See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/238192639

External Prestressing Technique for Strengthening of Prestressed Concrete

Structural Components

Article  in  Practice Periodical on Structural Design and Construction · May 2009

DOI: 10.1061/(ASCE)1084-0680(2009)14:2(90)

CITATIONS READS
7 1,838

6 authors, including:

S. Saibabu N. Lakshmanan
CSIR Structural Engineering Research Centre CSIR Structural Engineering Research Centre
16 PUBLICATIONS   38 CITATIONS    190 PUBLICATIONS   965 CITATIONS   

SEE PROFILE SEE PROFILE

A. Ramachandra Murthy Ramani Senthil

CSIR Structural Engineering Research Centre Anna University, Chennai
176 PUBLICATIONS   525 CITATIONS    32 PUBLICATIONS   98 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Seismic Performance evaluation of structures View project

MLP9941/CSIR-SERC/R&D Project View project

All content following this page was uploaded by S. Saibabu on 10 October 2016.

The user has requested enhancement of the downloaded file.

 External Prestressing Technique for Strengthening
of Prestressed Concrete Structural Components
S. Saibabu1; N. Lakshmanan2; A. Rama Chandra Murthy3; S. Chitra Ganapathi4;
R. Jayaraman5; and R. Senthil6

Abstract: This paper presents the details of an innovative external prestressing technique for strengthening of prestressed concrete
girders. The technique has been developed for anchoring the external prestressing to the sides of the end block. In the proposed technique,
transfer of external force is in shear mode on the end block and the required transverse prestressing force is smaller compared to
conventional techniques. In order to validate the technique, an experimental investigation has been carried out on two posttensioned end
blocks 共EB-1 and EB-4兲. Steel brackets are provided on either side of the end block for transferring external prestressing force and these
are connected to the anchor blocks by expansion-type anchor bolts. Performance of the end blocks has been studied for design, cracking,
and ultimate loads. A ductile failure is observed in EB-1 whereas a sudden failure is observed in EB-4. The slip and slope of the steel
bracket have been recorded at various stages during the experiment. These values are found to be smaller in EB-1 than in EB-4 due to a
higher anchorage depth of bolts. Finite-element analysis has been carried out by simulating the geometry, loading, material nonlinearity,
and test conditions. Linear and nonlinear static analysis has been conducted for the specified loadings and the responses have been
compared. From the analysis, it has been observed that the computed slope and slip of the steel bracket are in good agreement with the
corresponding experimental observations.
DOI: 10.1061/共ASCE兲1084-0680共2009兲14:2共90兲
CE Database subject headings: Prestressed concrete; Prestressing; Finite element method; Slopes; Slip; Concrete structures.

Introduction Bridge strengthening by external prestressing is to be carried out

The external prestressing technique is widely used in the con-
struction of various engineering structures and also as one of the
most efficient approaches for strengthening of existing structures.
In an external prestressing system, the prestressing tendons are
placed outside the concrete section and the prestressing force is
transferred to concrete through end anchorages and deviators. An
external prestressing system is simple to construct and easy to
inspect and maintain as compared to the internal tendon system.
Many bridges that were designed for earlier loading standards or
that have suffered damage or deterioration are not functioning
properly for the present traffic conditions. Poor reinforcement de-
tailing, design errors, and general wear and tear can impair struc-
tural performance. Corrosion of reinforcement, attack by
chemicals or pollution, overloading, impact damage from ve-
hicles, etc. can lead to loss of strength and loss of prestress.
1 hydration heat of concrete. Stoll et al. 共2000兲 studied experimen-
Scientist, SERC, CSIR Campus, Taramani, Chennai 600 113, India.
2 tally two full-scale bridge beams made up of high strength con-
Director, SERC, CSIR Campus, Taramani, Chennai 600 113, India.
3
Scientist, SERC, CSIR Campus, Taramani, Chennai 600 113, India
共corresponding author兲. E-mail: murthyarc@sercm.org, archandum@
yahoo.com
4
Scientist, SERC, CSIR Campus, Taramani, Chennai 600 113, India.
5
Scientist, SERC, CSIR Campus, Taramani, Chennai 600 113, India.
6
Assistant Professor, Structural Engg. Dept., Anna Univ., Chennai 25,
India.
Note. Discussion open until October 1, 2009. Separate discussions
must be submitted for individual papers. The manuscript for this paper
was submitted for review and possible publication on June 30, 2008;
approved on August 11, 2008. This paper is part of the Practice Periodi-
cal on Structural Design and Construction, Vol. 14, No. 2, May 1, 2009.
©ASCE, ISSN 1084-0680/2009/2-90–98/$25.00.
to regain the prestress lost. An external prestressing technique is
being widely used for enhancing strength of the existing/
distressed prestressed concrete girders to carry the required load-
ing or enhanced loading. If end blocks are not strong enough to
withstand the additional prestressing forces due to the proposed
external prestressing, it is susceptible to distress. A physical con-
cept of state of stress in the anchorage zone in the transverse
direction, that is, normal to planes parallel with the top and bot-
tom surfaces of the beam is tensile in nature over a length of end
block due to internal prestressing at a single point. Understanding
this state of stress due to external prestress is a complex phenom-
enon.
Ma et al. 共1999兲 provided the details of optimized anchorage
zone based on analytical and full-scale experimental studies.
Aparicio et al. 共2000兲 conducted ultimate load analysis with a
nonlinear finite element numerical model and measured the ther-
mal stresses and calculated the anchorage diaphragms due to the
crete and fiber reinforced polymer products for prestressing and
shear reinforcement. Miyamoto et al. 共2000兲 investigated the dy-
namic behavior of prestressed composite girder bridges, strength-
ened with external tendons. Aparicio et al. 共2002兲 presented the
results of a test program on externally prestressed concrete
beams. Five monolithic and three segmental beams were tested in
bending and in combined bending and shear. Choy et al. 共2002兲
studied the shear transfer mechanism of prestressed concrete en-
cased steel beams. Ghallab and Beeby 共2005兲 described various
factors that can influence the increase in the ultimate stress in
steel external prestressing tendons. Chen 共2005兲 conducted ex-
periments on four groups of prestressed steel-concrete composite

90 / PRACTICE PERIODICAL ON STRUCTURAL DESIGN AND CONSTRUCTION © ASCE / MAY 2009

Downloaded 05 May 2009 to 203.193.155.67. Redistribution subject to ASCE license or copyright; see http://pubs.asce.org/copyright

Fig. 1. Anchoring of external prestressing at the ends
beams with external tendons in negative moment regions. Lou
and Xiang 共2006兲 proposed a numerical model based on the finite
element method incorporating an arc length solution algorithm for
materially and geometrically nonlinear analysis of concrete beams
prestressed with external tendons. Results predicted by the analy-
sis were in good agreement with the experimental observations.
Czaderski and Motavalli 共2007兲 presented the details of a 40-year
old full scale concrete bridge girder with prestressed carbon fiber
reinforced polymer plates anchored using the gradient method.
Youakim and Karbhari 共2007兲 presented a simple method to cal-
culate the long-term prestress loss and the long-term change in
concrete stresses in continuously prestressed concrete members
with either carbon fiber reinforced polymer or aramid fiber rein-
forced polymer tendons.
In strengthening prestressed concrete beams by external pre-
stressing, the external tendons are anchored mostly to the ends of
the girder if the end regions are accessible for anchoring 共refer to
Fig. 1兲.
In this method, the external force is transferred to the member
in the compression mode. It may create bursting tensile stresses in
the end block. If the end blocks are not strong, it may lead to
cracking. Steel brackets with through-bolts along with laterally
prestressed anchor system is the another method to anchor exter-
nal prestressing tendons to the end block/web 共Fig. 2兲. This tech-
nique is widely employed in the field for retrofitting prestressed
concrete 共PSC兲 girders. In this method, transfer of prestress to the
member will take place through friction.
The required transversal prestressing force depends mainly on
the friction in the contact surface. The magnitude of the transver-
sal prestressing force fluctuates between 1.5 and 3.7 times the
value of the longitudinal prestressing depending upon the effec-
tiveness of the contact surface. Hence, it may create additional
tensile stress in the end block, which may lead to the development
of cracks. During execution of this method, active/passive rein-
forcement of an end block may be cut while placing transverse
prestressing bars through the end block. In addition, inducing
lateral prestressing is uneconomical and difficult for execution if
Tendon
Concrete Bracket
Tendon
Through-Bolts
Beam – Elevation End Section
Fig. 2. Anchoring of external prestressing at the sides using through
bolts
PRACTICE PERIODICAL ON STRUCTURAL DESIGN AND CONSTRUCTION © ASCE / MAY 2009 / 91
Downloaded 05 May 2009 to 203.193.155.67. Redistribution subject to ASCE license or copyright; see http://pubs.asce.org/copyright
Tendon
Steel Bracket
Tendon
Expansion-type
Anchor bolts
Beam – Elevation End Section
Fig. 3. Anchoring of external prestressing at the sides using
expansion-type anchor bolts
the girders are located at higher elevations. Further, loss of pre-
stress in short elements in the transverse direction is expected to
be high.
The writers propose an innovative method of side steel bracket
with shear key anchors. In the proposed method, the transfer of
external prestressing force will be in shear mode. Sleeve-type
expansion anchors have been proposed to anchor the external
prestressing tendons on the sides of the end block of the pre-
stressed girders through steel brackets as shown in Fig. 3. The
number of bolts and anchorage depth into the concrete block de-
pends upon the magnitude of required external force. The re-
quired transverse prestressing force is less compared to the value
obtained by the second method. Fouling with the existing rein-
forcement in the end block does not arise due to shorter anchor-
age depth of anchor bolts. In this method, required external
prestressing force will be transferred at a greater number of points
共through bolts兲, developing more uniform stress distribution in the
end block. This method is mainly suitable for continuous span
and suspended span bridges. Hence, this method is a simple and
safe solution for the anchorage of additional longitudinal pre-
stressing tendons. Tensile and shear forces will be developed in
the bolts due to external load.
To study the behavior of the end blocks during various stages
of transfer of prestress due to initial prestressing, existing pre-
stressing 共after the losses兲 during its service life and subsequently
due to proposed external prestressing for enhancing the service
life, an experimental investigation is planned on the end blocks.
This end block is subjected to the load, which will simulate the
equivalent forces that will be applied through anchorages. The
details of two PSC end blocks will be presented in subsequent
sections. Finite-element modeling and analysis have been carried
out by simulating the actual testing conditions. The analytical
responses such as out-of-plane deformation, slope of the steel
bracket, and slip have been compared and are observed to be in
good agreement with those measured.
Prestressed Concrete and Block
A posttensioned “I” beam of effective span 5 m 共16 ft, 5 in.兲 with
M45 concrete grade is designed. The depth of the beam is
650 mm 共25.59 in.兲 and flange width at top and bottom is
200 mm 共7.87 in.兲. Flange thicknesses at top and bottom are
90 mm 共3.54 in.兲 and thickness of web is 120 mm 共4.72 in.兲. The
end section of I-girder of length 1,020 mm 共40.16 in.兲 consisting
of a rectangular portion of 700 mm 共27.56 in.兲 and an I-shaped
segment of 320 mm 共12.60 in.兲 are chosen as the test specimens.
 BURSTING REINFORCEMENT

LIFTING HOOK
SPALLING REINFORCEMENT HANGER BARS 10mm Ø BAR

(a) 16 mm (b) 10 mm
SPIRAL REINFORCEMENT
650 Fig. 5. Torque-controlled expansion anchors of 共a兲 16 mm; 共b兲
10 mm diam
325

100
Expansion-Type Anchor Bolts
30 180 180 180 200 200 50
Two types of expansion-type anchors are used in the tests and are
1020
shown in Fig. 5. The capacities of anchor bolts specified by the
manufacturer are given in Table 1. These are used in the design of
Fig. 4. Reinforcement details of end block anchoring the external prestress of 188 kN accounting for 30%
loss of initial prestress.

Reinforcement Detailing Instrumentation

Schematic details of reinforcement of the test specimen are shown
in Fig. 4. The vertical bursting reinforcement consists of four
stirrups made of high strength deformed bars of 10 mm 共No.
3 bar兲 diam and is located at distances of 210 mm 共8.67 in.兲,
390 mm 共15.38 in.兲, 490 mm 共19.29 in.兲, and 570 mm 共22.44 in.兲
from the end, respectively. Two stirrups are provided in the rect-
angular beam portion. The horizontal bursting reinforcement con-
sists of 10 mm 共No. 3 bar兲 diam high strength deformed bars
located at top and bottom and connected to the vertical stirrups. In
order to resist high bearing stresses immediately behind the an-
chor plate, reinforcement in the form of spiral, 8 mm 共0.32 in.兲 in
diam, 25 mm 共0.98 in.兲 pitch, and 135 mm 共5.32 in.兲 outer diam
and extending over a length of 225 mm 共8.86 in.兲 共total of nine
turns兲 is provided concentrically in the end block. To cater the
spalling tensile stresses resulting at the surface of the jacking end
of the end block, a reinforcement mesh is formed with 10 mm
共No. 3 bar兲 diam bars and is located at 30 mm 共1.18 in.兲 from the
jacking end.
To measure out-of-plane displacement and vertical displacement
of the bracket, dial gauges D1 and D2 are positioned. Positions of
dial gauges are shown on the development surface of the end
block in Fig. 6. To measure vertical displacement 共slip兲 of the
bracket during external prestressing, dial gauges are positioned
parallel to the end block, one at the top of the steel bracket and
the other on the concrete at the same level. To measure out-of-
plane displacement 共to compute slope兲 of the bracket during ex-
ternal prestressing, dial gauges are positioned perpendicular to the
end block, one at the top of the steel bracket and the other at the
bottom of the steel bracket at each corner of the steel bracket.
Dial gauges have a least count of 0.01 mm. The locations of the
strain gauge are shown in Fig. 6. Positions of expansion-type
anchor bolts are identified with symbol “쎲.”
Table 1. Capacities of Bolts
Type of Characteristic Ultimate
Diameter of bolt force load 共kN兲 load 共kN兲

Experimental Program 16 mm Tension 50.4 71.3

共0.63 in.兲 Shear 101.1 128.6

Materials 10 mm Tension 29.5 37.7

共No. 3 bar兲 Shear 49.2 63.5
Concrete
Concrete of grade 45 MPa 共6.53 ksi兲 is used to cast end block 52mm DIA HOLE
specimens. The proportion of concrete ingredients obtained as per
design mix to achieve the target compressive strength of 45 MPa
is 1:1.83:2.74 and the water cement ratio is 0.4. The cement used D15 D3
D1 D5 D7 D6
D11 D9
D13 D2 D10
in the mix is 53-grade ordinary portland cement. River sand is D14
used as fine aggregate and angular shape crushed granite aggre-
gate of 10– 12.5 mm 共0.40– 0.49 in.兲 nominal size is used as
700
coarse aggregate.

Steel
D16 D4 D8 D12
High-yield strength deformed bars having yield strength of
415 MPa 共60.24 ksi兲 are used as reinforcement. The high tensile
rock-anchor rod of 36 mm 共No. 11 bar兲 diam having an average 320
tensile strength of 1,000 MPa 共145.16 ksi兲 is used for applying
the simulated internal prestressing force of 624 kN to the test
specimen. The expansion-type anchor bolts having design shear
strength of 80.9 kN and tensile strength of 33.6 kN are used for
transferring external prestressing force to the concrete member in
shear mode. Fig. 6. Instrumentation details

92 / PRACTICE PERIODICAL ON STRUCTURAL DESIGN AND CONSTRUCTION © ASCE / MAY 2009

Downloaded 05 May 2009 to 203.193.155.67. Redistribution subject to ASCE license or copyright; see http://pubs.asce.org/copyright

(a) Test specimen above the test floor
(b) Below the test floor
Fig. 7. Test setup: 共a兲 test specimen above the test floor; 共b兲 below
the test floor
As part of the experimental program, six cubes of size 100 mm
共3.94 in.兲 and six cylinders of size 100 mm 共3.94 in.兲 diam and
200 mm 共7.87 in.兲 height were cast along with the test specimen
in order to determine the compressive strength and split tensile
strength of concrete. The average compressive strength was ob-
tained as 36.92 MPa 共5.36 ksi兲 and 50.4 MPa 共7.32 ksi兲 after 7
and 28 days, respectively. Similarly, average split tensile strength
of cylinders after 7 and 28 days are 2.37 MPa 共0.34 ksi兲 and
4.18 MPa 共0.61 ksi兲, respectively.
Test Setup
Two rigid steel brackets are designed to transfer the external pre-
stressing force onto the test specimen. These brackets are con-
nected to the test specimen by expansion-type anchor bolts. Two
hydraulic actuators are fixed to the loading frame to apply exter-
nal force on the steel brackets. The test specimen itself is an-
chored to the test floor through a system of box beams and high
strength bolts. The expected end reaction from the beam due to
loading is applied to the test specimen through a horizontal jack.
The horizontal jack is fixed to a rigid pedestal anchored to the test
floor. The vertical load simulating the external prestress through
two vertical actuators is controlled from a hydraulic loading plant,
while the horizontal jack simulating the end reaction on the beam
is controlled through a separate electrically operated pump. The
test setup including the test floor is shown in Fig. 7.
Test Procedure
Internal full prestress force of 625 kN is applied vertically by
tensioning 36 mm 共No. 11 bar兲 diam anchor rod against the test
floor. The end reaction component of 125 kN corresponding to
PRACTICE PERIODICAL ON STRUCTURAL DESIGN AND CONSTRUCTION © ASCE / MAY 2009 / 93
Downloaded 05 May 2009 to 203.193.155.67. Redistribution subject to ASCE license or copyright; see http://pubs.asce.org/copyright
the design live load acting on the beam is applied horizontally
through 300 kN capacity jack. By assuming prestress loss of 30%
of initial prestress is induced in the test specimen by releasing
30% of initial prestress. A horizontal reaction is also reduced
correspondingly. For the test specimen EB-1, holes of 25 mm
共0.98 in.兲 diam and 125 mm 共4.92 in.兲 long are drilled on each
side of end block at four locations to insert expansion type bolts
of 16 mm 共0.63 in.兲 diam. Four bolts are arranged at an effective
horizontal spacing of 275 mm 共10.83 in.兲 and vertical spacing of
160 mm 共6.30 in.兲 on each side of the specimen. The effective
minimum edge distance provided for the bolts is 140 mm
共5.51 in.兲. The bolts were inserted through the steel brackets on
both sides and tightened by a torque meter to a specified value of
120 N m 共88.45 lb ft兲. The effective anchorage depth, designed
tensile, and shear loads of a single bolt are 120 mm 共4.72 in.兲,
33.6 kN, and 80.9 kN, respectively.
To understand the load transfer mechanism to the end block,
one more test specimen 共EB-4兲 was cast with small size bolt of
10 mm 共No. 3 bar兲 diam. The anchorage depth, designed tensile,
and shear loads of a single bolt are 90 mm 共3.54 in.兲, 19.70 kN,
and 39.4 kN, respectively. Eight bolts are arranged at an effective
horizontal spacing of 137.5 mm 共5.41 in.兲 and vertical spacing of
80 mm 共3.15 in.兲 on each side of the specimen. The effective
minimum edge distance provided for the bolts is 120 mm
共4.72 in.兲. The bolts are tightened by a torque meter to a specified
value of 50 N m 共36.86 lb ft兲.
To measure the vertical slip and horizontal displacement of the
steel bracket, dial gauges at the top and bottom of the steel
bracket are installed. External prestress is applied as a vertical
load on steel brackets in increments of 15 kN per actuator. For the
case of EB-1, reaction load is kept constant whereas reaction load
is varied proportional to the external load for the case of EB-4.
Reaction load is arrived at by equating the stresses due to external
load and reaction load to zero. The loads are applied until failure
of the end blocks. Corresponding to each stage of loading, the dial
gauge readings are recorded. During application of external load,
rotation as well as slipping of steel bracket against the concrete is
measured.
Discussion of Test Results
The first visible crack is observed on the top surface of the speci-
men, which extended throughout the width of the specimen for
EB-1. The corresponding load is 540 kN. The formation of a
crack indicates that the development of tension in the concrete
above the top row bolts is due to bending action caused by the
external loads. With further increase in load, the same crack
propagated vertically on either side of the vertical surfaces. Fur-
ther, inclined cracks are observed originating from the top row of
the bolts propagating away from the bolt towards the nearest
edges of the specimen. With a further increase in external load,
cracks started on both vertical faces of the specimen and widened
the already formed cracks. Pullout failure of the bolts along with
concrete surrounding bolts and concrete edge failure occurred si-
multaneously at ultimate load. The ultimate load for EB-1 is
795 kN. The failure pattern of the specimen 共EB-1兲 at ultimate
load is shown in Fig. 8
The design load for the group of 16 mm diam bolts is 245 kN.
The ratio of ultimate load and cracking load is 1.47, whereas the
ratio of ultimate load and design load is 3.25. The first visible
crack is observed on the surface of the specimen for EB-4 and the
corresponding load is 320 kN. The reason for early cracks in
EB-4 is attributed to the smaller diam of bolts, less anchorage
 (a) 16 mm dia bolt

(b) 10 mm dia bolt

Fig. 10. Deformed shape of bolts: 共a兲 16 mm diam bolt; 共b兲 10 mm
Fig. 8. Failure pattern of Specimen EB-1 at ultimate load diam bolt

Slips and Slopes

depth of bolts and concrete surrounding the bolts is subjected to Slips and slopes are an important observation as they indicate loss
more tension. The ultimate load for EB-4 is 570 kN. At ultimate of external prestress. The behavior of the EB-1 and EB-4 is simi-
load, concrete surrounding the bolts is pulled out suddenly due to lar up to design load. The average slip of EB-1 and EB-4 is
an increase in tension more than the concrete cone capacity. Fig. 0.3 mm 共0.012 in.兲 at their respective design loads. The average
slip of EB-1 at cracking and ultimate loads are 4.6 mm 共0.18 in.兲
9 shows the failure pattern of EB-4. The design load for the group
and 11.08 mm 共0.44 in.兲, respectively, while the corresponding
of 10 mm diam bolts is 252 kN. The ratio of ultimate load and
values of slip for EB-4 are 3.0 mm 共0.12 in.兲 and 9.87 mm
cracking load is 1.78, whereas the ratio of ultimate load and de-
共0.39 in.兲, respectively. It can be observed that the slip is smaller
sign load is 2.26. Deformation of the bolt in EB-1 and EB-4 at
by 30 and 50% at cracking load and ultimate load, respectively, in
ultimate load is shown in Fig. 10.
the case of EB-1 compared to that of EB-4. The reason may be
A ductile failure is observed in EB-1 whereas a sudden failure
due to more anchorage depth. Slope is calculated using the mea-
is observed in EB-4. Spalling and crushing of concrete are ob-
sured out-of-plane deformation. Slope of the steel bracket at ulti-
served more in EB-1 than in EB-4 at ultimate load. The test
mate loads for EB-1 and EB-4 are 0.74 and 1.0, respectively. It
results proved that the proposed anchoring system using
can be observed that the slope is smaller by 50 and 36% at crack-
expansion-type anchor bolts with 16 mm diam bolts with internal
ing load and ultimate load, respectively, in the case of EB-1 com-
prestressing and 10 mm diam bolts, transferred the 30% prestress
pared to that of EB-4. It can be observed that the slope of EB-4 is
loss 共187 kN兲 in shear mode as an external load to end block
less than the value of that of EB-1 at a given load. Slip and slope
specimens successfully. At ultimate load, EB-1 performed better
of the steel bracket under combination of loads for test specimens
than EB-4.
EB-1 and EB-4 are shown in Figs. 11 and 12, respectively.

Finite-Element Modeling and Analysis

Finite-element analysis 共FEA兲 is a numerical technique, widely

employed to the structures/structural components. FEA is a tool
that can simulate and predict the responses of reinforced and pre-
stressed concrete members. Many researchers attempted to model
and analyze the end block including anchor bolts using the gen-
eral purpose multiphysics finite element software ANSYS/
ABAQUS 共Bahaari and Sherbourne 2000; Fanning et al. 2000;
Jung et al. 2006; Kim et al. 2007兲. Finite-element modeling and
analysis is carried out by simulating the testing conditions of
EB-1 and EB-4. The popular finite element software ANSYS has
been used for modeling and analysis of the end block with anchor
bolts 共ANSYS 2002兲. Details of analysis pertaining to EB-1 are
provided herein.

Finite-Element Modeling Issues

A large number of different FE formulations have been used for
Fig. 9. Failure pattern of Specimen EB-4 the analysis of concrete structural components. These may be

94 / PRACTICE PERIODICAL ON STRUCTURAL DESIGN AND CONSTRUCTION © ASCE / MAY 2009

Downloaded 05 May 2009 to 203.193.155.67. Redistribution subject to ASCE license or copyright; see http://pubs.asce.org/copyright
 900 FE Modeling and Analysis of EB-1
800 Volume has been created as per the geometry described in the
Experimental Program and the Test Setup sections. Eight-noded
700 solid elements 共SOLID 45兲 have been employed to model con-
crete block, bolts, and bracket connection. The use of these ele-
600
ments provides the same number of integration point density as
the higher order elements but requires much less computational
Load, kN

500
effort. This element is defined by orthotropic material properties
400 that correspond to the element coordinate directions. Each node
has three translational degrees of freedom, namely, Ux, Uy, and
300 Uz in x, y and z directions, respectively. The element has plastic-
EB-1 ity, creep, swelling, stress stiffening, large deflection, and large
200
EB-4 strain capabilities. A bolt is idealized as square in shape instead of
100 circular to avoid modeling issues. The bolt is embedded for a
length of 125 mm 共4.92 in.兲 inside the concrete block. At the end
0 of the bolt, a link element has been created connecting the sur-
0.00 2.00 4.00 6.00 8.00 10.00 12.00 faces of bolt and concrete. The principal function of the link
Slip, mm element is to transfer the forces arising due to external prestress-
ing loads. An imaginary surface has been created between the
Fig. 11. Slip of the steel bracket concrete block and the bracket in order to allow the bolt move-
ment in the out-of-plane direction under the external prestressing
load.

categorized into facet plate/shell elements, thin-shell elements Material Properties

共Kirchoff assumptions兲, thin/thick shell elements 共Reissner- The following are the material properties considered for the con-
Mindlin theory兲, and three-dimensional elements. The choice of crete, steel, and bolt. For linear static analysis, modulus of elas-
an element for analysis of a structure/component depends on the ticity, Poisson’s ratio are the input values whereas for nonlinear
geometry and the purpose for which the results of the analysis are static analysis, multilinear elastic model available in ANSYS has
to be used. The following are some of the key issues w.r.t mod- been used. The material behavior is described by a piecewise
eling of PSC end block: linear stress-strain curve, starting at the origin, with positive stress
• FE modeling of posttensioned prestressed concrete end block; and strain values. Successive slopes can be greater than the pre-
• FE modeling of steel bracket and expansive-type anchor bolts; ceding slope; however, no slope can be greater than the elastic
• Material modeling of concrete, steel, and bolt; modulus of the material. The slope of the first curve segment
• Modeling of various loads such as compression load, reaction usually corresponds to the elastic modulus of the material, al-
load on end block, pretension in the bolt, and external load on though the elastic modulus can be input as greater than the first
steel bracket; slope to ensure that all slopes are less than or equal to the elastic
• Simulation of initial pretension and equivalent compression in modulus. Fig. 13 shows multilinear stress strain plot adopted for
concrete surrounding the bolt; steel and concrete material.
• Modeling end connections such as bolt and steel bracket and
Modulus of
bolt and concrete block; Material elasticity Poisson’s ratio, ␯
• Employing appropriate element types such as contact, gap,
solid, and link for effective load transfer and to simulate the 31,623 MPa
Concrete 0.12
realistic behavior; and 共4,590.36 ksi兲
• Simulation of material nonlinearity for larger load steps and 2 ⫻ 105 MPa
Steel 0.3
cracked concrete behavior. 共29,031.79 ksi兲
2 ⫻ 105 MPa
Bolt 0.3
共29,031.79 ksi兲

An interactive Newton-Raphson procedure has been employed

900 for nonlinear static analysis where the load is divided into a series
800 of load increments applied in several load steps. Before each
700 solution step, an out-of-balance load vector, which is the
600 difference between the restoring forces corresponding to element
Load, kN

500 stresses and the applied load is evaluated. A linear solution is then
400 carried out using out-of-balance loads and convergence is
300
EB-1
checked. When the convergence criterion is not satisfied, the
200 out-of-balance load is reevaluated, the stiffness matrix is updated,
EB-4
100 and a new solution is obtained. This iterative procedure continued
0 until the solution converges for given tolerance limits.
0.00 0.20 0.40 0.60 0.80 1.00 1.20
Slope, degrees Loading Conditions
As already stated, there are various loadings, namely, shear load
Fig. 12. Slope of the steel bracket in the form of reaction, pretension in the bolt, initial prestress, and

PRACTICE PERIODICAL ON STRUCTURAL DESIGN AND CONSTRUCTION © ASCE / MAY 2009 / 95

Downloaded 05 May 2009 to 203.193.155.67. Redistribution subject to ASCE license or copyright; see http://pubs.asce.org/copyright
 Fig. 13. Multilinear stress-strain plot: 共a兲 steel; 共b兲 concrete

external prestress. Initial prestress is applied as compressive load

on the top of the PSC block. The initial pretension force obtained
by applying torque on the bolt is applied to all the bolts on the
face of the element and an equivalent compressive force is ap-
plied to all the surrounding elements of the bolt to simulate the
head of the bolt. It is assumed that there is a perfect frictional
bond between bolt and concrete. Shear load, prestress load on the
block, and initial pretension to the bolt are maintained the same
throughout the analysis. External prestressing load is applied in-
crementally on the projection part of the bracket. The magnitudes Total Solid elements = 11362
Solid concrete elements = 8800
of the loads are given below Solid steel elements = 2562
Total no. of nodes = 15651
No. of equations = 46152
Shear load 共to PSC block兲: 231 kN

Prestress load 共on concrete block兲: 438 kN

Initial pretension 共to each bolt兲: 37.5 kN

Fig. 14. FE mesh and modeling characteristics
External prestress load 共on bracket兲: 0 to 400 kN

Boundary Conditions
All the degrees of freedom have been constrained on the bottom
surface of the block simulating the fixity condition of the test
floor. FE mesh along with modeling characteristics are shown in
Fig. 14.
Linear static and nonlinear static analysis has been carried out
depending on the magnitude of the external prestressing load. At
each incremental load, out-of-plane deformation and slip have
been noted down. Rotation of the bracket has been calculated by
using the computed out-of-plane deformation. Figs. 15 and 16
show the out-of-plane deformation and vertical deformation
共slip兲, respectively, up to an ultimate load of 400 kN. The com-
puted slope and slip values of the steel bracket have been com-
pared with the corresponding experimental observations. Figs. 17
and 18 show the slope and slip plots comparing computed and
experimental values. From Fig. 17, it can be observed that the
computed slopes and the corresponding experimental observa-
tions are in very good agreement with each other. From Fig. 18, it Fig. 15. Out-of-plane deformation at ultimate load

96 / PRACTICE PERIODICAL ON STRUCTURAL DESIGN AND CONSTRUCTION © ASCE / MAY 2009

Downloaded 05 May 2009 to 203.193.155.67. Redistribution subject to ASCE license or copyright; see http://pubs.asce.org/copyright
 450

400

350

300

External load, kN
250

200
Present study
150 Experimental

100

50

0
0 2 4 6 8 10 12
Slip, mm

Fig. 18. Slip of steel bracket under external prestress load for EB-1
Fig. 16. Vertical deformation 共slip兲 at ultimate load

ultimate load, EB-1 performed better than EB-4. Hence, it is re-

can be observed that there is reasonable agreement between the
computed and experimental slip values. The difference in the slip
may be attributed to the noncoincidence of nodal location and
experimentally measured point.
Summary and Concluding Remarks
An innovative external prestressing technique for strengthening of
prestressed concrete 共PSC兲 girders has been described. This tech-
nique is very much useful when the ends of the PSC girders are
inaccessible. The developed method relies on anchoring the ex-
ternal prestressing to the sides of the end block. The transfer of an
external force is in shear mode and the required transversal pre-
stressing force is smaller compared to conventional techniques. In
order to validate the proposed technique, an experimental inves-
tigation has been carried out on two posttensioned end blocks
共EB-1 and EB-4兲. The end blocks are embedded with expansion-
type anchor bolts as shear keys and steel brackets on either side
for transferring the external prestressing force. Performance of the
end blocks have been studied for design, cracking, and ultimate
loads. A ductile failure has been observed in EB-1 whereas a
sudden failure is observed in EB-4. Slip and slope of steel bracket
have been recorded at various stages during the experiment. At
450
400
350
External Load, kN
alized that longer size bolts for the system of anchoring external
prestressing force leads to better performance.
Finite-element modeling and analysis have been carried out by
simulating the geometry, loading, and material nonlinearity. Lin-
ear and nonlinear static analysis have been conducted for the
specified loadings and the responses have been compared. From
the analysis, it has been observed that the computed slope of the
steel bracket is in good agreement with the corresponding experi-
mental observations whereas slip is in reasonable agreement with
the corresponding experimental values.
Acknowledgments
The writers thank Dr. Nagesh R. Iyer, Director-Grade-Scientist
and Advisor共M兲, SERC for his valuable suggestions during the
course of the investigation. The writers also thank the staff mem-
bers of the Heavy Testing Laboratory for assistance in conducting
the experiments. This paper is published with the kind permission
of the Director, SERC, Chennai.
References
ANSYS. 共2002兲. ANSYS 6.0, theory and reference manual, ANSYS, Can-
onsburg, Pa.
Aparicio, A. C., Ramos, G., and Casas, J. R. 共2000兲. “Externally pre-
stressed high strength concrete viaduct.” J. Bridge Eng., 5共4兲, 337–
343.
Aparicio, A. C., Ramos, G., and Casas, J. R. 共2002兲. “Testing of exter-

300 nally prestressed concrete beams.” Eng. Struct., 24共1兲, 73–84.

Bahaari, M. R., and Sherbourne, A. N. 共2000兲. “Behaviour of eight-bolt
250
large capacity end plate connections.” Comput. Struct., 77共3兲, 315–
200 325.
150 Present study
Chen, S. 共2005兲. “Experimental study of prestressed steel-concrete com-
posite beams with external tendons for negative moments.” J. Constr.
Experimental
100 Steel Res., 61共12兲, 1613–1630.
50 Choy, S. C., Wong, Y. L., and Chan, S. L. 共2002兲. “Shear strength of
prestressed concrete encased steel beams with bonded tendons.”
0
Proc., Third Int. Conf. on Advances in Steel Structures, Hong Kong,
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
China, 543–549.
Slope, Deg.
Czaderski, C., and Motavalli, M. 共2007兲. “40-year-old full-scale concrete
bridge girder strengthened with prestressed CFRP plates anchored
Fig. 17. Slope of steel bracket under external prestress load for EB-1 using gradient method.” Composites, Part B, 38共7–8兲, 878–886.

PRACTICE PERIODICAL ON STRUCTURAL DESIGN AND CONSTRUCTION © ASCE / MAY 2009 / 97

Downloaded 05 May 2009 to 203.193.155.67. Redistribution subject to ASCE license or copyright; see http://pubs.asce.org/copyright
Fanning, P. J., Tucker, M., and Broderick, B. M. 共2000兲. “Non-linear
finite element analysis of semi-rigid bolted end-plate connections.”
Int. Conf. Comp. Struct. Tech., Leuven, Belgique, 397–403.
Ghallab, A., and Beeby, A. W. 共2005兲. “Factors affecting the external
prestressing stress in externally strengthened prestressed concrete
beams.” Cem. Concr. Compos., 27共9–10兲, 945–957.
Jung, J., Abolmaali, A., and Choi, Y. 共2006兲. “Finite-element analysis of
tapered steel and fiber-reinforced plastic bridge camera poles.” J.
Bridge Eng., 11共5兲, 611–617.
Kim, J., Yoon, J. C., and Kang, B.-S. 共2007兲. “Finite element analysis and
modeling of structure with bolted joints.” Appl. Math. Model., 31,
895–911.
Lou, T. J., and Xiang, Y. Q. 共2006兲. “Finite element modeling of concrete
98 / PRACTICE PERIODICAL ON STRUCTURAL DESIGN AND CONSTRUCTION © ASCE / MAY 2009
View publication stats Downloaded 05 May 2009 to 203.193.155.67. Redistribution subject to ASCE license or copyright; see http://pubs.asce.org/copyright
beams prestressed with external tendons.” Eng. Struct., 28共14兲, 1919–
1926.
Ma, Z., Saleh, M. A., and Tadros, M. K. 共1999兲. “Optimized post-
tensioning anchorage in prestressed concrete I-beams.” PCI J., 44共2兲,
56–73.
Miyamoto, A., Tei, K., Nakamura, H., and Bull, J. W. 共2000兲. “Behavior
of prestressed beam strengthened with external tendons.” J. Struct.
Eng., 126共9兲, 1033–1044.
Stoll, F., Saliba, J. E., and Casper, L. E. 共2000兲. “Experimental study of
CFRP-prestressed high strength concrete bridge beams.” Compos.
Struct., 49共2兲, 191–200.
Youakim, S. A., and Karbhari, V. M. 共2007兲. “An approach to determine
long-term behavior of concrete members prestressed with FRP ten-
dons.” Constr. Build. Mater., 21共5兲, 1052–1060.