for enhancing the ductility capacity of reinforced concrete relies on the
provision of sucient transverse reinforcement to conne the concrete in the potential plastic
hinge region. Eective connement of the core concrete however is predicated upon proper
detailing of the reinforcement, which requires the transverse reinforcement to be adequately
anchored and closely spaced in order to prevent buckling of the longitudinal reinforcement.
In the U.S., the current conning steel requirements for extended pile-shafts follow those
of reinforced concrete columns. For circular columns or piles conned by spiral or circular
hoops, the transverse steel in the plastic hinge region shall not be less than [1]:
j
s
=0.16
[
ce
[
ye
_
0.5 + 1.25
P
e
[
ce
A
g
_
+ 0.13(j
l
0.01) (1)
where the volumetric transverse steel ratio j
s
is dened as the volume of transverse steel per
unit volume of core concrete, and may be written as:
j
s
4A
sp
d
s
s
(2)
where A
sp
is the cross-sectional area of the transverse steel, d
s
is the core diameter formed
by the centerline of the spiral or circular hoops, and s is the center-to-center spacing of the
spiral or circular hoops. In Equation (1), [
ce
is the expected compressive strength of unconned
concrete, [
ye
is the expected yield strength of the reinforcement, P
e
is the design axial force, A
g
is the gross cross-sectional area of the column, and j
l
is the longitudinal reinforcement ratio.
The expected compressive strength of concrete may be taken as [
ce
=1.3[
c
, where the 30%
increase in compressive strength over its design value [
c
is in recognition of the conservatism
in current concrete mix design [1]. The higher compressive strength also recognizes the typical
strength gain beyond the specied 28-day compressive strength. A design concrete compressive
strength of [
c
=34.5 MPa is used in this paper. For the reinforcement used in the U.S., the
expected yield strength may be taken as [
ye
=1.1[
y
[1]. Thus, for grade-60 steel with yield
strength of 420MPa, the expected yield strength is [
ye
=462MPa. It should however be noted
that the expected yield strength of grade-A706 steel, which is equivalent to grade-60 steel,
has been reported by Caltrans to be slightly higher at [
ye
=475 MPa [2].
Figure 2 shows a plot of the conning steel ratio j
s
as required by Equation (1) for an
expected concrete compressive strength of [
ce
=44.8 MPa and expected steel yield strength of
[
ye
=475 MPa. Note that even though the x-axis in Figure 2 is plotted up to 0.15[
ce
A
g
, the
axial load level in extended pile-shafts is typically less than 0.1[
c
A
g
in the current design. It
can be seen from Figure 2 that the conning steel ratio varies in a fairly narrow range from
about j
s
=0.008 to 0.013 for axial force level up to 0.1[
c
A
g
. Although the actual concrete
compressive strength and steel yield strength may vary from that assumed in Figure 2, the
range of conning steel ratio is nonetheless considered representative of current design.
Copyright ? 2003 John Wiley & Sons, Ltd. Earthquake Engng Struct. Dyn. 2003; 32:19371954
1940 Y. H. CHAI AND S. T. SONG
Figure 2. Conning steel requirement per ATC-32 [1].
In addition to providing connement to the core concrete, the transverse reinforcement in the
plastic hinge region is also expected to restrain the longitudinal reinforcement from buckling,
which may occur in one of the two modes: (i) buckling between two adjacent layers of
transverse reinforcement, and (ii) buckling over a longer length involving multiple layers of
transverse reinforcement [3]. While the restraint against buckling between two adjacent layers
relies on close spacing of the transverse reinforcement, preventive measures against buckling
over multiple layers of transverse reinforcement rely on the provision of a sucient amount
of transverse reinforcement. For example, for restraint against buckling between two adjacent
layers of transverse reinforcement, the transverse steel spacing must be less than six times
the longitudinal bar diameter for reinforcement with ultimate-to-yield-strength ratio greater
or equal to 1.5, whereas for restraint against buckling over multiple layers of transverse
reinforcement, the transverse steel ratio must exceed 0.02% of the number of longitudinal
bars contained by the spiral or hoops [1, 3]. Although requirements against longitudinal bar
buckling may in some cases result in the transverse steel ratio being larger than that required
for connement-type failure, the primary focus of this paper is not on the buckling issues of
longitudinal reinforcement but rather on the requirements for connement of concrete.
Curvature ductility capacity
The curvature ductility capacity of a reinforced concrete section may be determined from
a moment-curvature analysis using the appropriate ultimate strain conditions for the section.
Copyright ? 2003 John Wiley & Sons, Ltd. Earthquake Engng Struct. Dyn. 2003; 32:19371954
ASSESSMENT OF SEISMIC PERFORMANCE OF EXTENDED PILE-SHAFTS 1941
In this paper, it is assumed that the ultimate condition of the pile section is governed by the
ultimate compressive strain of the concrete and not by fracture of the longitudinal reinforce-
ment. Such an assumption is reasonable since the plastic hinge length of the pile is longer
than that of a column due to the ability of the soil to spread the curvature distribution over
a longer length, which would result in a reduced curvature ductility demand and hence a
reduced tensile strain in the longitudinal reinforcement. The smaller tensile strain is not likely
to result in the fracture of the reinforcement under the current design level of ductility.
The current provision of transverse reinforcement is expected to signicantly enhance the
ultimate compressive strain of the concrete, which is in turn expected to increase the curvature
ductility capacity of the section. In this paper, the ultimate strain condition is dened using
the ultimate compressive strain c
cu
of the conned concrete, which is taken as [3]:
c
cu
=c
sp
+ 1.4
j
s
[
yhe
c
su
[
cc
(3)
where c
sp
is the ultimate compression strain of unconned concrete, which may be taken as
c
sp
=0.005 [2], j
s
is the conning steel ratio dened in Equation (2) [
yhe
and c
su
are the
expected yield strength and ultimate tensile strain of the transverse steel, respectively, and [
cc
is the compressive strength of the conned concrete, which is given by [4]:
[
cc
=[
ce
_
2.254
1 + 7.94
[
l
[
ce
2
[
l
[
ce
1.254
_
(4)
The term [
l
in Equation (4) corresponds to the eective conning pressure from the transverse
steel, which is given by [4]:
[
l
=k
e
[
l
(5)
where k
e
is the connement eectiveness coecient. For connement by spirals, the conne-
ment eectiveness coecient k
e
is given by [4]:
k
e
=
1
s
2d
s
1 j
cc
(6)
and s
1)
3z
p
J
u
M
max
(L
a
+L
f
)
3
(L
a
+L
m
)
+ 1 (8)
where j
a
, L
m
and L
f
are the aboveground height,
depth-to-maximum-moment and depth-to-xity normalized by the pile diameter, and z
p
is
the equivalent plastic hinge length L
p
normalized by the pile diameter D. The normalized
depth-to-xity L
f
depends on the exural rigidity of the pile, and the soil stiness, whereas
the normalized depth-to-maximum-moment L
m
depends on the exural strength of the pile
and the soil lateral strength. The kinematic relation in Equation (8) is similar to that used for
xed-base columns, such as that given by Priestley and Park [6], in that the local curvature
Copyright ? 2003 John Wiley & Sons, Ltd. Earthquake Engng Struct. Dyn. 2003; 32:19371954
1944 Y. H. CHAI AND S. T. SONG
ductility demand can be easily estimated for a given displacement ductility factor. However,
unlike the kinematic model for xed-base columns, the kinematic model of Equation (8) has
been extended to include the soil properties, particularly the lateral stiness and strength of
the soil.
Equivalent depth-to-xity
The kinematic relation of Equation (8) indicates that the local curvature ductility demand
depends on the parameter L
f
, which is an equivalent depth-to-xity normalized by the pile
diameter. The equivalent depth-of-xity may be interpreted as a stiness parameter charac-
terizing the lateral stiness of the soilpile system, the denition of which ensures the same
lateral stiness in the soilpile system and equivalent cantilever [7, 8]. For cohesive soils
where the lateral stiness of the soil is modeled by a constant modulus of subgrade reaction,
the normalized equivalent depth-to-xity is given by [5]:
L
f
=
_
3
_
4.24 + 6
a
+ 4.24
2
a
+
3
a
a
_
R
c
D
(9)
where R
c
is the characteristic length of the cohesive soilpile system, which is dened as:
R
c
4
_
EI
e
k
h
(10)
where EI
e
is the cracked exural rigidity of the pile, and k
h
is the constant modulus of
horizontal subgrade reaction (unit of force}length
2
). The constant modulus of horizontal sub-
grade reaction may be taken as k
h
=67s
u
, as recommended by Davisson [9], where s
u
is the
undrained shear strength of the cohesive soil.
For cohesionless soils where the lateral stiness of the soil is modeled by a linearly increas-
ing modulus of subgrade reaction, the normalized equivalent depth-to-xity is given by [5]:
L
f
=
_
3
_
7.2 + 9.6
a
+ 5.22
2
a
+
3
a
a
_
R
n
D
(11)
where R
n
is the characteristic length of the cohesionless soilpile system, which is dened as:
R
n
5
_
EI
e
n
h
(12)
where n
h
is the rate of increase of the modulus of horizontal subgrade reaction (unit of
force}length
3
). Guidance for selecting the value of n
h
for cohesionless soils is currently avail-
able, e.g. ATC-32 [1]. Note that the parameter
a
in Equations (9) and (11) is the aboveground
height L
a
normalized by the characteristic length of the soilpile system, and is given by:
a
L
a
R
c
or
L
a
R
n
(13)
for cohesive and cohesionless soils, respectively.
Copyright ? 2003 John Wiley & Sons, Ltd. Earthquake Engng Struct. Dyn. 2003; 32:19371954
ASSESSMENT OF SEISMIC PERFORMANCE OF EXTENDED PILE-SHAFTS 1945
Depth-to-maximum-moment
In adapting the equivalent cantilever model to extended pile-shafts, it is important to note that
the maximum bending moment should not be assumed to occur at the base of the cantilever but
rather at a depth above the equivalent depth-to-xity. The depth-to-maximum-moment, which
is essentially a strength parameter for the soilpile system, depends on the ultimate resis-
tance of the soil and the exural strength of the pile. The depth-to-maximum-moment denes
the location of the in-ground plastic hinge, and therefore inuences the ductility capacity of
the pile.
For cohesive soils, the depth-to-maximum-moment normalized by pile diameter, i.e. L
m
, is
given by the solution of the following quartic equation [5]:
M
max
=2L
a
L
m
+
_
1 +
14
9
L
a
_
L
m
2
+
7
81
(12 L
a
)L
3
m
7
108
L
4
m
(14)
where L
a
=L
a
}D is the normalized aboveground height, M
max
=M
max
}s
u
D
3
is the normalized
exural strength of the pile, M
max
is the exural strength of the pile, and s
u
is the undrained
shear strength of the cohesive soil. It should be noted that a parabolic ultimate soil pressure
distribution has been assumed when deriving Equation (14), and the solution for Equation
(14) is only valid for a normalized depth-to-maximum-moment of L
m
66. The solution range
however should cover most of the practical cases since the depth-to-maximum-moment is
typically less than six times the pile diameter except for very soft clays.
For cohesionless soils, the normalized depth-to-maximum-moment L
m
is given by [5]:
L
m
=
1
2
_
(L
a
)
2
a
+
_
(15)
where
=
3
_
4M
max
(L
a
)
3
+
_
8M
max
(2M
max
(L
a
)
3
) (16)
and M
max
=M
max
}K
p
D
4
is the normalized exural strength of the pile,
a
=0 to 8. For Caltrans plastic hinge length, the following properties were assumed: diam-
eter D=1.83 m, longitudinal reinforcement ratio j
l
=0.02, transverse steel ratio j
s
=0.0104,
expected compressive strength of concrete [
ce
=44.8 MPa, expected yield strength of lon-
gitudinal reinforcement [
ye
=475 MPa, and axial force P
e
=0.1[
c
A
g
. For cohesive soils, an
undrained shear strength of s
u
=20 kN}m
2
was used for the soft clays, whereas an undrained
shear strength of s
u
=200kN}m
2
was used for the sti clays. For cohesionless soils, a relative
density of D
r
=20% was used for loose sands, whereas a relative density of D
r
=80% was
used for dense sands.
It can be seen from Figure 4 that the current expressions for equivalent plastic hinge
length show a linear increase or near-linear increase in plastic hinge length with increased
aboveground height. The plastic hinge length for extended pile-shafts is seen to vary between
one and two pile diameters. The expression proposed by Caltrans (Equation (19)) is generally
longer than that proposed by Chai [5] or Budek et al. [11]. For Caltrans expression, soft
or loose soils generally result in a plastic hinge length slightly longer than that of sti or
dense soils. The Caltrans expression also shows a greater variation in plastic hinge length for
cohesive soils than for cohesionless soils. For cohesive soils, the plastic hinge length varies
Copyright ? 2003 John Wiley & Sons, Ltd. Earthquake Engng Struct. Dyn. 2003; 32:19371954
ASSESSMENT OF SEISMIC PERFORMANCE OF EXTENDED PILE-SHAFTS 1947
Figure 4. Normalized plastic hinge length of extended pile-shafts.
by about 0.25D for undrained shear strength from 20 to 200kN}m
2
, whereas the plastic hinge
length in cohesionless soils is practically the same for relative density from 20% to 80%.
RESULTS
In this study, the curvature ductility demand in extended pile-shafts is generated for a dis-
placement ductility factor of 3 to reect the overall ductility demand adopted in current design.
The undrained shear strength of cohesive soils was taken to vary from s
u
=20 to 300 kN}m
2
.
An undrained shear strength of 20kN}m
2
roughly represents a soft clay condition, whereas an
undrained shear strength between 100 and 200 kN}m
2
represents a very sti clay condition.
For an undrained shear strength greater than 200kN}m
2
, the cohesive soil can be classied as
hard clay [12]. For cohesionless soils, the relative density was varied from D
r
=20% to 85%,
where a relative density of 20% represents a loose sand condition and a relative density of
85% represents the boundary between dense and very dense sand [1].
Figures 5(a) and (b) show plots of the curvature ductility demand versus the undrained
shear strength of cohesive soils for a pile diameter of D=1.22 m. The curvature ductility
demands are determined for dierent normalized aboveground heights, longitudinal reinforce-
ment ratios, and axial load levels. It can be seen from these gures that the curvature ductility
demand decreases with increased undrained shear strength, i.e. decreasing curvature ductility
demand with increasing stiness of the soil. However, the rate of decrease of curvature duc-
tility demand is larger in the soft clay region than in the sti clay region. For the range of
Copyright ? 2003 John Wiley & Sons, Ltd. Earthquake Engng Struct. Dyn. 2003; 32:19371954
1948 Y. H. CHAI AND S. T. SONG
Figure 5. Curvature ductility demand in 1.22m diameter pile-shafts in cohesive soils and its comparison
with available curvature ductility capacity.
Copyright ? 2003 John Wiley & Sons, Ltd. Earthquake Engng Struct. Dyn. 2003; 32:19371954
ASSESSMENT OF SEISMIC PERFORMANCE OF EXTENDED PILE-SHAFTS 1949
normalized aboveground height 26L
a
610, the curvature ductility demand varies from 13 to
19 for soft clays, and from 11 to 14 for hard clays. The curvature ductility demand generally
increases with increased aboveground height, and is greater in the soft clay region than in the
hard clay region. The plots in Figure 5(a) also indicate that the curvature ductility demand
is only mildly aected by the longitudinal reinforcement ratio; only a small increase in cur-
vature ductility demand is noted for increasing longitudinal reinforcement ratio. For example,
for an axial force of P
e
=0.02[
c
A
g
in Figure 5(a), and for normalized aboveground height
of L
a
=8 and undrained shear strength of s
u
=100 kN}m
2
, the curvature ductility demand is
(j
[
)
dem
=13.7 for a longitudinal reinforcement ratio of j
l
=1%, whereas the curvature ductility
demand is (j
[
)
dem
=15.1 for a longitudinal reinforcement ratio of j
l
=4%. Cross-comparison
of the plots in Figures 5(a) and (b) also shows that the curvature ductility demand is not
very sensitive to the applied axial load, as the plots in Figure 5(a) are almost identical to
the corresponding plots in Figure 5(b). For example, for a longitudinal reinforcement ratio of
j
l
=2%, undrained shear strength of s
u
=100 kN}m
2
and normalized aboveground height of
L
a
=6, the curvature ductility demand is (j
[
)
dem
=13.8 for the axial force of P
e
=0.02[
c
A
g
in Figure 5(a), whereas the curvature ductility demand is (j
[
)
dem
=14.1 for the axial force of
P
e
=0.10[
c
A
g
in Figure 5(b). In this case, the increase in curvature ductility demand is less
than 3%.
It is also evident from Figures 5(a) and (b) that the curvature ductility capacity may be
smaller than the curvature ductility demand for the current level of displacement ductility,
depending on the longitudinal reinforcement ratio and aboveground height. In the case of a
low longitudinal reinforcement ratio of j
l
=1% in the top-left plot of Figure 5(a), the curvature
ductility capacity is (j
[
)
cap
=20.5, which is larger than the curvature ductility demand for the
entire range of undrained shear strength. For moderate longitudinal reinforcement ratios of
j
l
=2% or 3%, the reduction in curvature ductility capacity created a domain in the plot
where the curvature ductility capacity is smaller than the curvature ductility demand. For the
case of a high longitudinal reinforcement ratio of j
l
=4%, the curvature ductility capacity
falls below the curvature ductility demand for the entire range of undrained shear strength
and aboveground height. For the high axial compression of P
e
=0.10[
c
A
g
in Figure 5(b), the
curvature ductility demand varies in a manner similar to the curvature ductility demand for low
axial compression of P
e
=0.02[
c
A
g
in Figure 5(b). Note that the domain for (j
[
)
cap
(j
[
)
dem
becomes larger in Figure 5(b) than in Figure 5(a) due to the lower curvature ductility capacity
when subjected to a higher axial compression.
Figures 6(a) and (b) show similar plots of the curvature ductility demand, except for a
larger diameter of D=2.44 m. The larger diameter was selected to ascertain if pile diameter
has an inuence on the curvature ductility demand. It can be seen from Figure 6 that the
curvature ductility demand for D=2.44 m is almost identical to D=1.22 m, indicating that
the kinematic model upon which the curvature ductility demand is calculated is independent
of the pile diameter. The reduction in curvature ductility demand with increased undrained
shear strength and the increase of curvature ductility demand with increased aboveground
height, as noted for the diameter of D=1.22 m, are similarly noted for the larger diameter
of D=2.44 m. The relative insensitivity of curvature ductility demand to longitudinal rein-
forcement ratio and axial force level are also evident in Figures 6(a) and (b). The curvature
ductility capacity for D=2.44 m is also very close to that of D=1.22 m, resulting in similar
ratios between the curvature ductility demand and curvature ductility capacity for the two
pile diameters.
Copyright ? 2003 John Wiley & Sons, Ltd. Earthquake Engng Struct. Dyn. 2003; 32:19371954
1950 Y. H. CHAI AND S. T. SONG
Figure 6. Curvature ductility demand in 2.44m diameter pile-shafts in cohesive soils and its comparison
with available curvature ductility capacity.
Copyright ? 2003 John Wiley & Sons, Ltd. Earthquake Engng Struct. Dyn. 2003; 32:19371954
ASSESSMENT OF SEISMIC PERFORMANCE OF EXTENDED PILE-SHAFTS 1951
The curvature ductility demand in piles for cohesionless soils, and its comparison with
the curvature ductility capacity, is shown in Figures 7(a) and (b) for D=1.22 m and in
Figures 8(a) and (b) for D=2.44 m. The results are similar to that of cohesive soils in
that the curvature ductility demand decreases with increased soil stiness, as signied by the
increase in the relative density of the soil. The curvature ductility demand for cohesionless
soils is also not sensitive to the longitudinal reinforcement ratio or axial load level. Unlike
cohesive soils, however, the increase in curvature ductility demand with increased aboveground
height is more signicant in the dense sand region. The curvature ductility demand is also
independent of the pile diameter, as can be seen by comparing Figures 7(a) and (b) with
Figure 8(a) and (b). Similar to cohesive soils, the curvature ductility capacity for the low
reinforcement ratio of j
l
=1% exceeds the curvature ductility demand for the entire range of
relative density studied, and by a signicant margin for piles in the dense sand condition. For
moderate longitudinal reinforcement ratios of j
l
=2% or 3%, the curvature ductility capacity
falls below the curvature ductility demand in the low relative density region. Unlike cohesive
soils, however, the high longitudinal reinforcement ratio of j
l
=4% did not result in the
curvature ductility capacity falling entirely below the curvature ductility demand. However,
it should be noted that the rate of increase of the modulus of horizontal subgrade reaction
n
h
recommended by ATC-32 [1] was not reduced in this paper, even though testing of piles
in cohesionless soils indicated that the soil stiness may be signicantly lower than that
recommended in the literature [10]. A reduction in soil stiness will nonetheless increase
the curvature ductility demand, and further enlarge the domain where the curvature ductility
demand is larger than the curvature ductility capacity.
CONCLUSIONS
In this paper, the seismic performance of extended pile-shafts is assessed by comparing the
curvature ductility demand with the ductility capacity for a wide range of soil properties,
aboveground heights, axial load levels and longitudinal reinforcement ratios. The curvature
ductility demand was observed to decrease with increasing stiness of cohesive and cohesion-
less soils. The curvature ductility demand also increases with increased aboveground height.
The curvature ductility demand, however, is not sensitive to the longitudinal reinforcement ra-
tio or axial load level. Results from dierent pile diameters indicate that the kinematic model,
which is based on an equivalent cantilever concept, is independent of pile diameter for both
cohesive and cohesionless soils.
For longitudinal reinforcement ratios up to approximately 3%, the curvature ductility ca-
pacity available from the current level of conning steel is larger than 13, which is a value
currently recommended for design of extended pile-shafts. For a low longitudinal reinforce-
ment ratio of 1%, the available curvature ductility capacity is larger than the curvature ductility
demand for the undrained shear strength in the range of 206s
u
6300 kN}m
2
, and for the rel-
ative density in the range of 20%6D
r
685%. For increased longitudinal reinforcement to 2%
or 3%, however, the curvature ductility capacity may fall below the curvature ductility de-
mand for tall extended pile-shafts embedded in softer or loose soils. For a high longitudinal
reinforcement ratio of 4%, the curvature ductility capacity falls below the ductility demand
for the entire range of undrained shear strength of cohesive soils. For a longitudinal rein-
forcement ratio of 4% in cohesionless soils, however, the curvature ductility capacity only
Copyright ? 2003 John Wiley & Sons, Ltd. Earthquake Engng Struct. Dyn. 2003; 32:19371954
1952 Y. H. CHAI AND S. T. SONG
Figure 7. Curvature ductility demand in 1.22 m diameter pile-shafts in cohesionless soils and its
comparison with available curvature ductility capacity.
Copyright ? 2003 John Wiley & Sons, Ltd. Earthquake Engng Struct. Dyn. 2003; 32:19371954
ASSESSMENT OF SEISMIC PERFORMANCE OF EXTENDED PILE-SHAFTS 1953
Figure 8. Curvature ductility demand in 2.44 m diameter pile-shafts in cohesionless soils and its
comparison with available curvature ductility capacity.
Copyright ? 2003 John Wiley & Sons, Ltd. Earthquake Engng Struct. Dyn. 2003; 32:19371954
1954 Y. H. CHAI AND S. T. SONG
falls below the curvature ductility demand for tall extended pile- shafts or pile-shafts in loose
sand. Based on the results in this paper, it is recommended that an upper limit be placed on
the maximum longitudinal reinforcement in order to avoid a connement-type failure in the
pile. For sti clays with an undrained shear strength greater than 100 kN}m
2
, a limit of 3%
appears appropriate for an aboveground height less than six pile diameters. For cohesive soils
with an undrained shear strength less than 100 kN}m
2
, the limit should be reduced to 2%.
Similarly for cohesionless soils with relative density greater than 40%, a limit of 3% appears
appropriate for aboveground height less than eight pile diameters, whereas for cohesionless
soils with relative density less than 40%, the limit on longitudinal reinforcement ratio should
be reduced to 2%.
REFERENCES
1. Applied Technology Council ATC-32. Improved Seismic Design Criteria for California Bridges: Provisional
Recommendations, Redwood City, CA 94065, 1996.
2. SDC. Seismic Design Criteria Version 1.2, California Department of Transportation, Sacramento, CA 95816,
Dec 2001.
3. Priestley MJN, Seible F, Calvi GM. Seismic Design and Retrot of Bridges. Wiley Interscience: New York,
1996.
4. Mander JB, Priestley MJN, Park R. Theoretical stress-strain model for conned concrete. Journal of Structural
Engineering (ASCE) 1988; 114:18041826.
5. Chai YH. Flexural strength and ductility of extended pile-shafts Analytical model. Journal of Structural
Engineering (ASCE) 2002; 128:586594.
6. Priestley MJN, Park R. Strength and ductility of concrete bridge columns under seismic loading. ACI Structural
Journal 1987; 84:6176.
7. Dorwick DJ. Earthquake Resistant Design, 2nd edition. Wiley Interscience: New York, 1987.
8. Scarlat AS. Approximate Methods in Structural Seismic Design. E & FN Spon: London, 1996.
9. Davisson MT. Lateral load capacity of piles. Highway Research Record 1970; 333:104112.
10. Chai YH, Hutchinson TC. Flexural strength and ductility of extended pile-shafts Experimental study. Journal
of Structural Engineering (ASCE) 2002; 128:595602.
11. Budek AM, Priestley MJN, Benzoni G. Inelastic seismic response of bridge drilled-shaft RC pile}columns.
Journal of Structural Engineering (ASCE) 2000; 126:510517.
12. Das BM. Principles of Geotechnical Engineering, 2nd edition. PWS-Kent: Boston, 1990.
Copyright ? 2003 John Wiley & Sons, Ltd. Earthquake Engng Struct. Dyn. 2003; 32:19371954