Project Summary Report 1852-S 1

The University of Texas at Austin

C e n t e r f o rTransportation Research

PR

OJ

EC

T

S

UM

MA

RY

RE

PO

RT

CENTER FOR TRANSPORTATION RESEARCHTHE UNIVERSITY OF TEXAS AT AUSTIN

Project Summary Report 1852-SPrestressed Structural Lightweight Concrete Beams

Authors: J. A. Heffington, R. T. Kolozs, G. S. Sylva III, D. B. Thatcher, J. E. Breen, and N. H. BurnsDecember 2001

Summary: High Performance LightweightConcrete Prestressed Girders and Panels

A significant portion ofthe load carried by prestressedconcrete bridge girders is theself-weight of the girders anddeck. If all or part of thegirder and deck can be madeusing lightweight concrete,there is a potential for eco-nomic savings since the self-weight could be reduced by asmuch as 15-20%. This projectwas undertaken to explore theadvantages and disadvantagesof using high performancelightweight concrete (HPLC)pre-tensioned bridge girdersand deck panels.

With normalweight con-crete, high strength or highperformance is ordinarilydefined as concrete withstrengths ranging from 8000 to12,000 psi. In lightweightconcrete, high performance maymore often refer to strengthsgreater than or equal to6000 psi. The higher strength isachieved through a mix designwith a low water/cement ratioand admixtures. The possiblematerial savings must bebalanced against the higher costof the HPLC, and will varydepending on the specific jobconditions.

An additional possibleapplication of HPLC is inprecast pretensioned deck

panels. These panels compriseapproximately half the thick-ness of the deck. They arepositioned on top of thegirders, spanning the spacesbetween them. The panels actas formwork. They are left inplace when the rest of the deckis placed on top of them toform a composite system.This eliminates the difficultyand cost of getting underneaththe bridge in order to removetypical wooden or metalformwork. The precast pre-stressed panel serves as a stay-in-place form to support cast-in-place concrete. Lightweightconcrete panels could reduceoverall bridge dead load.

The overall objective ofthe research was to determinewhether high performancelightweight concrete is afeasible material for use inTexas highway bridges. Themain objective of the earlyportion of the project was todetermine whether high-strength high-performancelightweight concrete mixdesigns could be developedwith f c = 6000 psi and f

c =

8000 psi for use in prestressedconcrete girders. The equilib-rium unit weight desired wasto be less than 125 pounds percubic foot (pcf).

The intermediate portionof the project focused onstructural factors includingtransfer and developmentlength of prestressing strand inlightweight concrete, flexuraltesting of lightweight pre-stressed concrete beams, theuse of lightweight prestresseddeck panels, and the manufac-ture of the beams and the deckpanels. This phase includedfull scale testing of AASHTOType I (Texas Type A)precast pretensioned pre-stressed concrete beams withcomposite slabs.

The final phase was aneconomic and analytical studyof practical bridge applications.

What We Did...

The initial portion of theproject investigated mixdesigns for usable high-strength lightweight concretemixes for possible use inpretensioned bridge girders.Two distinct concrete mixeswere developed. One wasintended to have a 28-daystrength of 6000 psi while theother was intended to have a28-day strength of 8000 psi.(Subsequent results indicatedthat the 8000 psi design wasimpractical and the upper

Project Summary Report 1852-S 2

strength was re-rated as 7500 psi.)Both also were intended to have astrength of at least 3500 psi at oneday of age to expedite release ofprestress in precast plants.

To obtain these two mixes, anextensive laboratory mixing andtesting program was carried out.Thirty-five mixes were designedand fabricated. For each of themixes, mechanical behavior andworkability was determined. Fromthese mixes, two mix designs werechosen that combined the bestmechanical performance withadequate workability performance.

Using each mix, one highlycongested 20-foot long and twonormal 40-foot long beams wereplant cast. The 20-foot longbeams represented the highestdegree of congestion that mightreasonably be expected.

After the forms were removedand the transfer length testing wasperformed, the beams werebrought to Austin for testing thedevelopment length and flexuralcapacity of the beams. A group ofprestressed lightweight concretepanels were cast by a local fabri-cator. Transfer length tests wereperformed there.

Extensive comparative analy-ses of a typical bridge span usingAASHTO Type IV bridge girdersand decks made from variouscombinations of normal weightand lightweight concrete wereperformed using the TxDOTprestressed girder design program,PSTRS14. Three different com-posite deck combinations wereused in the analyses.

Finally, an economic andconstructability analysis wascarried out. Several Texas precastconcrete product manufacturerssupplied premium costs forlightweight concrete and beammanufacture. These studies

assessed the economic practicalityof using high performance light-weight concrete.

What We Found...The mix design phase of the

project was moderately successful.A highly dependable 6000 psi mixwas attained. Mix design isincluded in Report 1852-1. The6000 psi mix was controlled by the1-day strength requirement.4000 psi at 1 day was achievedwith 7.15 sacks of Type III cementwith 25% replacement of fly ash.The 28-day strength of the con-crete averaged 7200 psi in thelaboratory and 7800 psi in thefield. The fresh unit weight of theconcrete was 127 pounds per cubicfoot (pcf), which later decreased to118 pcf at equilibrium conditions.This concrete mix provided about30 minutes of working time underroom temperature and averagehumidity conditions. The concreteplaced well at the precast plant.

The desired 8000 psi wasattained in the laboratory, but fieldstudies indicated that 7500 psi wasa more practical maximum. Mixdesign is included in Report1852-1.

To reach 8000 psi at 28 days,10.5 sacks of cementitious mate-rial were required. The 1-daystrength was 5500 psi, easily sur-passing the 1-day strength require-ment. The 28-day strength of thismix in the laboratory was 8600 psibut in the field was probably only7500 psi, and the 180-day strengthof this mix in the field was 7900psi. The fresh unit weight of theconcrete was 129 pcf, dropping to122 pcf at equilibrium conditions.The concrete was somewhatdifficult to work in the laboratory.At room temperature and humidityconditions, this concrete givesonly about 20 minutes of workabil-

ity. At the precast plant, theworkers had trouble placing theconcrete. However, as shown inFigure 1, the final product wasvery satisfactory in appearance.

Overall, the performance ofthe nominal 8000 psi concrete wassomewhat disappointing. Theperformance in the field was notadequate for 8000 psi concrete. Itwould be acceptable for a nominal7500 psi mix.

The structural test phase of theproject was very successful. Itwas determined that the ACI andAASHTO code expressions fortransfer length (50 db) are aconservative estimate of thetransfer length for in. strands innormalweight concrete (37 db), butunderestimate the transfer lengthfor in. strands in lightweightconcrete (70 db). The transferlength at transfer of the prestress-ing force for 3/8-in. strands in thelightweight concrete deck panelsin this study was 43 db, about 10%longer than for normalweightconcrete panels. Clearly, themodulus of elasticity is an impor-tant factor in determining thetransfer length for both normal-weight and lightweight concrete.The actual development length for in strands in all girders was lessthan 60 in. The AASHTO/ACIdevelopment length equation isconservative for both normal-weight and lightweight concretegirders. The ultimate momentcapacity developed by the light-weight concrete girders and theload-deflection curves for thosegirders were virtually identical tothose for the normalweight con-crete girders. The full AASHTOultimate moment capacity of thegirders was developed with asubstantial reserve.

The comparative analysisstudy of this project was quite

Project Summary Report 1852-S 3

.

successful and showed manyunexpected results. It was deter-mined that higher prestress lossesand lower allowable tensilestresses dramatically reduced theefficiency of the lightweightconcrete girders. Elastic shorten-ing was determined to be the keyprestress loss variable in light-weight concrete. While light-weight girders using 7500 psiconcrete had no problem satisfyingallowable stress criteria and couldbe used for the standard TxDOTbridge section (110 feet span and8.5 feet beam spacing) used in theanalysis, lightweight concretegirders using 6000 psi concretemix design could not satisfyallowable stress criteria at release.Hence, the maximum span lengththat could be achieved would beless than 110 feet. Lightweightconcrete panels, contingent uponverification of allowable tensilestresses under fluid deck concretedead load capacity, could beimplemented and could potentiallyproduce savings in shipping andhandling costs as well as reductionof dead load transmitted to thegirders and substructure. Use ofthe standard TxDOT designprogram PSTRS14 showed thatrevisions were required to make itfully functional for the design oflightweight concrete beams.

The economic studies showedthat the premium cost of light-weight concrete ranged from $6/cyto $30/cy, with the average of allpremium costs equal to $18.50/cy.Comparisons of normalweightconcrete and lightweight concretebridge girders shipped approxi-mately 40 miles revealed 10 to15% higher unit cost for thelightweight girders.

Details of the complete testingprogram, comparative analyses,and an economic factors study are

provided in Report 1852-1 (Struc-tural Lightweight Concrete Pre-stressed Girders and Panels) andReport 1852-2 (Feasibility ofUtilizing High Performance Light-weight Concrete in PretensionedBridge Girders and Panels).

The Researchers Recommend...

The research team recom-mends that the use of lightweightconcrete precast deck panels beimmediately considered as analternate to normalweight concreteprecast deck panels. A checkshould be made of the standardpanels to determine in whichlengths the allowable tensile stressunder fluid deck concrete deadload governs. Panel ranges thatare sensitive to the tensile stresslimits could be required to benormalweight concrete or higherprestress levels could be specified.

Pretensioned lightweightconcrete girders can be used inapplications where economicanalysis and/or constructionlimitations such as crane capacitymake them desirable. It is neces-sary to check their shear capacityfor the longer transfer length notedand the reduced tensile strength ofthe lightweight concrete. Theseshould not be severe limits butmay require some additional shearreinforcement.

The increased losses due toelastic shortening which reducethe economic savings with light-weight concrete pretensionedapplications are not a factor inpost-tensioned applications.Lightweight concrete should beexamined for possible applicationin post-tensioned bridges.

Figure 1 Finish of Girder with 8000 psi Concrete Mix

Project Summary Report 1852-S 4

The University of Texas at AustinCenter for Transportion Research3208 Red River, Suite #200Austin, TX 78705-2650

Disclaimer

Research Supervisors: John E. Breen, P.E., (512) 471-4578, jbreen@mail.utexas.eduNed H. Burns, P.E., (512) 471-7298, nedburns@mail.utexas.edu

TxDOT Project Director: Tom Rummel, P.E., (512) 416-2254, trummel@dot.state.tx.usThe research is documented in the following reports:

Research Report 1852-1, Structural Lightweight Concrete Prestressed Girders and Panels,September 2001.Research Report 1852-2, Feasibility of Utilizing High-Performance Lightweight Concretein Pretensioned Bridge Girders and Panels, January 2002.

To obtain copies of the above reports, contact the Center for Transportation Research, The University ofTexas at Austin, (512) 232-3126, ctrlib@uts.cc.utexas.edu.

The results of this project gave TxDOT the basic tools (concrete batch designs and modifications to designprocesses) needed to design prestressed beams using lightweight concrete, should that prove necessary. It also demon-strated that fabrication of typical beams should be fairly straightforward using lightweight concrete, and those beamsshould behave satisfactorily under typical loading conditions. The research did, however, show that the use of thismaterial, for a number of reasons, would probably not be useful in most prestressed beam applications except in rarecircumstances.

The use of lightweight concrete deck panels will not be implemented at this time, due to unresolved questionsregarding economics and design constraints due to tensile stress limits for deck panel applications. Further work in thisarea is being contemplated to resolve these issues.

The research did provide useful results for implementation in certain specific instances for post-tensionedlightweight concrete structural elements on a case by case basis, although not as a standard process. The necessaryinformation was developed for designers to utilize this option should conditions warrant.

For more information, please contact Mr. Tom Yarbrough, P.E., RTI Research Engineer at (512) 465-7685 oremail at tyarbro@dot.state.tx.us.

This research was performed in cooperation with the Texas Department of Transportation and the U. S. Department ofTransportation, Federal Highway Administration. The content of this report reflects the views of the authors, who are responsiblefor the facts and accuracy of the data presented herein. The contents do not necessarily reflect the official view or policies of theFHWA or TxDOT. This report does not constitute a standard, specification, or regulation, nor is it intended for construction,bidding, or permit purposes. Trade names were used solely for information and not for product endorsement. The engineers incharge were John E. Breen, P.E. (Texas No. 18479) and Ned H. Burns, P.E. (Texas No. 20801).

Your Involvement Is Welcome!

For More Details...

TxDOT Implementation StatusFebruary 2002

US Postage PaidPermit No. 391

Austin TXThe University of Texas at AustinC e n t e r f o rTransportation Research