•;• Design-Construction Feature

4. The Kentucky River Bridge

:: Variable Depth Precast•• Prestressed Segmentalv Concrete Structure

Homer M. Walker, PEStructures Engineer-PartnerAmerican Engineering Co.Consulting EngineersLexington, Kentucky

T he first precast segmental post-tensioned concrete box girder

bridge in Kentucky was completed andthe entire project was opened to trafficon December 6, 1979. The river cross-ing was the critical time factor on theconstriction schedule for the project.

This is a notable feat considering thatdesign was started in April, 1976, andthat bids for the structure over theKentucky River were opened just 17months later. A Design Report, an En-

60

H. Hubert Janssen,PEChief Structural EngineerBVN-STS Consulting EngineersIndianapolis, Indiana

vironmental AsseCoast Guard Navirequired, in additicpreliminary plans,gation, final plan:views, conference;tractors.

Bridge constructmonths, despite asite on Decemberstrike during thestage, and several

Describes the preliminary studies, alternatestructure schemes, superstructure design,post-tensioning layout, epoxy joint details, segmentproduction and erection of the Kentucky RiverBridge--a variable depth precast prestressedsegmental structure, with a 323-ft (98.5 m) centerspan in which the segments were fabricated usingthe long-line casting method.

delays. Projects of this type commonlyrequire from 5 to 10 years from begin-ning of design to completion of con-struction. Therefore, the 44-monthschedule resulted in a considerable costsaving and made the highway availableto the traveling public at a much earlierdate.

Frankfort is the capital of the Com-monwealth of Kentucky. The bridgeover the Kentucky River is part of whatis called the Frankfort East-West Con-nector (KY 676, a bypass road) whichbrings together several outlying areasand a number of state office complexes.The need for a new crossing was em-

phasized during the 1978 flood whichclosed the downtown area with its rail-road and three highway bridges, leav-ing only the rural I-64 structures overthe Kentucky River to connect the city'ssuburban areas.

Site ProblemsEarly studies showed that an inter-

change was not possible in the limitedarea available in the narrow valley ap-proach to the west end of the bridge.Therefore, the intersection layoutwhich was ultimately selected deter-mined the alignment across the river.

The profile was set to provide 50 ft(15.25 m) of vertical clearance abovethe river pool stage, based on a steeltwo-girder system (with floor beamsand stringers) with a center span toprovide 300 ft (91.5 m) of horizontalclearance. These vertical and horizontalclearances were the minimum naviga-tional requirements acceptable to theU.S. Coast Guard.

A structure over the railroad some2000 ft (610 m) east of the river set thegrade at + 1.56 percent.

The west end of the structure wasdetermined by relocated Big EddyRoad, which passed under the bridge.The angle of this road crossing and theproximity of the intersection made ithighly desirable to set the abutmentnormal to the centerline of structure.

The east end of the structure was es-tablished by the soil slope stability re-quirement. The abutment could beeither normal or skewed to parallel theedge of water.

The hydrological study required thepiers to be parallel to the stream flow.This requirement caused considerableproblems in design and construction,but had the advantage of allowing ashorter center span while satisfyingnavigational requirements.

With the line and grade established,the roadway plans were completed andthe west approach section was let to

construction in September, 1976. Laterthe east approach contract was alsoawarded, so that roadway work (exceptfinal paving) was completed on bothsides of the river before the bridgecontract was awarded and constructionstarted.

Alternate Structure StudiesSteel plate girder framing plans, in-

eluding details for the connections atthe skewed piers and abutments wereprepared, analyzed, and a cost studywas completed. (Composite load factordesign was not allowed by the Depart-ment at that time.) Similar studies weremade for the concrete box girder struc-ture, with only the piers being skewed.

The first proposals for the concretestructure were based on a two-spline12-ft (3.66 m) constant depth box sec-tion with sloping sidewalls. When thesepreliminary plans were presented tothe Department of Transportation withcost studies which showed the concreteto be competitive with steel, the De-partment's interest and willingness toaccept the segmental concrete box girderconcept encouraged the design en-gineers to pursue this type of construc-tion further.

The Consulting Engineering firm ofBVN/STS had been established in In-dianapolis, Indiana, with a nucleus ofpersonnel and complete computer de-sign programs brought in from theDutch consulting firm of BVN. Theywere retained by the prime consultantand approved by the Kentucky De-partment of Transportation to prepare apreliminary report on the segmentalsuperstructure.

It was at this point that the variabledepth section was first proposed. Initialstudies indicated the cost of constantdepth and variable depth segmentswere approximately the same_ Sinceappearance was considered importantat this location, Charles G. Cook, Di-rector, Division of Bridges for the

62

Kentucky Department of Transportationchose the variable depth (with itsparabolic lines) primarily for aestheticreasons.

With the line and spans set, it waspossible for the Kentucky Departmentof Transportation, Division of Materials,Soils Section, to proceed with the sub-surface investigation in October, 1976.

More complete and detailed coststudies were made at this time. Theyshowed the initial cost of the steel to beconsiderably more than for the concretestructure. It was estimated that thesegmental erection time and thereforethe total construction time for the con-crete bridge would be the shorter of thetwo schemes.

This, combined with the cost ofpainting the steel periodically and theundesirable framing connections of thesteel plate girder caused the consultantto recommend the precast segmentalconcrete box girder structure. TheKentucky Department of Transportationconcurred with this recommendation onDecember 16, 1976, and the bridgemoved into the final design phase.

Construction PlansThis was prior to the time when it

became common practice to preparealternate plans for both steel and con-crete structures. The contract planswere prepared only for the segmentalconcrete bridge.

It should be noted that the AASHTOspecifications for segmental concretebox girders were not adopted untilmid-1977. However, every effort wasmade to comply with the current tenta-tive specifications.

In the course of preparing and re-viewing the preliminary plans, the con-sultants and the Kentucky Departmentof Transportation had decided the waythe bridge construction was to be han-dled. Since no Federal Highway Ad-ministration funds were involved in theproject, it was possible for the Plans

and Specifications to reflect thesechoices, without additional approvals.

Basic parameters were set for the fab-rication. Segments were to be precastusing the cross section specified, withthe variable depth, as shown. A strandsystem was to be used, with a choice ofpost-tensioning anchorages allowed.The segment manufacturer was re-quired to be experienced in this type ofconstruction.

The architectural details had beenworked out and agreed upon in thepreliminary stage. BVN/STS prepared acomplete superstructure design anddetail, with American EngineeringCompany reviewing the load andgeometric controls, coordinating thework and keeping the Kentucky De-partment of Transportation informed asdecisions were required. American En-gineering Company also prepared thesubstructure design, details, miscel-laneous drawings and notes requiredfor the Contract Plans.

Superstructure DesignThere are many segmental bridges,

either in service or being constructed.Much has been written about designconcepts, production of segments, anderection. Rather than repeatinggeneralities, this article intends to con-centrate on some unique aspects of thestructure, such as the variable depth,the treatment of the skew, the reductionof unbalanced moments during erec-tion, and the reduction in number ofsegments. Future deck repairs andsome structural details (such as post-tensioning layout, the shear keys, andthe shape of the cross girders) will alsobe discussed.

Constant depth box girders are usu-ally more economical than variabledepth girders up to span lengths of 300to 350 ft (91.5 to 106.7 m). The mainreason for this is the cost of forming. Inthe case of a variable depth, both innerand outer forms are more expensive be-

PCI JOURNAUJuIy-August 1981 63

0)A

2 323-0"

222'-5%'

234-6'8" 322 -1 I 222'-4%'*

ELEVATION — WESTBOUND SPLINE

---- --LIMITS OF BOX. WESTBOUND SPLINE-

LIMITS OF BOX-

Q BRG- ABUT. I & PIER I-I

222_5/$'

8'-0°

t SURVEY 6 OPEN JOINT-\

PLAN — TOP OF DECK (- PIER 2

323-0"

EASTBOUND SPLINE

BRG. ABUT.2^SURVEY DIMENSIONS

234-6%

222- 738* __ 322'-!I

234'-3% *

ELEVATION — EASTBOUND SPLINE

Fig. 1. Span arrangement to accommodate skewed piers. (Note: 1 ft = 0.305 m; 1 in. = 25.4 mm.)

^.- SURVEY

HALF SECTION AT ABUTMENT HALF SECTION AT PIER

Fig. 2. Cross section of superstructure. (Note: 1 ft = 0.305 m; 1 in. - 25.4 mm.)

cause they must be capable of quicklyadapting to different depths. Moreover,setup of the forms becomes more in-volved because of making the depthchanges each casting day.

The cost of continuity post-tensioningmay also be higher because the shal-lower depth does not allow placementof the tendons at as much eccentricityas would be possible with a constantdepth. Consequently, the prestressingforce will be higher, even after consid-ering the sizable reduction in bendingmoments by superimposed loads(which concentrate at the high sectionover the piers) relieving the midspansections.

Fabrication of the reinforcing barcages is also more complicated becauseof the depth variation, although muchcan be done to simplify this. Reinforc-ing bars, for example, that must fit thedepth of the section can be made in twoparts with an oversize lap, thus pro-viding ample tolerances.

Despite the skewed piers, this bridgeis straight, that is, the abutments and alljoints between precast segments are atright angles to the centerline of the

structure. This aspect greatly simplifiesdesign, detailing, manufacture of seg-ments, erection of the superstructureand construction of the abutments.

In order to achieve this, the followingmeasures were taken. First, the endspans of each of the adjacent structuresare different (see Fig. 1). Secondly, thebearings are not placed on the center-line of the pier (see Fig. 3). Eccentricplacement of the bearings an the piersas shown hardly changes the loadingconditions, and only the effect of plac-ing the bearings close to the edge of thepier requires further investigation. Ad-ditional skew could have been pro-vided by also placing the bearings at askew to the pier segment. This wouldcause additional bending in thesuperstructure which must be takeneither in torsion of the cross girder orlongitudinal bending of the box girder.

The segments were cast level andtilted to the cross slope of the roadwayin the field (see Fig. 2). Thesuperstructure length is measuredalong the roadway. This combination oflength along the grade, variable depthon the grade, cross slope changing the

PCI JOURNAL/July-August 1981 65

Fig. 3. Bearing arrangement on pier cap.

ABUTMENT SIDE

STAGE 1

STAGE 2

17 SEGIMENTS 17 SEGMENTS

TEMPORARY STRUT 0HAVING FLAT JAC.ASSEMBLY "S"MAX CAP 200 STAGE 3

STAGE 4

1

ERECTION PROCEDURE-CANTILEVERS

Fig. 4. Sequence of erection procedure and location of strut.

offset from survey line to center of boxfrom top to bottom as the depth variedand the skewed pier line made the lay-out geometry quite complicated.

Due to the height of the piers and thelength of the center span, expansionbearings were only needed at theabutments. The pier bearings were ofthe spherical pot type which accommo-date rotations and small movements in

the tranverse directions of the deck.Fig. 3 shows the typical bearing ar-rangement on the pier cap.

Transportation load limits madenecessary the smaller lengths of thedeeper segments. The shallower sec-tions could be made longer thereforereducing the total number of segmentsto be manufactured and handled. Seg-ments close to the piers, having a heavy

66

SECTION AT PIER

234'-5%"

t BRG. ABUT

222-6%

BRG. ABUT.I - Q BRG PIER t

Nf ERMEDIATE DIAPMRASM

323- 0"

E RRGG PIER 2

TENDON GROUP 2 i TENDON GROUPI

YID SPAN DIAPHRAGM-5 TENDONS OF 12-270 L STEEL STRAND.INSTALL AND STRESS TO 34R/ TENDONPRIOR TO PLACING B" CONCRETE OVERLAY.

ELEVATION

0C-0CIIzC-LC

C

CSN

SD

4'-0'

GVERL AYE D

STRESSd EACH SIDE

Ml0 5PANONLY

DSECTION AT MID SPANCAST - IN - PLACE JOINT

INTERYEOIATE CIA

1 SECTION AT END SPAN} CAST-IN-PLACE JOINT

SECTION AT ABUTMENT

FUTURE CONCRETE OVERL

LICATIDN OF FUTURE

--3 - 3" GALV. STEEL PIPE —9

y n

SECTION A - A SECTION B - B

I.D. GALS.STEEL PIPE GAL

STEELV

PIPEyR

NSECTION C - C

ii Fig. 5. Provisions for future tendon placement inside box at diaphragms. (Note: 1 ft = 0.305 m; 1 in. = 25.4 mm.)

Fig. 6. Draped tendons in web walls. (Note: 1 ft = 0.305 m; 1 in. = 25.4 mm.)

bottom slab, and which are up to 16 ft(4.88 m) in height were made 5 ft 3 in.(1.6 m) long; whereas, segments closerto midspan are 7 ft (2.13m) long (seeFigs. 6 and 7).

Although it was not done here, othervariations such as web thickness couldhave been made. The number of seg-ment types should be limited toperhaps two or three and the variationsshould not require major changes of theform. Making short segments in a formdesigned to make longer segments didnot cause any problems on this project.

It is quite common to design piers insuch a way that they are capable ofsustaining erection loads. The criticalcondition is the erection of the last un-balanced segment, combined withsome working loads. In the erection ofthis structure, the controlling loadingcase would have necessitated a largerpier footing size with a considerable in-crease in the cost of the cofferdam, ex-cavation, tremie seal, concrete, andother items.

Rather than increase the footing itwas decided to use an auxiliary strut inthe tail spans at approximately 100 ft(30.5 m) from the pier. This strut con-sisted of a single H pile, with a jackingarrangement (see Figs. 4 and 25), andwas a very economical solution to thisproblem. A counterweight was placedon the deck above the strut in order toprovide positive bearing forces at alltimes.

Future Deck RepairsDuring one of the many meetings

between the Kentucky Department ofTransportation engineers and the con-sultants that took place in the designphase, concern was expressed for thedurability of the deck and the possibil-ity of its future repair or replacement.Decks of segmental bridges have con-siderable resistance against deteriora-tion, since the air-entrained concrete isfactory produced and therefore ofsuperior quality and of strengths oftenin excess of the 5500 psi (37.9 MPa) re-quired for this project.

In addition, the Kentucky RiverBridge deck is also post-tensioned inboth directions. An epoxy coating onthe top mat of the reinforcement and 2in. (51 mm) of concrete cover aid inpreventing damage by chloride ionswhich may eventually penetrate intothe deck.

Nevertheless, the possibility of deckreplacement was considered, and asimple solution was developed. Re-moval of large portions of the deck isnot possible, but the addition of a newdeck overlay is quite easy. If required,the plans provide for the removal of thebituminous wearing surface and thedeteriorated concrete down to the topreinforcing bar gnat, This removal ofweight is equivalent to about 6 in. (152mm) of concrete. An 8 in. (203 mm)structuraloverlaycould be added withoutexceeding the substructure capacity.

68

DnC-0C3]zELC

ID

N

(OW

ALL SEGMENTS AREMATCHCAST.

IT BUILT TO THE UNDERSIDE OF BRIDGEILE AND HAVE THE CASTING CORRECTIONSPORATED INTO ITS CONSTRUCTION.

RAIL- I rSUPPORTS

LONG LINE METHOD

11a5 3 IA a7 `-O" _

PRODUCTION SEQUENCE 1

HORIZONTAL SOFFIT

BULKHEAD W/ 1ADJUSTABLE DEPTHOLDSEGMENT ULKHEAD FASTENED

TO SIDE FORM FORII CASTING OF SEGMENTS

r

-' TWO SETS OF WHEELS ALLOINST LONGITUDINAL AND TRANSVEI

OINIRISMOVEMENT OF SIDE FORM F

STRIPPING AND CONTINUOUSFABRICATION OF SEGMENTS.

CASTING FROM PIER TO MID SPAN SPLICE (REPEAT 4.TIMES

CASTING FROM PIER TO ABUTMENT (REPEAT 4 TIMEjC)CD

Fig. 7. Segment casting bed arrangement and production sequence. (Note: 1 ft = 0.305 m; 1 in. = 25.4 mm.)

Additional post-tensioning tendons tocarry a full 8-in. (203 min) overlay hasbeen provided for in the design. Thisadditional post-tensioning is to beplaced in polyethelene pipes inside thebox girders, The provisions for thishave been included in this contract bydesigning and detailing for anchoragein the abutment, pier and midspandiaphragms. This post-tensioning couldalso be used to upgrade the load carry-ing capacity of the structure. Details ofthe additional post-tensioning areshown in Fig. 5.

Tendon Layout, Shear Keysand Epoxy Joints

The essentials of the post-tensioninglayout are the use of draped tendonsanchored in the webs as shown in Figs.6 and 21. Although more complicated toinstall than straight tendons placed inthe deck and the bottom slab, thestructural advantages are sufficientlybeneficial to warrant their use. Thevertical component of inclined ten-dons (approximately 45 percent of thetendon force) reduces the externalshear forces, A reduction of as much as900 kips (4003 kN), or 50 percent of thetotal shear force, has been achieved inthe high shear area of the structure bymerely draping the tendons.

This reduction in shear force is ben-eficial, both for the vertical epoxiedjoint, which is relieved, as well as forhigh principle stress zones which willremain uncracked. Another advantageof the draped tendons is the reductionof shear forces acting on the keys dur-ing erection. Other methods to takeshear, such as by reinforcing bars, arefeasible but more costly, and less effec-tive..

The Kentucky River Bridge segmentsare designed with large, single keys,the intent of which are to serve as erec-tion aids only. These keys are simplydesigned as corbels, carrying a verticalforce. This vertical force may be equal

to the weight of up to four segments, orsay, 200 tons (181 t) depending on thespeed of erection and rate of hardeningof the epoxy, which may be slow in lowtemperatures. Corbels for such highload carrying capacities must be care-fully designed. By using draped ten-dons, however, the shear forces actingon the corbels during erection can bereduced to zero, or even become nega-tive. This simplifies key design consid-e rably.

The use of an epoxy resin in thejoints is essential in the design of pre-cast segmental bridges. Durability andstructural adequacy are the most im-portant reasons for this. The epoxy jointhas been extensively tested and hasbeen found to perform very well to theextent that no weakness was observed.Good quality control on product andapplication, however, is required forobtaining good joints in the field. Someof the measures that can be taken to in-sure quality are:

1. Testing of the epoxy productsprior to selection, and also upon deliv-ery of each shipment.

2. Color coding of resin and hardenerto have a visual aide in mixing.

3. Training and supervision of themixing and application of the productin accordance with manufacturers' in-structions.

4. Daily verification of hardening.5. Spot checks on cores drilled

through joints of suspect quality.The occurrence of inferior quality

epoxied joints on this project wasminor, and where present, could beeasily and satisfactorily repaired. Qual-ity control was improved after the pub-lication of major problems on otherprojects. This improvement was not oneof cost, but rather of efforts born out ofgreater awareness.

Those responsible for inspection ofprojects like this, should be thoroughlyfamiliar with all aspects of the designand construction. This can be achievedbest by participation of the designers of

70

the structure in the inspection of theproject on a regular schedule.

Construction PhaseThe, bid opening, advertised for

September 15, 1977, was preceeded bya Pre-Bid Conference on August 4,1977. This allowed prospective biddersto ask questions and make suggestionswhich resulted in a number of minorchanges to the Contract Plans.

There were four bids from generalcontractors on the project, with propos-als furnished to the contractors fromseveral segment manufacturers, post-tensioning suppliers, and segmenterectors. All bids were over the en-gineer's estimate, with the major over-age being on the erection. It appearsthat all contractors (several boughtplans but did not bid) were concernedwith the possibility of dropping a seg-ment, its replacement cost and the timedelays effect on the constructionschedule. Other factors were the lim-

ited storage space at the site and theearly completion date. The proposalwas based on 300 working days, witherection to start early in the spring of1979 (the minimum temperature on theepoxy controlling the schedule) andcompletion of the structure scheduledfor December of 1979.

S. J. Groves and Sons Company ofMinneapolis, Minnesota, submitted thelow bid of $6,596,000 ($522,600 underthe next lowest hid) from theirSpringfield, Illinois District Office.After award of the contract, a number ofalternates were considered by the con-tractor, including placing their owncasting bed near the site and subcon-tracting the erecting. They chose to buythe segments, post-tensioning materialsand expertise from Construction Prod-ucts Corporation of Henderson, Ken-tucky, and do the erecting with theirown crews and equipment.

The contractor planned to erect bothsplines on the west side of the riverfirst, with a barge-mounted crane in the

r {

a e^

'•' l' ^..-..

Fig. 8. Early construction of long-line casting bed in Henderson, Kentucky.

PCI JOURNALJJuIy-August 1981 71

Fig. 9. First segments produced showing view of adjustable bulkhead.

center span and a crane on land for theend span. His first job was to site gradea work area between the relocated roadand the edge of water and drive steelpiles between the pier location and thetop of the bank. Steel pontoons ap-proximately 6 x 8 x 40 ft (1.83 x 2.44 x12.20 m) were mounted on the piles,forming a work platform along the cen-terline of structure, from which the piercould be constructed and later the seg-ments could be erected.

The pier footings are on extremelyhard limestone approximately 36 ft (11in) below the normal water surface. Thecontractor designed and fabricated afigure eight, two-level steel compres-sion ring from welded plate girders sal-vaged from a highway bridge. The de-tails were worked out so that the framecould be lifted up and floated from thewest pier to the east pier. The framewas held in place with six round po-sitioning piles.

Once it was in position and at theproper elevation, the sheet piles weredriven, the cofferdam dewatered andthe excavation completed. Before thepier footings were cast, and follow-

ing the basic erection concepts shownon the Contract Plans, the contractordesigned and detailed the temporarystruts to rest on the pier footings and tieto the columns. These were placed be-fore backfilling and flooding of thecofferdam. The struts would supportthe superstructure cantilever during theunbalanced phase of the segment erec-tion.

Working with small crews, the con-tractor completed the west pier andmost of the west abutment in the sum-mer of 1978. The cofferdam was movedto the east pier where the layout wasduplicated and construction continuedon the substructure. Concurrently withthis on-site work, the segment produc-tion was taking place. Figs. 9 and 10show how the first segments were fab-ricated.

Long-Line Bed is ChosenPrior to the bid, an analysis was made

on the cost of a long-line forming sys-tem vs. the more conventional short-line system. On a construction cost onlycomparison, both systems seemed to be

72

Fig. 10. First segments were produced by placing concrete with a bucket. Later theconcrete was pumped into place.

competitive. Upon award of a contractto supply the segments, several otheritems entered into the decision to pro-ceed with the long-line method.

1. The long-line bed in this case wasable to handle the entire structure dueto the fact that the "corrections" orcasting curves were similar for eachfree cantilever. This may not always bethe case, however. With the correctionsbeing similar, the casting of one soffiteliminated the very costly bottom car-riages which would have had to be ca-pable of adjusting to an 8-ft (2.44 in)differential segment depth.

2. Incorporated in the long-line bedwas a means of casting the pier segmentin between two adjacent segments (seeFigs. 7, 11, 12, and 13). Thus, casting ofthe pier segments (with the heavydiaphragm) was done simultaneouslywith the casting of the normal can-tilever segments, and in effect removedthe very difficult pier (and also theabutment) segments from the critical

Fig. 11. The pier segments were built withwood forms to remove them from thecritical production schedule.

Ar

:' 6

PCI JOURNAUJuly-August 1981 73

Fig. 12. A previously cast segment was brought back and positioned as in a short-linebed in order to cast the pier segment.

path of the production sequence. Thiswas extremely important on this projectdue to the relatively tight constructionschedule.

3. The manufacturer had alreadyproduced several bridges with theshort-line setup and was well aware ofthe system. It was felt that the directlabor costs of segment productionwould be less using the long-linemethod. The long-line method requiredthe positioning of the moveable steelbulkhead versus the exacting work re-quired of positioning the "old" segmenton a daily basis, which is required bythe short-line system.

Fabrication of SegmentsWith the long-line method justified, a

foundation was designed and installedin a short period of time (see Fig. 8),using elements regularly produced inthe manufacturing plant. The basicelements included prestressed concretepiling driven approximately 60 ft (18.3

m) to support 40 tons (36 t) each. Pre-stressed panels 4 ft (1.2 m) wide and 21ft (6.4 m) long spanned between eachbeam (column) line and provided pos-itive reinforcement and a bottom fonts.These were topped with 2 in. (51 mm)of structural concrete that became the"soffit" form of the superstructure.

While the foundation was being built,the segment steel forms were orderedand the details incorporated into thefoundation construction. Also duringthis period, the shop drawings werecompleted and reviewed by the con-sultant.

Fig. 14 shows a section of the castingbed where segments could be producedduring inclement weather. Fig. 15shows a segment being placed in thestorage yard.

Production began (in June, 1978)with casting of the segment adjacent tothe pier and proceeding towards theriver splice. When this run was com-pleted a second run was begun with thesegment adjacent to the pier and pro-

74

4.F ',r

Fig. 13. The heavy pier diaphragm was also formed with wood.

ceeding towards the tailspan splice.While the second run was being cast,the first segment cast was brought backto the bed and positioned so that thepier segment could be cast between itstwo adjacent segments. This procedureeliminated the pier segments from thecritical path of the casting operations.

The segment casting operation con-sisted of four basic operations:

1. Place reinforcing cage and relatedwork

2. Place steel bulkhead3. Place inner mandrel and close

side forms4. Cast concreteThe reinforcing bar cage was prefab-

ricated in a near-by building. It wasthen carted to the casting bed andplaced on the soffit with a crane. Thesteel bulkhead was then placed at apredetermined position on the soffit.The bulkhead had interchangeablepanels to account for varying segmentdepth. The inner mandrel was mountedon a frame that straddled the soffit and

utilized the same rail as the side formfor longitudinal movement.

The sides of the inner mandrel had a"slip" adjustment to vary with both adepth change from joint to joint and avarying bottom slab slope. The sideforms required only occasional adjust-ment and had such a capability. Theside forms traveled on a set of rails oneach side of the soffit.

A concrete pump was used to cast thesegments and this worked very effi-ciently. Casting time of a typical seg-ment was approximately one hour. Asthe segment casting preceeded alongthe soffit, the top slab was transverselypost-tensioned in a separate operation.Grouting of the transverse tendon ductswas done in the storage yard.

Transportation of the very tall seg-ments from the precast plant to the jobsite posed some special considerations.Due to restrictive overpass clearancesby land transport, 36 segments had tohe transported by barge on the Ohioand Kentucky Rivers to the job site {see

PCI JOURNAL/July-August 1981 75

Fig. 14. A work shed was placed over the bed so that production continued in badweather. Average production was over four segments per week until all 244 weremade.

Fig. 15. Once the length of the bed was produced the segments were stored in the yardto await delivery.

76

Fig. 16. A special trailer was built to transport the deeper segments.

r

Fig. 18. The pier segments arrived in Frankfort in April, 1979.

^ T a.

Figs. 17 and 18). Nine segments couldbe accommodated safely on the bargeand the trip took about 3 days depend-ing upon the river traffic.

Those. segments that ranged from 12to 14 ft (3.66 to 4.27 m) tall were trans-ported on a special trailer designed forthis project (see Fig. 16).

The special trailer supported thesegments under the cantilever wingadjacent to the web, allowing a segmentto be hauled with just 6 in. (152 mm) ofground clearance. Limited storage areason the job site necessitated close coor-dination between the contractor and theprecaste r.

Fig. 19. The first segment was placedMay 1, 1979. Note the vertical tie downrod on the side away from the secondsegment. The concrete risers (with theflat jacks) support the structure until theload is transferred to the pot bearingswhich are already in place.

78

Segment Erection 23). Pot hearings were grouted in placeErection of the superstructure seg-

ments was begun May 1, 1979, with thefirst of a pair of three-segment pierunits (see Figs. 19 and 20). Each seg-ment was joined to the previous seg-ment by means of an epoxy resin on thematch cast joint face and six temporarypost-tensioning bars (see Fig. 22). Thetemporary bars provided a uniformcompression on the epoxied joint face.

Due to the variable depth, and thusvariable section properties at each joint,the forces on the temporary post-tensioning bars were varied to keep theuniform compression of 50 psi (0.034MPa) minimum on the joint face. Per-manent post-tensioning consisted oftwelve '-in. (13 mm) diameter 270Kstrand and Freyssinet anchors. Most ofthe tendons were placed as individualstrands by a machine that pushed thestrand through the conduit.

Flat jacks were used to set the struc-ture to elevation and to grade (see Fig.

after the balanced cantilevers werelined up (see Fig. 24).

The unbalanced moments encoun-tered in erecting the free cantileverover 100 ft (30.5 in) from the piers werecontrolled by a strut assembly whichwas placed on the abutment side (seeFig. 25). Segments were then hung onthe abutment side of the cantilever first,alternating sides until the closuresplices were reached. Several segmentsin the tailspans were erected onfalsework, from the abutment to theclosure splice to the abutment.

Both splines were erected concur-rently (for minimum movement of ma-terials and equipment) with the westside (Pier 1) being placed first. Whilethe land crane placed the segments onfalsework (these went very quickly), theriver crane erected the first segments atPier 2.

Erection on the east side proceededmuch faster, with the crews trained and

PCI JOURNAUJuIy-August 1981 79

Fig. 21, The hydraulic jack, shown here pulling draped tendons in the web wall,tensioned twelve 1/2-in. (13 mm) diameter 270K strands and had a piston to seat theanchorage cone.

Fig. 22. High strength steel rods were used to hold segments in place until tendonswere placed and stressed.

80

Fig, 23. Flat jacks were used to set the structure to elevation and to grade.

Em

r

Fig. 24. The pot bearings were grouted in place after the balanced cantilevers werelined up.

PCI JOURNAL/July-August 1981 81

Fig. 25. An erection strut was placed in each end span to stabilize the cantilevers.

Fig. 26. The cast-in-place splices were formed and supported on the superstructure.

82

Fig. 27. The barrier walls were formed and cast in place in the late fall of 1979.

Fig. 28. The final cleanup just prior to final completion of the structure.

PCI JOURNALJJuly-August 1981 83

working a regular routine. Four seg-ments (two each spline) were the nor-mal schedule. On several days six seg-ments were erected and eight wouldhave been possible except for the un-certain delivery schedule caused by thetruckers' strike.

With erection crews on the east side,carpenter crews formed the closure

splice in the west end span (see Fig.26). These were cast in place and thepositive moment tendons were placedand tensioned. Then the east end spansplices and last the midspan spliceswere cast in place.

Barrier walls were formed and cast inplace in the late fall of 1979 (see Figs.27 and 28).

CLOSING REMARKS

All those who had a part in the designand construction of the Kentucky RiverBridge are very pleased with the finalresults. Fig. 29 shows an overall shot ofthe finished structure.

Through the entire project, there wasa spirt of cooperation as if everyonerealized that it was only by working to-gether that the structure could be com-pleted on schedule.

The Kentucky Department of Trans-portation, Division of Construction, wasresponsible for all plant and on-site in-spection, with all decisions made bythem. However, the Consultant Agree-ment included design, checking shopand erection drawings and a per diemarrangement for superstructure inspec-tion.

Thus for a small cost, they could havethe advice of experts who were familiarwith the design and construction pro-cedures. As erection got into full swingthese inspections were scheduled on aregular basis.

The consultants were on top of eachcrisis as it developed and were able towork with the Kentucky Department ofTransportation, the segment supplier,and the contractor in finding a satisfac-tory solution to the problem.

It is believed that four "first time inthe United States" features are in-cluded in this project. The 323-ft (98.5m) center span is the longest span todate to be built of precast segments

erected in free cantilever, althoughseveral longer spans are under con-struction or have been designed. Theskewed pier arrangement is unique tothis structure. This is the first precastvariable depth structure, althoughnumerous cast-in-place structures havethis feature. The long-line bed has notbeen used before.

Certainly, all of these features wereproved to be feasible on this project.The longer spans and variable depthhave already been included in otherprojects under construction. The long-line bed proved to be a fast, efficientand economical method of productionwhich will become common practice inthe future. The skewed piers andcast-in-place splices are expensive, andwe will eliminate them wherever pos-sible on future jobs.

Access to the inside of the abutmentsand box girders was provided for in-spection. In fact, a lighting system wasinstalled for this purpose. The Ken-tucky Department of Transportationplans to watch the structure closely butno provisions were made for a check ofthe stresses on the bridge.

The structure appears to be func-tioning properly. In the last 20 monthsthere has been no maintenance re-quired on the project. The increasinguse of this highway indicates that theproject was much needed and thetimely completion was justified.

84

Fig. 29. The finished structure with the State Capitol buildings in the background.

CREDITSConsulting Engineer: American Engineering Company,

Consulting Engineers, Lexington, Kentucky.

Segmental Consultant: BVN-STS Consulting Engineers,Indianiapolis, Indiana.

General Contractor: S. J. Groves and Sons Company,Springfield, Illinois.

Precast Concrete Manufacturer: Construction ProductsCorporation, Prestressed Concrete Division, Henderson,Kentucky.

Owner: Bureau of Highways, Division of Bridges, KentuckyDepartment of Transportation, Frankfort, Kentucky.

PCI JOURNAL/July-August 1981 85