土木工程外文文獻(xiàn)

MaterialsandStructures?

RILEM

10.1617/s11527-010-9700-yOriginalArticleImpactofcrackwidthonbond:confinedandunconfinedrebarDavid

W.

Law1

,Denglei

Tang2,Thomas

K.

C.

Molyneaux3andRebecca

Gravina3(1)

SchooloftheBuiltEnvironment,HeriotWattUniversity,Edinburgh,EH144AS,UK(2)

VicRoads,Melbourne,VIC,Australia(3)

SchoolofCivil,EnvironmentalandChemicalEngineering,RMITUniversity,Melbourne,VIC,3000,AustraliaDavid

W.

Law

Email:Received:14

January

Accepted:14

December

Publishedonline:23

December

Abstract

Thispaperreportstheresultsofaresearchprojectcomparingtheeffectofsurfacecrackwidthanddegreeofcorrosiononthebondstrengthofconfinedandunconfineddeformed12and16

mmmildsteelreinforcingbars.ThecorrosionwasinducedbychloridecontaminationoftheconcreteandanappliedDCcurrent.Theprincipalparametersinvestigatedwereconfinementofthereinforcement,thecoverdepth,bardiameter,degreeofcorrosionandthesurfacecrackwidth.Theresultsindicatedthatpotentialrelationshipbetweenthecrackwidthandthebondstrength.Theresultsalsoshowedanincreaseinbondstrengthatthepointwhereinitialsurfacecrackingwasobservedforbarswithconfiningstirrups.Nosuchincreasewasobservedwithunconfinedspecimens.Keywords

Bond

-

Corrosion

-

Rebar

-

Cover

-

Crackwidth

-

Concrete1

IntroductionThecorrosionofsteelreinforcementisamajorcauseofthedeteriorationofreinforcedconcretestructuresthroughouttheworld.Inuncorrodedstructuresthebondbetweenthesteelreinforcementandtheconcreteensuresthatreinforcedconcreteactsinacompositemanner.However,whencorrosionofthesteeloccursthiscompositeperformanceisadverselyaffected.Thisisduetotheformationofcorrosionproductsonthesteelsurface,whichaffectthebondbetweenthesteelandtheconcrete.Thedeteriorationofreinforcedconcreteischaracterizedbyageneralorlocalizedlossofsectiononthereinforcingbarsandtheformationofexpansivecorrosionproducts.Thisdeteriorationcanaffectstructuresinanumberofways;theproductionofexpansiveproductscreatestensilestresseswithintheconcrete,whichcanresultincrackingandspallingoftheconcretecover.Thiscrackingcanleadtoacceleratedingressoftheaggressiveagentscausingfurthercorrosion.Itcanalsoresultinalossofstrengthandstiffnessoftheconcretecover.Thecorrosionproductscanalsoaffectthebondstrengthbetweentheconcreteandthereinforcingsteel.Finallythecorrosionreducesthecrosssectionofthereinforcingsteel,whichcanaffecttheductilityofthesteelandtheloadbearingcapacity,whichcanultimatelyimpactupontheserviceabilityofthestructureandthestructuralcapacity[12,25].Previousresearchhasinvestigatedtheimpactofcorrosiononbond[2–5,7,12,20,23–25,27,29],withanumberofmodelsbeingproposed[4,6,9,10,18,19,24,29].Themajorityofthisresearchhasfocusedontherelationshipbetweenthelevelofcorrosion(masslossofsteel)orthecurrentdensitydegree(corrosioncurrentappliedinacceleratedtesting)andcrackwidth,orontherelationshipbetweenbondstrengthandlevelofcorrosion.Otherresearchhasinvestigatedthemechanicalbehaviourofcorrodedsteel[1,11]andthefrictioncharacteristics[13].However,littleresearchhasfocusedontherelationshipbetweencrackwidthandbond[23,26,28],aparameterthatcanbemeasuredwithrelativeeaseonactualstructures.Thecorrosionofthereinforcingsteelresultsintheformationofironoxideswhichoccupyalargervolumethanthatoftheparentmetal.Thisexpansioncreatestensilestresseswithinthesurroundingconcrete,eventuallyleadingtocrackingofthecoverconcrete.Oncecrackingoccursthereisalossofconfiningforcefromtheconcrete.Thissuggeststhatthelossofbondcapacitycouldberelatedtothelongitudinalcrackwidth[12].However,theuseofconfinementwithintheconcretecancounteractthislossofbondcapacitytoacertaindegree.Researchtodatehasprimarilyinvolvedspecimenswithconfinement.Thispaperreportsastudycomparingthelossofbondofspecimenswithandwithoutconfinement.2

Experimentalinvestigation2.1

SpecimensBeamendspecimens[28]wereselectedforthisstudy.Thistypeofeccentricpulloutor‘beamend’typespecimenusesabondedlengthrepresentativeoftheanchoragezoneofatypicalsimplysupportedbeam.Specimensofrectangularcrosssectionwerecastwithalongitudinalreinforcingbarineachcorner,Fig.

1.An80

mmplastictubewasprovidedatthebarunderneaththetransversereactiontoensurethatthebondstrengthwasnotenhancedduetoa(transverse)compressiveforceactingonthebaroverthislength.Fig.

1

BeamendspecimenDeformedrebarof12and16

mmdiameterwithcoverofthreetimesbardiameterwereinvestigated.Duplicatesetsofconfinedandunconfinedspecimensweretested.Theconfinedspecimenshadthreesetsof6

mmstainlesssteelstirrupsequallyspacedfromtheplastictube,at75

mmcentres.Thisrepresentsfourgroupsofspecimenswithacombinationofdifferentbardiameterandwith/withoutconfinement.Thespecimenswereselectedinordertoinvestigatetheinfluenceofbarsize,confinementandcrackwidthonbondstrength.2.2

MaterialsThemixdesignisshown,Table

1.ThecementwasTypeIPortlandcement,theaggregatewasbasaltwithspecificgravity2.99.ThecoarseandfineaggregatewerepreparedinaccordancewithAS1141-.MixingwasundertakeninaccordancewithAS1012.2-1994.Specimenswerecuredfor28

daysunderwethessianbeforetesting.Table

1

ConcretemixdesignMaterialCementw/cSand10

mmwashedaggregate7

mmwashedaggregateSaltSlumpQuantity381

kg/m30.49517

kg/m3463

kg/m3463

kg/m318.84

kg/m3140

±

25

mmInordertocomparebondstrengthforthedifferentconcretecompressivestrengths,Eq.

1isusedtonormalizebondstrengthfornon-corrodedspecimensashasbeenusedbyotherresearcher[8].(1)whereisthebondstrengthforgrade40concrete,τexptlistheexperimentalbondstrengthandfcistheexperimentalcompressivestrength.ThetensilestrengthoftheΦ12andΦ16

mmsteelbarswasnominally500

MPa,whichequatestoafailureloadof56.5and100.5

kN,respectively.2.3

ExperimentmethodologyAcceleratedcorrosionhasbeenusedbyanumberofauthorstoreplicatethecorrosionofthereinforcingsteelhappeninginthenaturalenvironment[2,3,5,6,10,18,20,24,27,28,30].Thesehaveinvolvedexperimentsusingimpressedcurrentsorartificialweatheringwithwet/drycyclesandelevatedtemperaturestoreducethetimeuntilcorrosion,whilemaintainingdeteriorationmechanismsrepresentativeofnaturalexposure.Studiesusingimpressedcurrentshaveusedcurrentdensitiesbetween100

μA/cm2and500

mA/cm2[20].Researchhassuggestedthatcurrentdensitiesupto200

μA/cm2resultinsimilarstressesduringtheearlystagesofcorrosionwhencomparedto100

μA/cm2[21].Assuchanappliedcurrentdensityof200

μA/cm2wasselectedforthisstudy—representativeofthelowerendofthespectrumofsuchcurrentdensitiesadoptedinpreviousresearch.However,cautionshouldbeappliedwhenacceleratingthecorrosionusingimpressedcurrentastheaccelerationprocessdoesnotexactlyreplicatethemechanismsinvolvedinactualstructures.Inacceleratedteststhepitsarenotallowedtoprogressnaturally,andtheremaybeamoreuniformcorrosiononthesurface.Alsotherateofcorrosionmayimpactonthecorrosionproducts,suchthatdifferentoxidationstateproductsmaybeformed,whichcouldimpactonbond.Thesteelbarsservedastheanodeandfourmildsteelmetalplateswerefixedonthesurfacetoserveascathodes.Sponges(sprayedwithsaltwater)wereplacedbetweenthemetalplatesandconcretetoprovideanadequatecontact,Fig.

2.Fig.

2

AcceleratedcorrosionsystemWhentherequiredcrackwidthwasachievedforaparticularbar,theimpressedcurrentwasdiscontinuedforthatbar.Thespecimenwasremovedforpullouttestingwhenallfourlocationsexhibitedthetargetcrackwidth.Averagesurfacecrackwidthsof0.05,0.5,1and1.5

mmwereadoptedasthetargetcrackwidths.Thesurfacecrackwidthwasmeasuredat20

mmintervalsalongthelengthofthebar,beginning20

mmfromtheendofthe(plastictube)bondbreakerusinganopticalmicroscope.Thelevelofaccuracyinthemeasurementswas±0.02

mm.Measurementsofcrackwidthweretakenonthesurfacenormaltothebardirectionregardlessoftheactualcrackorientationatthatlocation.Bondstrengthtestswereconductedbymeansofahandoperatedhydraulicjackandacustom-builttestrigasshowninFig.

3.TheloadingschemeisillustratedinFig.

4.Aplastictubeoflength80

mmwasprovidedattheendoftheconcretesectionunderneaththetransversereactiontoensurethatthebondstrengthwasnotenhancedbythereactive(compressive)force(actingnormaltothebar).Thespecimenwaspositionedsothatanaxialforcewasappliedtothebarbeingtested.Therestraintsweresufficientlyrigidtoensureminimalrotationortwistingofthespecimenduringloading.Fig.

3

Pull-outtest,16

mmbarunconfinedFig.

4

Schematicofloading.Note:onlytestbarshownforclarity3

Experimentalresultsanddiscussion3.1

VisualinspectionFollowingtheacceleratedcorrosionphaseeachspecimenwasvisuallyinspectedforthelocationofcracks,meancrackwidthandmaximumcrackwidth(Sect.

2.3).Whileeachspecimenhadameantargetcrackwidthforeachbar,variationsinthiscrackwidthwereobservedpriortopullouttesting.Thisisduetocorrosionandcrackingbeingadynamicprocesswithcrackspropagatingatdifferentrates.Thus,whileindividualbarsweredisconnected,oncethetargetcrackwidthhadbeenachieved,corrosionandcrackpropagationcontinued(tosomeextent)untilallbarshadachievedthetargetcrackwidthandpullouttestsconducted.Thisresultedinarangeofdataforthemaximumandmeancrackwidthsforthepullouttests.Thevisualinspectionofthespecimensshowedthreestagestothecrackingprocess.Theinitialcracksoccurredinaveryshortperiod,usuallygeneratedwithinafewdays.Afterthat,mostcracksgrewataconstantrateuntiltheyreached1

mm,3–4

weeksafterfirstcracking.Aftercrackshadreached1

mmtheythengrewveryslowly,withsomecracksnotincreasingatall.Fortheconfinedandunconfinedspecimensthesurfacecrackstendedtooccuronthesideofthespecimens(asopposedtothetoporbottom)andtofollowthelineofthebars.Inthecaseoftheunconfinedspecimensingeneraltheseweretheonlycrackwhileitwascommoninthecasesofconfinedspecimenstoobservecracksthatwerealignedverticallydowntheside—adjacenttooneofthelinks,Fig.

5.Fig.

5

TypicalcrackpatternsDuringthepull-outtestingthemostcommonfailuremodeforbothconfinedandunconfinedwassplittingfailure—withtheinitial(pre-test)crackscausedbythecorrosionenlargingunderloadandultimatelyleadingtothesectionfailingexhibitingspallingofthetopcorner/edge,Fig.

6.Howeverforseveraloftheconfinedspecimens,asecondmodeoffailurealsooccurredwithdiagonal(shearlike)cracksappearinginthesidewalls,Fig.

7.Theappearanceofthesecracksdidnotappeartoberelatedtothepresenceofverticalcracksobserved(inspecimenswithstirrups)duringthecorrosionphaseasreportedabove.Fig.

6

Longitudinalcrackingafterpull-outFig.

7

Diagonalcrackingafterpull-outThebarswereinitially(precasting)cleanedwitha12%hydrochloricacidsolution,thenwashedindistilledwaterandneutralizedbyacalciumhydroxidesolutionbeforebeingwashedindistilledwateragain.Followingthepull-outtests,thecorrodedbarswerecleanedinthesamewayandweighedagain.ThecorrosiondegreewasdeterminedusingthefollowingequationwhereG0istheinitialweightofthesteelbarbeforecorrosion,Gisthefinalweightofthesteelbarafterremovalofthepost-testcorrosionproducts,g0istheweightperunitlengthofthesteelbar(0.888and1.58

g/mmforΦ12andΦ16

mm

bars,respectively),listheembeddedbondlength.Figures

8and9showsteelbarswithvaryingdegreeofcorrosion.Themajorityexhibitedvisiblepitting,similartothatobservedonreinforcementinactualstructures,Fig.

9.However,asmallnumberofothersexhibitedsignificantoverallsectionloss,withamoreuniformlevelofcorrosion,Fig.

8,whichmaybeafunctionoftheaccelerationmethodology.Fig.

8

Corroded12

mmbarwithapproximately30%masslossFig.

9

Corroded16

mmbarwithapproximately15%massloss3.2

BondstressandcrackwidthFigure

10showsthevariationofbondstresswithmeancrackwidthfor16

mm

barsandFig.

11forthe12

mm

bars.Figures

12and13showthedataforthemaximumcrackwidth.Fig.

10

Meancrackwidthversusbondstressfor16mmbarsFig.

11

Meancrackwidthversusbondstressfor12

mmbarsFig.

12

Maximumcrackwidthversusbondstressfor16

mmbarsFig.

13

Maximumcrackwidthversusbondstressfor12

mmbarsThedatashowaninitialincreaseinbondstrengthforthe12

mmspecimenswithstirrups,followedbyasignificantdecreaseinbond,whichisinagreementwithotherauthors[12,15].Forthe16

mmspecimensanincreaseonthecontrolbondstresswasobservedforspecimenswith0.28and0.35

mmmeancrackwidths,however,adecreaseinbondstresswasobservedforatthemeancrackwidthof0.05

mm.The12

mmbarswithstirrupsdisplayedanincreaseinbondstressofapproximately25%fromthecontrolvaluestothemaximumbondstress.Anincreaseofapproximately14%wasobservedforthe16

mmspecimens.Otherresearchers[17,24,25]havereportedenhancementsofbondstressofbetween10and60%duetoconfinement,slightlyhighertothatobservedintheseexperiment.Howevertheloadingtechniquesandcoverdepthshavenotallbeenthesame.Variationsinexperimentaltechniquesincludeashorterembeddedlengthandalowercover.Thevariationontheproposedempiricalrelationshipbetweenbondstrength,degreeofcorrosion,barsize,cover,linkdetailsandtensilestrengthpredictedbyRodriguez[24]hasbeendiscussedindetailinTangetal.[28].Theanalysisdemonstratesthattherewouldbeanexpectedenhancementofbondstrengthduetoconfinementofapproximately25%—correspondingtoachangeofbondstrengthofapproximately0.75

MPaforthe16

mmbars(assessedata2%sectionloss).Forthe12

mmbarsthecorrespondingeffectofconfinementisfoundtobeapproximately35%correspondingtoa1.0

MPadifferenceinbondstress.Theexperimentalresults(14and25%,above)are60–70%ofthesevalues.Bothsetsofdataindicatearelationshipshowingdecreasingbondstrengthwith(visiblesurface)crackwidth.Aregressionanalysisofthebondstrengthdatarevealsabetterlinearrelationshipwiththemaximumcrackwidthasopposedtothemeancrackwidth(excludingtheuncrackedconfinedspecimens),Table

2.Table

2

Bestfitparameters,crackwidthversusbondstrength

Unconfined12

mmConfined12

mmUnconfined16

mmConfined16

mmMeancrackwidth

R20.9200.6370.6720.659

Slope(m)?3.997?3.653?2.999?8.848

Intercept(b)7.5608.1226.4968.746Maximumcrackwidth

R20.9370.8550.7140.616

Slope(m)?2.719?2.968?1.815?5.330

Intercept(b)7.8058.4036.7079.636Therewasalsoasignificantlybetterfitfortheunconfinedspecimensthantheconfinedspecimens.Thisisconsistentwiththeobservationthatintheunconfinedspecimensthebondstrengthwillberelatedtothebondbetweenthebarsandtheconcrete,whichwillbeaffectedbythelevelofcorrosionpresent,whichitselfwillinfluencethecrackwidth.Inconfinedspecimenstheconfiningsteelwillimpactuponboththebondandthecracking.3.3

CorrosiondegreeandbondstressItisapparentthat(Fig.

14)forcorrosiondegreeslessthan5%thebondstresscorrelatedwell.However,asthedegreeofcorrosionincreasedtherewasnoobservablecorrelationatall.Thiscontrastswiththerelationshipbetweentheobservedcrackwidthandbondstress,whichgivesareasonablecorrelation,evenascrackwidthsincreaseto2and2.5

mm.Apossibleexplanationforthisvariationisthatintheinitialstagesofcorrosionvirtuallyallthedissolvedironionsreacttoformexpansivecorrosionproducts.Thisreactionimpactsonboththebondstressandtheformationofcracks.However,oncecrackshavebeenformeditispossiblefortheironionstobetransportedalongthecrackandoutoftheconcrete.Asthebondhasalreadybeeneffectivelylostatthecrackanyironionsdissolvingatthecrackandbeingdirectlytransportedoutoftheconcretewillcauseanincreaseinthedegreeofcorrosion,butnotaffectthesurfacecrackwidth.Thelocation,orientationandchemistrywithinthecrackwillcontroltherelationshipbetweenbondstressanddegreeofcorrosion,whichwillvaryfromspecimentospecimen.Hencethelargevariationsincorrosiondegreeandbondstressforhighlevelsofcorrosion.Fig.

14

Bondstressversuscorrosiondegree,12

mmbars,unconfinedspecimenSignificantlylargercrackwidthswereobservedfortheunconfinedspecimens,comparedtotheconfinedspecimenswithsimilarlevelsofcorrosionandmasslost.Thelargestobservedcrackforunconfinedspecimenswas2.5

mmcomparedto1.4

mmfortheconfinedspecimens.Thisisasexpectedandisadirectresultoftheconfinementwhichlimitsthedegreeofcracking.3.4

EffectofconfinementTheunconfinedspecimensforboth16and12

mmbarsdidnotdisplaytheinitialincreaseinbondstrengthobservedfortheconfinedbars.Indeedtheunconfinedspecimenswithcracksalldisplayedareducedbondstresscomparedtothecontrolspecimens.Thisisinagreementwithotherauthors[16,24]findingsforcrackedspecimens.IncrackedcorrodedspecimensFangobservedasubstantialreductioninbondstrengthfordeformedbarswithoutstirrups,whileRodriguezobservedbondstrengthsofhighlycorrodedcrackedspecimenswithoutstirrupswereclosetozero,whilehighlycorrodedcrackedspecimenswithstirrupsretainedbondstrengthsofbetween3and4

MPa.InuncorrodedspecimensChananotedanincreaseinbondstrengthduetostirrupsofbetween10and20%[14].HoweverRodriguezandFangobservednovariationduetothepresenceofconfinementinuncorrodedbars.Thedataisperhapsunexpectedasitcouldbeanticipatedthatthecorrosionproductswouldleadtoanincreaseinbondduetotheincreaseininternalpressures,causedbythecorrosionproductsincreasingtheconfinementandmechanicalinterlockingaroundthebar,coupledwithincreasedroughnessofthebarresultinginagreaterfrictionbetweenthebarandthesurroundingconcrete.However,thesepressureswouldthenrelievedbythesubsequentcrackingoftheconcrete,whichwouldcontributetothedecreaseinthebondstrengthascrackwidthsincrease.Apossiblehypothesisisthatduetothelevelofcover,threetimesbardiameter,theeffectofconfinementbythestirrupsisreduced,suchthatithaslittleimpactonthebondstressinuncrackedconcrete.However,oncecrackinghastakenplacetheconfinementdoeshaveabeneficialeffectonthebond.Itmayalsobethatthecompressivestrengthoftheconcretecombinedwiththecoverwillhaveaneffectonthebondstressesforuncorrodedspecimens.Thedatapresentedherehasacoverofthreetimesbardiameterandastrengthof40

MPa,otherresearchrangesfrom1.5tofourtimescoverwithcompressivestrengthsfrom40to77

MPa.3.5

Comparisonof12and16

mmrebarThemaximumbondstressfor16

mmunconfinedbarswasmeasuredat8.06

MPaandforthe12

mm

barsitwas8.43

MPa.Thesebothcorrespondedtothecontrolspecimenswithnocorrosion.Theunconfinedspecimensforboththe12and16

mm

barsshowednoincreaseinbondstressduetocorrosion.Fortheconfinedspecimensthemaximumbondstressforthecontrolspecimenswere7.29

MPaforthe12

mm

barsand6.34

MPaforthe16

mmbars.Themaximumbondstressforbothsetsofconfinedspecimenscorrespondedtopointoftheinitialcracking.Themaximumbondstresseswereobservedatameancrackwidthof0.01

mmforthe12

mmbarsand0.28

mmforthe16

mmbars.Thecorrespondingbondstresseswere,8.45and7.20

MPa.Overallthe12

mmbarsdisplayedhigherbondstressescomparedtothe16

mmbarsatallcrackwidths.Thisisattributedtoadifferentfailuremode.The16

mmspecimensdemonstratesplittingfailurewhilethe12

mmbarsbondfailure.3.6

EffectofcastingpositionTherewasnosignificantdifferenceofbondstrengthduetothepositionofthebar(toporbottomcast)oncecrackingwasobserved,Fig.

15.Forcontrolspecimens,withnocorrosion,however,thebottomcastbarshadaslightlyhigherbondstressthanthetopcastbars.Theseobservationsareinagreementwithotherauthors[4,11,15,22].Itisgenerallyacceptedthatuncorrodedbottomcastbarshavesignificantlyimprovedbondcomparedtotopcastbarsduetothecorrosionproductsfillingthevoidsthatareoftenpresentundertopcastbarsasthecorrosionprogresses[14].Thecorrosionalsoactsasan‘a(chǎn)nchor’,similartotheribsondeformedbars,toincreasethebond.Overall,themeanvalueofbondstressforallbars(corrodedanduncorroded)locatedinthetopwerewithin1%ofthemeanbondstressofallbarslocatedinthebottomofthesection—forbothunconfinedandconfinedbars.Thisisprobablyduetothelevelofcover.Theresultsreportedpreviouslyareonspecimenswithonetimescover[14].However,atthreetimescoveritwouldbeanticipatedthatgreatercompactionwouldbeachievedaroundthetopcastbars.Thustheareaofvoidswouldbereducedandthustheeffectofthecorrosionproductfillingthesevoidsandincreasingthebondstrengthwouldbereduced.Fig.

15

Bondstressversusmeancrackwidthfor12

mmbars,topandbottomcastpositions,confinedspecimen4

Conclusions?

Arelationshipwasobservedbetweencrackwidthandbondstress.Thecorrelationwasbetterformaximumcrackwidthandbondstressthanformeancrackwidthandbondstress.?

Confinedbarsdisplayedahigherbondstressatthepointofinitialcrackingthanwherenocorrosionhadoccurred.Ascrackwidthincreasethebondstressreducedsignificantly.?

Unconfinedbarsdisplayedadecreaseinbondstressatinitialcracking,followedbyafurtherdecreaseascrackingincreased.?

Topcastbarsdisplayedahigherbondstressinspecimenswithnocorrosion.Oncecrackinghadoccurrednovariationbetweentopandbottomcastbarswasobserved.?

The12

mmbarsdisplayedhigherbondstressvaluesthan16

mmwithnocorrosion,controlspecimens,andatsimilarcrackwidths.?

Agoodcorrelationwasobservedbetweenbondstressanddegreeofcorrosionwasobservedatlowlevelsofcorrosion(lessthan5%).However,athigherlevelsofcorrosionnocorrelationwasdiscerned.Overalltheresultsindicatedapotentialrelationshipbetweenthemaximumcrackwidthandthebond.Resultsshownhereinshouldbeinterpretedwithcautionasthisvariationmaybenotonlyduetovariationsbetweenacceleratedcorrosionandnaturalcorrosionbutalsoduetothecomplexityofthecrackingmechanisminreality.References1.AlmusallamAA()Effectofdegreeofcorrosiononthepropertiesofreinforcingsteelbars.ConstrBuildMater15:361–368

2.AlmusallamAA,Al-GhataniAS,AzizAR,Rasheeduzzafar(1996)Effectofreinforcementcorrosiononbondstrength.ConstrBuildMater10(2):123–129

3.AlonsoC,AndradeC,RodriguezJ,DiezJM(1998)Factorscontrollingcrackingofconcreteaffectedbyreinforcementcorrosion.MaterStruct31:435–441

4.Al-SulaimaniGJ,KaleemullahM,BasunbulIA,Rasheeduzzafar(1990)Influenceofcorrosionandcrackingonbondbehaviorandstrengthofreinforcedconcretemember.ACIStructJ87(2):220–231

5.AndradeC,AlonsoC,MolinaFJ(1993)Covercrackingasafunctionofrebarcorrosion:Part1.Experimentaltest.MaterStruct26(9):453–464

6.AndradeC,AlonsoC,RodriguezJ,CasalJ,DiezJM(1995)Relationbetweencorrosionandcracking.InternalreportofBrite/EuramProjectBE-4062,C.E.C.

7.AndradeC,Alonso