-
nfth
otoShin4-95-897ki-A
Received 1 December 2009Revised 2 September 2011Accepted 2
September 2011Available online 4 November 2011
exural capacity was developed. The main structural
characteristics of the developed pile include (1) the
damaged concrete-pile foundations in reclaimed ground have
beenconducted to clarify the damage-process mechanism therein.Based
on a three-dimensional numerical simulation, Uzuokaet al. [5]
reported that precast, prestressed concrete piles in a liq-ueed
soil failed during this severe earthquake due to a lack of
used in these piles, it cannot contribute to increasing the
exuralstrength of the pile. In this paper, a new method for
increasing theexural strength of concrete piles was developed,
wherein unbond-ed prestressing steel bars are incorporated at the
center of the cross-section of the reinforced concrete piles. As
shown in Fig. 1b, in com-parison to conventional concrete piles
that are subject to small axialforces, the neutral axis of the
developed pile is much closer to thecentroidal axis because of the
prestress that is provided by the unb-ondedprestressed steel bars;
thus, the compression region increasesin cross-section and can
improve the exural strength of the pile. In
Corresponding author. Tel.: +81 3 52862694; fax: +81 3
52863485.E-mail addresses: [email protected] (M. Akiyama),
abe.satoshi.ha
@obayashi.co.jp (S. Abe), [email protected] (N. Aoki),
[email protected]
Engineering Structures 34 (2012) 259270
Contents lists available at
Engineering
lseku.ac.jp (M. Suzuki).1. Introduction
In the seismic design of concrete bridge systems, a plastic
hingemust be introduced at the bottom of each bridge pier rather
than atthe pile foundation. This is an important concept in
capacity designto help guarantee the rehabilitation of the bridge
after a largeearthquake [1,2]; however, if concrete bridge systems
areconstructed on strata (such as very soft soil) that can
experiencesoil liquefaction in a severe earthquake, it is difcult
to preventyielding of the pile foundation. Numerous structures with
precastconcrete-pile foundations in reclaimed ground were
seriouslydamaged by the 1995 HyogokenNambu Earthquake in
Japan[3,4]. Field investigations and numerical analyses of some of
the
exural strength and ductility capacity. In order to
maximizepost-event operability and minimize the repair costs of
bridges, in-creased attention should be paid to improving the
exural strengthand ductility capacity of precast concrete piles
that are driven intoliqueable soil.
Fig. 1adepicts the straindistribution in a concretepile that is
sub-jected to a bending moment at the point where the strain of the
ex-treme compression ber reaches 0.0035. Because the dead
load-induced axial force that acts on the pile foundations of most
bridgesin Japan is so small, as shown in Fig. 1, the neutral axis
of the pile thatis subjected to the bending moment becomes closer
to the extremecompressionber as thebendingmoment increases.
Therefore, eventhough concrete with a compressive strength of over
100 MPa isKeywords:Precast pileHigh-strength concreteHigh-strength
steelFlexural strengthCarbon-ber sheet0141-0296/$ - see front
matter 2011 Elsevier Ltd.
Adoi:10.1016/j.engstruct.2011.09.007neutral axis is constantly near
the centroidal axis of the pile, even if the longitudinal
reinforcement yieldsdue to a exural moment, because the pile has a
high axial compressive force that is induced by pre-stressed steel
bars, and hence, the concrete in the compression region can
contribute to increasing theexural strength of the pile; and (2)
the exural strength of the pile increases because the
high-strengthconcrete is conned by high-strength spirals and
carbon-ber sheets in combination with concrete inll-ing, and,
together, these modications provide a sufciently high
lateral-connement pressure.The results of bending tests demonstrate
that the proposed prestressed reinforced concrete pile with
carbon-ber sheets and concrete inlling had a much higher exural
capacity than a conventional precastconcrete pile. In addition, an
analytical approach is presented that can be used to obtain the
relationshipbetween the bending moment and the curvature of the
proposed pile. Even if concrete bridge systems areconstructed on
strata that can experience soil liquefaction, such as very soft
soil, bridge foundations thatuse the proposed piles could remain
undamaged under the design seismic action.
2011 Elsevier Ltd. All rights reserved.Article history: In this
study, a prestressed reinforced concrete pile that uses
high-strength material to increase the pilesFlexural test of
precast high-strength reiunbonded bars arranged at the center
of
Mitsuyoshi Akiyama a,, Satoshi Abe b, Nao Aoki c, MaDepartment
of Civil and Environmental Engineering, Waseda University, 3-4-1
Okubo,b Technical Research Institute, Obayashi Corporation, 4-6
Shimokiyoto, Kiyose, Tokyo 20cEast Nippon Expressway Company
Limited, 3-3-2 Kasumigaseki, Chiyodaku, Tokyo 100dDepartment of
Civil and Environmental Engineering, Tohoku University, 6-6-06
Arama
a r t i c l e i n f o a b s t r a c t
journal homepage: www.ell rights reserved.orced concrete pile
prestressed withe cross-section
yuki Suzuki d
juku-Ku, Tokyo 169-8555, Japan58, Japan9, Japanza-Aoba, Aobaku,
Sendai 980-8579, Japan
SciVerse ScienceDirect
Structures
vier .com/ locate /engstruct
-
.003
03
e dis
tra
trai
l for
g Stthe developed pile, concrete with a compressive strength
of100 MPa, a longitudinal reinforcement with a yield strength of700
MPa and a transverse reinforcement with a yield strength of1450 MPa
was used. In addition, carbon-ber sheets were used toprevent
spalling of the cover concrete. The effects of test variables,such
as the number of longitudinal bars, the prestress level andthe
presence or absence of carbon-ber sheets, on the exuralstrength of
the proposed pile were investigated through a series ofstatic
bending tests. An analytical approach that can be used to ob-tain
the relationship between themoment and curvature of the pro-posed
pile is presented.
This paper primarily investigates exure of the developed
pileunder monotonic loading with special emphasis on the
exuralstrength; hence, the effects of the experimental variables on
theductility capacity and residual displacement of the pile are
outsidethe scope of the current investigation. These effects will
be exam-ined in a future study using cyclic load tests.
2. Prestressed reinforced concrete piles using
high-strengthmaterials and carbon-ber sheets
A total of 17 piles, each 400 mm in diameter and 4000 mm
long,were tested under a static bending test. Table 1 summarizes
thematerial properties of the proposed piles. Examples of a
cross-sec-tion of the proposed pile are provided in Fig. 2. In
order to produce
0
0.0
Small axial stress
Axial stress provided by unbonded prestressing steel bars
xb,1 < xb,2 where, xb,1 and xb,2 are thbending
(a)
(b)
s
s
Fig. 1. Effect of the piles axia
260 M. Akiyama et al. / Engineerinthe proposed pile, (1) a
hollow pile was molded from high-strengthconcrete using centrifugal
force. The centrifugal forcewas applied asshown in Fig. 3 for 15.5
min. All of the piles were removed from thesteelmold after being
steam-cured for the rst 12 h andwere subse-quently air-cured until
testing. Then, (2) prestressing steel barswitha sheathwere inserted
into the center of the pile. (3) For some spec-imens, concrete
inlling (denoted Con-B in Fig. 2) was placed inthe hollowof the
pile,whereas (4), for other specimens, carbon-bersheets were
attached to the pile surface. Finally, (5) prestress wasprovided by
tightening the nutswith a torquewrench before releas-ing the jack
under themonitoring of a strain gauge thatwas attachedto the
prestressing steel bars. Some of the specimens did not haveconcrete
inlling and/or carbon-ber sheets. The sheath was notlled with grout
in any of the piles.
As shown in Fig. 2, because the ratio of the area of the
concretecover to that of the core concrete inCon-A is not small, it
is importantto prevent spalling of the concrete cover that is used
to increase theexural strength, even after the cover is damaged.
Over the past dec-ade, there has been an increasing interest in the
use of ber rein-forced polymer (FRP) in the repairing, retrotting,
strengthening,and new construction of concrete components. FRP can
provide lat-eral conning pressure to the internal concrete of FRP,
and themodeling of sections that are conned by FRP reinforcement
hasbeen examined [68]. In this study, unidirectional
carbon-bersheets were used to investigate the effect of conning the
concretecover with such sheets on the resultant exural strength of
the pile.The carbon-ber sheet can only affect the piles response to
a cir-cumferential stress. Therefore, the carbon-ber sheets crack
in re-sponse to tensile stress that is caused by a bending
moment.
Previous studieshave shownthat the axial load level
signicantlyaffects the exural behavior, especially the ductility,
of high-strength reinforced concrete columns [911]. This should be
consid-ered in determining the target prestress level of specimens,
which isexpressed as the axial force that is provided by the
unbonded pre-stressing steel barsdividedby the cross-sectional area
of thepile. Be-cause, in terms of capacity design, a bridge
foundation would bedesigned to remain undamaged under design
seismic action, in thisstudy target prestress
levelsweredeterminedby taking into accountonly the number and the
specied tensile strain of prestressed steelbars. As shown in Table
1, target prestress levels ranged from 0 to21.0 MPa. To insure
clearance between prestressing steel bars, therenumber of
prestressing steel bars arranged at the center of the cross-section
must be no more than three for the pile with a diameter of400 mm.
In addition, themanufacturer of the prestressing steel barsspecies
that the maximum tensile strain during providing the pre-stress
should be less than 65% of its designed yielding strain.
There-fore, the maximum prestress level provided by three
prestressing
Small axial stress 5
5 Axial stress provided by unbonded prestressing steel bars
xb,1
xb,2
tances from the extreme compression fiber to the neutral axis
of
in
n
ce on the strain distribution.
ructures 34 (2012) 259270steel bars for a pile with a diameter
of 400 mm is approximately21 MPa. Because the prestress was
provided just before testing, thelong-term loss of prestress in the
proposed pile could not be evalu-ated from this experiment.
The main structural characteristics of the developed pile
arethat (1) the neutral axis is constantly near the centroidal axis
ofthe pile, even if longitudinal reinforcements yield due to a
exuralmoment; this occurs because the pile has a high axial
compressiveforce that is provided by the unbonded prestressed steel
bars. Thus,the concrete in the compression zone can contribute to
increasingthe exural strength of the pile. (2) In addition, the
strain of theprestressed steel bars that are subjected to a bending
moment isso low that it is elastic until the ultimate state of the
pile isreached; this is because the unbonded prestressed steel bars
arearranged at the center of the piles cross-section. Even
thoughthe prestressed steel bars cannot resist the bending moment,
theresidual displacement can be reduced by the prestressed steel
barsafter the load is removed. (3) High-strength longitudinal bars
andspirals are used to increase both the exural and shear
strengthsof the pile. (4) The exural strength of the pile increases
due tothe fact that the high-strength concrete is conned by
high-strength spirals and carbon-ber sheets and due to the use
of
-
ssing
(MPa) (mm) (%)
ere
y th
cons
g Stconcrete inlling; these modications together provide a
suf-ciently high lateral-connement pressure.
For maximum exural strength, the proposed pile should have
Table 1Test specimens.
Notation ofspecimena
f 0c0b (MPa) Longitudinal bar Diameter of prestre
steel bar (mm)Con-A
Con-B
fylc
(MPa)Diameter qgd
(%)
D4-1 94.1 787 22.2 2.9 32.0D4-2 D4-3 107 D4-4 D4-5 114 D4-6 D4-7
110 796 31.8 5.9D4-8 D4-9 108 41.0 787 22.2 1.8D4-10D4-11 102 37.0
778 40.0D4-12D4-13 98.7 37.2 787D4-14 97.3 38.1D4-15 796 31.8
3.8D4-16 88.0 39.7 778 22.2 1.8D4-17
a D4 indicates a diameter of 400 mm.b The average compressive
strength that was obtained from three cylinders that w
molded by centrifugal force and concrete inlling, respectively,
as shown in Fig. 2.c Yield strength of the longitudinal bar.d Ratio
of the area of the longitudinal bar to the cross-sectional area.e
The prestress level, which was calculated as the axial force that
was provided bf Yield strength of the spiral.g Spacing of the
spiral.h Volumetric ratio of the spiral.i Type-Ameans that
carbon-ber sheets were attached to the concrete in the
carbon-ber sheets were attached to the entire pile.
M. Akiyama et al. / Engineerinboth concrete inlling and
carbon-ber sheets; however, pileswithout concrete inlling and/or
carbon-ber sheets were testedin order to investigate the effect of
these modications on thebehavior of the pile. All piles have a
sufcient number of high-strength spirals to exhibit a exure failure
mode. In order to permitoptimal shear design of the proposed piles,
the effects of the num-ber and yield strengths of spirals on the
shear strength of the pro-posed pile should be investigated.
3. Experimental procedure
3.1. Specimen properties and materials
The arrangement of the longitudinal bars and unbonded
pre-stressed steel bars in the proposed pile are shown in Fig. 2.
Asshown in the gure, the spacing of the unbonded prestressed
steelbars and the specied thickness of the Con-A depends on the
diam-eters of the unbonded prestressed steel bars. Table 1 depicts
theaverage thickness of Con-A, measured after the hollow pile
wasmolded by the centrifugal force.
Type I Ordinary Portland cement was used in all of the
concretemixtures. Crushed gravel was used as the coarse aggregate,
where-in the maximum aggregate size, Gmax, was 15 mm. For the
concretethat was used in the Con-A, silica fumes were used to
obtain highstrength, workability and the reduction of ne-particle
segrega-tion. Highly owable concrete was used as the concrete
inlling.The concretes compressive strength, f0c0, was measured as
theaverage of three identical cylinders, each of which had a
diameterof 100 mm and a height of 200 mm. These cylinders were
testedunder axial loading at the time the corresponding pile was
tested.
The yield strengths of the longitudinal bars and spirals
areshown in Table 1. The yield strengths of the unbonded
prestressedsteel bars were 1150 and 1230 MPa for the 32- and
40-mm-diam-eter bars, respectively. The tensile strength of the
carbon-bersheet was 4620 MPa; a single layer with a sufcient
overlap lengthto anchor the sheet (=100 mm) was used. The
carbon-ber sheets
tant-moment region + 150 mm, as shown in Fig. 3, whereas Type-B
means that0.0 1440 60 1.29 75.6 20.6 79.6 0.0 78.9 9.8 81.8
20.3 80.4 19.5 120 0.65 84.9 19.8 60 1.29 83.1 20.3 120 0.65
80.4 12.6 60 1.29 89.6 12.1 120 0.65 89.4 6.3 60 1.29 66.2
12.7 68.9 Type-A21.0 66.6 20.4 83.7 Type-A13.4 78.0 Type-B13.7
66.9 Type-A20.1 66.4 Type-A
100 mm in diameter and 200 mm in length. Con-A and Con-B
indicate concrete
e prestressed steel bars divided by the sum of the areas of
Con-A and Con-B.Prestresslevele (MPa)
Spiral Thickness ofCon-A (mm)
Carbon-bersheeti
fysf sg qsh
ructures 34 (2012) 259270 261were bonded onto the concretes
surface with epoxy resin. Theminimum curing time for bonding of the
carbon-ber sheets beforethe test was 10 days.
3.2. Testing procedure and instrumentation
Each specimen was tested under a monotonically increasingload
until failure using a four-point bending setup, as shown inFig. 4.
Deection was measured using ve linear variable-differen-tial
transducers (LVDTs). The deection distribution was approxi-mated by
a cubic function under the boundary condition thatdeection on the
supports is zero. The averaged curvature couldbe obtained by
differentiating the approximated deection func-tion for the
constant-moment region.
The cracking behavior of the pile without the carbon-bersheets
was visually observed. Electrical strain gauges were at-tached to
the following surfaces: concrete, prestressing steel
bars,longitudinal bars, spirals and carbon-ber sheets. When
prestresswas applied to the pile, the strain gauge on the
prestressed steelbars was controlled such that a specied prestress
was obtained.After the steel bar-provided prestress was introduced,
the mea-sured compression strains of the concrete and longitudinal
barswere almost equal to the values that were computed from the
gi-ven prestress.
4. Experimental results and discussion
4.1. General observations
Because the specimens had a sufcient number of spirals, all
thepiles exhibited a exural failure mode only within the
constant-moment region, even though minor shear cracks were
observed
-
Concrete infilling (Con-B)
400
240 115
80 80
Direction of loading
400 260 137
70 70
Longitudinal bar
Concrete molded by centrifugal force (Con-A)
Prestressing steel bar
Spiral
Carbon-fiber sheet
Units: mm
Direction of loading
Units: mm
Concrete molded by centrifugal force (Con-A) Prestressing steel
bar
Spiral Longitudinal bar
(a) Sample cross-section of a proposed pile that uses
prestressed steel bars where each has a diameter of 32 mm.
(b) Sample cross-section of a proposed pile that uses
prestressed steel bars where each has a diameter of 40 mm.
Specimen: D4-1
Specimen: D4-14
Fig. 2. Sample cross-sections of the proposed piles.
Metallic mold for creating the pile
Step 1250 rotations per minute for 6 min.
Step 2400 rotations per minute for 4 min.
Step 3850 rotations per minute for 2 min.
Step 41,000 rotations per minute for 3 min.
Step 51,550 rotations per minute for 0.5 min.
Fig. 3. Molding the pile using centrifugal force.
262 M. Akiyama et al. / Engineering Structures 34 (2012)
259270
-
500
000
500
rime
g Stoutside the constant-moment region. Fig. 5 shows the crack
pat-terns of piles D4-3 and D4-5, which were constructed without
car-bon-ber sheets or concrete inlling, at the yield point of
thelongitudinal bars and at the maximum loading point. These
twospecimens had almost the same strength concrete and steel
barsbut different amounts of applied prestress. As the prestress
appliedto the pile was increased, the number of cracks decreased.
Thecrack length of the pile with prestress that was provided by
theprestressed steel bars was shorter than the crack length of the
pilewithout prestress because the pile with a high prestress had
amuch larger compression zone in the cross-section.
The piles without carbon-ber sheets and concrete
inllingexhibited brittle behavior after the concrete cover in the
con-stant-moment region was cracked and spalled. The piles
withoutcarbon-ber sheets but with concrete inlling did not show
brittlebehavior even after the concrete cover was spalled. The
longitudi-nal bars in the piles with a wide spacing of spirals
buckled afterspalling of the concrete cover. The maximum loads on
these spec-imens were observed during spalling of the concrete
cover. Thepiles with carbon-ber sheets exhibited ductile behavior.
Thesepiles experienced maximum loading during the rupture of the
car-bon-ber sheets due to the expansion of the cover concrete. Fig.
6
Prestressing steel bar
Fig. 4. ExpePile
1,260
500
Nut
2,
Steel plate 150
M. Akiyama et al. / Engineerindepicts an example of the
appearance of the constant-moment re-gion at the point of maximum
loading of specimens D4-5 (withoutconcrete inlling and carbon-ber
sheeting) and D4-16 (with con-crete inlling and carbon-ber
sheeting).
4.2. Effects of the test variables on the exural strengths of
theproposed piles
Fig. 7 depicts the relationship between the load and deectionat
the midspans of piles D4-3, D4-4 and D4-5. These piles did nothave
carbon-ber sheets or concrete inlling. The three pilesexhibited
almost the same material strengths and structural prop-erties but
were subjected to different amounts of prestress. Theoccurrence of
specic events, such as the rst yielding of the lon-gitudinal bar
and spalling of the cover concrete, is indicated inFig. 7. The
tests conrmed that the load at cracking increased withincreasing
prestress; however, the peak loads of these three pileswere almost
the same, and the capacity of the pile with the highestprestress,
pile D4-5, decreased by about 30% after spalling of theconcrete
cover. Because the cover concrete of this pile had a
largecompressive strain, spalling of the cover occurred before
yieldingof the longitudinal bar.Fig. 8 depicts the results for
piles D4-9, D4-11 and D4-13, whichhad concrete inlling but no
carbon-ber sheets. The tests con-rmed that concrete inlling can
prevent a sudden decrease in loadafter spalling of the concrete
cover. It also conrmed that the max-imum loads of these piles did
not depend on the amount ofprestress.
Fig. 9 depicts the effects of different prestressing levels on
theexural strength of piles D4-14 and D4-16. These piles had
bothconcrete inlling and carbon-ber sheets, and the only
differencebetween them was the magnitude of the initial prestress.
Becausethe cover concrete was conned by the carbon-ber sheets, it
wasnot spalled by the compressive stress that was caused by the
bend-ing moment. As shown in Figs. 7 and 8, unlike piles without
car-bon-ber sheets, the use of a concrete cover in piles with
carbon-ber sheets can contribute to an increased exural strength.
As aresult, these piles had larger maximum loads. Because the
neutralaxis of a pile without initial prestress or with a low
prestress levelis closer to the extreme compression ber, the
tensile strain of pre-stressing steel bars in such a pile increases
as the bending momentincreases. Even for piles without initial
prestress, before loadingthe nuts were tightened with a torque
wrench to the point just be-fore the tensile strain of the
prestressing steel bars would begin to
LVDT
1,000 1,260
4000
500 500
Carbon-fiber sheet
-kN actuator
500
Units: mm
150
ntal setup.
ructures 34 (2012) 259270 263increase. As described hereinafter,
this increase in the tensile straincould result in there being no
difference in the compressive axialforce provided by the
prestressing steel bars among piles as thebending moment increases.
Therefore, as shown in Figs. 79, theprestress level does not have
an appreciable inuence on exuralstrength; however, Naaman and
Alkhairi [12] pointed out thatthe increment in the tensile stress
of unbonded tendons dependson the specimen length. In their
experiments, shorter specimensshowed larger increments of tensile
stress. The effect of the initialprestress level on the exural
strength of piles with differentlengths should be investigated.
The effects of the use of carbon-ber sheeting on the
exuralstrengthsof pilesD4-13andD4-14are shown in Fig. 10.
Theonlydif-ference between these two piles is the presence or
absence of car-bon-ber sheets. Pile D4-14, which had carbon-ber
sheets,possessed amuchhigherexural strength.Asmentionedabove,
uni-directional carbon-ber was used in the test; therefore, this
sheetcan only serve to reinforce the circumferential stress of the
pile. Be-cause the proposed pilewith carbon-ber sheets and concrete
inll-ing had a much higher exural strength than did the
conventionalprecast concretepile (seeAppendix), it is expected that
theproposedpile will be able to prevent yielding of the pile
foundation understrong earthquake excitation (Japan Road
Association [13]).
-
he l
g StThe effect of the amount of longitudinal bars on the piles
ex-ural strength is indicated in Fig. 11. If the piles had neither
concreteinlling nor carbon-ber sheets, the differences in their
maximumloads were not very large, as shown in Fig. 11a, and this is
becausespalling of the cover concrete occurs early due to the large
pre-stress. When spalling of the cover concrete is prevented by
the
Fig. 5. Crack patterns at the yielding of t
264 M. Akiyama et al. / Engineerinpresence of a carbon-ber
sheet, the amount of longitudinal barsaffects the exural strength,
as shown in Fig. 11b. Fig. 12 depictsthe effect of spiral spacing
on exural behavior. From Fig. 12a, itcan be seen that it is not
important to have a smaller spiral spacingfor piles without
concrete inlling because these piles exhibit brit-tle behaviors
after the concrete cover has become spalled. Even inpiles to which
concrete inlling was added to prevent brittlebehavior, the buckling
of the longitudinal bars of piles with largerspiral spacings was
clearly observed, and such buckling is indicatedby the sudden load
drops depicted in Fig. 12b.
Fig. 7 through Fig. 12 present data that are related to the
ductil-ity capacity of the piles. Because the proposed pile has an
increasedexural strength, it must also have sufcient shear strength
to pre-vent a brittle failure mode. As shown in Fig. 12, it is also
necessaryfor the proposed pile to have a smaller spiral spacing in
order toprevent buckling of the longitudinal bars. Although the
pile is de-signed to prevent yielding of the pile foundation and
not as asource of hysteretic energy dissipation, further research
is neededto identify the optimal combination of prestress level,
concrete andrebar strengths, in addition to the amount of
carbon-ber sheetingthat is needed to insure the adequate ductility
of piles under strongexcitation.
Fig. 13 depicts the relationship between the load and strain,
ez,on a carbon-ber sheet at the midspan of pile D4-12. Based
onexperimental tests of concrete columns with carbon-ber
sheetingunder concentric loading, Kawashima et al. [14] reported
thatstrain on the carbon-ber sheets increased at a great rate
whenez was larger than 0.0010.002 because of the expansion of
theinternal concrete. Similarly, in the measurements, the impact
ofez on the carbon-ber sheets of the pile increased sharply
afterreaching a value of 0.0010.002, and the longitudinal
barsyielded. This result indicates that a high conning pressure
wasprovided to the internal concrete by the carbon-ber sheets
andthat it contributed to increasing the exural strength of the
pile.
5. Analytical evaluation of the experimental results
ongitudinal bar and at a maximum load.
ructures 34 (2012) 259270The momentcurvature relationships
presented in this studyare derived from a cross-section layer model
that takes into ac-count the constitutive laws of the materials. In
order to describethe compression behavior of concrete, the
stress-averaged strainmodel presented by Akiyama et al. [15] is
used. In previousstressstrain models [1620], the experimental
longitudinal strainthat was used was expressed by the change in
gauge length dividedby the original gauge length. In tests that
have been reported in theliterature, the total length or diameter
of the specimen is used asthe gauge length. This strain should be
dened as the averagedstrain because the core concrete has a certain
fracture zone, andthe descending stress-averaged strain curve
depends on the gaugelength. Based on the constant-fracture-energy
concept for com-pression, Akiyama et al. developed a formalized
stress-averagedstrain model that uses the compressive fracture
energy and effec-tive conning pressure. This model is applicable to
reinforced con-crete columns that consist of concretes with
compressive strengthsof up to 130 MPa and transverse reinforcement
yield strengths ofup to 1450 MPa. Regardless of the gauge length
and cross-sectionaldimensions, this model agrees well with most
test results reportedin the literature.
The effective conning pressures are given by:
pe ke;vqw f s;c at a peak stress f cc of the confined
concrete1
p0e ke;vqw f y;h at 0:5 f cc after the peak stress 2
where f s;c Es 0:45ec0 6:39ke;vqwfc0
0:881( ) fyh 3
-
g StM. Akiyama et al. / Engineerinec0 0:0028 0:0008kb 4
Kb 40fc0 1:0 5
Fig. 6. Appearance of a constant-moment region at a maximum
loading.
Deflection at the midspan (mm)
Load
(kN)
D4-3 (fpe = 0N/mm2)D4-4 (fpe = 9.8N/mm2)D4-5 (fpe =
20.3N/mm2)
Yielding of the longitudinal bar Spalling of the cover
concrete
0 50 1000
400
800
1200
Fig. 7. Relationship between the load and deection in a pile
without carbon-bersheets or concrete inlling.400800
1200
Load
(kN)
D4-11 (fpe = 6.3N/mm2)
Yielding of the longitudinal bar Spalling of the cover
concrete
D4-9 (fpe = 12.6N/mm2)D4-13 (fpe = 21.0N/mm2)
ructures 34 (2012) 259270 265fc0 0:85f 0c0 6
ke;v 2s02 10s0ds 15d2s15d2s 1 qcc
for a circular cross-section 7
where qw is the area ratio of the transverse reinforcement,
which isdened as the total cross-sectional area, As, of the spirals
with spac-ing s divided by the area, sd, where d is the core
dimension that ismeasured from the center to center of a spiral;
ke,v is the effectiveconnement coefcient; fs,c is the stress in the
spiral at the peakstress; qcc is the ratio of the area of the
longitudinal steel to the areaof the core; ds is the core dimension
measured from center to centerof a spiral in the circular column; s
is the clear spacing between spi-rals; fyh is the yield strength of
the spiral; and f 0c0 is the compressive
0 50 1000
Deflection at the midspan (mm)
Fig. 8. Relationship between the load and deection in a pile
without carbon-bersheets and with concrete inlling.
D4-16(fpe = 13.7N/mm2)D4-14(fpe = 20.4N/mm2)
Deflection at the midspan (mm)
Load
(kN)
0 50 1000
400
800
1200
Yielding of the longitudinal bar
Fig. 9. Relationship between the load and deection in a pile
with carbon-bersheets and concrete inlling.
-
N)
400
g St
Lo
ad (k
400 Yielding of the longitudinal bar Spalling of the cover
concrete
D4-13 1,200
D4-14
800
266 M. Akiyama et al. / Engineerinstrength of plain concrete
that has been measured from a cylinderwith a diameter of 100 mm and
height of 200 mm.
The compressive fracture energy, Gf,c, is given by:
Gf ;c Gfc0 1 157 pefc0
77:3 pe
fc0
2( )8
Gfc0 134 93:3kb 9In this study, the effective conning pressures
pe and p0e in Eqs.
(1) and (2) were modied by taking into consideration whether
ornot the pile had carbon-ber sheets and/or concrete inlling. Whena
pile without a carbon-ber sheet had concrete inlling, the origi-nal
equations indicated in Eqs. (1) and (2) were used to obtain
theeffective conning pressure for the spiral-conned concrete in
re-gions (b) and (c) in Fig. 14, whereas the concrete cover in
region (a)in Fig. 14 was treated as plain concrete.
0 Deflection at the midspan (mm)
0 50 100
Fig. 10. Effect of carbon-ber sheeting on the exural
strength.
Yielding of the longitudinal bar
0 Deflection at the midspan (mm)
0 50 100 0 50 100
1,200
Load
(kN)
800
400
D4-6 D4-8
D4-16D4-15
(b)(a)
Yielding of the longitudinal bar Spalling of the cover
concrete
Fig. 11. Effect of longitudinal bar number on the exural
strength.1,200
Load
(kN)
800
Yielding of the longitudinal bar Spalling of the cover
concrete
D4-6 D4-5
Yielding of the longitudinal bar Spalling of the cover
concrete
D4-10 D4-9
Buckling of the longitudinal bar of D4-10
(b)(a)ructures 34 (2012) 259270When a pile did not have both a
carbon-ber sheet and concreteinlling, the concrete in region (a) of
Fig. 15 was treated as plainconcrete. The effective conning
pressure of the concrete in region(b) of Fig. 15 was modied to take
into account the reducedconning pressure due to the presence of the
hollow core usingthe factor f, as proposed by Kohashi et al.,
wherein this was basedon their experimental results with hollow
reinforced concrete col-umns under concentric loading [21,22]. The
effective conningpressure of a pile lacking a carbon-ber sheet and
concrete inllingwas determined as follows:
pe fke;vqwfs;c at the peak stress f cc of the confined
concrete10
p0e fke;vqwfyh at 0:5 f cc after the peak stress 11
0 Deflection at the midspan (mm)
0 50 100 0 50 100
Fig. 12. Effect of spiral spacing on the exural strength.
0
Gauge B
Gauge C
Gauge A
Load
(kN)
Gauge A Gauge B Gauge C
Yielding of the longitudinal bar Occurrence of damage of strain
gauge
Strain of the carbon-fiber sheet0.015
400
800
1,200
0.005 0.010
Fig. 13. Relationship between the load and the strain of the
carbon-ber sheets inpile D4-12.
-
f 2:0 tD
1 eF1100Ps 0 < t D2
12
F1 1fyh=200 F22 F2 13
F2 2:0 f 0c0 60F2 4 f 0c=30 60 < f 0c0 < 120
14
Ps As=t s 15where t and D are the thickness and diameter of the
pile,respectively.
For concrete columns with carbon-ber sheets under
concentricloading, some models for estimating the conning pressures
thatare provided by the carbon-ber sheets have been presented
[68,14]. Based on equations presented by Kawashima et al. [14],
Region (a)
Region (b)
Regions (a) and (b): Con-A
Fig. 15. Cross-section of a pile without concrete inlling or a
carbon-ber sheet.
Region (c)
Region (a)
Region (b)
Regions (a) and (b): Con-A Region (c): Con-B
Fig. 16. Cross-section of a pile with concrete inlling and a
carbon-ber sheet.
Region (a)
Region (b)
Region (c)
Regions (a) and (b): Con-A Region (c): Con-B
Fig. 14. Cross-section of a pile with concrete inlling and
without a carbon-bersheet.
0 0.5 1 1.50
50
100
Axial strain (%)
Axi
al S
tres
s (M
Pa)
Stress-averaged strain relation of concrete in Region (b) in
Fig. 15 without concrete infilling or a carbon-fiber sheet.
Stress-averaged strain relation of concrete in region (b) in
Fig. 14 with concrete infilling and without a carbon-fiber
sheet
Stress-averaged strain relation of concrete in region (b) in
Fig. 16 with concrete infilling and a carbon-fiber sheet
Fig. 17. Relationship between the stress and averaged strain in
conned concrete.
M. Akiyama et al. / Engineering Structures 34 (2012) 259270
2671000
0
2
4
6
D4-11 (fpe = 6.3N/mm2)
Incr
emen
t of S
tress
(MPa
)
Deflection (mm)50
D4-9 (fpe = 12.6N/mm2)
D4-13 (fpe = 21.0N/mm2) Fig. 18. Effect of the initial prestress
on the increment of stress in a prestressedsteel bar during
loading.
-
the effective conning pressure applied to the concrete in
region(a) of Fig. 16 is given as follows:
pe qCFeCFtECF at the peak stress f cc of the confined
concrete16
p0e qCF fCF at 0:5 f cc after the peak stress 17
qCF 4n tCF
D18
where qCF is the area ratio of the carbon-ber sheets, eCFt is
thestrain at the peak stress (=0.0015), ECF and fCF are the modulus
ofelasticity and tensile strength of the carbon-ber sheet,
respec-tively, n is the number of carbon-ber sheets that are
wrappedaround a pile, and tCF is the thickness of a carbon-ber
sheet.
The effective conning pressure applied to the concrete in
re-gions (b) and (c) in Fig. 16 is provided by the carbon-ber
sheetsand spirals and is given as follows:
pe ke;vqwfs;c qCFeCFtECFat the peak stress f cc of the confined
concrete 19
p0e ke;vqwfyh qCF fCF at 0:5 f cc after the peak stress 20The
stress-averaged strain model that was proposed by Akiy-
ama et al. [15] requires the use of the element length Lm
overwhich the compressive fracture energy Gf,c is dissipated. In
the ana-lytical evaluation of the concentric compression test
results, Lmmust be the same as the gauge length. Gf,c is kept
constant in thestrain-localized element, regardless of Lm, under
the condition thatLm is larger than the compressive fracture zone.
Therefore, the
Experimental result Computed result
Given by concrete in region (b) in Figs. 14 to 16
Given by longitudinal bars
Given by concrete in region (a) in Figs. 14 to 16
Given by concrete in region (c) in Figs. 14 and 16
M
omen
t (kN
m)
0
400
800 4-4D2-4D1-4D
D4-5 D4-6 D4-7
D4-9 D4-10 D4-11
0
400
800
0.1
800
D4-3
D4-8
D4-12
-15
0.1
Computed result assuming without prestressing
268 M. Akiyama et al. / Engineering Structures 34 (2012)
2592700
400
0
400
800 D4-13
D4-14
D4
0.1
0.1 Curvatu
Fig. 19. Comparison of the experimental and co
D4-16
D4-17 re (m-1)
mputed momentcurvature relationships.
-
post-peak portion of the stress-averaged strain curve changes
withLm, and it is steeper for larger values of Lm; however, there
havebeen no reports describing how to determine Lm in the
analyticalevaluations of concrete components that have been
subjected tosimultaneous axial loading and bending. In order to
minimize thedifferences between the experimental and computed
results ofthe momentcurvature relationship, Lm was set to 500
mm.Fig. 17 depicts the relationship between the stress and
averagedstrain of the concrete in region (b) of Figs. 1416 withLm =
500 mm. It was conrmed that having both concrete inllingand the
carbon-ber sheets enhanced the strength and ductilitycapacity of
the conned concrete.
Even if the stress-averaged strain relation of the tensioned
con-crete is considered in the section analysis, this may provide a
neg-ligible contribution to the relationship between the moment
andcurvature, that is, the tensile strength of the concrete is
ignored.A bi-linear model was used to demonstrate the steels
stressstrainrelationship. Because the unbonded prestressed steel
bars wereplaced at the center of the cross-section, the prestressed
steel barsdid not resist the bending moment. In the exural
analysis, thebending moment that was carried by the prestressed
steel barswas ignored.
Fig. 18 depicts the relationship between the increment in
thestress of the prestressed steel bars and the deections of
D4-9,D4-11 and D4-13 during loading. The increment in stress of
theprestressed steel bars that occurred before the ultimate
exuralload was reached was not negligible for the pile with the
lower
7.10 and 5.75 MPa, respectively. In comparison to the initial
pre-stress of these piles, the increments are signicant. In
addition,as shown in Fig. 12b, the piles with fewer spirals (i.e.,
widerspiral spacings), such as D4-8 and D4-10, exhibited buckling
ofthe longitudinal bars. The results for these specimens did
notindicate good correlations between the experimental and
analyt-ical results. It should be noted that if the tensile strain
of theprestressed steel bar does not increase with the bending
mo-ment, a higher exural strength could not be expected for thepile
with a lower prestress level, as shown in the computed re-sults for
D4-1, D4-3, and D4-11. The effects of the test parame-ters on the
increase in the tensile strain of the prestressed steelbars need to
be examined. For the specimens with higher pre-stress levels and
smaller spiral spacings, the analytical andexperimental
momentcurvature curves agree very well.
The data presented in Figs. 17 and 19 conrm that the carbon-ber
sheets and concrete inlling signicantly affected the
exuralstrengths of the piles. Even though the carbon-ber sheeting
thatwas used in this experiment cannot itself resist the bending
mo-ment and the contribution to the momentcurvature relation thatis
provided by concrete inlling [region (c)] is small, as shown inFig.
19, the utilization of carbon-ber sheets and concrete inllingcan
provide the conning pressures to concrete in regions (a) and(b) and
improve the behavior of concrete in regions (a) and (b). As
aresult, the piles with carbon-ber sheets and concrete inlling
(i.e.,D4-12, 14, 15, 16, and 17) show better structural
performances.Fig. 19 shows the computed results for the piles with
carbon-ber
0
M. Akiyama et al. / Engineering Structures 34 (2012) 259270
269prestress level, as shown in the gure; however, because this
incre-ment could not be quantied in this study, the sectional
analysiswas conducted by assuming that the initial compressive
force thatwas provided by the prestressed bars acted as an external
axialforce on the centroid. For the pile with the lower prestress
level,this will produce an underestimate of the exural
strength.
Fig. 19 depicts a comparison of the experimental and com-puted
momentcurvature relationships. The increments in thestress of the
prestressed steel bars of D4-3 and D4-11 were
0
250
500
0.00 0.05 0.1
Mom
ent (
kN m
)
(a)
5-4D
Experimental results Computed results (Lm=300 mm)Computed
results (Lm =500 mm)Computed results (Lm =700 mm) Curvature
(m-1)
Fig. 20. Effect of Lm on the computedsheets and concrete inlling
under the assumption that piles donot have prestressing steel bars.
These results show that it is nec-essary for the proposed piles to
have prestressing steel bars to in-crease the exural strength.
The element length, Lm, cannot be determined based onmechanical
considerations. Fig. 20 shows the effect of Lm on thecomputed
momentcurvature relationship. With longer Lm, thestress-averaged
strain relation in the post-peak region proposedby Akiyama et al.
[15] has a steeper descent if the conning
0
250
500
0.00 0.05 0.10
Mom
ent (
kN m
)
(b)
9-4D
Experimental results Computed results (Lm =300 mm) Computed
results (Lm =500 mm) Computed results (Lm =700 mm) m(erutavruC
-1)
momentcurvature relationship.
-
pressure applied to the concrete is low. Therefore, the
computedresult with Lm = 700 mm for pile D4-5 lacking a carbon-ber
sheetand concrete inlling exhibits brittler behavior, as shown
inFig. 20a. Since the stress-averaged strain relation has no
descend-ing branch and becomes independent of Lm as the conning
pres-sure increases, the computed momentcurvature relations withLm
= 300, 500, and 700 mm are almost the same as those shownin Fig.
20b. Although the best t with the experimental resultsfor piles
with diameters of 400 mm can be obtained by setting
(DAM105, A-D22) with a diameter of 400 mm are 3.9%, 0.47%,
4.1and 105 MPa, respectively. It should be noted that the
prestressingtendons of a conventional pile are bonded with concrete
and arenot arranged at the center of the pile but, instead, are
arrangedalong the circumference of the spiral, like longitudinal
bars. Thebending moment at the rst yield of a longitudinal bar is
suggestedby the manufacturer to be 170 kN m. If this pile is tested
underload by using a similar four-point setup, as shown in Fig. 4,
the cal-culated load at the rst yield of the longitudinal bar is
270 kN.
270 M. Akiyama et al. / Engineering Structures 34 (2012)
259270Lm = 500 mm, the appropriate magnitude of Lm to reproduce
theexperimental results may depend on a piles diameter and
struc-tural details. Further investigation is needed to determine
Lm. Theanalytical method to obtain the relationship between the
bendingmoment and the curvature will be used in the seismic design
of theproposed pile.
6. Conclusions
(1) In order to prevent the yielding of a pile foundation due to
asevere earthquake, the design of a new precast, high-strength, and
reinforced concrete pile that was prestressedwith unbonded bars
that were arranged at the center of pilecross-section was presented
in this study.
(2) The effects of concrete strength, number of longitudinal
bars,prestress level and presence or absence of concrete inllingand
carbon-ber sheets on the exural strength of the pro-posed pile were
experimentally investigated. The test resultsindicate that it is
necessary for the proposed pile to haveboth concrete inlling and
carbon-ber sheeting to increasethe pile exural strength and prevent
brittle failure.
(3) Sectional analyses of the proposed piles were conducted
toobtain the momentcurvature relationships. The effectiveconning
pressures were estimated by taking into consider-ation the presence
or absence of carbon-ber sheets andconcrete inlling. By using the
stress-averaged strain rela-tionship for the conned concrete with
the effective conn-ing pressures, the analytical and experimental
momentcurvature relationships agreed well with one another.
(4) Further research is needed to investigate the effects of
com-bined axial and shear forces on the ductility capacity
andplastic hinge behavior of the proposed pile. In these tests,the
structural details of pile heads that have been embeddedin concrete
footings should be examined so that design fea-tures can be
identied that will prevent the failure of pile-concrete footing
joints that are subjected to cyclic loading.
Appendix A. The exural strengths of conventional piles thatare
used in Japan
In Japan, the area ratio of the longitudinal bars, volumetric
ratioof the spirals, prestress level and specied concrete strength
for aconventional precast, prestressed, and reinforced concrete
pileReferences
[1] Priestley MJ, Seible F, Calve GM. Seismic design and retrot
of bridges. JohnWiley & Sons, Ltd.; 1995.
[2] Akiyama M, Matsuzaki M, Dang DH, Suzuki M. Reliability-based
capacitydesign for reinforced concrete bridge structures. Struct
Infrastruct Eng 2010.doi:10.1080/15732479.2010.507707.
[3] Matsui T, Oda K. Foundation damage of structures Special
Issue onGeotechnical Aspects of the January 17, 1995 HyogokenNambu
Earthquake.Soils Foundat 1996;1:189200.
[4] Tokimatsu K, Mizuno H, Kakurai M. Building damage associated
withgeotechnical problems Special Issue on Geotechnical Aspects of
the January17, 1995 HyogokenNambu Earthquake. Soils Foundat
1996;1:21934.
[5] Uzuoka R, Sento N, Kazama M, Zhang F, Yashima A, Oka F.
Three-dimensionalnumerical simulation of earthquake damage to
group-piles in a liqueedground. Soil Dyn Earthquake Eng
2007;27:395413.
[6] Karabinis AI, Rousakis TC. Concrete conned by FRP material:
a plasticityapproach. Eng Struct 2002;24:92332.
[7] Rousakis TC, Karabinis AI, Kiousis PD. FRP-conned concrete
members: axialcompression experiments and plasticity modelling. Eng
Struct2007;29:134353.
[8] Li G. Experimental study of FRP conned concrete cylinders.
Eng Struct2006;28:10018.
[9] Lgeron F, Paultre P. Behavior of high-strength concrete
columns under cyclicexure and constant axial load. ACI Struct J
2000;97(4):591601.
[10] Azizinamini A, Baum KSS, Brungardt P, Hateld E. Seismic
behavior of squarehigh-strength concrete columns. ACI Struct J
1994;91(3):33645.
[11] Ho JCM, Pam HJ. Inelastic design of low-axially loaded
high-strengthreinforced concrete columns. Eng Struct
2003;25:108396.
[12] Naaman AE, Alkhairi FM. Stress at ultimate in unbonded
post-tensioningtendons: part 2 Proposed methodology. ACI Struct J
1991;88(6):68392.
[13] Japan Road Association. Design specications of highway
bridges, Part Vseismic design. Tokyo, Japan: Maruzen; 2002.
[14] Kawashima K, Hosotani M, Yoneda K. Carbon ber sheet retrot
of reinforcedconcrete bridge piers. Proc Int Workshop Ann Commemor
Chi-Chi Earthquake,Nat Center Res Earthquake Eng, Taipei, Taiwan
2000;2:12435.
[15] Akiyama M, Frangopol DM, Suzuki M. Stress-averaged strain
model forconned high-strength concrete. ACI Struct J
2010;107(2):17988.
[16] Razvi R, Saatcioglu M. Connment model for high-strength
concrete. J StructEng ASCE 1999;125(3):2819.
[17] Cusson D, Paultre P. Stress-strain model for conned
high-strength concrete. JStruct Eng ASCE 1995;121(3):46877.
[18] Bing L, Park R, Tanaka H. Stress-strain behavior of
high-strength concreteconned by ultra-high- and normal-strength
transverse renforcements. ACIStruct J 2001;98(3):395406.
[19] Lokuge WP, Sanjayan JG, Setunge S. Stress-strain model for
laterally connedconcrete. J Mater Civil Eng ASCE
2005;17(6):60716.
[20] Hong K-N, Han SH, Yi S-T. High-strength concrete columns
conned by low-volumetric-ratio lateral ties. Eng Struct
2006;28:134653.
[21] Kohashi H, Nakatsuka T, Yamato S. Strength-deformation
characteristics ofcylindrical concrete conned with circular lateral
reinforcement. Pro JapanConcrete Inst 1999;21(3):2416 (in
Japanese).
[22] Kohashi H, Nakatsuka T, Yamato S, Yamamoto H. The effects
of materialstrength on strength-deformation characteristics of
cylindrical concreteconned with circular lateral reinforcement.
Proc Japan Concrete Inst2000;22(3):21722 (in Japanese).
Flexural test of precast high-strength reinforced concrete pile
prestressed with unbonded bars arranged at the center of the
cross-section1 Introduction2 Prestressed reinforced concrete piles
using high-strength materials and carbon-fiber sheets3 Experimental
procedure3.1 Specimen properties and materials3.2 Testing procedure
and instrumentation
4 Experimental results and discussion4.1 General observations4.2
Effects of the test variables on the flexural strengths of the
proposed piles
5 Analytical evaluation of the experimental results6
ConclusionsAppendix A The flexural strengths of conventional piles
that are used in JapanReferences