Seismic Performance of PrecastSegmental Bridges: Segment-to-Segment Joints Subjected to HighFlexural Moments and Low Shears
Sami Megally, Ph.D.Assistant Project Scientist
Department of Structural EngineeringUniversity of California at San Diego
La Jolla, California
Frieder Seible, Ph.D., P.E.Dean of Jacobs School of EngineeringandProfessor of Structural EngineeringUniversity of California at San DiegoLa Jolla, California
Robert K. Dowell, Ph.D., RE.Principal
Dowell-Holombo Engineering, Inc.San Diego, California
andAssistant Adjunct Professor
Department of Structural EngineeringUniversity of California at San Diego
La Jolla, California
A three-phase research project is currently inprogress to investigate the seismic performance ofprecast segmental concrete bridges. This paperpresents experimental and analytical results of thefirst phase, which focused on segment-to-segmentjoints subjected to high flexural moments and lowshears. Four large-scale units were tested underfully reversed cyclic displacements. The major testvariable was the ratio of internal to external post-tensioning of the superstructure. This paper alsopresents results of nonlinear finite elementmodeling of the test units. All test units couldundergo significant seismic displacements and jointopenings before failure. It was found that the use ofexternal tendons improves ductility anddisplacement capacity and minimizes permanentdisplacements and permanent openings ofsegment-to-segment joints. It was also found thatinternally bonded and external (unbonded) tendonsshould not be combined in high seismic zones.
precast segmental bridge construction has many advantages over the conventional cast-in-place construction method. However, the popularity of precast
segmental bridges in high seismic zones is hampered because of severe restrictions on their construction.
The AASHTO Guide Specifications for Design and Construction of Segmental Concrete Bridges,1 referred to as theAASHTO Guide Specifications throughout this paper, per-
80 PCI JOURNAL
450’
Fig. 1. Prototype structure: (a) Elevation; and (b) Cross section.
mits the use of precast segmental construction in high seismic zones (ZonesC and D) provided that the precastsegments are epoxy bonded. The sameAASHTO Guide Specifications’ alsorequires that external (unbonded) tendons shall provide not more than 50percent of the total post-tensioning.
This requirement was not based onany experimental or analytical evidence but was adopted in the secondedition of the AASHTO Guide Specifications’ to match the current policyof the California Department of Transportation (Caltrans). In addition, precast segmental bridge constructionwithout mild steel reinforcementcrossing the segment-to-segmentjoints is not recommended in currentpractice in high seismic zones such asCalifornia.
The above-mentioned recommendations and restrictions of current practice are justified due to lack of experimental research data regarding seismicperformance of precast segmentalbridges. Thus, a comprehensive research project has been developed byASBI (American Segmental Bridge
Institute), Caltrans and the Universityof California at San Diego (UCSD) toinvestigate the seismic performance ofprecast segmental bridges. The large-scale experimental research project iscurrently in progress at UCSD. Thisresearch project is funded by Caltransand it consists of the following threephases:
Phase I: To investigate the seismic performance of segment-to-segment joints in superstructures with different ratios of internal to externalpost-tensioning. Only superstructurejoints close to midspan in regions withhigh positive flexural moments andlow shears were considered in thisphase. Phase I consists of the following two parts:
(1) Phase I-A: Superstructures with100 percent internal post-tensioning(Test Units 100INT and 100INTCIP;see Table 1).
(2) Phase I-B: Superstructures withdifferent ratios of internal to externalpost-tensioning (Test Units 100INT,100EXT and 5OINTJ5OEXT). TestUnits 100EXT and 5OINT/5OEXTwere identical to Test Unit 100INT,
with the only difference being the percentage of external post-tensioning.Test Unit 100EXT, with 100 percentexternal post-tensioning, does not satisfy the current requirement of theAASHTO Guide Specifications’ thatnot more than 50 percent of post-tensioning should be achieved by externaltendons.
• Phase II: To investigate the seismic performance of superstructuresegment-to-segment joints close to thesupports in regions with high shearsand negative flexural moments.
• Phase III: To investigate the performance of superstructure-columnsystems under gravity loads combinedwith seismic forces. Seismic performance of cast-in-place and precastsegmental hollow rectangular columnswill also be investigated in Phase III.
The first two phases of the projecthave been completed, whereas thethird phase is currently in progress.Results of Phase I-A test units werepresented in an earlier paper,2 whileexperimental and analytical results ofPhase I-B are presented in this paper.The experimental and finite element
(b) Cross section
March-April 2003 81
Fig. 2. Joint andsegment numbering
of test units.
results of Phase II are presented in acompanion paper.3
The major objectives of the researchprogram were to investigate: (1) jointbehavior in terms of opening and closure under repeated cyclic loads, (2)development of crack patterns and (3)modes of failure.
As part of the research project,three-dimensional finite element models of the test units were developed.The models allowed for concretecracking and crushing, opening andclosure of segment-to-segment jointsand inelastic characteristics of prestressed and nonprestressed steels. Thefinite element models were validatedwith the experimental results presented in this paper.
The main objective of the analyticalmodels described in this paper is tocapture important characteristics ofjoint behavior. These finite elementresults will be used in Phase III of theresearch project to aid development ofseveral comprehensive global finite element bridge models to be used foranalytical parametric studies.
PROTOTYPE STRUCTURETest units of Phase I of this experi
mental program are based on the prototype structure shown in Fig. 1. Thesuperstructure consists of three 100 ft(30.48 m) interior spans and 75 ft(22.86 m) exterior spans with a totallength of 450 ft (137.2 m). The superstructure is post-tensioned withharped-shape tendons (see Fig. la).
Because of its short spans, the prototype superstructure is built using thespan-by-span construction method.Fig. lb shows the cross section of theprototype superstructure.
The prototype structure was designed according to the AASHTOGuide Specifications,1the AASHTOPCI-ASBI Segmental Box GirderStandards for Span-by-Span and Balanced Cantilever Construction,4andthe AASHTO Standard Specificationsfor Highway Bridges.5 More detailsabout the design of the prototypestructure are given elsewhere.6
EXPERIMENTAL PROGRAMThis section gives a description of
the Phase I test units and method ofconstruction as well as the test setupand loading sequence. While construction of the test units and setup havebeen described in an earlier paper,2theyare repeated here for completeness.
Description of Test Units
The critical location of the prototypestructure for positive flexural momentunder dead load and seismic forceswas found to be approximately atmidspan.6This is because onset ofjoint opening under the combined effects of dead load, longitudinal andvertical seismic forces was determinedto occur near the midspan.6 The testunits model the middle third of an interior prototype span in which the tendons are horizontal (see Fig. la). The
test units were designed at two-thirdsscale of the prototype structure.
Four units were constructed andtested. Fig. 2 shows a typical test unit,simply supported at its ends. Each testunit consisted of six precast segments,which were epoxy bonded at their interfaces. The test zone consisted offour 6 ft(1.83 m) long by 4 ft (1.22 m)deep precast segments (Segments 2 to5 in Fig. 2). Each test unit was supported at its ends by precast end segments (Segments 1 and 6 in Fig. 2).
Fig. 3 shows the cross section andreinforcement of precast Segments 2through 5 (see Fig. 2) in the Phase I-Btest units. Half of the prototype boxgirder superstructure section was modeled and idealized in the shape of anequivalent I-section to simplify thetest setup.
The test matrix is given in Table 1. Atotal of four units were built and tested;the test units were grouped into twophases, or two test series. Test Series I-A (Phase I-A) consisted of two testunits (Units 100INT and 100INTCIP,Table 1). The test variable in Phase I-Awas the presence of mild steel reinforcement crossing the segment-to-segment joints. Experimental and finite element analysis results of the Phase I-Atest units were presented in an earlierpaper2 as mentioned previously.
This paper is concerned with TestSeries I-B (Phase I-B), which consisted of two test units (Units 100EXTand 5OINT/5OEXT) in addition to TestUnit 100INT of Phase I-A; the latterwas the control specimen of the Phase
End Endsegment Test Zone 5egm
End supportframe (fixedbottom)
End supportframe (pinnedbottom)
82 PCI JOURNAL
I test series. The test variable in PhaseI-B was the ratio of internal to externalpost-tensioning (see Table I and Fig.3). The internally bonded tendon provided 100 percent of the post-tensioning in Test Unit 100INT.
Unit 1 OOEXT was identical to Unit1 OOINT except that two external (unbonded) tendons, placed symmetrically about the centerline of the web,provided 100 percent of the post-tensioning (see Fig. 3). Thus, Unit1 OOEXT did not satisfy the AASHTOGuide Specifications requirement thatnot more than 50 percent of post-tensioning should be provided by externaltendons in high seismic zones.’
An internally bonded tendon provided 50 percent of the post-tensioning in Unit 5OINTI5OEXT, with the remaining 50 percent provided by twoexternal (unbonded) tendons (see Fig.3). Thus, Unit 5OINT/5OEXT satisfiedthe above-mentioned AASHTO GuideSpecifications’ upper limit for the allowable percentage of external post-tensioning.’
Sikadur 31, SBA (Segmental BridgeAdhesive) slow-set epoxy was usedfor bonding of the precast segments ofeach test unit. For Units 100INT,100EXT and 5OINT/5OEXT, theepoxy was applied to the entire crosssection of the segment-to-segmentjoints. Specifications of the epoxy aregiven in Table 2, in which 0comp afldabOfld are the compressive and bondstrengths, respectively.
Each test unit was post-tensionedwith sixteen 0.6 in. (15.2 mm) diameter strands with an ultimate tensilestrength of 270 ksi ( 1860 MPa). Theprestressing force was equal for alltest units and was calculated so thatthe concrete stresses resulting frompost-tensioning were the same as forthe prototype structure.
Table 2 gives the measured concretematerial properties of the Phase I-Btest units. In Table 2,f is the concretecompressive strength on the day oftesting.
Construction of Test Units
As mentioned above, each test unitconsisted of six precast segments.Segments 1, 3 and 5 (see Fig. 2) werecast at the same time. This was fol
lowed by construction of Segments 2,4 and 6, which were match-castagainst Segments 1, 3 and 5.
The segments of each test unit wereassembled on a wooden platform atthe UCSD Structures Laboratory. Theepoxy was applied to the joint surfaces. After application of the epoxyand placement of each segment in itsfinal position, the test unit was temporarily post-tensioned by means ofhigh-strength ASTM A 722 prestressing steel bars.2 The temporary pre
stressing forces in the high-strengthbars were determined so that the entiresegment-to-segment joint surfaceswould have a minimum compressivestress of 40 psi (0.28 MPa).7
After epoxy bonding of the precastsegments, each test unit was post-tensioned with a jacking force of 720 kips(3203 kN). The effective prestressingforce at time of testing was estimated,by hand calculations and computertime-step analyses, to be about 600kips (2669 kN). The difference be-
100% Internal post-tensioning (Unit 100INT)
Detail A4’ - 8”
Section View
Fig. 3. Cross section of test units (Phase I-B).
Table 1. Test matrix.
Phase Test unit Test unit Test unit(Test series) number designation description
1 100INT 100% Internal post-tensioninglA*
2 100INTCIP100% Internal post-tensioning and
cast-in-place deck closure joints
1 100tNT 100% Internal post-tensioningI-Bt 3 100EXT 100%_External post-tensioning
4 5OINT/5OEXT 50% Internal + 50% External post-tensioning
Results of the Phase I-A test units were presented in Reference 2.t Phase I-B consists of Test Units IOOBXT and 5OINT/5OEXT in addition to Test Unit 100INT of Phase I-A.
Table 2. Material properties.
f’ (on day of testing) Epoxy properties*
-
-- ksi (MPa) —— (Sikadur 31, SBA) ksi (MPa)Segment Segment Segment
Test unit No. 1 & 5 No. 3 No. 2, 4 & 6 (3100TNT 5.11 (35.2) 7.21 (49.7) 6.96(48.0) 2.0 (13.8) 1.0 (6.9)10OEXT 5.92 (40.8) 5.92(40.8) 6.79 (46.8) 2.0 (13.8)
5OINT/5OEXT 8.01 (55.2) 8.01 (55.2) 6.17 (42.6) 2.0 (13.8) 1.0 (6.9)* Epoxy properties are those specified by the manufacturer (slow-set epoxy).
March-April 2003 83
Fig. 4. Test setup:(a) Elevation; and
(b) Three-dimensional view.
tween jacking and effective prestressing forces was due to losses from anchor set, elastic shortening, creep andshrinkage of the concrete as well as relaxation of the prestressing steel.
The temporary prestressing force inthe bottom slab was released after permanent post-tensioning of the testunit, whereas the temporary prestress
ing force in the top slab was releasedonly after vertical loading of the testunit was applied to simulate prototypedead load stresses. The stressing andloading sequence was designed toavoid cracking of the units before thetest. After permanent post-tensioning,the wooden platform, supporting thesegments during assembly, was re
moved and the test unit was mountedon the two end supports.
Test SetupFig. 4a shows a schematic elevation
view of the test unit and the loadframe. Each test unit was simply supported by a steel pin and steel links at
End supportframe (fixedbottom)
(2) Endsupport blocks5’- 0’ long
(a) Elevation
Hingedsupport
Rollersupport
(b) Three-dimensional view
84 PCI JOURNAL
its ends. At one end, the steel linkswere fixed at their bottom ends to restrain horizontal movement of the testunits. At the other end, the steel linkswere pinned at their bottom ends(rocker links) to allow rotation of theframe legs and horizontal expansionand contraction movements of the testunits. The loads were transferred fromthe test units to the steel links bymeans of steel pins set inside horizontal steel pipes that were cast into theend segments at the neutral axis of thetest units, allowing the ends of the testunits to rotate freely.
Four vertical servo-controlled hydraulic actuators were used to applyexternal loads to each test unit to simulate the effect of highway loadingand vertical seismic displacements onthe superstructure. Fig. 4b shows apictorial drawing of the test setup. Asat the midspan joint of the prototypespan, the midspan joint of each testunit was subjected to zero shearingforce and the highest flexural moment.
At the beginning of the test, eachunit was loaded in the downward direction to a prescribed level so that thestresses at the midspan joint wereequal to the prototype structurestresses under combined dead load,superimposed dead load, as well asprestressing primary and secondary effects. This load level is referred to asthe reference load level throughoutthis paper. Each test was conducted asfollows:
Stage I (Service Load Conditioning) — Only the two interior actuatorswere used in load control during thistest stage. Each test unit was loaded tothe reference load level at P 74.5kips (331 kN), where P is the load pereach actuator. The temporary prestressing force in the top slab was released at this stage.
Cycling the load P between 112 and65 kips (498 and 289 kN) 100,000times followed this. The upper andlower load limits provided the samemidspan stresses as the prototypestructure under maximum and minimum service loads.2 This testing stagewas performed to investigate the effects of service loading on seismicperformance of the superstructure.
Stage II (Seismic Test) — All fouractuators were used in displacement
control with actuator forces maintained equal throughout this test stage.Each test unit was loaded to the reference load level of P = 40.5 kips (180kN) per actuator, or a total load of 162kips (721 kN). Each test unit was thensubjected to fully reversed cyclic vertical displacements at midspan withincreasing amplitude to failure.
For each target displacement up to 4in. (102 mm) magnitude, three cycleswere completed. Beyond 4 in. (102mm) displacements, only one cyclewas performed at each displacementlevel. The displacement history duringthe seismic test (Test Stage II) isshown in Fig. 5.
Electrical resistance gauges wereused to measure strains in the concreteand prestressing steel. Vertical displacements along the span, joint openings at various locations, vertical sliding between the precast segments ateach joint and support displacementswere measured by means of linear potentiometers.
EXPERIMENTAL RESULTSThe major experimental results are
presented in this section. These experimental results include crack patterns,modes of failure, load-displacementresponse, performance of joints,strains in prestressing steel, stresses inexternal tendons and flexural momentcapacity.
Crack Patterns andModes of Failure
All test units were subjected to verylow tensile stresses at the midspanjoint during Test Stage I (ServiceLoad Conditioning).2Thus, no jointopenings were observed in any of thetest units during Test Stage I; becauseof this linear-elastic behavior of thetest units, only the results of the seismic tests (Test Stage II) will be discussed here.
Test Unit 100INT — The firstcrack occurred under downward loading at the midspan joint (Joint J3, Fig.2) during the first loading cycle to0.25 in. (6.35 mm) displacement (seeFig. 5). Opening of Joint J4 (see Fig.2) also occurred during the same displacement cycle. The midspan totaldisplacement at onset of cracking ofJoints 13 and J4 was 0.30 in. (7.62mm). The total displacement is thesum of the seismic displacement (seeFig. 5) and the displacement at the reference load level ( 0.11 in. = 2.79mm).
Joint J2 (see Fig. 2) opened underdownward loading during the 0.5 in.(12.7 mm) displacement cycle. Themidspan total displacement at onset ofJoint J2 opening was 0.53 in. (13.5mm). Thus, the three interior Joints J2,J3 and 14 (see Fig. 2) opened duringdownward loading.
A few additional flexural cracks oc
87
65
—‘4
j-l-2-3
-5-6-7-8
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28Nuner ofloadmg cycle
Fig. 5. Loading protocol of the seismic test (Test Stage II).
March-April 2003 85
Fig. 6. Failure at the midspan joint: (a) Test Unit 1 OOINT; (b) Test Unit 100EXT; and(C) Test Unit 5OINT/5OEXT.
curred inside Segments 3 and 4 (seeFig. 2) during subsequent downwardloading cycles. However, the widthsof these cracks were very small andinelastic deformations of the test unitwere concentrated mainly at midspanJoint 33.
Midspan Joint J3 was the only jointthat opened during upward loading ofUnit 100INT, and joint opening occurred during the 0.5 in. (12.7 mm)displacement cycle. Because Unit1 OOINT did not have any mild steel reinforcement crossing the joints, theopening of midspan Joint J3 increasedsignificantly under upward loadingduring subsequent displacement cycles.2
Shear cracks occurred in the webunder downward loading between thesupported ends of the test unit and theload application points (zero shearingforce at midspan). Shear crackscrossed the epoxy-bonded joints withno vertical sliding between adjacentprecast segments.
Under downward loading, themidspan Joint J3 opened significantlywith increased applied displacementuntil rupture of the prestressingstrands at the midspan joint at a displacement of about 4.8 in. (122 mm)and a total load of 490 kips (2180 kN).Fig. 6a shows the midspan joint afterfailure of Test Unit 100INT, as well asa close-up view of the rupturedstrands.
Test Unit 100EXT — The firstcrack occurred under downward loading at the midspan joint (Joint J3, Fig.2) during the first cycle to 0.25 in.(6.35 mm) displacement (see Fig. 5).The total midspan displacement atonset of Joint J3 cracking was about0.32 in. (8.13 mm). During subsequentloading cycles, widening of Joint J3was observed with no openings in anyof the other joints.
Midspan Joint J3 was also the onlyjoint that opened during upward loading of Unit I OOEXT. Joint 33 openedunder upward loading during the 0.5in. (12.7 mm) displacement cycle; adrop in the overall load occurred afterthis. During the 2.0 in. (50.8 mm) upward displacement cycle, the bottomslab came into contact with the external tendons; the load carrying capacityincreased during subsequent upward
CAL10099
PNECAST E14ALIBRVGAS
(JOlT ‘
00101410 1ICTAL 0000 490 NIPS
a0ER 0.0 (9
CIC00
909011000 1 2000
PI,LC1 IlL
00151.9 TOOl
IOTA L 7 ‘2 LIZ 11 ‘2GDER7”
I 2001
21117911 ((‘12 4 0 II(2211
‘
-
— Al-MI!. 1/2(1(1- -
86 PCI JOURNAL
Table 3. Summary of test results.
* Values given for downward loading direction.t 4 and were determined for the 3 in. (76.2 mm) displacement cycle.
loading cycles as a result of this contact. There were no adverse effects ofthe bottom slab bearing on the external tendons.
Shear cracks occurred in the webunder downward loading in Segments2 and 5 only (see Fig. 2); there weresignificantly fewer shear cracks thanobserved for all other test units, whichhad internally bonded tendons. Noshear cracks were observed insideSegments 3 and 4 (see Fig. 2).
Onset of concrete deck crushing occurred during the 4 in. (102 mm)downward displacement cycle. A peakdownward load of 417 kips (1855 kN)was attained at a midspan displacement of 3.53 in. (89.7 mm). Unlike theexplosive failure resulting from rupture of strands in Unit 100INT, theload carrying capacity of Unit100EXT dropped gradually with eachdisplacement cycle beyond the peakload.
The test was terminated at amidspan downward displacement ofabout 6.6 in. (168 mm), when the displacement capacity of the hydraulicactuators was reached. Fig. 6b showsthe midspan joint of Unit 100EXT atthe maximum displacement levelreached during the test.
Test Unit 5OINT/5OEXT — Performance of Unit 5OINT/5OEXT wassimilar to that of Unit IOOINT. Thefirst crack was observed inside Segment 3 at about 15 in. (381 mm) awayfrom the midspan Joint J3 (see Fig. 2);this crack occurred during the firstdownward displacement cycle to 0.25in. (6.35 mm) amplitude.
Opening of the midspan Joint J3 occuffed under downward loading during the first cycle to 0.5 in. (12.7 mm)displacement. The total midspan displacement at onset of Joint J3 crackingwas about 0.49 in. (12.4 mm). Opening of Joints J2 and J4 (see Fig. 2) oc
ç M7 Mci*Test unit kips(kN) in. (mm) in.(mm) 4J’j (percent)
.kip-ft (kN-m) kip-ft (kN-rn)MJMcak
IOOINT 490 (2180) 4.8 (l22)f 1.17 (29.7) 8.36 4.21 1.60 3126 (4238) 2993 (4058) 1.04100INTCIP 480 (2135) 5.9 (150) 0.53 (13.5) 3.79 8.75 3.33 3062 (4151) 2974 (4032) 1.03
IOOEXT 417 (1855) 6.6 (168) 0.14 (3.56) 1.00 2.63 1.00 2688 (3644) 2732 (3704) 0.985OINT/5OEXT 451 (2006) 4.1 (104) 0.82 (20.8) 5.86 3.87 1.47 2894 (3924) 2867 (3887) 1.01
Vertical displacennt, A (nTn)
-175 -150 -125 -100 -75 -50 -25 0 25 50 75 100 125 150 175 200
-7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8Vertical displacement, A (in.)
(a) Test Unit 100INT
Vertical displacement, A (nut)
-175 -150 -125 -100 -75 -50 -25 0 25 50 75 100 125 150 175 200
2500
2000
1500
1
5000
01-
-500
-1000
-1500
2500
2000
1500
1
5000
0
-500
-1000
-1500
Onset of deckcompression failure
Onset of cracking
Reference load level
Actuator displacement
/ apacityreache
600
500
400
300
•200
l00C
-100
-200
-300
cracking11 44
‘VA
-7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7Vertical displacement, A (in.)
(b) Test Unit 1 OOEXT
8
Fig. 7. Load versus vertical displacement measured at 6 in. (152 mm) from midspan:(a) Test Unit 100INT and (b) Test Unit 100EXT.
March-April 2003 87
curred during subsequent downwardloading cycles.
Midspan Joint J3 was the only jointthat opened during upward loading ofUnit 5OINT/5OEXT. It opened duringthe 0.5 in. (12.7 mm) displacementcycle. As in Unit 100INT, shear
cracks occurred in the web of Unit5OINT/5OEXT under downward loading between the supported ends of thetest unit and the load applicationpoints.
Under downward loading, themidspan Joint J3 opened significantly
with increased applied displacementuntil rupture of some of the internallybonded tendon strands occurred at themidspan joint at a displacement of 4.1in. (104 mm) and a peak load of 452kips (2011 kN). Fig. 6c shows themidspan joint at onset of failure of theinternally bonded tendon in Unit5OINT/5OEXT.
During the 5.0 in. (127 mm) displacement cycle, more internallybonded strands ruptured accompaniedby a significant drop in the load. Thetest continued for two displacementcycles after this, with only the externaltendons intact.
Table 3 summarizes the experimental peak loads and maximum displacements of all test units, including Unit100INTCIP of Phase 1-A, underdownward loading. In Table 3, P isthe peak total load and Au is the maximum displacement measured at 6 in.(152 mm) from midspan. Other experimental results given in Table 3 willbe explained in a following section.
(oad-Disptacement Response
Fig. 7 shows the history of total applied load versus vertical displacement, measured at 6 in. (152 mm)from midspan for the Series I-B(Phase I-B) test units. The sign convention in Fig. 7 is positive for downward loading and displacement. Fig. 7indicates that performance of Test Series I-B units was similar under upward loading. All of the test unitsfailed under downward loading, asmentioned in the previous section.
The load-displacement envelopesfor all test units, including Unit100INTCIP of Phase I-A, are plottedin Fig. 8, for downward loading direction only, to enable comparisons ofload-displacement responses for thedifferent test units. The maximum displacements, Au reported in Table 3 forall test units, correspond to the loadlevels indicated by the solid circlesshown in Fig. 8.
The following observations can beseen from Fig. 8:
1. All test units could undergo significant inelastic displacements beforefailure.
2. Failure of the test units with internally bonded tendons (Units 100INT,
2O0
•0
0
Cs
Vertical displacement, z (mm)
-175 -150 -125 -100 -75 -50 -25 0 25 50 75 100 125 150 175 200
2500
2000
1500
1000-
5000
0
-500
-1000
-1500
-7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8
Vertical displacement, (rn.)
(c) Test Unit 5OINT/5OEXT
Fig. 7 (cont.). Load versus vertical displacement measured at 6 in. (152 mm) frommidspan: (C) Test Unit 5OINT/5OEXT.
Vertical displacement, , (mm)
-25 0 25 50 75 100 125 150 175 200600
I I I
100% Intemal PT 100% Internal PT 2500(Cast-In-Place deck closure joints)
500
a. 2000
400 ‘5 \ 100%ema1PT-a, - ,
/_—.‘° 1500’-
50% Internal300 +50% Eemal PT ‘. l
1000--‘ 200 Referenceloadlevel ik
-__
100 -I 500
0 0-1 0 1 2 3 4 5 6 7 8
Vertical displacement, Is (in.)
Fig. 8. Envelopes of load versus vertical displacement curves (downward loadingdirection only).
88 PCI JOURNAL
100INTCIP and 5OINT/5OEXT) wasexplosive. The load carrying capacitywas completely lost upon failure ofUnits IOOINT and 100INTCIP. A significant portion of the load carryingcapacity of Unit 5OINT/5OEXT waslost upon rupture of the internallybonded tendon. However, Test Unit5OINT/5OEXT had a residual load carrying capacity, which slightly exceeded the reference load level, because the external tendons were stillintact.
3. Failure of Unit 5OINT/5OEXToccurred at a relatively low displacement compared to all other test units.The internally bonded tendon in Unit5OINT/5OEXT was carrying significantly higher forces than the externaltendons as will be shown later, resulting in relatively early rupture of theinternal tendon. The curves shown inFig. 8 indicate that a combination ofinternally bonded and external (unbonded) tendons, as currently allowedin the AASHTO Guide Specifications,’ results in less desirable seismicperformance.
4. Failure of Unit 100EXT was ductile, as the load carrying capacitydropped gradually with increasing displacements beyond the peak load. UnitI OOEXT had the highest displacementcapacity among all test units.
5. The load carrying capacity ofUnit 100EXT was less than the capacities of the other test units because oflower stresses in the external tendonsand because of the geometry changeat midspan section due to the testunit’s vertical displacement with theexternal tendons remaining essentiallyhorizontal.
Flexural Moment VersusOpening of Joints
Opening of the joints occurred inthe concrete cover adjacent to theepoxy layer between the bonded precast segments, rather than in the epoxylayer itself.2 As mentioned earlier,opening of all joints was measured bymeans of linear potentiometers. Examples of joint opening results are shownin Fig. 9, which shows joint opening atmidspan (measured at the bottom surface) versus midspan flexural momentfor the Phase I-B test units.
surface (Phase I-B test units).
In any load-displacement cycle, permanent joint opening is that measuredduring unloading at the reference loadlevel (represented by the horizontalsolid line in Fig. 9). Fig. 9 indicatesthat Unit 100EXT had the lowest values of permanent joint opening amongall test units. This indicates that with100 percent external post-tensioning,
the joints will be closed almost completely following major seismicevents.
Flexural Moment VersusJoint Rotation
Fig. 10 shows midspan joint rotation versus flexural moment for the
Joint opening, 6bol (m)
3500
3000
2500
‘ 20000
. 1500
10000
500
0
-500
-1000
-1500
-2000
0
C.’-
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5Joint opening, 8bot (in.)
Fig. 9. Flexural moment versus midspan joint opening measured at the bottom
3500
3000
2500
..-.. 2000
‘1500
1000
500
10u- -500
-1000
-1500
-2000-0.08 -0.06 -0.04 -0.02 0.00 0.02 0.04 0.06 0.08
Joint mtation, 0 (rad.)
4500
4000
3500
3000
2500
2000
1500
l000
.500
0C)
-500 t
-1000
-1500
-2000
-2500
Fig. 10. Flexural moment versus midspan joint rotation (Phase I-B test units).
March-April 2003 89
Phase I-B test units. The joint rotationwas obtained from the joint openingsmeasured at the top and bottom surfaces of the test units. The linear potentiometers that measured midspanjoint openings fell off Test Units100EXT and 5OINT/50EXT duringtesting.
The potentiometers that measuredjoint openings fell off Unit 100EXTduring the 5 in. (127 mm) displacement cycle; however, the joint rotationwas extrapolated from the vertical displacements that were measured by thevertical potentiometers (see the greencurve in Fig. 10). Fig. 10 indicates thatwith 100 percent external post-tensioning, rotational capacity of the jointcould reach about 0.070 radians,which was significantly higher thanthe joint rotational capacity of allother test units.
The potentiometers fell off Unit5OINTI5OEXT at the end of the 4 in.(102 mm) displacement cycle; itmeans just after onset of failure of thetest unit by rupture of the internallybonded tendon. The maximummidspan joint rotation of Unit5OINT/5OEXT was about 0.032 radians, which was slightly less than thejoint rotational capacity of Unit100INT with 100 percent internalpost-tensioning.
Strains in Prestressing Steel
Fig. 11 shows the strain historymeasured at midspan Joint J3 in: (1)the internally bonded tendon in UnitIOOTNTCIP (Phase I-A), (2) one of theexternal tendons in Unit 100EXT, (3)the internally bonded tendon in Unit5OINTI5OEXT, and (4) one of the external tendons in Unit 5OINT/5OEXT.Unfortunately, all strain gaugesmounted on the prestressing steel ofUnit 100INT were damaged duringpost-tensioning.
Although this paper focuses on theresults of the Phase I-B test units,strains in the internally bonded tendonof Unit 100INTCIP (Phase I-A) arepresented in Fig. 11 to compare withprestressing steel strains in the PhaseI-B test units. In Fig. 11, Cycle 1 represents the time when strains wererecorded before starting Test Stage II(the seismic test).
The yield strain level, representedby a horizontal dashed line in Fig. 11,corresponds to the 0.2 percent offsetyield strain definition. Values of themaximum displacements at midspanare also shown in Fig. 11 on the corresponding loading cycles.
The figure shows that strains in theinternally bonded tendons in Units100INTCIP and 5OINT/5OEXT increased with applied loading until the
Table 4. Stresses in the external(unbonded) tendons at peak load.
Test f,. q. (1 fp Tt
unit — ksi(MPa) ksi (MPa)IOOEXT 243.0(1676) 245.5(1693)
5OINT/5OEXT 243.0 (1676 243.9 (1682)
strain gauges malfunctioned underdownward loading. The strainsrecorded before malfunctioning of thestrain gauges exceeded 0.013 and0.020 in the internal tendons of UnitsIOOINTCIP and 5OINTI5OEXT, respectively.
The following can be observed fromthe results shown in Fig. 11:
1. Significant plastic deformationsof the internally bonded tendon inUnit 1 OOINTCIP occurred during the1.5 in. (38.1 mm) displacement cycles,whereas in Unit 5OINTI5OEXT plasticdeformations in the internally bondedtendons started to occur at the relatively low displacement of 0.75 in.(19.1 mm). Also, the strain in prestressing steel increased with upwarddisplacement and reduced after unloading to the reference load level(zero seismic displacement).
2. Plastic deformations of the external tendons occurred at substantiallyhigher displacements than for internally bonded tendons.
3. The internally bonded tendon inUnit 5OINT/5OEXT carried a substantially higher force than the external tendons in the same test unit (see the strain
plots for the two tendons in Fig. 11).4. At the same midspan displace
ment levels, strains in the internallybonded tendon of Unit 5OINT/5OEXTwere substantially higher than thestrains in the internal tendon of Unit1 OOINTCIP. Joints 32 and 14 of Unit5OINT/5OEXT (see Fig. 2 for jointnumbers) had smaller rotations, at thesame midspan displacement levels,than Joints J2 and 14 in Units 100INTand 100INTCIP. Thus, to get the samemidspan displacements in Unit5OINT/5OEXT, rotations of Joint 33were relatively high compared toUnits 100INT and 100INTCIP. Thisresulted in higher strains in the internally bonded tendon at midspan ofUnit 5OINT/5OEXT. Also, the internaland external tendons in Unit5OINT/5OEXT did not participate in
Seismic test
25,000
20,000
In
15,00000
10,000
5,000
/Malfunctioning of
strain gauges
— Internal Tendon (Unit IOOINTCIP)External Tendon (Unit 100EXT)
— Internal Tendon (Unit 5OINT/5OEXT)— External Tendon (Unit 5OINTI5OEXT)
2”__A._ 3” 6”
1”0.75”
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28Number of loading cycle
Fig. 11. Measured strains in the prestressing steel at midspan.
90 PCI JOURNAL
the force resistance in parallel but rathersequentially with the internally bondedtendon carrying most of the loading upto its failure.
5. The high strains and yielding of theinternally bonded tendon at low displacements in Unit 5OJNT/5OEXT indicate that for good seismic performance,combination of internally bonded andexternal (unbonded) tendons should beavoided in high seismic zones; the current AASHTO Guide Specifications1should be revised in this aspect.
Stresses in External (Unbonded)Tendons at NominalFlexural Resistance
The experimental results showedthat strains in the external (unbonded)tendons of Units 100EXT and5OINT/5OEXT were less than strains inthe internally bonded tendons. Reduction in strains and stresses of externalunbonded tendons is considered in theAASHTO Guide Specifications,1which gives the following equation tocalculate stresses in external tendons,f3 at nominal flexural resistance:
In U. S. units:
effective stress in prestressingsteel after losses
d = distance from extreme compression fiber to centroid of prestressing steel
c, = neutral axis depth from the extreme compression fiber
4 = tendon length between anchorages
N3 = number of support hingescrossed by the tendon
cycle (downward loading direction only).
The effective stress, based on measured strains, at the beginning of theseismic test (Test Stage II) was about181 and 179 ksi (1248 and 1234 MPa)in the external tendons of Units100EXT and 5OINT/5OEXT, respectively. The second term on the right
(ksi) hand side of Eq. (1) gives the changein tendon stresses.
The AASHTO LRFD Bridge Design Specifications8gives an equation
(Ia) similar to Eq. (1) to calcu1atef3;however, the numeral 900 in the righthand side of Eq. (la) is replaced by
(MPa)
913.5. Similarly, the numeral 6207 inEq. (ib) is replaced by 6300.
According to the AASHTO LRFDSpecifications,8or when Eq. (1) isused, the stress in unbonded tendonsshould not be taken greater than theyield stress, f (f,, = 0.90f for lowrelaxation strands,8wheref is the ultimate tensile stress). Values of f3calculated using Eq. (1), withf3f3,are given in Table 4 for the externaltendons of Units 100EXT and5OINTI5OEXT.
Strains in external tendons, 2,, at
Vertical displacement, z (mm)40 60
600
500
400
300C
t200
100
00.0 0.5 1.0 1.5 2.0 2.5
Vertcal displacement, t (in.)
2500
2000
1500
1000
500
03.0 3.5 4.0
Fig. 12. Load versus vertical displacement for the 3 in. (76.2 mm) displacement
= fpe + 900
In metric (SI) units:
fps = fpe + 6207
where
fpe
(lb) Concrete solidelements
Compression onlyspring elements
Fig. 13. Finite element model.
March-April 2003 91
the peak load for Units 100EXT and5OINT/5OEXT were measured. Thetendon stresses, Test’ given in Table4, were calculated from the measuredtendon strains using the followingstress-strain relationship equation9(assuming elastic modulus, E5 = 29,000ksi = 200 GPa):
Post-EarthquakePermanent Displacements
Table 4 indicates that the experimental values of tendon stresses werevery close to those calculated using
Eq. (1) given by the AASHTO GuideSpecifications’ (with f f,,). Itshould be mentioned that the yieldstress upper limit on f5 calculatedusing Eq. (1) was most likely inadvertently left out of the AASHTO GuideSpecifications.’
MacGregor et al.1° conducted an ex
periment on a three-span precast segmental box girder bridge model withexternal (unbonded) tendons. Eq. (1)is based on the method proposed byMacGregor et a!.’° on prediction ofstresses in unbonded tendons. Their
test specimen was subjected to mono-tonic loading and not reversed cyclicloading. To the authors’ knowledge,no previous experimental research hasbeen conducted on the performance ofprecast segmental bridges under re
versed cyclic loading.
One important aspect for desirableseismic performance of bridges is tominimize permanent displacements ofthe superstructure after an earthquakeevent. Fig. 12 shows the applied loadversus displacement during the downward loading portion of the 3 in. (76.2mm) displacement cycle for all testunits of Phase I.
The displacement, measured at thereference load level in the unloadingportion of any of the curves shown inFig. 12, represents the permanent displacement, 4 Values of 4 measuredafter 3 in. (76.2 mm) maximum vertical displacement for all test units, aregiven in Table 3.
The values of 4 are normalized tothe permanent displacement of UnitIOOEXT, Ref (1Ref = 0.14 in. = 3.56mm), and values of the ratio (4JzlRef)
are given in Table 3 for all test units.Comparison of the (4J/-Ref) values indicates that permanent displacementscan be minimized by the use of 100percent external post-tensioning. Thisis because strains in the external tendons are significantly less than strainsin the internally bonded tendons (see
(2) Fig. 11).The strains in external (unbonded)
pp pp Strand rupture \ Strand rupture
+4 44 monotonic anaiysisY, / (experiment)
H4. hi ‘2 / Strand rupture44,
.“ /,/...‘ (cyclic analysis)
iA ,Reference load level
Vertical displacement, A (mm)-150 -125 -100 -75 -50 -25 0 25 50 75 100 125 150 175 200
6002500
500
_________________________
2000400
__________________________
1500300
1000 ,—.m200
.o 100 500CS 0
-100 . 500Dowmard displacement -
-200 — Finite element (cyclic) -1000Finite element (Monotonic)
-300 Experiment-1500
-400 I I I I I I I I
-6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8Vertical disp1acemet, A (in.)
(a) Test Unit 1 OOINT
A
ss DX) Onset of compression Compression failure
.L.L .L.L failure (experiment) (mOnOtofliC analysis)
__!!_!-!;_‘-,T--_. \
_LLjLJ-
Reference load level
100 1’-’ Compression failure(cyclic analysis)
Vertical displacement, A (mm)-150 -125 -100 -75 -50 -25 0 25 50 75 100 125 150 175 200
600 1I 2500
500
2000400
___________________
- 1500300
1000200 Z
500 °CS0
0 —
—
_____________ __________________
CSCS” 0
I-
-100 Dowmard displacement -500
-200 — Finite element (Cyclic) -1000Finite element (Monotonic)
300 Experiment• -1500
-400 I p I
-6 -5 -4 -3 -2 1 0 1 2 3 4 5 6 7 8Vertical displacement, A (in.)
(b) Test Unit 1 OOEXT
Fig. 14. Load versus displacement analysis results: (a) Test Unit 100INT and
(b) Test Unit 100EXT.
f,,
29,0.975
92 PCI JOURNAL
tendons are spread over the whole tendons’ length, whereas strains inbonded tendons are concentrated overa much shorter distance. Thus, for thesame joint rotation, the strains in thebonded tendons are much higher thanfor the unbonded tendons. Inelasticstrains in internally bonded tendonsresult in loss of the prestressing force,large permanent displacements andpermanent joint openings.
Equivalent ViscousDamping Coefficient
The equivalent viscous damping coefficient, ‘, can be considered a measure of the energy dissipation capability of the test units. For comparisonpurposes, the coefficient was determined by relating the area within theload-displacement hysteresis ioop ofthe 3 in. (76.2 mm) displacementcycle to the elastic strain energy at thesame cycle.” The values of aregiven in Table 3 for all four test unitsof Phase I.
The equivalent viscous damping coefficients for all test units were normalized to that of Unit I OOEXT, Ref
(ef = 2.63 percent), and the ratios of(‘2’Ref) are given in Table 3. The values of (C/Ref) indicate that superstructures with internally bonded tendonsare able to dissipate more energy thansuperstructures with external tendons.
Despite the enhancement in energydissipation capability, use of cast-in-place deck closure joints (UnitI OOINTCIP) may not be economicalbecause they would complicate theprecast segmental concept and wouldresult in slower construction andhigher costs. Also, energy dissipationcapability is more important in thecolumns since column plastic hingingis anticipated whereas bridge superstructures are designed to remain essentially elastic during major seismicevents. Thus, energy dissipation capability of the superstructure may not bean important design issue.
Flexural Moment Capacity
The experimental peak flexural moment at midspan, Mesj, is given inTable 3 for all four test units of PhaseI. The calculated flexural moment ca
pacity, Majc, as well as values of theratio (MeiIMaic) are also given inTable 3. The flexural moment capacity, M01, was calculated according toprovisions of Article 9.17 of theAASHTO Standard Specifications5and Article 11.2 of the AASHTOGuide Specifications.’ Stress in theexternal tendons at ultimate load wascalculated using Eq. (1) according toprovisions of Article 11.2 of theAASHTO Guide Specifications,’ butassuming that the tendon stress wouldnot exceed the yield stress, f =
243 ksi = 1676 MPa).Values of the ratio (MTest/MCalc)
were close to 1.00 for all test units, indicating that the flexural moment capacity of precast segmental bridge superstructures can be reasonablyestimated using provisions of Article9.17 of the AASHTO Standard Specifications5 and Article 11.2 of theAASHTO Guide Specifications.’
The ratio (Mresr/Mcaic) was slightlyless than 1.00 for Unit 100EXT,which had only external tendons. Thisslightly non-conservative estimate offlexural moment capacity for this testunit was due to the change in geometry of the cross section as the test unitwas displaced vertically while the external tendons remained essentially
horizontal. This resulted in reductionof the internal moment arm betweenthe tendons and centroid of the compressive stress block at the midspansection; this reduction in the momentarm was not considered in the calculation of Maic.
In calculating the flexural momentcapacity of Unit 5OINT/5OEXT,stresses in the internally bonded tendons were different than stresses in theexternal tendons. Tendon stresses werecalculated using provisions of theAASHTO Standard Specifications5andthe AASHTO Guide Specifications.’
FINITE ELEMENT ANALYSISThree-dimensional finite element
models of the test units were developed. The finite element models andresults of Units I OOINT and 1 OOINTCIP were presented in an earlierpaper;2 however, results of Unit100INT will be briefly presented inthis paper for completeness, and because Unit 1 OOINT is one of the PhaseI-B units.
Decription of theFinite Element Models
Detailed finite element models weredeveloped for all test units (Fig. 13).
Vertical displacennt, A (mm)-175 -150 -125 -100 -75 -50 -25 0 25 50 75 100 125 150 175 200
600 I
PP pp
Li‘V
Reference load level
Rupture of strands Onset of rupture of(monotonic analysis) (experiment)
/s1’”NRupture of strands/ / ‘(cyclic analysis)
500
400
300
200
1000
00
-100
-200
-300
-400
2500
2000
1500
1000 z•500
0
0F-
-500
-1000
.1500
:::z
Dowmard
displacement
— Finite element (Cyclic)——— Finite element (Monotonic)
Experiment
-7 -6 -5 -4 -3 -2 -1 0 1 2Vertical displacennt, A (in.)
3 4 5
(c) Test Unit 5OINT/5OEXT
7 8
Fig. 14 (cont). Load versus displacement analysis results: (C) Test Unit 5OINT/5OEXT.
March-April 2003 93
Analyses were performed using thegeneral-purpose finite element program ABAQUS,12 interfaced with theANACAP’3concrete material model.
The concrete was modeled as 3-D,eight-node, solid brick elements withstrain-hardening and strain-softeningcapabilities in compression, and tension cutoff with cracks that do notheal upon closure.’3 Confinement effects were assumed to be negligibleand the unconfined concrete strengthwas taken as 7.5 ksi (51.7 MPa). Themodel was developed in a similar wayto the test units, with no solid elements crossing the joints between precast segments and no connection between solid elements on either side ofthe joints.
The joints were free to open by providing double nodes and compression-only springs at all nodes in the crosssection at locations of joints. Prestressing steel was modeled by trusselements and connected to the concrete nodes at each 12 in. (305 mm)cross section, representing bondedstrands.
External tendons were also modeledby truss elements, connected to thesolid blocks in the precast end segments (Segments 1 and 6 in Fig. 2).Beyond 2.0 in. (50.8 mm) displacement in the upward loading direction,
the bottom slab of Test Units 100EXTand 5OINT/5OEXT was in contactwith the external tendons; this contactwas achieved by gap elements in thefinite element models.
All mild steel reinforcement wasmodeled as 1-D sub-elements to thesolid concrete elements. No mild steelreinforcement crossed the joints ofUnits 100INT, 100EXT and5OINT/5OEXT. At the joints, the prestressing steel was not connected tothe center nodes, but to nodes at sections 12 in. (305 mm) on either sideof the centerline. This represented anidealized unbonded length at thejoints of 24 in. (610 mm) and allowedfor strain penetration on both sides ofthe joint. Loading was applied to themodels in displacement control asshown in Fig. 5.
Finite Element Results
Test Unit 100INT — There wasvery close agreement between theanalysis and measured load-displacement responses (see Fig. 14a). Analysis results showed that the model behaved very similarly to the test unit interms of ultimate load, displacement atfailure, and shape of the hysteresisloops in both the upward and downward loading directions. Rupture
stress and strain of the prestressingsteel were assumed as 270 ksi (oo 1860MPa) and 0.04, respectively, whichappeared to be reasonable based onUnit 100INT load-displacement resuits presented in Fig. l4a.
Fig. 15 shows the prestressing steelstress versus loading cycle number atmidspan (Joint J3, see Fig. 2) fromanalysis of Test Unit 100INT. It isclear from Fig. 15 that beyond 0.75 in.(19.1 mm) displacement, plastic deformations of the tendon occurred andthe initial prestressing force reducedsignificantly, and was completely lostfollowing 4 in. (102 mm) of downward displacement.
Test Unit 100EXT — The shape ofthe monotonic load-displacement results for the upward loading directionmatched the test results very well (seeFig. l4b). However, under downwardloading, the monotonic load-displacement curve matched the test resultsonly in the pre-peak range (see Fig.14b). The failure mode was concretecrushing of the deck, as observed inthe experiment. In the cyclic analysis,the model could not capture degradation of concrete strength with increased downward displacements andthe failure mode according to the finite element analysis was crushing ofthe deck, as determined from straincontour values. The unloading stiffness of the analysis did not match thatof the test very well.
Fig. 15 shows the prestressing steelstress versus loading cycle number atmidspan (Joint J3, see Fig. 2) fromanalysis of Test Unit 100EXT. Fig. 15shows that the external tendonsyielded during the 3 in. (76.2 mm)first displacement cycle; the initialprestressing force started to reduceduring the same displacement cyclebut not dramatically, compared to theinternally bonded tendon of Unit100INT. The initial prestressing forcewas not completely lost, even at highmidspan displacements through theend of the test.
Test Unit 5OINT/5OEXT — Thecyclic load-displacement analysis resuits for Unit 5OINT/5OEXT matchedthe test results very well (see Fig.14c). The shape of the monotonicload-displacement curve matched thetest results only in the downward load-
300
280
260
240
220
200
180
1609
140
120
l00
80
60
40
20
0
2000
1800
1600
1400
1200
1000
800
600
400
200
0
00
GOGO0)
GO
a0
Internally bonded tendon(Unit 5OINT/5OEXT)
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28
Number of loading cycle
Fig. 15. Stress history analysis results of prestressing steel at midspan.
94 PCI JOURNAL
ing direction. The failure mode fromboth monotonic and cyclic analyseswas rupture of the internally bondedtendon, as observed in the experiment.The finite element model could reasonably capture the residual load carrying capacity of the test unit afterrupture of the internally bonded tendon, as seen in Fig. 14c.
Fig. 15 shows the stress in the internally bonded tendon at midspan versusloading cycle number from analysis ofUnit 5OINT/5OEXT. The figure showsthat yielding of the internally bondedtendon occurred during the 1.0 in.(25.4 mm) displacement cycle, whichwas early compared to all other testunits. This was due to the high force inthe internally bonded tendon of thistest unit as evidenced from the measured strains shown in Fig. 11.
The initial prestressing force startedto reduce during the 0.75 in. (19.1mm) displacement cycle and was completely lost after a maximum displacement of 1.5 in. (38.1 mm), which wasagain sooner than the other test units(see Fig. 15). This indicates that thecombination of internally bonded andexternal tendons in high seismic zonesmay result in premature yielding ofthe internally bonded tendon and premature loss of the initial prestressingforce. This may have important consequences in the design of precast structures, where under a severe earthquakethe prestressing force is diminished orlost altogether before rupture of thestrands.
DESIGN IMPLICATIONSThis paper presents the experimen
tal and analytical results of Phase I ofthe research project on seismic performance of precast segmental bridges atthe University of California at SanDiego (UCSD). Seismic design recommendations can be made uponcompletion of the three phases of theresearch project. However, based onthe results presented in this paper, thefollowing can be implied for seismicdesign of precast segmental bridge superstructures:
Combination of internally bondedtendons with external (unbonded) tendons, as currently allowed by theAASHTO Guide Specifications’ may
result in premature yielding, loss ofthe initial prestressing force and rupture of the internally bonded tendons.In perspective of seismic design, combination of internally bonded and external tendons is not recommended.
Use of only external tendonswould improve the seismic performance in terms of ductility, displacement capacity, post-earthquake permanent displacements and permanentjoint openings.
The flexural capacity of precastsegmental bridge superstructures iswell-predicted using provisions of Article 9.17 of the ASSHTO StandardSpecifications5and Article 11.2 of theAASHTO Guide Specifications.1However, with external post-tensioning, designers should pay attention tothe change in the internal moment armbetween external tendons and the extreme compression fiber of the crosssection. This change in moment armshould be considered in special casessuch as in seismic design where largeductility and displacements are required in the superstructure. Thestresses in the external tendons shouldnot exceed the yield stress, f (f =
0.90 for low relaxation strands8).Finite element analyses showed
that the effective prestressing force ininternally bonded tendons could reduce after a major seismic event, especially if the superstructure segment-to-segment joints are subjected tosignificant openings or rotations during the seismic event. External tendons proved to be a good alternative tointernally bonded tendons in whichcase less reduction in the effectiveprestressing force is expected at highdisplacement levels.
CONCLUSIONSA three-phase research project is
currently in progress at the Universityof California at San Diego to investigate the seismic performance of precast segmental bridges. Results of alarge-scale experimental program andfinite element study of the seismic performance of segment-to-segment jointssubjected to high flexural momentsand low shears (Phase I) are presentedin this paper. The following conclusions can be drawn from this study:
1. Crack patterns for all test unitswith internally bonded tendons weresimilar under downward loading. Onlythe midspan joint opened during testing of the unit with 100 percent external post-tensioning.
2. The segment-to-segment jointscan experience significant repeatedopenings and closures under reversedcyclic loading without failure, evenwith no mild steel reinforcementcrossing the joints. Precast segmentalsuperstructures can undergo significant seismic displacements withoutfailure.
3. Test Units 100INT and5OINT/5OEXT experienced explosivefailures as a result of rupture of the internally bonded tendon. With 100 percent external post-tensioning, the explosive failure was avoided and theload carrying capacity dropped gradually in Unit 100EXT with increaseddisplacements in the post-peak range;the failure initiated by concrete crushing in the deck.
4. Ductility and displacement capacity can be substantially enhanced byuse of 100 percent external post-tensioning. Use of only external tendonswill also minimize post-earthquakepermanent displacements of the superstructure as well as permanent openings of the segment-to-segment joints.
5. Combination of internally bondedand external (unbonded) tendons inprecast segmental bridge superstructures, as currently allowed by theAASHTO Guide Specifications,’should be avoided in high seismiczones. This is because internallybonded and external (unbonded) tendons do not participate in the force resistance in parallel, but rather sequentially with the internally bondedtendons carrying most of the loadingup to their failure.
6. Finite element analyses showedthat under severe earthquakes, the prestressing force in the internally bondedtendons could diminish under repeatedcycling in the inelastic strain range.Premature loss of the prestressingforce may occur if internally bondedand external (unbonded) tendons arecombined in the superstructure. Lossof prestressing force in external tendons is less severe for a given displacement or ductility level.
March-April 2003 95
ACKNOWLEDGMENTS
Caltrans is acknowledged for funding this research underContract No. 59A0051. Members of the AASHTO-PCIASBI Technical Research Committee are acknowledged forparticipating in the development of this research project.
J. Muller International, San Diego, prepared the construction drawings of the test units. Sika Corporation donated the
epoxy. Dywidag-Systems International (DSI), USA, donated the post-tensioning strands, high-strength bars andhardware.
The authors wish to express their appreciation to the PCIJOURNAL reviewers for their valuable and constructivecomments.
REFERENCES
1. AASHTO, Guide Specifications for Design and Constructionof Segmental Concrete Bridges, American Association of StateHighway and Transportation Officials, Second Edition, Washington, DC, 1999.
2. Megally, S., Seible, F., Garg, M., and Dowell, R. K., “SeismicPerformance of Precast Segmental Bridge Superstructures withInternally Bonded Prestressing Tendons,” PCI JOURNAL, V.47, No. 2, March-April 2002, pp. 40-56.
3. Megally, S., Seible, F., and Dowell, R.K., “Seismic Performance of Precast Segmental Bridges: Segment-to-SegmentJoints Subjected to High Flexural Moments and High Shears,”to be published in May-June 2003 PCI JOURNAL.
4. AASHTO-PCI-ASBI, Segmental Box Girder Standards forSpan-by-Span and Balanced Cantilever Construction, American Association of State Highway and Transportation Officials, Precast/Prestressed Concrete Institute and AmericanSegmental Bridge Institute, 1997.
5. AASHTO, Standard Specifications for Highway Bridges, 13thEdition, American Association of State Highway and Transportation Officials, Washington, DC, 1983.
6. Megally, S. H., Garg, M., Seible, F., and Dowell, R. K., “Seismic Performance of Precast Segmental Bridge Superstructures,” Structural Systems Research Project Report SSRP
200 1/24, Department of Structural Engineering, University ofCalifornia at San Diego, La Jolla, CA, May 2002, 317 pp.
7. ASBI, Recommended Contract Administration Guidelines forDesign and Construction of Segmental Concrete Bridges,American Segmental Bridge Institute, Phoenix, AZ, March1995, pp. 109-116.
8. AASHTO, LRFD Bridge Design Specifications, Second Edition, American Association of State Highway and Transportation Officials, Washington, DC, 1998.
9. Collins, M. P., and Mitchell, D., Prestressed Concrete Structures, Response Publication, Toronto, Canada, 1997.
10. MacGregor, R. J. G., Kreger, M. B., and Breen, I. E.,“Strength and Ductility of a Three-Span Externally Post-Tensioned Segmental Box Girder Bridge Model,” Research Report 365-3F, University of Texas at Austin, Austin, TX, January 1989, 324 pp.
11. Chopra, A. K., Dynamics of Structures, Prentice Hall, Inc.,Upper Saddle River, NJ, 1995, 729 pp.
12. Hibbitt, Karlson, and Sorenson, ABAQUS User’s and TheoryManual, Version 5.8, 1999.
13. ANATECH Consulting Engineers, ANA CAP-U User’s andTheory Manual, Version 2.5, 1997.
96 PCI JOURNAL