Special Report

Segmental Bridge Constructionin Florida — A Review

and Perspectiveby

Alan J. Moreton, P.E.State Structures EngineerFlorida Department of TransportationTallahassee, Florida

36

SYNOPSISThis paper offers an overview of the precast concretesegmental bridges designed and built in the state of Floridaduring the last ten years. The article summarizes variousstatistical structural parameters, segment manufacturing anderection methods, construction times, costs, and reviewsproblems typically encountered. Also included is a discussionof current industry and nationwide design and constructionpractices and some suggestions for possible improvements.

CONTENTS

Synopsis............................................37

1. Introduction ......................................38

2. Precast Segmental Bridges ........................38

3. Florida's Segmental Bridges .......................44

4. Structural Parametrics .............................44

5. Casting Yard Operations ...........................48

6. Rejected Segments .............................. 49

7. Erection Operations ...............................50

8. Some Typical Problems ...........................52

9. Time ............................................55

10. Costs ...........................................57

11. Administration Processes — Design, Construction

andShop Drawings ...............................60

12. Actions by the Florida Department of Transportation ...63

13. Benefits of Segmental Bridges ......................64

14. Summary ........................................65

15. Conclusions .....................................66

References..........................................66

Acknowledgments....................................67

Appendix............................................68

PCI JOURNAUMay-June 1989 37

1. INTRODUCTION

M ode 01 concrete segmental bridgeswere made possible as a result of

postwar developments in post-tension-ing systems and materials technology.This was spurred on by the need formuch reconstruction and new infra-structure in Europe following WorldWar 11. After some initial developmentin J urope, these systems were intro-duced into North America in the '60sand '70s and have now become quitecommonplace.

Early post-tensioned concrete seg-mental bridges were cast-in-place incantilever using traveling formworkwith spans of up to 400 ft (122 m.). Thisremains the preferred method for largespans such as the Houston Ship Channel[at 750 ft (229 in)] and the planned con-crete alternate for the Acosta Bridge inJacksonville [at 630 ft (192 m)1.

Precast segmental bridges were a nat-ural development for efficiency, stan-dardized mass production, speed oferection, the elimination of expensiveformwork in deep valleys and overnavigable waterways and, particularly,to afford solutions for restricted con-strnction access in congested urban orenvironmentall y sensitive areas. Themost notable example of the latter is theLinn Cove Viaduct' on Grandfather17ountain in North Carolina. Thisbridge was constructed entirely From thetop, including piers and foundations, inorder to preserve the delicate environ-ment of this scenic region. It is the "pre-cast type of segmental bridge that hasfound many applications in Florida overthe last 10 years. So far, 31 major struc-tures have been built, including theSunshine Skyway Bridge.

2. PRECAST SEGMENTAL BRIDGES

Precast segmental bridges are socalled because they are made of indi-vidual precast units or "segments"carefully manufactured in a precast con-crete plant, either on or off the site. Thesegments are later erected and securedtogether by longitudinal post-tensioningto form each span or cantilever.

They fall into the following generalcategories according to their method ofconstruction:

(a) Span-by-Span — Where all thesegments of'one span between piers areerected on a special supporting truss organtry and are longitudinally post-ten-sioned together after making smallcast-in-place closure joints at one orboth ends next to the pier segments.Examples include the Long Key andSeven Mile Bridges in the Florida Keys(Figs. 1 and 2) 2.3 and the high level ap-proaches to the new Sunshine Skyway

Bridge over '1 ampa Bay.'.s(h) Balanced Cantilever — Where

segments are erected sequentially incantilever on each side of an initial seg-ment placed on top of the pier. Stabilityis provided by a temporary tower orother support near the permanent pier.Cantilevers are joined by cast-in-placeclosures at niidspan. There are manyexamples of this throughout the state,including Ramp I at the Florida "l urn-pike/1-75 Interchange (Fig. 3).8

(c) Progressive Cantilever - Wheresegments are erected in cantilever inone direction, starting at one end of thebridge and progressing overall the piersin sequence. Additional intermediatetemporary piers or towers with cablestays are needed to facilitate construc-tion in cantilever from one pier to thenext. Examples include the Linn CoveViaduct in North Carolina and the Fon-

38

Fig. 1. Long Key Bridge — A precast segmental, post-tensioned, span-by-spanstructure. (Courtesy of Figg and Muller Engineers Inc., Designer)

PC i JOURNAL/May-June 1989 39

M

Fig. 2. Seven Mile Bridge — A precast segmental, post-tensioned, span-by-spanstructure. (Courtesy of Figg and Muller Engineers Inc.. Designer)

tenoy Bridge in France.7In all these systems, the precast seg-

ments are usually made in a special formor "casting cell" where a new segmentis cast against its older neighbor toachieve a perfectly mating or "snatchcast" joint. To date, only the "short line"or "single cell" casting machine hasbeen used (as opposed to the "long linebed" system) in Florida. In the long linebed system, the entire soffit of thebridge is laid out in the casting yard andeach segment is made in turn in itsproper place.

When erected in the bridge, the jointsbetween segments are coated withepoxy to fill any surface imperfectionsand provide a tightly bedded joint. Thisalso helps to seal and protect internalpost-tensioning tendons. Because exter-nal post-tensioning tendons were usedin the first span-by-span bridges in theFlorida Keys, the segments were notjointed with epoxy but were left dry.However, all subsequent structures, in-chiding the similar span-by-span ap-proaches to the Skyway, have epoxy fil-led joints between segments.

Temporary post-tensioning bars areused to secure each segment tightly toits neighbor prior to installing andstressing the permanent longitudinalpost-tensioning tendons which providethe structural capacity and continuity ofthe superstructure. Schematic illustra-tions of typical span-by-span and bal-anced cantilever construction areshown iii Figs. 4 through 8.

Other types of bridge constructionsystems are sometimes referred to as"segmental." These include:

• Incremental launching• Partial precast and cast-in-place

cousin iction• Post-tensioned segmental I-girders• Cast-in-place post-tensioned

bridges• Precast wet joint segmentalThese systems share many common

features, especially post-tensioning andspecial erection systems of falsework,towers or travelers, etc. However, theyare different in techniques from precastsegmental bridges and, with the excep-tion of post-tensioned I-girders, havefound few applications in Florida.

40

Fig. 3. Ramp I — A precast segmental, post-tensioned, balanced cantilever viaduct.(Courtesy of Beiswenger Hoch and Associates, Designer)

continuous deck

P.T. at pier P.T.,_ at midspan

Section

Fig. 4. Typical span-by-span superstructure.

PCI JOURNALMay-June 1989 41

o - :-r--

L.driiied Shafts

C.I.P. footer

V-pier column pier = =_ -

4 drilled shafts-

precast box pier

Fig. 5. Typical span-by-span substructures.

P.T.at pier T. at midspan

Section

Fig. 6. Typical balanced cantilever superstructure.

42

r C.I.P. joint

^caat-In-place piers

__

(

pler with features -

driven piles or spread footings

II

Fig. 7. Typical balanced cantilever substructures.

Span-by-Span anchorage

external P.T. deviation saddle

Balanced Cantilever cantilever P.T.

Ifl nsr I ii /1

mldspan closure ) continuity P.T. top and btm.

Fig. 8. Typical post-tensioning.

PCI JOURNALMay-June 1989 43

3. FLORIDA'S SEGMENTAL BRIDGES

The Long Key Bridge was the firstprecast segmental bridge constructed inFlorida; since then, a total of 31 bridgeshave been built, including the cablestayed Sunshine Skyway Bridge overTampa Bay (Fig. 9). Several more havebeen and are being designed. (The Sun-shine Skyway is an exceptional structurein its own right and will not be dis-cussed in detail here. However, certaininformation and experience from theSkyway project have been incor-porated — particularly that which per-tains to the high level span-by-span ap-proaches which are typical of othersegmental, rather than cable stayed,construction. Also, the Sunshine Sky-way main span unit incorporates bal-anced cantilever spans of 240 ft (73 m)on either side of the cable stayed sec-tions. This is a somewhat unique appli-cation which would not necessarily re-flect costs or other data typical of a nor-mal cantilever of this span.j

Generally, Florida's segmentalbridges are either span-by-span or bal-anced cantilever. Typical features arehighlighted in Table 1. In general,span-by-span construction has been

used only on straight structures, overwater with spans between 118 and 143 ft(36 and 44 m). For larger spans up to 225ft (69 m) and especially for curved inter-change viaducts on land sites, balancedcantilever has been used.

Substructure and foundation typesfollow a similar trend, with lighter pre-cast and cast-in-place substructures inspan-by-span applications, as opposedto more massive cast-in-place substruc-tures required by balanced cantileverconstruction. The difference arises fromthe span lengths and the fact thatcantilever construction usually requiressignificant out-of-balance erection ef-fects be carried into the foundations.There is no out-of-balance effect withspan-by-span construction, so the foun-dation loads and moments are muchless. Also, frequent economic use hasbeen made of drilled shafts in most ofthe span-by-span structures as opposedto more commonl y used driven piles.Although partly dictated by site andground conditions, either foundation sys-tem could have been used. This also re-flects the different philosophies of desig-ners.

4. STRUCTURAL PARAMETRICS

Structural economy in materials andefficiency has been achieved in seg-mental construction from the funda-mental principle of continuity in super-structures as opposed to traditional sim-ple span girders. Continuity permits ageneral reduction in structural deadload with savings in substructures andfoundations, particularly in span-by-span systems.

The use of continuous construction as ameans to structural efficiency and econ-omy is also applied in other structuralsystems such as steel plate and box gir-

ders and the newly developed Floridabulb tee precast post-tensioned girdersystem, recently introduced on the EauGalliee and Howard Frankland Bridges,

Including the Skyway approaches, fivespan-by-span structures, totaling over2,600,000 sq ft (241,000 i2 ) of bridgedeck, have been built. Twenty-sixbalanced cantilever bridges, represent-ing another 1,400,000 sq ft (130,000 m2),have been built. These were built using7500 precast deck segments and severalhundred precast pier segments (Tables2, 3 and 4).

44

Fig. 9. Sunshine Skyway Bridge — A precast segmental, post-tensioned, cable stayedcantilever bridge with span-by-span high level approaches. (Courtesy of Figg and MullerEngineers Inc., Designer)

Table 1. Precast segmental projects.

No. of Balanced Span-Projects Fridges cantilever by-span Principal features

Long Key 1 • Span-by-spanSeven Mile 1 • Over waterNiles Channel 1 • StraightChannel Five 1 • Precast/cast-in-placeRamp 1 1 • substructuresSkyway Approaches 1 •

builtMostly drilled shafts

17511595-1 5 • 5 built = 2.6 M4 sq ftPalmetto 5 • 1 bid not builtAirport 4 • 1 in design44111595-M 27511595-2 9 • Balanced cantilever

Interchange sitesI95/1595-DEF 8 a 11 hi d but CurveddSouth Fork New River 2 • not built

Cast-in-placeHoward Frankland 1 • • substructures

Mostly driven pilesPort of Miami 1 • 26 built = 1.4 M sq ftEdison 1 • 1 a design lU bid not builtGolden Glades n • Several in design

PCI JOURNAL/May -June 1989 45

Table 2. Superstructure parametrics.

P:rrrnrretersSpan-

by-stunBalanced

cantilever

Span ranges, f} 118-1.43 71-224Box depth, ft 7-8 6.5-9.3Segrnent weight, tons

Typical 64- 75 38-58Pier 40-83 48-98Ahutmentlexpansion joint 34 - 52 25-72

Equivalent solid concretethickness, ft 1.3 - 1.6 1.6-2.0Reinforcing liars, psf 4.5-9.4 8.4- 18,(1Transverse post-tensioning, psi 0.5 - 0.8 0.6 - 1.1Longitudinal post-tensioning, psi 1.5.2.8 3.6 - 5.5

Metric(Sl)convertiion Frrinrs:3.28tt= I.iiI)rn; 1 tort(l1S)= 0.91 hour;i psf = 489 kgmrr.

Table 3. Substructure parametrics.

ParameterSpan-

by-spanBalancedcantilever

Span ranges, ft 118-143 71-224Height ranges, ft 20-152 24-97

Ratio of so lid substructure (percent) 0.94 2.02spanucd void

Reinforcing bars in substructure,lb per cu yd 135 154

Post-tensioning in substructure,lb per cu yd 7.3 t)

PoundationsDrilled shafts, percent 84 0Driven piles. percent 16 80Spread footings, percent 0 20

Metric (SO conversion factor,: 3.214 ft = I m: I II •rru yd - ().59 kit' iii'.

In span-by-span construction, mostdecks are 38 to 43 ft (11.6 to 13.1 m)wide for two lanes of' traffic with shoul-ders and harriers. Superstructure seg-ments are typically 18 11 (5.5 nm) long and7to8ft(2.1to2.4 in) deep lbr spans from118 to 143 ft (36 to 44 m) and weighabout 65 tons (59 tonnes) each,

The average equivalent solid concretedeck thickness is about 1.4 It (426 nlnl)and the weight of reinforcing bars isbetween 4.5 and 9.41 psf (22 and 46

kg/m 2 ), depending upon whether trans-verse post-tensioning is used or not.However, higher quantities of' rein-Forcement in more recent projects are areflection ref' more stringent design cri-teria introduced by Florida's Depart-ment of Transportation in 1984.

When used, transverse post-tensioning averages 0.70 psf (3.42kg/rn 2 ). Longitudinal post-tensioningranges from 1.5 to 2.8 psf (7.3 to 13.7kghn2). This is primarily a function of

46

Table 4. Casting operations.

Structure

Number ofsuperstructure

segments__—

Number ofcatstingcclls

Average sustainedproduction rate,*

segments per weektypical Other

LongKey 734 3 1 17Seven Mile 2154 5 1 284ilesChannel 276 3 1 17Channel Five 299 3 1 17Ramp 1 201 2 1 6Skyway Approaches 584 2 1 10Skyway M 5U 333 2 1 5175/1595-1 567 2 1 10Palmetto 658 5 2 24Airport 286 2 1 12US 44111595 385 2 1 11I75/1595-2 1316 7 3 32

After mobil bait iou.cei ieanrir i g period.

Fig, 10. Seven Mile Bridge— Precast vertically post-tensioned, segmentalbox piers.

the span lengths but also reflects a littlemore conservatism in later structures.

In balanced cantilever construction,the segment widths are typically around42 ft 9 in. (13.03 m) at the top slab forsingle boxes with two lanes, shouldersand barriers, and up to 2 x 28 ft (2 x 6.56m) for twin boxes in wider bridges (Fig.4). Segment weights range from 25 to 98tons (23 to 89 tonnes) but are typically50 to 60 tons (45 to 55 tonnes). Equiva-lent solid concrete deck thicknesses

range from 1.56 to 2.04 ft (475 to 622mm), although this does not necessarilymatch spans which range from 120 to225 ft (36.6 to 68.6 in).

Reinforcing bar steel varies from 8.4 to18.0 psf (41.1 to 88.0 kg/rn s ) with a typi-cal range from 9 to 14 psf (44.0 to 68.5kglm 2 ). Transverse post-tensioningtypically averages about 0.85 psf (4.2kg/m 2)• Longitudinal post-tensioningamounts vary from 3.6 to 5.5 psf (17.6 to26.9 kg/m 2 ). The variations in reinlorc-

F'CIJOURNAUMay-June 1989 47

ing bar and post-tensioning quantitiesreflect both structural requirements forthe spans and more conservatism in theFDOT design criteria of 1984. Also,some structural depths were restrictedby highway clearances dictating less ef^ficient sections, thus requiring high rein-forcing bar and post-tensioning content.

A review of substructure data showsclearly how much lighter substructuresare in general for span-by-span as op-posed to cantilever construction — for

reasons discussed in Section 3. Com-pare, for example, the average ratio ofsolid substructure to spanned void vol-nines ibr span-by-span at about 0.94 per-cent and cantilever construction at about2.0 percent, i.e., almost a 1:2 variation.

Vertically post-tensioned substruc-tures have only been used on precastbox piers to date, for the Seven Mile(Fig. 10), Channel Five and SkywayBridges. These are typically efficient,high level span-by-span strictures,

5. CASTING YARD OPERATIONS

On all projects but one, the primecontractor elected to produce the seg-ments himself by establishing a precastyard at or near the site. In only one case,that of the Seven Mile Bridge, was theprime contractor already in the precastbusiness. This contractor had his ownproduction facilities in Tampa Bay, from

where the segments were barged to theFlorida Keys.

Casting facilities have been geared tothe overall size of the project and con-tract period on an anticipated peak pro-duction rate of one typical segment perday per casting cell and one pier orabutment type segment every 2 or 3

Ie

Fig. 11. Pier segment in casting cell — 175/1595 Phase 2 interchange.

48

days, In general, these rates have beenachieved.

The normal complement of castingmachines was two to four cells for typi-cal segments and one for pier segments(Fig. 11). The expansion joint, abutmentor other nonstandard segments wereusuall y made by adapting a typicalcasting cell with wooden bulkheads orsimilar formwork.

Generally, it would require about 3to 4 months to establish the casting yardfacilities. This was followed by a fewweeks of learning by the crews andmodifications and improvement to thefacilities until a consistent productionrate was achieved.

It should be emphasized that projectshave run considerably smoother wherethe contractor engaged experiencedpersonnel or sought advice in the plan-ning and acquisition of his castingfacilities. Also, this was especially truewhen good quality forms were used.

Concrete has been produced by batchplants at the casting yard and/or deliv-ered by truck, depending upon avail-ability and prices in the locality. Seg-ments have rarely, if ever, been lost clueto failure of concrete production anddelivery.

Travel lifts have been most efficientfor handling segments in the castingyard. Segments have been lifted eitherby special frames attached by threadbars through the top slab, C-frames, or

6. REJECTED

Experience shows that success insegmental bridge construction dependsalmost entirely upon the casting opera-tions.

Attention to workable and construct-ible details, good planning of castingwork, and quality workmanship paydividends. This is illustrated in Table 5,which lists rejected segments and suM-

PCI JOURNAL'May-June 1989

Table 5. Rejected segments.

ProjectN umber

lostPercent

losSummary of reasons for rejecting segments:

. ' ods/honeycombing*• Post-tensioning ducts displacedLong Key 4 0.5

Seven Mile 3 0.2 • Heinfbrcing bar post-tensioning conflicts*NilesChannel 0 0 • Weakfonns*Channel Five 1 0.4 • Dimensional tolerancesRamp 1 1 0.5 • Geometric alignment controlSkyway Approach 3 0.5 • Improper handling/storageSkyway Main Spans 1 0.3 • Curinglthermal1751I595-1 18 3.2 • Low strengthPalmetto 19 2.9 • Damaged shear keysAirport 0 {) • AccidentUS 44111595 1 0.3 • Weather175/1595-2 3 0.3 *\lajority of lossr¼ forthe,e re,isu1is.

7. ERECTION OPERATIONS

Various erection systems have beenused in Florida (Table 6). These havefallen into the broad categories ofground lased crane for cantilever erec-tion at interchange sites and truss or

Fig. 13. Overhead gantry for span-by-spanerection at Seven Mile Bridge.

gantry for span-by-span erection overwater. The latter was the first methodintroduced with the construction of theKeys Bridges. An underslung truss wasused fir Long Key, Niles Channel andChannel Five, and an overhead gantryfor Seven Mile (Fig. 13).

This technique was subsequentlyused for the Skyway approaches. Sinceall cantilever structures except the mainspan unit of the Skyway have been overland, regular cranes have been the mostsuitable. The balanced cantilevers of theSkyway main span unit were erected bybeam and winch devices with stabilitybeing provided by steel girders span-ning to the previously erected structure.These devices were also used for themain span segment erection (Fig. 14).An incidental use of the beam andwinch system was required on the 175/1595 Phase 2 project to erect some seg-ments underneath an existing higherlevel bridge.

Rates of erection have been geared tothe project size and contract duration.After a few weeks of learning period,span-by-span construction typically ranat three spans per week and in somecases, at Long Key and Seven Mile,

50

Fig. 14. Main span segment erection at Sunshine Skyway using beam and winchequipment.

Table 6. Erection methods.

Parameter Span-by-span Cantilever

Site Water LandDelivery Barge LowboyErection equipment

Truss plus floating crane 33 percent*Overhead gantry 67 percent*Crane on ground 91 percent*Beath and winch 9 percenttStability towers Not required UsedFalsework Not required Used

Learning period 3 to 4 weeks VariesSustained rate 3 spans per week 4 to 6 segments

perday per cantilever(2 cranes)

Iii, percentage is !rased upon tli total nun herol segments erected per bridgetype.

!'his figure includes the segments Lii the Sunshine Skyway main spans.

PCI JOURNAL'May-June 1989 51

Fig. 15. Balanced cantilevers under construction at US 441/1595 Interchange. (Designer:Greiner Engineering Sciences Inc.)

achieved one span per day, i.e., up tofive spans in one week.

Balanced cantilever erection rateshave varied widely. On at least twoprojects, Ranip I and US 441/I595, therate of erection was not so critical tooverall completion and it proceededwell, on time and with very few minorcomplications. Other cantilever projectsencountered slow learning periods anddifficulties through poorly alignedpost-tensioning ducts, geometric align-ment control, etc. However, on more re-cent projects, such difficulties were

avoided and erection progressed rapidly,particularly at the US 441/1595 and I75/1595 Phase 2 interchanges.

The former project involves two verylong curved segmental cantilever via-ducts and is approximately 6 monthsahead of schedule (Fig. 15). On the lat-ter project, which involves nine similarsegmental bridges, the contractorachieved a substantial incentive bonusfor opening a section ahead of schedule.At the time oIthis writing, he is likely tofinish the entire interchange 9 monthsahead of schedule,

8. SOME TYPICAL PROBLEMS

The following problems have beenencountered to some extent at varioustimes. Fortunately, their recurrence isfar less likely today. While the followingproblems are cited, it must he remem-bered that they are not exclusive tosegmental construction, Many of these

(and other) problems occur in othertypes of construction. In the construc-tion industry as a whole, problems arenot unusual and they are routinely re-solved.

Honeycombing — This occasionallyoccurred in web walls and congested

52

reinforcing bar zones but was usuallycosmetic and easily repaired by cuttingback to sound concrete and filling withhigh strength, nonshrink, cement mortaror fine aggregate concrete. Only on rareoccasions was it so severe as to requireengineering analysis and/or total rejec-tion of the segment.

Damaged shear keys — Occasionaldamage or loss of whole keys mightoccur during stripping or handling.Usually, this happened with new crewsearly in the projects, and it was avoidedby taking more care. In the KeysBridges, the loss of two or three shearke ys was determined not to be detri-mental to erection and they were re-paired by dry packing after erection butbefore post-tensioning.

Top riding surface finish — A fairlygood riding surface was achieved in allthe Keys Bridges as a result of accurateworkmanship in finishing the top sur-face in the casting yard. However, somesubsequent cantilever bridges had less-er quality and required grinding. Thepoor quality surface was due to im-proper attention to finishing work in thecasting cell. A good quality riding sur-face has been achieved on both themajor projects nearing completion atUS4411I595 and at 175/1595 Phase 2. Arotary screed (Fig. 16) was used at theformer project and a straight vibratoryscreed at the latter project (Fig, 17), Inboth cases, the screed was followed bythe use of a straightedge and a light ap-plication of bull floats. Skilled concretegangs were employed on both of theseprojects.

Concrete materials — Generally,these have been satisfactory and therehave been no more problems with qual-ity control than with other methods ofconstruction.

Misaligned post-tensioning ducts —Problems with misalignments have not

arisen in span-by-span constructionusing external tendons passing throughdeviation saddles and anchor zones.There have been some problems in

cantilever construction which causedexcessive friction and consequentl y, re-duced elongations or, in the worst cases,wire breakage. These were attributedto:

— Inadequate duct supports, stiffen-ers and seals

— Inaccurate placement of ducts—Trapping ducts between reinforc-

ing bars— Damage from concrete placement

and consolidation— Detailing too tightBlocked post-tensioning ducts -

These were clue to either inadequateseals against cement grout during con-creting or from crossflow of post-tensioning grout after stressing. (Bothare avoidable with care and attention toworkmanship.)

Handling — In the Keys, a few boxpier segments were lost when a liftingcable failed and one segment fell onothers in storage in the casting yard. Amain span segment of the Skyway waslost due to a failure of the gantry.

Stacking and storage — Improperdouble stacking of a pair of segments forSeven Mile caused cracking and the re-jection of one of them. The problem wascaused by uneven settlement. Currentpractice permits double stacking undercontrolled conditions using a three pointsupport, two tinder one web and one inthe center of the other web. An analysisof the loaded segment must show ac-ceptable stresses.

Cracking — Occasionally cracking hasoccurred due to curing temperatures orshrinkage. Cracks of a structural typehave sometimes occurred for whichthere was usually an explanation. Forexample:

— Spalls due to concentrated localpost-tensioning effect from anchor-ages

— Flexural cracks due to inadequatebond on epoxy coated reintorcingbars

Epoxy coated reinforcing bars —These have exhibited an inability to

PCI JOURNALMay-June 1989 53

Fig. 16. Use of a rotary screed, followed by straightedge and light application of bull floatsfor a good finish.

Fig. 17. Use of a vibratory screed, followed by straightedge and light application of bullfloats for a good finish.

.ri4

bond to concrete and control shrinkageand flexural strains between the rein-forcing bars and the concrete. (Evidencefor this comes from field experience andobservations on various structures, somebeing segmental bridges since epoxybars were first used in these. This alsoinvolves questions about the adequacyof the corrosion protection, which is stillunder investigation by the Departmentof Transportation and beyond the scopeof this paper.)

Alignment and cambers —Therehave been no difficulties with align-ments and cambers in span-by-spanconstruction. Horizontal and verticalalignment errors were encountered, butsatisfactorily resolved, in some canti-lever construction. On one occasion thecontractor made the last two closures outof sequence, resulting in a cusp in thevertical profile.

Weak forms — Some forms could notwithstand concreting operations andgave way at joints, causing voids andother problems. This was cause for re-jecting several segments on one projectand the problem was solved only bystrengthening the forms. Forms must berobust and able to withstand concretepressures and much abuse; joints shouldbe good with reliable, tight seals. Weak,flexible forms result in cement pasteleakage, honeycombing, bulging andgeneral loss of tolerance.

Slippage of post-tensioningwedges — This has occurred occa-sionally with some systems. it has been

readily rectified by using differentwedges and/or changing the post-ten-sioning jacks.

Detailing — Many difficulties havearisen from inadequate allowances forreinforcing bar sizes, bending andplacement of tolerances, conflicts be-tween reinforcing bars and post-ten-sioning ducts, and so on. These comefrom many sources relating to basic de-sign detailing, shop drawing prepara-tion, reviews, and general coordinationor lack thereof.

As a general observation, many of theabove-mentioned problems have alsobeen encountered in other types of con-strtiction. Examples include honey-combing, misaligned and blockedpost-tensioning ducts, miscellaneouscracking, excessive camber growth,post-tensioning wedge slippage, inade-quate tolerances, sweep and warp ingirders, etc.

These and similar problems are a factof life in all types of construction and willcontinue to be dealt with as a matter ofroutine. Experience shows that a "prob-lem" becomes much more severe if indi-viduals and organizations are not ade-quately prepared and experienced to re-solve the particular problem as soon as itarises.

By their nature, segmental bridges in-volve tile use of precast concrete andpost-tensioning operations. Conse-quently, problems like those describedhere can be anticipated and — by careand forethought — avoided.

9. TIME

The initial introduction of segmentalconstruction with the span-by-spansystems in the Florida Keys achievedsome quite remarkable rates of progressand span erection. Reports were com-mon of two, three and occasionally morespans being erected in one week usingthis method. The factory-style quality

control of the precasting also paid off inthe quality of the final riding surface.

Introduction of the cantilever type ofstructure brought some unfortunate timedelays with some of the first projectsconstructed. These delays arose primar-ily out of problems due to workmanship,inadequate equipment and lack of at-

PCI JOURNALiMay-June 1989 55

Fig. 18. Segmental cantilever bridges under construction at 175x'1595 Phase 2 project.

(Contractor: Harbert Westbrook Joint Venture. Designer: Beiswenger, Hoch andAssociates)

Fig. 19. 175/1595 Phase 2 project under construction.

56

Table 7. Construction time summary.

BridgeContract

time Comments

Lung Key 128 Bridge on time*Seven Mile 84 6 months aheadNiles Channel 135 Bridge on time/aheadChannel Five 100 On tirneRamp 1 100 On timeSkyway Approaches 105 Approaches on time — Late on main spans175/1595-1 135 LatePalmetto 117 LateAirport 213 LateUS 441/1595 — On timelahead1751[595-2 — 2 months ahead at an interim deadline --

9 months ahead at completion

"Approach area was casting ti•ard for Long K. Niles Channel and Channel Five Bridges.

tention to detail. Each problem causedsome delay which spread over thewhole construction period. Unfortun-ately, most problems could have beenaverted by proper care and attention.Mostly, they can be attributed to inex-perience.

This situation changed significantlyfor the better with the latest projectsnearing completion on the I595;namely, US 441/1595 and 175/I595Phase 2 (Figs. 18 and 19). In both cases,an experienced contractor used well-trained crews under proper supervision.The design and shop drawing reviewswere also done under experiencedgroups.

The net result was a considerable im-

provement in all respects. There werevery few problems of the sort whichtroubled previous cantilever segmentalprojects.

A summary of construction times andcomments is presented in Table 7. Ingeneral, projects ran better with experi-enced contractors and supervisiongroups.

Shop drawing preparation and re-views influenced time on some seg-mental projects, Improvements havebeen made in the whole process formore recent projects. Consideration isbeing given to future changes aimed atsimplifying and reducing much re-peated effort, especially for individualsegment drawings.

10. COSTS

Contract dates and total project andbridge bid costs are shown in Table 8. Adetailed breakdown of average unitcosts and square foot prices, correctedfor inflation to 1987, are given in Table9. For the data available and within thevagaries of inflation, segmental bridgesaverage $44 to $52 per sq ft (8474 to$560 per sq m) at 1987 prices.

It is interesting that on two recentinterchange projects, which includedbridge alternates in segmental concreteand steel, namely the I95/1595 Inter-change and South Fork New River proj-ects, the segmental bridges were lowerin price than the steel. 'These projectswere large and bridge constructionamounted to only one-third of each total

PCI JOURNAL/May-June 1989 57

Table 8. Project bids.

Bridt;c Construction

Totalproject

Smillion

Segmentalportion$million

BuiltLongKey 1179-7181 15.1 14.5Seven Mile 10179-7182 43.4 43.0Niles Channel 4181 - 4/83 7.9 7.7Channel Five 5181 - 1183 10.4 9.0Ramp! 2182-5184 22.3 4.6Skyway Approaches 4183- 10/87 71.117511595-1 7183 -11/87 10.2 8.8Palmetto 3184 - 9186 9.4 7.8Airport 10184 - 12187 7.8 5.3US 441/1595 10/86- 60.2 11.9175/1595-2 4/87- 51.1 27.1

Not built195/1595-IIEF 6187- (104) 31.2*South Fork New River 6187 - (60) 19.8*Howard Frankland 8/87 - (44.9)1 (46.5)

'Segmental prices of next low I al given. Ii Ftal project low Lid ill(/ tided forbuilding steel alternate at H percent higher than segnirntal.t Bulb tee wax law I)id by 3.5 tx'rre•nt.

Table 9 Cost breakdown.

Bridge

Surfaceareasgft

(1000s)

Cost8 persgft

(1987)

1987 Prices

Span-by-spa's avg $44 per sq ftCantilever avg $52 per sq ftLong Key 468 38

Seven Mile 1376 34 Substructure concrete 8254 per en ydNiles Channel 176 50 Superstructure segments 8456 per cu ydChannel Five 190 55Ramp I 91 61 Post-tensioning $1.1.6 per lbSkyway Approaches 416 — Beintorcirtg bars `b0.42 per lbI75/1595-1 252 4148Palmetto 198 48Airport 125 68US 441/159517511595-2

177537

5056 Sletric (SD conversion factors:

195![595-DEF 561 10.76 sq ft = I m';1 to }-d = 0.765 ms; Jib = 0.454 kg.

South Fork 46New River 433 44

Ilowartl Fraiikland 1060

58

Table 10. Average bridge costs in Florida.*

Types ofbridges

Percentage ofall bridge

construction $ per sq ftStandard AASI-ITO girderoverpass four spans with piercaps and columns 25 15 (484)

AASHTO girder simple spantrestle bridge with pile bents 40 29 (312)

Major structures of all types,large spans, long bridges,simply supported orcontinuotrsin steel orconcrete includingsegmental 25 42 (452)

Other miscellaneous bridges,bascules, etc. to -

Overall average 36.4 (392)

lt,rseil upon square footage nrl,ri i Ige type conxtnmcted.tI igiires in parentheses are in $ per sq irr,

project. The bridges were built in steelbecause of the influence on the total hidof the remaining two-thirds of each proj-ect, which involved considerable road-way, retaining wall, embankment andcomplex utility work, etc. Incidentally,the traffic and construction plan for thesteel alternate of the 195/I595 projecthas largely been worked to that devisedfor the segmental alternate.

All costs presented here are based onbid prices. There have been exceptionalcircumstances on a few projects, whereadditional costs have been incurred forcorrecting problems above and beyondthe routine normally encountered withsegmental or post-tensioned concreteconstruction. These relate to questionsof detail, corrosion protection and mate-rials, and to general design, construction

and specification issues. They do notrelate to segmental construction as amethod for designing and buildingbridges.

On this basis, it is fair to compare costswith other types of construction and tonote that on recent projects, segmentalconstruction has been very competitiveand most successful.

For comparison purposes, costs from arecent survey of 75 bridges built over 3years up to fiscal year 1986-87 are sum-marized in Table 10. The averages inthis table have not been corrected farinflation to 1987, so may be a few per-cent low. Nevertheless, it is clear thatsegmental bridges are competitive con-sidering they are generally used forlonger, more costly, spans and particu-larly on curved viaducts.

PCi JOURNAL/May-June 1989 59

11. ADMINISTRATION PROCESSES -- DESIGN,CONSTRUCTION AND SHOP DRAWINGS

The following section discusses someof the problems encountered within thesegmental industry as a whole and offersrecommendations for future improve-ments, It is based upon the author's ex-perience on many segmental projectsover the last 12 years and represents hisown views which are not necessarilythose of the Florida Department ofTransportation, its agents, or any otherorganization. These ideas and proposalsare intended to promote discussion,thereby leading to a better understand-ing and general improvement in thisarea.

11.1 Current PracticeThe practice to date has been for con-

tract documents to require and for con-tractors to produce many (often hun-dreds) shop drawings detailing each andevery precast segment. The production,submittal, review and correction of shopdrawings is an awesome burden for thecontractor, designer and client. It in-variably leads to delays, differences ofopinion, professional posturing andclaims. All this is quite unnecessary andserves no good purpose, The time hascome to overhaul the shop drawing pro-cess as it relates to precast segmentalbridges. But this involves more thanshop drawings. It is necessary to start atthe very beginning, clearly establishing:

1. The functions, roles and responsi-bilities of the Client, Designer,Construction Engineering and In-spection Agency, Contractor andContractor's Engineer.

2. The limits of professional liabilityattached to the engineering aspectsof each function and party.

3. The scopes of services to be pro-vided by the Designer during de-sign, shop drawing review andconstruction.

4. The scopes of services to be pro-vided by Construction Engineer-ing Inspectors during construction(especially "Engineering" func-tions).

5. The contractual and engineeringobligations of the Contractor andhis Engineer.

6. The communication and adminis-tration process, especially iden-tifying routes for paper flow andresponsible decision makers.

7. The function, responsibility andjurisdiction of the DOT andFHWA and their agents.

8. The complete integration andmutual agreement of designscopes, design guidelines and cri-teria, specifications and any spe-cial provisions attached to thecontract.

Most, if not all, of these exist and arein force at this time both in Florida andother states. Difficulties have been ex-perienced in the past when there havebeen ambiguities and differences.

11.2 Recommendations forImprovement

The author believes that a major im-provement would be possible in precastsegmental work by adopting some or allof the fol lowing:

11.2.1 Design Plans1. Organize the plans for the conven-

ience of the Contractor who hasto fabricate and erect compo-nents.

2. Show details in full, either on ornext to the sheet showing thecomponent, and to a large scale.

3. Ensure that all reinforcing barcages can be assembled easily

60

from simple bar shapes, avoidingas much as possible closed loopsand multiple bends.

4. Ensure that reinforcing barbending diagrams are shown infull in the plans adjacent to thecomponent to which they applyor on the next sheet(s).

5. Ensure that reinforcing barlengths and bends are accordingto normally accepted industrypractice, amply allowing forbending tolerances.

6. Ensure that the reinforcing barswill fit inside the concrete dimen-sions, recognizing that there areconstruction tolerances (in thespecifications) on concrete thick-ness and covers. Do not forgetthat a ribbed reinforcing bar isphysically larger than its nominaldiameter.

7. Completely dimension post-ten-sioning duct alignments througheach segment - This is best doneby quoting offsets vertically andlaterally from known control linesor surfaces at regular intervals ofno more than 2 to 3 ft (0.61 to 0.91in) where small radii and reversecurves occur.

8. In anchorage zones, allow for thelargest commercially availableanchorage likely to be used withthe tendons concerned. Then, ifthe Contractor elects to use asmaller anchorage, it can easily heaccommodated with only a minorchange to the very localized de-tail.

9. Ensure that all reinforcing barsare bent to avoid post-tensioningducts. This implies an assumptionabout the size of duct likely to beused. In today's practice, mostpost-tensioning systems use ductsof similar size according to thenumber and size of strands in thetendon.

10. Clearly state on the plans that, inthe event of a conflict between

the post-tensioning duct andreinforcing bar, the post-tension-ing duct alignment must takeprecedence and that the rein-forcing bar shall be repositionedor replaced according to the di-rection of the Engineer.

11. Show the erection sequence thatwas assumed in the design, espe-cially the sequences in which clo-sures are made. Show where andat what stage of construction tem-porary supports or heavy equip-ment are placed on or removedfrom the structure.

12. Specify a sequence for post-ten-sioning.

13. Show the assumed ages of seg-ments at the time of erection andall material properties assumed inthe design.

14. Provide deflections at time of"infinity" and at the time of erec-tion of the segments according tothe assumed erection sequence,supports, equipment and times.Also, quote acceptable timing andload variations,

15. Quote assumed stiffness of tem-porary supports and erectionequipment loads.

16. Provide an envelope of stressesfor the top and bottom fibers re-sulting from all loads.

17. Quote the maximum loads, mo-ments and shears in both direc-tions that can he safely taken bysubstructure piers and foun-dations.

18. Quote all required loads, move-ments and settings for bearingsand expansion joints.

19. Ensure that the drawings agreewith the Contract specifications.

The emphasis in these should beupon constructibility for the conven-ience and benefit of the Contractor andClient. It should be apparent that, ifthey are properly included in the designplans, then the need for shop drawingsrepeating and redetailing much of this

PCI JOURNAL'May-June 1989 61

information is greatly reduced. Ideally,it should be possible to build a structurefrom the design plans with the excep-tion of only a few items (as listed in thenext section). In practice, there will al-ways be a need for some flexibility inconstruction techniques which requireminor adjustments to the design details.

It is the author's view that these pro-cedures place little additional workupon the Designer. It is more a case ofasking the Designer to organize hisplans and produce details which complywith normal industry construction prac-tice. In effect, this is also asking detail-ers (and for that matter, design en-gineers) to pay attention to practical de-tailing! It might require a little more ef-fort in the future than in the past, but itwill avoid the need for the Contractor toduplicate and then carry to completionthat which was previously produced inlarge part by the Designer.

11.2.2 Shop DrawingsC,i-,-en the above re-emphasis on con-

structible design plans, shop drawingsshould now be required only for:

1. Minor changes to details atanchorages to accommodate theContractor's elected post-ten-sioning system.

2. Details of the post-tensioninghardware components them-selves (man iifacturer's standarddrawings).

3. Details covering inserts or liftingholes.

4. Details covering localizedstrengthening for supports orspecial equipment placed in lo-cations not already allowed for inthe design plans.

5. Calculations pertaining to an%localized strengthening for same.NIajor redistribution of supportsor significant changes to erectionequipment loads would consti-tute a need for at least global re-analysis of the structure by the

Contractor and checking by theEngineer).

6. Checks and details for specialhandling, storage or stacking ofsegments.

7. Drawings and calculations pre-pared Under the seal of a regis-tered professional (structural),en-gineer for temporary falsework,special erection equipment, clo-sure devices or other itemsneeded for construction accord-ing to the Contractors methods,with the exception of regular con-struction equipment such ascommerciall y available cranes.

8. Casting cells and similar equip-ment.

9. Geometry control method in ahandbook or manual format.

10. An erection control manualquoting the sequence of opera-tions in great detail for the erec-tion of each segment, stressin g ofeach temporary and permanenttendon, movement of equipmentand introduction or removal ofsupports and devices, etc. Thismanual is drawn up for the benefitof the field erection and construc-tion supervision personnel andshould he reviewed and approvedby the Engineer for compliancewith the intent of the design.

11. Casting curves. Oily if the castingand erection operations differsignificantly (the Designer shouldquote some allowable time andload variations in the plans) fromthe design should the Contractorhe required to reanal y ze thestructure according to his own se-quences, methods and timings inorder to devise his own values ofdeflections. In all other cases, thecasting curves should be pro-duced from the deflectionsquoted on the plans. The Con-tractor should include his castingcurves in or with his geometrycontrol manual.

62

11.2.3 Comments

The above suggestions lar improve-ments to "administration processes"would require the mutual cooperation ofall sides of the industry: clients, states,the Federal Highway Administration,consultants, contractors, etc., and would

probably he best pursued through theauspices of recognized professional andindustrial organizations_

The Florida Department of Trans-portation has found benefits to otherareas of operations and to other types ofstructures, not just segmental, throughpursuing these kinds of improvements.

12. ACTIONS BY THE FLORIDA DEPARTMENTOF TRANSPORTATION

Since the introduction of the firstsegmental bridges, the FDOT hasgained considerable experience and hastaken many positive steps toward im-proving its own operations and thisfield, For example, it has:

1. Issued a "Design Criteria forSegmental Bridges" (1953/4)which has since been incorpor-ated in the Department's Struc-tures "Design Guidelines." (Partsof the original criteria have alsobeen incorporated in the Post-Tensioning Institute's report ofFebruary 1988 entitled "Designand Construction Specificationsfor Segmental Concrete Bridges,"prepared for the National Co-operative Highway ResearchProgram.)

2. Clarified and improved specifi-cations (Special Provisions).

3. Introduced "Designer ServicesDurih g Construction" on allmajor bridge projects, regardlessof type, in order to have immed-iate assistance available on anydesign related issues which mayarise during the course of theproject in addition to the needs ofnormal shop drawing reviews.

4. Tightened qualifications requiredof the Designer, the ConstructionEngineering and InspectionAgency and the Contractor,

5. Published a "Guide to the Con-

struction of Segmental Bridges"(1987) for use by construction en-gineering and inspection person-nel. An improved version is being,prepared which separates theguide into "segmental" and"post-tensioning" manuals, thelatter covering all types of post-tensioning construction.

6. Gained "hands-on" training forFDOT field engineers and in-spection staff.

7. Prepared a "Structures DetailingManual" for all bridge types.

8. Developed a "generic" segmentalspecification (unfinished).

9. Introduced a generic technicalscope of services for "constrt.ic-tion engineering and inspection"fir all types of structures.

1{) . Written a three-dimensionalgeometry control desktop com-puter program for the castingcontrol of precast segmentalbridges. (This is based upon theauthors own work originallyundertaken for checking thegeometry control of the LinnCove Viaduct.)

11. Also, the Department's own de-sign and construction consultantshave gained experience.

Many of these items were also of ben-efit in other areas, for example, inpost-tensioned structures of all typesand administrative procedures.

PCI JOURNAL^May-June 1989 63

Fig. 20. The Sunshine Skyway Bridge at nignt a brightluture for segmental bridges.

(Courtesy of Figg and Muller Engineers Inc.)

13. BENEFITS OF SEGMENTAL BRIDGES

Much of the foregoing has concen-trated on the problems and areas in needof improvement, overlooking the hene-fits of segmental construction.

The benefits include:1. Precast production off site under

factory-controlled conditions.2. Concurrent production on and off

site, i.e., substructure and foun-dation construction proceeds con-currently with segment manu-facturing.

3. Rapid erection systems.4. Overhead construction; may

avoid obstacles, which is par-ticularly valuable in congestedurban areas and sensitive en-vironments.

5. Preserves the environment,6. Avoids extensive falsework.7. Requires minimal onsite form-

work and cast-in-place work.8. Affords great flexibility in con-

stnictfon operations. This is valu-able for maintenance of traffic onlarge urban interchanges andother congested areas.

9. Bridge construction is placed offthe critical path.

10. Competitive concrete construe-tion (especially in Florida).

11. Efficient for large span concretestructures.

12. The traditional segmental box istorsionally rigid which makes itideal for curved bridges.

13. Substructures require less spacethan "conventional" beam con-struction, especially with highskew crossings. This offers greatadvantages in restricted locations.

14. Aesthetically attractive.For more detailed information on the

design and construction of segmentalbridges, the reader should refer to spe-cialist literature 7' ''°

64

14. SUMMARYOver the last 10 years, Florida has

been a leading state in the design andconstruction of precast segmentalbridges. With the exception of the mainspan portion of the New Sunshine Sky-way cable stayed bridge (Fig. 20), theseprecast segmental bridges mostly fallinto two groups: either straight span-by-span over water or curved balancedcantilever viaducts at major inter-changes. The spans involved generallyrange from 100 ft (30 m) to well over 200ft (60 m), covering the intermediate spanrange beyond the limits of normal pre-cast girder construction.

Span-by-span structures have not yetexceeded 143 ft (43.6 m). Recently, thenew bulb tee was successful against thespan-by-span segmental alternate for theHoward Frankland Bridge by a marginof about 3 percent. This initial resultindicates competition in the shorterspan ranges. However, in excess of thisspan length and on curved viaducts,segmental cantilever and steel are likelyto remain more effective. Balancedcantilever bridges have been successfulin interchange applications, especiallybecause they can readily accommodatethe varying alignments and span lengthstypical of such locations.

Substructures are typically lighter byhalf for span-by-span compared to bal-anced cantilever structures. This is be-cause of the basic difference in con-struction methods; most span-by-spanstructures have been founded on drilledshafts, whereas most cantilever struc-tures are founded on driven piles. Thisreflects design philosophies as much asconstruction methods and geologicalconditions, It should be noted thatspan-by-span construction affords moreopportunity to standardize pier shaftsand so develop very efficient systems.In cantilever construction, piers tend tovary more in height and constructionload requirements and each pier tendsto he unique.

Contractors elected to use single cellcasting machines, and most preferred toestablish their own precast yards. Allcasting operations went through aperiod of mobilization, usually 3 to 4months, followed by several weeks oflearning before production reached asustained rate of one segment per cellper day. Most operations achieved thisrate. Casting operations were geared tothe size and duration of the project, mostbeing a completely new operationwriting off the cost of the forms and yardon the job. Generally, robust equipmentand forms saved time and money de-spite the higher initial outlay.

Erection operations for span-by-spanbridges were either by truss or gantry.Balanced cantilevers were generallyerected by cranes standing on theground. All projects experienced alearning period of a few weeks or spansbefore achieving a sustained erectionrate. Typically, span-by-span constric-tion proceeded at three spans per weekand balanced cantilever at four seg-ments per day per cantilever. Higherrates were achieved occasionally.

Problems in segmental constructiongenerally center upon attention to detailand quality of workmanship. The mostsignificant factor, perhaps, is the abilityto readily assemble the reinforcing barcage and post-tensioning ducts withoutconflicts or misplacements and then en-suring that all the reinforcement in theducts and other embedments remain inplace during concreting. Misplaced andblocked post-tensioning ducts preventsuccessful construction. Attention togood workmanship and inspection paysdividends. Special care is needed inconcrete placement, consolidation andfinishing to ensure a good qualit y seg-ment.

Successful erection depends almostentirely upon the quality of the seg-ments. These problems are not justpeculiar to segmental bridges; they have

PCI JOURNAL'May-June 1989 65

also occurred in other precast, post-ten-sioned AASHTO and hulk tee beamconstruction. As experience growswithin the industry and the profession,such problems are less frequent. How-ever, there is a need to educate and in-form designers and detailers aboutpractical constr uctible details and toenforce good workmanship througheducation and specifications.

There is also a need to address (na-tionwide) administration processes re-lated to design, shop drawings, inspec-tion and construction practices sincemuch wasted effort has been involved.The author considers that much im-provement is possible by the adoption ofappropriate standards and practices.Such measures will make the entire ad-ministrative process more efficient.

15. CONCLUSIONS

Experience in Florida has shown thatit is possible to complete segmentalstructures on time and ahead ofschedule. Of the 11 major projects con-taining 31 bridges so far constructed,eight projects went well in casting anderection, and three others were delayedfor a variety of reasons, mostly con-nected with inexperience. The latestprojects are proceeding very well andwill be completed ahead of schedule.

Segmental construction has success-

fully demonstrated its competitivenessin Florida against other alternates. Forthe span ranges and applications in-volved, it is likely to remain competitivefor the foreseeable future.

While it is always possible to makeimprovements, Florida's experiencesand successes clearly demonstrate thatthe learning and development phase haspassed and that segmental technologyhas a place in the future of bridge engi-neering.

REFERENCES

1. Muller, Jean M., and Barker, James M.,"Design and Construction of Linn CoveViaduct," PCI JOURNAL, V. 30, No. 5,September-October 1985, pp. 38-53.

2. "Segmental Bridge Design in the FloridaKeys," Concrete International, V. 2, No.8, August 1980, pp. 17-22.

3. Blaha, William, "Seven Mile BridgeSpeeds Towards Completion in theFlorida Keys," Concrete Products Ma go-rifle, V. 84, December 1981, pp. 20-25.

4. "Sunshine Skyway," Report by FloridaDepartment of Transportation, June1984.

5. "High Level Innovation" (SunshineSkyway), Engineering News Record,September 11,1986.

6. "Turning the Corner in Bridge Design

and Construction" (Ramp 1), Road.+'Magazine, November 1984.

7. Podolny, Walter, Jr., and Muller,Jean M., Construction and Design ofPrestressed Concrete SegmentalBridges, John Wiley & Sons, New YorkN.Y., 1982.

8. "Test Girder Shows Strength," Engi-neering News Record, December 19,1985.

9. Precast Segmental Box Girder BridgeManual, published by the PrestressedConcrete Institute and the Post-Ten-sioning Institute, 1978, 116 pp.

10. Heins, C. P., and Lawrie, R. A., Design ofModern Concrete Highway Bridges,John Wiley & Sons, New York, N.Y.,1984.

66

ACKNOWLEDGMENTSThe majority of the information pre-

sented in this paper was obtained by theauthor from public records and col-leagues within various offices of theFlorida Department of Transportation.Additional information was obtainedthrough questionnaires and conversa-tions with many firms and individualsinvolved. Wherever possible, effortswere made to contact riot only the Firmsbut also the persons with firsthandknowledge and experience in the vari-ous projects. The author also drew uponhis own personal involvement in severalof the projects.

Thanks are extended to all colleagues,firms and individuals for their coopera-tior► and assistance with information.Special thanks are owed to Mrs. LauraHelms and Mrs. Karla Beard for theirpatience and dedication in the prepara-

tiou of this paper.The author acknowledges the fol-

lowing organizations for their assistancewith information for this paper:Beiswenger, Hoch and Associates

- Capeletti Brothers, Inc.Cianhro Corporation

- D.R.C. Consultants, Inc.- Figg and Muller Engineers, Inc.- Florida Department of Transportation- Harbert-Westbrook — Joint Venture-Heningson, Durham and Richardson- Hubert Janssen, Inc.- Misener Marine- Parsons, Brinckerhoff, Quade &

Douglas-PaschenlAmerican Bridge JointVenture

- Post, Buckley, Schuh & Jernigan, Inc.- Tony Gee & Quandel (Alfred Benesch)- Volker-Stevin Construction Inc.

NOTE: Discussion of this article is invited. Please submityour comments to PCI Headquarters by February 1, 1990.

PCI JOURNAL'May-June 1989 67

APPENDIXA more detailed breakdown of the in-

formation summarized in Tables 1through 9 is provided in Tables Althrough A9.

Table Al. Precast segmental projects............ 69

Table A2. Superstructure parametrics. .......... 70

Table A3. Substructure parametrics . ............ 71

Table A4. Casting operations . .................. 72

Table A5. Rejected segments . ................. 73

Table A6. Erection methods . ................... 74

Table A7. Construction time summary. .......... 75

TableA8. Project bids . ........................ 76

TableA9. Cost breakdown. ...... .............. 77

68

Table Al. Precast segmental projects.

Bridge type Bridge over Geometry Substructure

Canti- Span-No. of Driven Drilled SpreadProject bridges lever by-span Land Water Curved Straight CIP Precast piles shafts footings

BuiltLong Key 1 • • • •Seven Mile 1 • • • • • •Niles Channel I • • • •Channel Five 1 • • • • •Rampl 1 • • • • • •Skyway Approaches 1 • • • • •175/1595 Phase 1 5 • • • • • •Palmetto 5 • • . • •Airport 4 • • • • •US 441/1595 (M)175/1595 Phase 2 {u)

29

••

••

••

••

•• •

Total 3I

Bid butnot built195/1595 (DEF) 8 • . . . •South Fork New River 2 • • • • • • •Howard Frankland 1 • • • • • • •

Total 11

In DesignPortofMiami 1 • • • • •Edison 1 • • • • • •Golden GIades (N) • • •

Table A2. Superstructure parametrics.

Frans- I,ongi-

Bridge Segment weights Rein-versepost-

tudinalpost- Struchiretylx

tip,ruplan No. of {tons) forcing ten- ten-Abut-Spans (ft) Depth area 'o, of seg- EST"* bar sinning sioning Canti- by-

Range AvgProject (ft-in. (ft2) bridges ments F'ypical Pier ment ft2/ft2 lb/ft2 lb/ft' Ib1Ft 2 ]ever span

Long Key 118 7a) 468,358 1 734 fib 62 49 1.34 4.48 0.63 1.50 •SevenMile 135 741 1,376,257 1 2154 65 68 52 1.35 6.85 0 2.21 •\ilesChannel 118 7-41 175,634 1 276 66 66 49 1.35 4.83 0.63 1.70 •Channel Five 135 7-0 190,126 1 299 65 68 52 1.35 (6.85? (0.63) (2.21) •]tamp 1 120-224 194 9-4 91,271 1 201 58 94 52167 1.75 8.58 0.90 4.76 •SkywayApproach 135 84) 415,534) 1 584 75 83 34140 1.42 6.66 0.50 1.87 •175/1595 Phase 1 100-200 160 7-3 251,680 5 567 52 70 38/61 1.61 8.36 0.97 3.8.2 aPalmetto 84-215 155 8-0 197,724 5 658 38 48 25135 1.66 8.93 0.87 3.83Airport 85-16.1 121 7-3 124,520 4 286 54 70 62 1.68 11.06 1.07 3.63 •u544111595(M) 124-224 159 8-3 177,199 2 385 56 98 71 1.71 14.50 1.10 3.59 •1751 I Single box 85-206 137 7-3 342,471 4 750 52 65 57172 L56 14.00 (1.86 5-50 •1595(2) I` Twin boxes 60-183 118 6.6 194,472 5 tilt 40 18 39 1.86 18.00 0.56 5.45 •

643!195/ j Single box 71-205 161 8-3 4.20,606 7 89(i 56 90 61 1.63 0.82 5.27 •595 1 'l'winboxes 83-14)1 1211 6-6 140,608 I 258 40 54 39 2.04 - 0.62 5.09 •

8-31 bySouth Fork New River 95-301) - I6-6 433,249 2 984 50/88 117 54188 1.77 - 0.72 5.66 •1loward Approach - 143 8-{) 839,982 l 1584 70 -13 40 1.61 9.43 0.81) 2.69 •Irtnkland J Main Spans 143-231 1141 219,540 J 1

.544 55 48 45154 1.79 8.40 0.77 3.67 •

fort of Miami 90-195 - 9-3 268,906 1 496 75 100 95 1.89 - 0.84 - •Edison - 143 8-) 593,0411 1 1242 64 40 1.59 7.79 0.73 2.83 •C,nlden Glades (10(4- - - (N) - - - - - - •

225)

quivalent wdid concrete thickness of superstructure.Metric (Si/conversion factors: 3.'38 ft - 1 in; 1 ton (U,S,) = 0.91 ton ee;1 psf - 4.89 kg/inns.

Table A3. Substructure parametrics.

Post-SC:S" tensioning Bridge type Substruc-

Plan Substructur Heinforcingbars in htretypeTSB Span-Spans (R) Height (ft) area concrete (per in substructure substructure Canti- by- Pre-

Range Avg Rance avg( Total lb lblytl5Project (Ru) (y'r191 cerit) (1b+ yd$1 lever " part CII' cast

1, rug Key 116 25 468,358 3,011 0.69 499,11)0 100 it • •Seven ( Main Spans 1,35 20-70 50 (145,000) (3,400) 1.27 (410,0(1(1) 121 10 • •Mile 1l Approach 135 20 1231,000 (4,400) 0.63 - - 0 • •\ile.sCh,umel 118 20-40 25 175,635 1,030 1.27 ..-.. - 0 • •Channel Five 135 20-70 50 190,I26 {3,400) 0.57 (410,0(10) 121 10 •Rainp I 120-224 194 35-49 42 91,271 2,561 1.80 396,620 155 0 • +Skyway Approach 135 45 . 152 98 415,530 17,993 1-18 2,357,056 131 24 •175/1595 Phase 1 100-200 160 28-60 42 251,680 8,036 2.05 1.182,591 147 0 • •Palmetto 84-215 155 30-51 41 197,724 4,844 1.61 816,48() 168 0 • •Airport 85-162 121 24-40 32 124,520 3,524 2.39 570,538 162 0 • •1-S441/I595(M) 1.24-224 159 25-81 54 177,199 7,312 2.06 1.060,184 145 {l • •IZX Single box 85-206 157

3{^A7

59 342,471 15,636 2.09 2 ,411,683 157 0 • •1595(2) Twin hoxes 60-1993 118 33 194,272 5,15.2 2.17 755.8163 117 0 • •

1951 Single box 71-20.5 161 -i47)

420,606 15,200 I • •1595 Twinboxes 83-140 120 - 140,608 3,946 1^Jfi

III • •South Fork New River 95-300 - - (63) 433,249 12,184 1.24 No data • •Howard Approach 143 143 - - 839,982 4,718 - • • +[rank1:V slain Spans 143-231 - - 219.54(1 6,1)5.4 - . • • u

Port o I Miami 96-195 - - - 268,906 - - • •Edison - 143 593,040 (3,01101 No data • •Golden Clades (100- - - - - . •

.>rin}

`Ratio of url iJ concern Vn substructure to void sp.<nued by the bridge_Metric (S l lconversion factors: 3.288 = 1 m: 1 1&yd 5 = 0.50 kg/nit',

Table A4. Casting operations.

SuperstructureNUMbers and types of segments No, of casting cells

Learning period/phase

Sustained productionrate after learning phases

Abutment' Abutment/ Appox. No, of Typicals Others TotalProject 'typical Pier Other Total typical Pier Other Weeks segments per week per week per week

Long Key 618 90 26 734 3 1 — — — 15 2 17Seven Mile 1850 228 76 2154 5 — 1 — — 25 2-3 28Niles Channel 232 34 10 276 3 1 — — 15 2 17Channel Five 288 3 1 — — 15 2 17Ramp1 187 10 4 201 (1) 1 — — — — —Skyway Approach 504 64 16 584 2 1 2 l 10Skyway Main Spans 319 14 — 333 2 1 2 10J 5

175/1595 Phase 1 187 10 4 567 2 1 1 20 18 5-10 2 10Palmetto { Single box 461 25 12 498 3 1 -

20 19 15-25 (24)Twin boxes 144 8 8 160 2 1 -Airport 258 20 8 286 2 1 — 4-6 10 2 12US44111595 (M) 356 25 4 385 2 1 — 4-6 4-6 10 1 11175/ 1 Single box 689 47 14 750. 4 1 1 29-351595 (2) Twin boxes 550 46 20 616 3 1 _. 3-5 15 — — (>50 ihex)

Notes:1. Learning phase is approximate estimate after mobilization.2. Sustained weekly production rates are averages.3. All segments for .Niles Channel arid Channel Five were made at Long Key using same equiprnent.4. Precast substructure segments are not included.

Table A5. Rejected segments.

Project Superstructure segments Iost and reasonsAttrition—

segments lostitotalLoss

(percent)Long Key 2—Voidslhoneycombing 2 —Accident 41734 0.5Seven Mile 2 — Voids/honeycombing 1 —Improper stacking 3/2154 0.2Niles Channel 0 0 0Channel Five I — Voids/honeycombing (cement/material problem) 1/299 0.4Ramp I 1 — Accident 11201 0.5Skyway Approach I —Damaged shear keys 1 - Hurricane damage 1—Other 3/584 0.5Skyway Main Spans 1 — Collapse of gantry 11333 0.3175/I595 Phase 1 18 — Various reasons —Voids/honeycombing/displaced post-tensioning ducts 18/567 3.2Palmetto 19 — Various reasons —Voidsfhoneycombingldisplaced post-tensioning ducts 19/658 2.9

/weakout of tolerance concrete thicknessforms (4 or5 had to be discardedwhen earlier segments were rejected)

Airport 0 —But poor riding surface finish 0 0L'S4411I595 (M) 1 —Accident 1/385 0.217.5/1595 Phase 2 1 — Thermal shock I — Low strength 1 — Strands grouted 3/1366 0.3

before being stressed

C)

0C11z

CvC-C

W

Table A6. Erection methods.

ConstructionSegmentdelivery Erection equipment/supports Erection time and completion

BeamCanti- Span- Low- Stability False- Truss/ and Learning Sustained Bridge

Project lever by-span boy Barge

.

Crane towers work Gantry winch period rate completion

On timeLong Key • • 3-4 weeks 3 spans/weekSeven Mile • • • 3-4 weeks 3 spans/week 6 months aheadNilesChannel • • • None 3 spans/week On timeChannel Five • • • None 3 spans/week On timeRamp I a • • a • — — On timeSkyway Approach • • • 2 spans 21/2 spans/week On time

Main Spans a • • • • 3 cycles — Late175/1595 Phase 1 . • • • • Many weeks 10-15 seg/week LatePalmetto • • • • • A few weeks 15-25 seg/week LateAirport • • • • • Many weeks Late1. 1 5441/1595(M) a a • • • 4-6 weeks > 10 seg/week Ahead175/1595 Phase 2 • • • • • • 2-3 weeks > 30 seg/week 9 months ahead

0C-

0Cz

am

CDm

Table A7. Construction time summary.

Project

Originalcalendar

days_-

Extendedcalendar

days

Finalcalendar

days

Timeused

percent Comments

Long Key 915 1054 1352 128 Long Key completed on time, approach was casting yard for Niles andChannel Five

Seven Mile — — — <100 Completed 6 to 7 months ahead of scheduleNiles Channel 420 546 739 135 Stricture completed ahead of scheduleChannel Five — -- — <100 Structure completed on timeRamp 1 675 --- 675 100 Structure completed on timeSkyway 900 1416 1480 105 Approaches went well, time lost on main span cable stays, etc.175/1595 Phase 1 920 1149 1539 134 Delayed due to structural problems, repaired okPalmetto 730 742 868 117 DelayedAirport 490 585 1244 213 DelayedUS 441/1595 (M) 1300 1300 — <100 Bridges on time or ahead175/1595 (U) 1052 1052 <100 60 days ahead at interim deadline, full completion approximately 9

months ahead of schedule195/1595 (DEF) Steel alternate built due to total project bid but segmentalSouth Fork New River Steel alternate built bridge prices were actually cheaper than steelHoward Frankland Bulb tee alternate 3 percent lower than segmental

Table A8. Project bids.

Actual Total project SegmentalStart finish bid amount lridge portion

Project date (late (S) ($) bid Comments

BuiltLong Key January 79 July 81 15,097,276 14,50(1,000±Seven Mile October 79 July 82 43,394,764 43,000,000±Niles Channel April 81 April 83 7,906,574 7,700,000±Channel Five May 81 January 83 10,363,912 9,000,000±Ramp 1 February 82 May 84 22,344,172 4,628,0(0Skyway Approach

and Main Spans April 83 October 87 71,132,079 71,132,079175/1595 Phase 1 July 83 November 87 10,176,199 8,765,000Palmetto (larch 84 September 86 9,445,663 7,825,000Airport October 84 December 87 7,793,829 5,292,178US 441/1595 (M) October 86 (July 89?) 6(1,243,919 11,938,00() Project ahead of scheduleI75/1595 Phase 2 April 87 (July 89?) 51,132,584 27,054,000 Project ahead of schedule

Not built *Segmental price of second low hid195/1595 (DEF) — — (104 million) 31,193,000)* Steel alternate built at $35,641,918tSouth Fork New River — — (60 million) 19,753,000)* Steel alternate built at $21,109,633Howard Frankland — (44.5 million) 46,500,000) Bulb tee alternate built at x+44,500,00(1

In designPort of Miami (June 89) $21 million estimate t Steel bridge built since it was oul) partEdison (Sept 90) $25 million estimate of the overall project hid which includedGolden Glades (]9905) roadway, r.smps, walls, utilities, etc.

C)C0

z

Dl

C-C7CD

Table A9. Cost breakdown.

Structure I, I',' 'vft' hreakllowi, L,lit 11T1ccs kl(luste{I to I^l.47

Imper til l,- Sub- Post- IH[•urI'la+!Contract til,a' Sc L lr,rl, I.,I ,re:( strut- ,tru[- 7fth . tructrlr tied!- ten- Iorcinu

start lu,t'- b,- bridge h,r it•, total tore hire Adjusted concrelr lent lOuiuul hltrProject date ICC C'I spa- 11 I„III I()IIs IItIHl 1 sllt s,fl' 's It I1, 1987 .'es-'f1'( t1\•(1•1 • l/i1, 1111]

1311iIt

Long; hey Jan 79 • 1-1.5 188.1 II. I - :38

Seven like Oct 79 • 1341 1 371.3 31. - 34\ilex Ch:umel Apr h1 • 7.7 175.0+ 43.61 - 511

Clrannel Fite \tavhl • 9.11 100.1 47..E - - 55

Ramp I Feb ti-! • .63 81.:3 501." 111.7 111 -14 I,l -3411; 577 1.55 u t4

SIswavApproach Alir83 • - 115.5 - -- -175r1595 Phase 1 Jill 83 • h 77 '51.7 '3 . 1.t :37.4 7.3 41 'SU7 3t1I 1.40 11.51

Pahrretto S 1.tr84 • 7.3 197.7 '19. , ' 't9.4 1(1.= 4h 159 -149 1.111 11.43

Airport Oct 85 • 5.:t9 11-4.5 -11.5 tech 15.7 48 225 :371 1.13 4I.36

I. S 441/ 1595 i \1 , Oct tit; • 11.9.1 177.:t 87.I IS.11 19.4 tits 334 1s36 1.1141 11..141

175;1595 Ahr87 • 7I45 53h.7 511.1 :36.J 13.5 51) 31111 1119 1I11) I1. }

li,d but not built19511595 (1)141, 1 iii h • 11.19 51,1 55.. .37.9 17.7 51•i :3511 lsis 41.85

Smith Fork \ew Riser Ji,,, ts7 • 19.75 333 .45.t . - - -147 -I Iowand t'rauIklmicl Aiugh • • .tl} 511 111,595 .43.9 - -- -14

Port [,f \Iuunl 31959, • 11 I !hh.slh1lox(>II 11110: • 1:t5' 5931 - - -

Can lc , cr 3 : 54 1.51, IN I I.^1^Slttut „-hi • -spa,,, :13