THE COMPOSITE U-BEAM BRIDGE SUPERSTRUCTURE

THE COMPOSITE U-BEAMBRIDGE SUPERSTRUCTURE

John R. SalmonsUniversity of Missouri—ColumbiaColumbia, Missouri

William J. KagayUniversity of Missouri—ColumbiaColumbia, Missouri

As a result of an expanding bridgeconstruction program and rapidly in-creasing construction costs, there is agreat demand for more efficient,more economical bridge structures.Precast, prestressed composite con-crete construction in general offersthese features. The composite U-beam construction was developedand evaluated to show even greaterefficiency and economy and the re-sults of these studies are presentedin two successive papers ("TheStructural Performance of the Com-posite U-Beam Bridge Superstruc-ture" will appear in the next issue ofthe PCI JOURNAL).

GENERAL DESCRIPTIONThe system consists of a series of

precast, prestressed U-beams fur-nished with an interior top slab formand used in conjunction with thecast-in-place concrete deck. The U-beams are set side by side on sup-porting bent beams or abutments,similar to setting conventional pre-cast, prestressed box beams. The legs

of the U-beams extend upward withcorrugated metal arches fitted be-tween them to serve as stay-in-placeforms for the cast-in-place top slab.The resulting bridge deck is verysimilar to a multi-cell box super-structure. A typical cross-section ofthe composite U-beam bridge deckis shown in Fig. 1.

Beam dimensions. The beam dimen-sions were standardized (Fig. 2) inorder to realize increased economythrough repetition in precasting. Thewidth of the member was fixed at 5ft. (1.5 m) with a variation in depthlimited to 4-in. (10 cm) incrementsfor beam heights of 20, 24, 28, 32and 36 in. (51, 61, 71, 81 and 91 cm).This variation in depth will accom-modate AASHO HS20-44 loadingover a span range of approximately30 to 80 ft. (9 to 24 m) depending up-on the thickness of the cast-in-placetop slab. The usual range in roadwaywidths can be provided with thestandard-width beams and by vary-ing the width of the overhangs.

20ҟ PCI Journal

A composite U-beam bridge superstructure was developedin order to realize increased economy with precast, prestressedbridge construction. The concept of the proposed system ispresented, and direct cost comparisons with existingbridges are made.

Interior voids. The voids in thebridge deck do not necessarily ex-tend continuously through the entirespan length, but may be interruptedprior to reaching a pier or at inter-mediate points between piers. Thisprovides for diaphragms as well assolid bearing ends at the supports.This also furnishes a medium

through which mild-steel reinforce-ment or post-tensioning rods can beplaced transversely to provide later-al ties. The stay-in-place slab formscan be fabricated from a variety ofmaterials. The corrugated metalforms proved to be quite satisfac-tory; the springing line edges of theforms are securely attached to the U-

26'- ROADWAY

5-U BEAMS AT 15'-O"=25'-O"

HALF SECTION NEAR CENTER I HALF SECTION NEAR INTERIOROF SPAN SUPPORT

Fig. 1. Typical cross-section of the composite U-beam bridge deck

May-June 1971ҟ 21

TED

I" 5"MIN..-//

I--IҟS^• Ii CORRUC

STEEL

PRECASTU-BEAM

MONOLITHICALLYCAST-IN-PLACETOP SLAB

5.21"n

J 2I2r

3'R

4° ..

B

AC

5"MIN.

3'0°MAX.

5'- 0"

A 20" 24" 28° 32" 36"

B 4.67° 4.50" 4.33° 4.16" 4.00"

Fig. 2. U-beam dimensions and details

beam uprights when shipped fromthe plant.

STRUCTURAL PERFORMANCE

Since a full scale bridge using thecomposite U-beam superstructurehas not yet been constructed, struc-tural evaluation was based on lab-oratory investigations. Three typesof tests were conducted—two test se-ries were with single units while thethird test was with a multi-unit sys-tem.

Single-unit tests. The single-unit testseries was a model-and-prototype

22

study using a 5-ft. (1.5 m) wide, 18-in. (45 cm) deep, 36-ft. (11 m) longprototype member and one-halfscale models. Three members ofeach size were tested and the resultswere used for model-to-prototypecorrelation and general verificationof the design procedure. The be-havior of the prototype could be re-liably predicted from the model andthe design procedure was substan-tiated.

Multi-unit test. The third test was ofa five unit, one-half scale bridgedeck. The interaction of the units,the transverse load distribution and

PCI Journal

15"ҟ 26'-0" ROADWAYҟ 15"9"ҟ6I ҟ 13`-0"ҟ "flii r I1

DRY PACKEDҟICEMENT MORTARҟI -

s^ I BITUMINOUS SURFACE 2 " ± 14TIE BOLTS

F:.

I J 111

98'

8"

the composite behavior of the unitsmaking up the system were studied.Again, the test results were consis-tent and predictable. In general thestructural performance was as antic-ipated and the bridge can be de-signed according to conventionalprocedures and standard practices

of precast, prestressed composite de-sign. The detailed results of thestructural evaluation phase of thestudy are available in research re-ports(1'2,3'5)

ECONOMIC EVALUATIONBefore any new or different struc-

tural system can be adopted it must

16"ҟ 28'-0" ROADWAY I6'

IQҟi6['ҟ14'- 0" 14'- 0" 16' I 2"

BRIDGE A-2416 (H15 LOADING)

I62" 26'-0" ROADWAY I6"

Ii3-0" 13'- 0" 6"ҟI IOi"

^oҟi 2d SYMM. ABOUT

12E20.7 iҟI 12C20.7ҟz[^I

21WF55 32"xe"fC,TYR 21WF55

3'-4q" 7'- 4" j 3'- 8" I 3'- 8"ҟjҟ7'ҟ4" jHALF SECTION AT CTR. OF SPANҟHALF SECTION AT INT. BENT

BRIDGE A-2039(H15 LOADING)

3'-2" i 3'-2" , 3'-2" 1 3'-2" :19" 119" . 3'-2" 1 3'-2" 3'-2" 1 3'-2" i

BRIDGE A -2141 (H15 LOADING)

Fig. 3. Superstructures of completed Missouri highway system bridgesMay-June 1971 23

N

C,C,

30-9"

' I #5®12" CTR. #5@ 12'CTR. ALT WITH #5 @ 12'CTR. BENTS 2'&'#6 SPAN I-2,4-5,#5 SPAN 2-3,3-4 &#6 @12"CTR. BENT 3 I

#5 #5 @ 8'CTR. SYMM. ABOUT C I °I

#8 SHEAR CONNECTOR @2:-O"C R. I61N./FT CROWN --

L ^4^

4/2 j 6-lU-BEAMS @ 5'-O"= 30'-O" 'I y2°HALF SECTION NEAR CENTER OF SPAN HALF SECTION NEAR INT. BENT

2" END MIDSPANҟI" 4°

j3n #4

#4 CI =42' BEAM 52"SPA• C3=69' BEAM

2

12CI-5- 3iifi STR.

#4 STIRRUP C3-7-71s STR.3"R

#q

I I ICI-16-7ig4'STR., 3-22-74s 4'STR.I #4

Fig. 4. Re-designed Bridge A-2416 (H15 loading) 5'-0"

34"

I-

mCD10

28'- 9"

102 6" 26'- 0" ROADWAY 6"l0tjF j I #5 @ 12"CTR. #5' 6" CTR.

#5 I . I #6 I t I I i---#5 #5 Cd 8"CTR. I SYMM. ABOUT C I' I IS"

R ,.. I #8 SHEAR CONNECTOR 'CTR. 3/16 1 N./FT.CROWN I ___ 9,r. ^ ^^^ r y t ^r 434

I ri'. Jl ii , 0

1 22 1 2'i 5-U-BEAM @ 5'-0" 625'-O" 'a 22y2HALF SECTION NEAR CENTER OF SPAN HALF SECTION NEAR INT. BENT

END MIDSPAN2° In 4n'--

Is" ' \\ #4

#4 \\ oJ3-^s4STR. ^2 5

#4STIRRUP 12 26"

#q 3°R

rye 20-1e STR. #45'-0"

Fig. 5. Re-designed Bridge A-2039 (H15 loading)

C,

o`

28`- 6"g" g" 26'- 0" ROADWAY 6

" #5 @ 6"CTR.I #5 @ 12 CTR. iI ^ I

#5 I #5@ 8"CTR. SYMM. ABOUT E I i 18"' I #8 SHEAR CONNECTO 2' Cu CT: 461N.

/FT. CROWN I --

21" j 5-U-BEAM @ 5'-O" 25'-Oa j 21"HALF SECTION NEAR CENTER OF SPAN I HALF SECTION NEAR INT. BENT

2 END MIDSPAN'

I" 4'1

#

4 ]4•-^6"'OSTR. r12I 20"

#4 STIRRUP 3"R^_ - 14"12 #4 18-^'16"STR. #4

rrr5-Fig. 6. Re-designed Bridge A-2141 (H15 loading)

exhibit superior structural perfor-mance, increased economy, or both,over existing methods of construc-tion. Without the benefit of fieldconstruction, an economic studymust of necessity be theoretical, butsince one of the prime considerationsis cost, an intensive effort was madeto realistically evaluate the cost ofthe system.

Method of evaluation. One methodof obtaining a realistic cost evalua-tion of the U-beam system was tocompare actual costs on completedbridges with estimated costs on thesame bridges re-designed with theproposed system.

Three typical bridge structures,designed with conventional super-structures and recently constructedin Missouri, were used as the basisfor comparison. They were selectedto give a representative range ofspan length and to consider the mostcommon types of bridge superstruc-tures used for shorter spans in theMissouri highway system.

Existing bridge dimensions. The firststructure considered is a three-spanprecast slab structure designated A-2141. It has equal spans of 34 ft. (10m) and a 26 ft. (7.9 m) roadway. Thesecond is a continuous compositesteel I-beam structure with a 26 ft.(7.9 m) roadway and three spans of35-43-35 ft. (11-13-11 m), designatedA-2039. The third bridge, designatedA-2416, has four spans of 43-70-70-43ft. (13-21-21-13 m) and a 28 ft. (8.5m) roadway. It consists of a voidedcast-in-place slab deck. Each ofthese structures was designed forAASHO H15 loading. The cross-sec-tions of the superstructures areshown in Fig. 3.

The bridges considered in thisstudy were built in late 1966 and1967. The Missouri State Highway

Commission made available actualunit costs on these three structures,along with superstructure quantities,to be used in the cost analysis.

U-beam re-design. The superstruc-ture of each of the three bridges wasre-designed using the proposed com-posite U-beam bridge deck system.The design procedures were similarto that used for the design of normalprestressed composite bridge beams,where the dead load is simply sup-ported and the live loads resisted bythe continuous structure.

The dead load of the U-beams andof the cast-in-place top slab are sup-ported by the beams as simple mem-bers. However, after the top slab hascured, and if this top slab is contin-uous and reinforced over the sup-ports, any additional loads appliedto the deck will be resisted by theresulting continuous, composite sys-tem. For the AASHO truck loadingsconsidered in the re-designs, the de-sign moments were determined forthis continuous condition. The re-sulting cross-sections of the re-de-signed bridge decks are presented inFigs. 4, 5 and 6.

U-beam production costs. Wherelengths and weights do not exceedpermissible limits, the most econom-ical prestressed concrete membersare usually precast and preten-sioned. Since the U-beams are notdrastically different from some com-monly used precast, pretensionedbuilding members, prestressed con-crete producers could be expectedto estimate the production and deliv-ery cost of the U-beams with a rea-sonable degree of accuracy.

The initial re-designs of the threebridge decks was made sufficientlydetailed to allow a realistic produc-tion cost estimate. Several pre-stressed concrete producers in Mis-

May-June 1971ҟ 27

TABLE 1. Cost per square foot of precast, prestressed U-beams

ProducerCastingmethod

Size of units—depth/length

20"/33' 26"/34' 26"/42' 34"/42' 34"/69'

No. 1 Inverted $3.15 $3.50 $3.75 $4.26 $4.51No. 1 Upright 3.25 3.60 3.50 4.05 4.10No. 2 Upright 3.15 3.54 3.46 3.71 4.03No. 3 Upright 3.21 3.52 3.81 4.45 4.51

souri, Kansas, Iowa and Nebraskawere asked to make cost estimatesfor the prestressed concrete mem-bers. At the same time these produc-ers were asked to make recommen-dations for any changes to the U-beam configuration or to the methodof production.

Three prestressed concrete pro-ducers responded to the request withboth estimates and recommenda-tions. The prices obtained are listedin Table 1.

The cost estimates furnished bythe producers were made for themembers delivered within a 100-mile(161 km) radius of the plant. Theywere also asked to consider a rea-sonable volume of production, rath-er than estimating on the basis ofcustom members. Variations in theprice estimates for each of the mem-bers considered was quite small.

Manufacture of U-beams. From theinitial conception of the proposedsystem, it was anticipated that themost economical method of produc-tion was casting the beams in an in-verted position. However, two of thethree producers immediately recom-mended that it would be more eco-nomical to cast the U-beams in theupright position because of highhandling and inverting costs for the

inverted position. The third produc-er was requested to provide esti-mates for both methods of produc-tion. His figures show that theinverted casting method is more eco-nomical in the shorter spans and theupright casting method more eco-nomical in the longer spans.

It was concluded that the uprightmethod of casting should be usedfor beam production. This allows thehorizontal shear connector to be aconventional connector of U-shapedreinforcing bars extending from thetop of the legs. This type of connec-tor is used in standard precast, pre-stressed I-beam bridge construction.Additional minor recommendationsfrom the producers were receivedand incorporated into the designs.

Construction costs. In order to com-plete a cost analysis of the proposedbridge deck, an evaluation of the on-site construction costs was required.To accomplish this in a realisticmanner, two general contractingcompanies, with considerable expe-rience in bridge construction, wereconsulted. Since detailed plans for acomplete bridge were not available,they were asked to furnish unit costsand methods of estimation for theproposed system. As would be ex-pected, each company used methods

28ҟ PCI Journal

Table 2. Summary of superstructure costs

Bridge Cost of superstructureSavings

$/sq.ft.

U-beamcost

%

Materials

%

Labor

%

Equipment andoverheadProposed system Actual bridge

Total $/sq.ft. $/sq.ft.A-2141 High

AverageLow

$19,882.4019,284.2318,760.58

6.836.636.45

7.387.387.38

.55

.75

.93

41.842.142.5

70.671.572.5

14.314.715.0

15.113.812.5

A-2039 HighAverageLow

24,820.9123,956.6823,156.60

7.447.186.94

8.228.228.22

.781.041.28

41.541.842.2

71.672.874.0

13.313.714.1

15.113.511.9

A-2416 High 58,756.60 8.74* 10.33 1.59 51.2 74.0 10.7 15.3Average 54,376.58 8.12* 10.33 2.21 52.1 76.6 11.0 12.4Low 50,088.78 7.51* 10.33 2.82 52.8 79.2 11.3 9.5

*For Bridge A-2416, the cost of the superstructure was increased by $0.38/sq.ft. to account for the bent beams whichwere an integral part of the voided slab superstructure.

I-

N

CDV

CO

of estimating that were slightly dif-ferent and resulted in cost estimateswith some variation. On the otherhand, many of the unit prices and la-bor estimates were very consistent.

Both companies considered thefollowing items in estimating cost:

1. Erection cost which includedboth labor and equipment.

2. Forming costs for slab over-hangs, curbs and parapets. Thisitem contained forming mate-rial as well as labor rates, andincluded a working space out-side the slab overhangs.

3. Concrete, including labor andmaterial cost.

4. Overhead, equipment, insur-ance and supervision.

In addition, reinforcing steel wasconsidered in the estimates. How-ever, the Bridge Division of the Mis-souri State Highway Commissionfurnis'. ed unit prices for this itemwhich included material and place-ment labor.

The contractors were also con-sulted regarding the completeness ofthe estimate. Each indicated thatthis procedure would provide ascomplete an estimate as possiblewithout detailed design drawingsand knowledge of the locations ofthe structures.

In general, material and laborcosts were the same for both con-tractors. The three principal differ-ences in the final estimate were inthe erection of the precast, pre-stressed U-beams, forming of theslab overhang, and in overhead.These differences partly account forthe variation in superstructure pricespresented.

SUMMARY OF SUPERSTRUCTURE COSTS

To complete the economic evalua-tion of the proposed composite U-beam bridge system, beam costs and

construction costs were combinedfor the three re-designed bridge su-perstructures. These combined costswere subsequently broken down intototal cost and percentages of totalcost according to the following divi-sions:

1. Total costs of the U-beam sys-tems and comparative sq. ft.costs of the original and re-de-signed structures

2. Percent for U-beams3. Percent for on-site construction

materials, including U-beams4. Percent for construction labor5. Percent for equipment, over-

head and insuranceSince four estimates for each beam

and two estimation methods wereavailable, several total cost figuresare possible. For comparison pur-poses, high, low and average valueswere considered. Costs and percent-ages are presented in Table 2.

Cost evaluation. From the priceslisted in Table 2, it can be seen thatfor the shorter span bridges, A-2141and A-2039, the cost of the originalsuperstructures exceeds the averageestimated cost of the proposed sys-tem by 10 and 13 percent, respec-tively. These values would indicatethat, even with the conservative na-ture of the estimate, the proposedsystem does not appear to be signif-icantly more economical than pres-ent types of construction. However,the estimated average cost of theproposed system for bridge A-2416is about 21 percent less than the ac-tual cost of the original voided slabsuperstructure. It appears that formedium span ranges the proposedsystem has an economic advantage.

Table 2 indicates that beam costaccounts for about 45 percent of to-tal superstructure cost. The esti-mates for these members were made

30 PCI Journal

Table 3. Cost per square foot of U-beam superstructureunder two types of loading

BridgeA-2141

BridgeA-2039

BridgeA-2416*

H15-44 HS20-44 H15-44 HS20-44 H15-44 - HS20-44

High $6.83 7.00 $7.44 7.54 $8.36 8.49

Average 6.63 6.76 7.18ҟ7.30 7.74ҟ7.87

Low 6.45 6.56 6.94 7.05 7.13 7.23

*Bent beam costs not included

without the benefit of production ex-perience and costs would be ex-pected to lower as this experience isgained. Because of the nature of thesystem, any reduction in the squarefoot cost of the beams results in al-most the same (about 87 percent) de-crease in the square foot cost of thesuperstructure.

The distribution of costs, as ex-pressed by the percentages of thetotal superstructure costs, should al-so be considered. For most bridgeconstruction, a minimum of 30 to 40percent of the total cost of the struc-ture is on-site labor cost, as com-pared to 10 to 15 percent for theproposed system. As a result, fluctua-tions in the labor market and varia-tions between local labor wagescales should have a relatively lighteffect upon the overall cost of thestructure.

Effect of bridge Ioads. In addition tothe AASHO H15 loading, all threesuperstructures were designed forHS20 loading. This resulted in an in-crease of cost averaging 2 percent.The superstructure costs for the twotypes of loading are given in Table3.

Construction time. An item whichcannot be considered from the infor-mation presented in this paper is thereduced construction time which ispossible with the proposed bridgesystem. An evaluation of the effect ofconstruction time involves many fac-tors which are known only after thebridge site has been selected andconstruction is considered alongwith the total highway project. How-ever, there are many instances whenthe construction of a bridge, orbridges, is critical to the completionof a project, whereby the construc-tion time becomes an important eco-nomic consideration.

CONCLUSIONS

The proposed U-beam bridge sys-tem has been shown to be structur-ally sound and economical for 30 to80 ft. (9 to 24 m) spans. Plans arepresently being made in Missouri toconstruct some trial structures usingthe proposed system. This construc-tion should not only help to verifythe cost estimates, but should alsoprovide useful information by whichsome of the unknown factors previ-ously mentioned can be evaluated.,

.May-June 1971 31

ACKNOWLEDGMENTS

This investigation was part of astudy sponsored by the MissouriState Highway Commission in coop-eration with the U.S. Department ofTransportation, Federal HighwayAdministration. Sincere appreciationis expressed to Wilson ConcreteCompany, Nebraska PrestressedConcrete Company, PrestressedConcrete of Iowa, Inc., Tobin Con-struction Company and WilkersonConstruction Company for their as-sistance with costing the proposedbridge system.

REFERENCES1. Salmons, J. R. and Poepsel, J. R., "In-

vestigation of a Prestressed-PrecastComposite Bridge System," MissouriCooperative Highway Research ProgramReport 68-8, University of Missouri, De-

partment of Civil Engineering, May1968.

2. Salmons, J. R. and Mokhtari, S., "ModelStudy of Prestressed-Precast CompositeBridge System," Missouri CooperativeHighway Research Program Report 68-9, University of Missouri, Departmentof Civil Engineering, June 1968.

3. Salmons, J. R. and Mokhtari, S., "Studyof Precast-Prestressed Model BridgeSlab," Missouri Cooperative HighwayResearch Program Report 68-14, Uni-versity of Missouri, Department of CivilEngineering, November 1968.

4. Salmons, J. R. and Kagay, W. J., "Eco-nomic Evaluation of the Proposed Pre-cast-Prestressed Bridge System," Mis-souri Cooperative Highway ResearchProgram Report 69-1, University ofMissouri, Department of Civil Engineer-ing, January 1969.

5. Salmons, J. R., "Study of a ProposedPrecast-Prestressed Composite BridgeSystem," Final Report, Missouri Coop-erative Highway Research Program Re-port No. 69-2, University of Missouri,Department of Civil Engineering, April1970.

Discussion of this paper is invited. Please forward your discussion to PCI Headquartersby Sept. I to permit publication in the Sept.-Oct. 1971 issue of the PCI JOURNAL.

32ҟ PCI Journal