www.elsevier.com/locate/jcsr
SeismicbehaviorofconnectionscomposedofCFSSTCsandsteel–concrete
compositebeams—experimentalstudy
JianguoNiea,KaiQinb,C.S.Caic,∗
aKeyLaboratoryofStructuralEngineeringandVibrationofChinaEducationMinistry,DepartmentofCivilEngineering,TsinghuaUniversity,
bChinaDevelopmentBankJiangsuBranch,Nanjing,JiangsuProvince,China,210024
Beijing,100084,China
cDepartmentofCivilandEnvironmentalEngineering,LouisianaStateUniversity,BatonRouge,LA70803,USA
Received3November2006;accepted11December2007
Abstract
Inordertoinvestigatetheseismicbehaviorofconnectionscomposedofconcrete-filledsquaresteeltubularcolumnsandsteel–concretecompositebeams,fourteencruciformconnectionspecimensweretested.Thestrength,deformation,andenergydissipationcapacityofthesecompositeconnectionswereanalyzed.Thetestresultsshowedthatthestrengthofconnectionswithinteriordiaphragmsisadequate,buttheirductilityislow.Also,thedeformationcapacityofconnectionswithanchoredstudsisgood,buttheirstrengthislow.Incomparison,theconnectionswithexteriordiaphragmshaveadequatestrength,goodductility,andhigh-energydissipationcapacity,andasaresult,itcanbeconcludedthattheyaremoresuitableforapplicationsinmomentresistingframesinseismicregions.c2008ElsevierLtd.Allrightsreserved.
Keywords:Connections;Tubularcolumns;Compositebeams;Interiordiaphragms;Exteriordiaphragms;Hystereticbehavior
1.Introduction
Inthebuildingconstructionindustry,concrete-filledtubularcolumnsaregainingpopularityallovertheworld.Concrete-filledsteeltubularcolumnspossessmanyadvantagescomparedtoconventionalsteelorconcretecolumns.Someoftheseadvantagesinclude:(1)thesteeltubeprovidesformworkfortheconcretecore;(2)theconcretepreventslocalbucklingofthesteeltubewall;(3)thesteeltubeprohibitsconcretespalling;and(4)compositecolumnsaddsignificantstiffnesscomparedtotraditionalsteelframeconstructions.Althoughtheconcrete-filledsquaresteeltubularcolumns(CFSSTCs)areinferiortotheconcrete-filledcircularsteeltubularcolumnsintermsofbearingcapacity,theyaresuperiorinmanyotheraspects,suchasbeam–columnconnectionconstructability.Therefore,theyareincreasinglyusedforhigh-risebuildingsinmanycountries.Recently,becauseoftherequirementforpracticalapplica-tions,someresearchworkswerecarriedoutonthefollowingrelatedconnections.Morinoetal.[1]testedtensubassemblies
∗Correspondingauthor.Tel.:+12255788898;fax:+12255788652.
E-mailaddress:cscai@lsu.edu(C.S.Cai).
c2008ElsevierLtd.Allrightsreserved.0143-974X/$-seefrontmatter
doi:10.1016/j.jcsr.2007.12.004
undersimulatedseismicloading.AllspecimensconsistedofCFSSTCsandH-shapedsteelbeams.Sasakietal.[2]testedthreeconnectionsbetweenCFSSTCsandsteelH-beamswithinteriordiaphragms.Basedonthetestresults,formulaeofflex-uralcapacityandshearcapacityoftheconnectionswithin-teriordiaphragmswereproposed.Shimetal.[3]testednineCFSSTCtosteelH-shapedbeamconnections.Allspecimensshowedstablehystereticbehaviorsandenergydissipationca-pacities.Luetal.[4]testedfiveconnectionspecimenswithin-teriordiaphragms.Testresultsshowedthatthesubassemblieswithsteelbeamsrigidlyframedtotheconcrete-filledrectan-gulartubularcolumnsexhibitedadequateloadbearingcapacityandgoodductilityifahighweldingqualitywasguaranteed.Koester[5]testedfifteenhalf-scalepanel-zonespecimensandsixfull-scalesplit-teeconnections.Basedonthetestresults,thepanel-zonebehaviorofconnectionsbetweenconcrete-filledrectangularsteeltubesandwideflangebeamswasstudied.Riclesetal.[6]testedtenfull-scalecompositeconnectionscomposedofCFSSTCsandwideflangebeamswithdifferentstiffeningdetails.Testresultsshowedthatthesplit-teeconnectionsandextended-teeconnectionswithtaperedteeflangesprovidedadequatecyclicjointstiffnessandstrength
J.Nieetal./JournalofConstructionalSteelResearch64(2008)1178–11911179
Fig.1.SpecimenCFRTJ-1connectiondetails(unit=mm).
foraweak-beam-strong-columnsystem.Nishiyamaetal.[7]testedtenconnectionspecimenswithdifferentconfigurations,combinationsofmaterialstrength,andloadingconditions.Itwasconcludedthatthecombinationofthestrengthofsteeltubeandconcreteaffectedthemaximumstrengthandyieldbehaviorofthesubassemblies.AlthoughtheconnectionscomposedofCFSSTCsandsteelbeamswerewidelystudied,littleresearchwasconductedtostudythemechanicalpropertiesofconnectionscomposedofCFSSTCsandsteel–concretecompositebeams.Therefore,itisnecessarytoinvestigatetheseismicbehaviorofthistypeofconnections.
AsapartofaseriesofinvestigationsontheseismicbehaviorofconnectionsforCFSSTCsandsteel–concretecompositebeams,experimentalresearchwasconductedatTsinghuaUniversity.Testswereperformedtoevaluatethebehaviorofvariousconnectiondetails.Asapartoftheresearchprogram,fourteenconnectionspecimensweretestedundercyclicloading.Basedonthetestdata,ananalysisofthestrength,stiffness,ductility,andenergydissipationcapacityofthesespecimenswasconducted.2.Experimentalprogram
Threeconnectionspecimenswithinteriordiaphragms,sevenconnectionspecimenswithexteriordiaphragms,andfourconnectionspecimenswithanchoredstudsweretestedundercyclicloading.Thespecimenscaleis1:2.5.
InChinesedesigncodes,connectionswitheitherinteriordiaphragmsorexteriordiaphragmsarerecommendedforCFSSTCswithsteelbeams.Sincetheexperimentalresearchofthesetypesofconnectionswasinadequate,CFRTJ-1–CFRTJ-10weredesignedandtestedtoinvestigatetheseismicbehavioroftheconnectionsandtheinfluencesofdifferentparameters(suchastheaxialloadratioanddimensionsoftheexteriordiaphragms).Theconnectionswithanchoredstudswerenotreportedinanyreferences.SpecimensCFRTJ-11–CFRTJ-14werethusdesignedandtestedtovalidatethefeasibilityofthistypeofconnections.Duetothepagelimitation,adetaileddescriptionofeachconnectionisdifficulttoinclude.Theconnectiondetailsandbeamandcolumndimensions(suchasZ1–Z5,ZL1–ZL8)areshowninFigs.1–9,andtheparametersofthesearespecimensgiveninTables1and2,wheren=NCFSST/fyAs+fcAcistheaxialloadratioofCFSSTs,NCFSSTistheaxialloadontheCFSSTs,fyistheyieldstrengthofsteeltube,Asisthesectionareaofthesteeltube,fcisthestrengthoftheconcretecore,andAcisthesectionareaoftheconcretecore.
Thebeamswereweldedfromsteelplates.Verticalstiffenerswereweldedtothebeamsinordertopreventtorsionofthesebeamsduringtests.Studswereweldedtothefoursteelplatesfirst,andthenthefourpiecesofsteelplateswereweldedtogethertoformasquaretube.Theinteriordiaphragms,exteriordiaphragms,andbeamswereweldedtothecolumnsafterwards.The45mm×6mm×250mmsteelanglethatisusedtoweldthelongitudinalrebarstoensurethetensionforcetransfermechanismoftheserebarswasweldedbetweenthetopflangeofthebeamandthecolumnflange.Thelongitudinalrebarswerearrangedwithaspacingof70mm,andthetransverserebarswerearrangedwithaspacingof50mm.TheC40(40MPa)concretewascastintothesquaretubestoformtheCFSSTCs,andC35(35MPa)concretewascasttoformtheslabs.Thedimensionsandspacingsoftheanchoredstudsforthecolumnsareasfollows:diameterof10mm,heightof60mm,andspacingof60mm.Thedimensionsandspacingsofthestudsforthesteel–concretecompositebeamsareasfollows:diameterof8mm,heightof30mm,andspacingisnotedinFigs.1–9.
ThematerialpropertiesforthespecimensaresummarizedinTables3and4.Theyieldstrengthandultimatestrengthofthestudsarefy,stud=344MPaandfu,stud=470MPa,respectively.ThetestsetupisshowninFig.10,where1=counterforceframe,2=hydraulicactuator(5000kN),3=hydraulicactuator(600kN),4=hinge,5=horizontalbrace,and6=specimen.Duringtesting,anaxialloadwasappliedatthetopoftheCFSSTCsthroughactuator2,andfourhydraulicactuators(actuator3)wereusedtosupplythecyclicloading.
1180J.Nieetal./JournalofConstructionalSteelResearch64(2008)1178–1191
Fig.2.SpecimenCFRTJ-2connectiondetails(unit=mm).
Fig.3.SpecimenCFRTJ-3connectiondetails(unit=mm).
Themethodofquasi-statictestingwasadoptedwiththeloadingprocedureas:(1)Applyinganaxialloaduptothedesignlevelonthetopofthecolumnandkeepingitconstantduringthetest;(2)Imposingverticalloadsattheendsofthebeamsusingforce-controlbeforethespecimenyields,andthenusingdisplacement-controlafterthespecimenyields;(3)Repeatingloadingonlyonceateachcontrolpointbeforethespecimenyieldsandrepeatingtheloadingtwiceateachcontrolpointtoobtainthedegradedcurveofrestoringforceafterthespecimenyields.3.Testresults3.1.Failuremodes
FailuremodesforthetestspecimensaresummarizedinTable5,andthetypicalfracturephenomenaofthespecimensareshowninFig.11.Duringthetest,loadswereappliedtothetwobeamsindifferentdirections,asshowninFig.11(a).Positiveloadingmeansthattheleftbeamisloadeddownwardwhiletherightbeamisloadedupward.Negativeloadingisjusttheoppositeofpositiveloading.Thetubeflangesarereferredtoasthetubesidesthatareconnectedtothebeams,whilethetubewebsarereferredtoastheothertwotubesidesthatarenormaltothetubeflanges.
Fortheconnectionspecimenswithinteriordiaphragms(CFRTJ-1–CFRTJ-3),themainfailuremodeswereeithertheflexuralfailureofthebeamsorconnections.AstheflexuralcapacitiesofZL2andZL4arelessthanthoseoftheconnections,yieldingandlocalbucklingoccurredinthosetwobeams,asshowninFig.11(b).Forotherbeamsofconnectionswithinteriordiaphragms,theflexuralcapacitiesofbeamswerelargerthanthoseoftheconnections;therefore,theinteriordiaphragmsandthesteeltubeflangesyielded.Atthesametime,someweldsbetweenthesteeltubeflangeandbeamwebfractured.
J.Nieetal./JournalofConstructionalSteelResearch64(2008)1178–11911181
Fig.4.SpecimenCFRTJ-4,5,6,7connectiondetails(unit=mm).
Fig.5.SpecimenCFRTJ-8,9,10connectiondetails(unit=mm).
1182J.Nieetal./JournalofConstructionalSteelResearch64(2008)1178–1191
Fig.6.SpecimenCFRTJ-11connectiondetails(unit=mm).
Fig.7.SpecimenCFRTJ-12connectiondetails(unit=mm).
Fortheconnectionspecimenswithexteriordiaphragms(CFRTJ-4–CFRTJ-10),themainfailuremodewasshearfailureofthepanelzone.Sheardeformationdevelopedsufficientlyduringthetestprocess.TheextentofsheardeformationandyieldingofthepanelzoneisevidentinFig.11(c).Thisyieldingwasaccompaniedbydiagonalshearcracksintheconcretewithinthepanelzone.Attheendofthetest,thefractureandlocalbucklingoccurredinpartoftheexteriordiaphragms,asshowninFig.11(d).Partoftheconcreteslabwascrushed,andtherebarswerepushedout,asseeninFig.11(d).Simultaneously,fractureoftheweldsbetweenthesteeltubeflangeandweboccurredinpartoftheCFSSTCs,asshowninFig.11(e).Somechipsoftheconcretecorefellthroughtheripofthesteeltube;fromthisevidenceitcouldbeconcludedthattheconcretecorehadbeencrushedalready.Fortheconnectionspecimenswithanchoredstuds,themainfailuremodewasflexuralfailureoftheconnections.ThepunchingshearfailureofthesteeltubeflangeoccurredinCFRTJ-11andCFRTJ-12atthepositionofthebottomflangeofthecompositebeams,asshowninFig.11(f).Fractureoftheweldsbetweenthesteeltubeflangeandweborripofthesteeltubeflangeoccurredinotherspecimenswithanchoredstuds.Duringthetest,fractureofthebeamflange-to-columngrooveweldsoccurredinsomespecimens,asshowninFig.11(g).Fractureofotherwelds,suchastheweldsbetweentheangleandthesteeltubesorweldsbetweentheexteriordiaphragmsandthesteeltubesalsooccurred.Asaresult,thehystereticcharacteristicswereinfluenced.Therefore,itisimportanttoensurethequalityofweldsinconnectionapplications.
Atthebeginningofthetest,theconcreteslabscrackedundernegativemoment.Withtheincreaseinloading,thewidthsofthecracksgrewlargerandtheregionsofthecracksgrewwider.Finally,thediagonalcracksoccurredinthemiddleof
J.Nieetal./JournalofConstructionalSteelResearch64(2008)1178–11911183
Fig.8.SpecimenCFRTJ-13connectiondetails(unit=mm).
Fig.9.SpecimenCFRTJ-14connectiondetails(unit=mm).
theconnectionslabs,asshowninFig.11(h),andthetransversecracksoccurredonbothsidesoftheconcreteslabs.3.2.P–∆curves
TheP–∆hysteresisloopsofthespecimensareshowninFig.12,inwhichPistheloadappliedtotheendofthebeamand∆isthedisplacementattheendofthebeam.Thehystereticcurvesarecompared,andobservationscanbemadeasfollows:(1)Thehysteresisloopsoftheconnectionswithinteriordiaphragms(CFRTJ-1–CFRTJ-3)areinashuttletype,andasaresult,theenergydissipatedperloopisgood.Sincethedeformationissmall,theductilityofthistypeofconnectionsispoor,andthecumulativedissipatedenergyisalsosmall.(2)Sincetheanchoredstuds(CFRTJ-11–CFRTJ-14)arenoteffectiveintermsoftensiontransfermechanismcomparedtointeriordiaphragmsandexteriordiaphragms,thestrengthofthistypeofconnectionsislessthanthatoftheothertwotypesofconnections.ThenegativeflexuralcapacitiesarelargerthanthepositiveonesforspecimensCFRTJ-11–CFRTJ-13duetotheexistenceoftheangle.Therefore,thenegativehysteresisloopsshowednearlylinearcharacteristicswhencyclicloadswereapplied.Afterthepositiveloadsdecreasedto0.85Pu+(Pu+isthepositiveultimatestrengthoftheconnections),thecyclicloadswereterminated.Then,thestaticloadswereappliedtothebeamsinthenegativedirection,andthemonotonycurveswereobtained,asshowninFig.12(k)–(m).Whenthecyclicloadswereappliedtothesubassemblies,thepositivehysteresisloopspresentedpinchinganddegradedphenomenaforthepunchingshearfailuremode,andasaresult,theenergydissipatedperloopisnotenough.Consequently,althoughtheductilityofCFRTJ-11andCFRTJ-12isgood,thecumulativedissipatedenergiesareclosetothoseoftheconnectionswithinteriordiaphragms.
(3)Thehysteresisloopsoftheconnectionswithexteriordiaphragms(CFRTJ-4–CFRTJ-10)showninFig.12(d)through
1184
Table1
ParametersoftestspecimensSpecimenCFRTJ-1CFRTJ-2CFRTJ-3SpecimenCFRTJ-4CFRTJ-5CFRTJ-6CFRTJ-7CFRTJ-8CFRTJ-9CFRTJ-10SpecimenCFRTJ-11CFRTJ-12CFRTJ-13CFRTJ-14
Type
AnchoredstudsType
ExteriordiaphragmType
Interiordiaphragm
J.Nieetal./JournalofConstructionalSteelResearch64(2008)1178–1191
BeamZL1+ZL2ZL3+ZL42ZL5Beam2ZL52ZL52ZL52ZL52ZL52ZL52ZL5Beam2ZL62ZL72ZL72ZL8
ColumnZ1Z2Z3ColumnZ4Z5Z4Z4Z4Z4Z4ColumnZ4Z4Z4Z4
n0.30.30.3n0.30.30.10.50.30.50.1n0.30.30.30.21
SymmetryAsymmetryAsymmetrySymmetryBet(mm)200200200200160160160Note
Note
TherearenoanchoredstudsbutonlyinteriordiaphragmsinZ1∼Z3.
Beb(mm)160160160160120120120
Note
ThereareninerowsofanchoredstudsinZ4andsevenrowsofanchoredstudsinZ5.
ThereareninerowsanchoredstudsinZ4andeighttriangularstiffeningdiaphragmsinCFRTJ-13.
Note:nistheaxialloadratio;Betisthewidthoftopexteriordiaphragm;andBebisthewidthofbottomexteriordiaphragm.
Table2
GroupingofspecimensaccordingtoparametersforcomparisonobjectivesParameterType
SpecimenCFRTJ-3CFRTJ-4CFRTJ-11CFRTJ-4CFRTJ-6CFRTJ-7CFRTJ-8CFRTJ-9CFRTJ-10CFRTJ-4CFRTJ-5CFRTJ-4CFRTJ-8CFRTJ-6CFRTJ-10CFRTJ-7CFRTJ-9CFRTJ-1CFRTJ-2CFRTJ-3
Beams
CFRTJ-11CFRTJ-12CFRTJ-13CFRTJ-14
ComparisonobjectiveInteriordiaphragmExteriordiaphragmAnchoredstuds0.30.10.50.30.50.1
NinerowsanchoredstudsSevenrowsanchoredstudsBet=200mmBet=160mmBet=200mmBet=160mmBet=200mm
Beb=160mmBeb=120mmBeb=160mmBeb=120mmBeb=160mm
Axialloadratio
Axialloadratio
Rowsofanchoredstuds
DimensionsoftheexteriordiaphragmsDimensionsoftheexteriordiaphragmsDimensionsoftheexteriordiaphragmsInteriordiaphragmsandbeams
Bet=160mmBeb=120mm
10mm+8mm+10mm,ZL1+ZL28mm+6mm+8mm,ZL3+ZL48mm+8mm,2ZL5
2ZL62ZL7
2ZL7(withstiffeningdiaphragms)2ZL8
Fig.12(j)arestableandplentiful.Also,theyhaveenoughstrength,goodductility,andhigh-energydissipationcapacity.Asaresult,itcanbeconcludedthattheconnectionswithexteriordiaphragmsaresuperiortotheconnectionswithinteriordiaphragmsoranchoredstudsintermsofbothmechanicalpropertiesandhysteresischaracteristics.
J.Nieetal./JournalofConstructionalSteelResearch64(2008)1178–1191
1185
Fig.10.Testsetup.
Table3
MaterialpropertiesofsteelTypet(mm)fy(MPa)fu(MPa)δ(%)Plate
639353921.2852660814.31046158918.91243056019.01648359119.1Rebar
6
340
476
31.6
Note:tisthethicknessofsteelplateorthediameterofrebar;fyistheyieldstrengthofsteel;fuistheultimatestrengthofsteel;andδistheelongationratioofsteel.
Table4
Materialpropertiesofconcretefcu,150(MPa)TypeCFRTJ-1CFRTJ-2CFRTJ-3C3552.051.656.4C4058.559.654.0TypeCFRTJ-4CFRTJ-5CFRTJ-6C3545.445.143.8C4053.457.656.1TypeCFRTJ-7CFRTJ-8CFRTJ-9C3552.851.840.8C4054.550.745.1TypeCFRTJ-10CFRTJ-11CFRTJ-12C3544.744.448.0C4047.042.344.9
TypeCFRTJ-13CFRTJ-14C3547.9NoslabC40
47.1
53.4
Note:fcu,150isthecubestrengthofconcrete.
TheskeletoncurvesoftheP–∆hysteresisloopsareshowninFig.13.Thecurvesarecompared,andobservationscanbemadeasfollows:
Table5
SummaryoffailuremodesFailuremodeSpecimen
FlexuralfailureCFRTJ-1(ZL1),CFRTJ-2(ZL3),CFRTJ-3,CFRTJ-11–14
modeofconnectionShearfailureCFRTJ-4–10
modeofconnectionFlexuralfailureCFRTJ-1(ZL2),CFRTJ-2(ZL4)
modeofbeam
Fig.11.Typicalfractureofthespecimens.
(1)Thestrengthoftheconnectionwithamiddleinteriordiaphragm(CFRTJ-1andCFRTJ-2)ishigherthanthatoftheconnectionwithoutamiddleinteriordiaphragm(CFRTJ-3).Therefore,thecontributionofamiddleinteriordiaphragmshouldbetakenintoaccountinthecalculationofflexuralcapacityoftheconnections.
1186J.Nieetal./JournalofConstructionalSteelResearch64(2008)1178–1191
(a)CFRTJ-1.(b)CFRTJ-2.(c)CFRTJ-3.(d)CFRTJ-4.
(e)CFRTJ-5.(f)CFRTJ-6.(g)CFRTJ-7.(h)CFRTJ-8.
(i)CFRTJ-9.(j)CFRTJ-10.(k)CFRTJ-11.(l)CFRTJ-12.
(m)CFRTJ-13.(n)CFRTJ-14.
Fig.12.P–∆curvesofthespecimens.
(a)CFRTJ-1–3(ID).(b)CFRTJ-4–10(ED).(c)CFRTJ-11–14(AS).(d)CFRTJ-3,4,11.
Fig.13.ComparisonofP–∆skeletoncurvesofthespecimens(ID=connectionswithinteriordiaphragms,ED=connectionswithexteriordiaphragms,AS=connectionswithanchoredstuds).
(2)Thestrengthsoftheconnectionswithexteriordiaphragmsarenotsignificantlyinfluencedbytheaxialloadratioandthedimensionsoftheexteriordiaphragms,whilethedeformationcapacitywasobviouslyinfluencedbythesetwoparameters.Withanincreaseoftheaxialloadratioorweakeningoftheexteriordiaphragms,theductilityofthespecimensdecreased.
(3)ThestrengthandductilitywereinfluencedbytheparameterBb/Bcfortheconnectionswithanchoredstuds,inwhichBbisthewidthofthebeam,andBcisthewidthofthecolumn.WhenBb/Bc<0.85andtherearenoanglecases,apunchingshearfailuremodeoccurred.Underthiscondition,thestrengthislow,whiletheductilityisgood.WhenBb/Bc0.85
orthereareangles,fractureoftheweldsbetweenthesteeltubeflangeandweboccurred.Forthisfailuremode,thestrengthishigh,whiletheductilityispoor.3.3.Qj–γjcurves
TheQj–γjhysteresisloopsofthespecimensareshowninFig.14,inwhichQjistheshearforceofthepanelzone,andγjisthesheardeformationofthepanelzone.Thehystereticcurvesarecompared,andobservationscanbemadeasfollows:(1)Thehysteresisloopsoftheconnectionswithinteriordiaphragmswereinashuttletype.Theirsheardeformationdevelopedinsufficiently,asthefailuremodesofthistypeofconnectionwereflexuralfailureofthebeamsorconnections.
J.Nieetal./JournalofConstructionalSteelResearch64(2008)1178–11911187
(a)CFRTJ-1.(b)CFRTJ-2.(c)CFRTJ-3.(d)CFRTJ-4.
(e)CFRTJ-5.(f)CFRTJ-6.(g)CFRTJ-7.(h)CFRTJ-8.
(i)CFRTJ-9.(j)CFRTJ-10.(k)CFRTJ-11.(l)CFRTJ-12.
(m)CFRTJ-13.(n)CFRTJ-14.
Fig.14.Qj–γjcurvesofthespecimens.
(a)CFRTJ-1–3(ID).(b)CFRTJ-4–10(ED).(c)CFRTJ-11–14(AS).(d)CFRTJ-3,4,11.
Fig.15.ComparisonofQj–γjskeletoncurvesofthespecimens.
ThepanelzoneofCFRTJ-2wasalmostintheelasticstage.ThemaximumsheardistortionsofCFRTJ-1andCFRTJ-3werelessthan12×10−3rad.
(2)Thehysteresisloopsoftheconnectionswithexteriordiaphragmsareplentiful,stable,andenergyconsumptive.Thefailuremodeofthistypeofconnectionwasashearfailureofthepanelzone,andasaresult,thesheardistortionsofCFRTJ-4–CFRTJ-10developedsufficientlyduringthetest.Theaveragesheardeformationoftheconnectionswithexteriordiaphragmsis27.5×10−3rad.
(3)Thehysteresisloopsoftheconnectionswithanchoredstudsarealmostlinear,andthemaximumsheardistortionsare
lessthan3×10−3rad.Therefore,itcanbeconcludedthatthepanelzoneofthistypeofconnectionswasalmostintheelasticstage.
TheskeletoncurvesoftheQj−γjareshowninFig.15.Thecurvesarecompared,andobservationscanbedrawnasfollows:(1)Thesheardeformationoftheconnectionswithinteriordiaphragmsoranchoredstudsdevelopedinsufficientlyduringthetest,whilethesheardeformationoftheconnectionswithexteriordiaphragmsdevelopedsufficiently.
(2)Fortheconnectionswithexteriordiaphragms,theshearcapacityofthepanelzonewasinsignificantlyinfluencedbytheaxialloadratioandthedimensionsofexteriordiaphragms,
1188J.Nieetal./JournalofConstructionalSteelResearch64(2008)1178–1191
Fig.16.Definitionofcyclicstiffness.
whilethesheardeformationcapacitywasobviouslyinfluencedbythesetwoparameters.Withtheincreaseoftheaxialloadratioorweakeningoftheexteriordiaphragms,theultimatesheardistortionsofthespecimensdecreased.
(3)Theshearstiffnessofthesethreetypesofconnectionsisalmostthesame.Consequently,itcanbeconcludedthattheshearstiffnessofthepanelzoneismainlybasedonthecontributionofthesteeltubeandconcretecore,whiletheinteriordiaphragms,exteriordiaphragms,andanchoredstudshavelessinfluencesontheshearstiffness.3.4.Stiffnessdegradation
Duringthetestthestiffnessdecreasedwiththecyclicloadingforthereasonofcumulativedamnification.Thestiffnessofspecimensundercyclicloadingcanbeevaluatedbytheindex-cyclicstiffness[8],whichcanbecalculatedasfollows:Kl=
ni=1
(1)Thecyclicstiffnessofthespecimensdegradedsteadily
duringtheentiretestprocess.Itisdifferentfromtheconcreteconnections,whosestiffnessdegradationmainlyoccursaftertheconcretecracks.
(2)Thestiffnessoftheconnectionwithamiddleinteriordiaphragmishigherthanthatoftheconnectionwithoutamiddleinteriordiaphragm.Therefore,thecontributionofthemiddleinteriordiaphragmshouldbetakenintoaccountinthestiffnessanalysisoftheconnection.
(3)Atthebeginningofthetest,thepositivestiffnesswashigherthanthenegativeonebecauseofthecompositeeffectoftheconcreteslabs.Afterseveralhysteresisloops,thewidthsofthecracksgrewlargerandtheregionsofthecracksgrewwiderintheconcreteslabs,andasaresult,thepositivestiffnessvalueisalmostthesameasthenegativeone.
(4)Thecyclicstiffnessoftheconnectionswithexteriordiaphragmsdegradedmoresteadilythanthatoftheconnectionswithinteriordiaphragmsoranchoredstuds.3.5.Energydissipationcapacity
TheEn–nhcurves,Ec–nhcurves,andhe–nccurvesofthespecimensareshowninFigs.18–20,respectively,whereEnistheenergydissipatedperhemicycleofthehysteresisloops,nhisthenumberofhemicycles,Ecisthecumulativedissipatedenergyofthespecimens,heistheequivalentdampingratioofthespecimens,andncisthenumberofcycles.Thesecurvesarecompared,andobservationscanbemadeasfollows:
(1)Theenergydissipationcapacitiesofthespecimensincreasedwiththeincreaseinhemicyclesofthehysteresisloops.Aftertheconnectionsreachedtheirultimatecapacities,thestrengthswentdown,whiletheenergydissipationcapacitiesstillgrew.
(2)Theenergydissipatedperhemicycleoftheconnectionswithinteriordiaphragmswasclosetothatoftheconnectionswithexteriordiaphragms.Sincetheductilityofthespecimenswithinteriordiaphragmswaslow,theircumulativedissipatedenergywaslessthanthatoftheconnectionswithexteriordiaphragms.
Pji/
ni=1
∆ij
whereKl=cyclicstiffness;Pji=thepeakloadoftheith
cyclewhenthedeformationiscontrolledas∆j,asshowninFig.16;∆ij=thepeakdisplacementoftheithcyclewhenthedeformationiscontrolledas∆j;andn=thenumberofcycleswhenthedeformationiscontrolledas∆j.
TheKl–∆curvesofthespecimensareshowninFig.17.Thesecyclicstiffnessdegradationcurvesarecompared,andobservationscanbemadeasfollows:
Fig.17.ComparisonofKl–∆curvesofthespecimens.
(a)CFRTJ-1–3(ID).(b)CFRTJ-4–10(ED).(c)CFRTJ-11–14(AS).(d)CFRTJ-3,4,11.
Fig.18.ComparisonofEn–nhcurvesofthespecimens.
J.Nieetal./JournalofConstructionalSteelResearch64(2008)1178–11911189
(a)CFRTJ-1–3(ID).(b)CFRTJ-4–10(ED).(c)CFRTJ-11–14(AS).(d)CFRTJ-3,4,11.
Fig.19.ComparisonofEc–nhcurvesofthespecimens.
(a)CFRTJ-1–3(ID).(b)CFRTJ-4–10(ED).(c)CFRTJ-11–14(AS).(d)CFRTJ-3,4,11
Fig.20.Comparisonofhe–nccurvesofthespecimens.
Fig.21.Deformationanalysisofanexteriorconnection.
(3)Theenergydissipatedperhemicycleoftheconnectionswithanchoredstudswaslessthanthatoftheconnectionswithinteriordiaphragmsorexteriordiaphragms.Sincetheductilityofthespecimenswithanchoredstudswashigh,thecumulativedissipatedenergywasalmostthesameasthatoftheconnectionswithinteriordiaphragms,butitwasstilllessthanthatoftheconnectionswithexteriordiaphragms.
(4)Theenergydissipatedperhemicycleoftheconnectionswithexteriordiaphragmswasgoodandtheirductilitywashigh.Asaresult,thecumulativedissipatedenergyofthespecimenswithexteriordiaphragmswasmuchhigherthanthatoftheothertwotypesofconnections.
(5)Theenergydissipationcapacityoftheconnectionswithexteriordiaphragmswasobviouslyinfluencedbytheaxialloadratioandthedimensionsofexteriordiaphragms.Withtheincreaseoftheaxialloadratioorweakeningoftheexteriordiaphragms,thecumulativedissipatedenergydecreased.
(6)Theaveragefinalequivalentdampingratiosoftheconnectionswithinteriordiaphragms,exteriordiaphragms,andanchoredstudswere0.317,0.389,and0.304,respectively.Comparingthisdatawiththeresultsoftheconcreteconnectionsandsteelreinforcedconcreteconnections[9],itcanbeconcludedthattheenergydissipationcapacityoftheconnectionscomposedofCFSSTCsandsteel–concretecompositebeamsisclosetothatofthesteelreinforcedconcrete
connections,anditismuchhigherthanthatoftheconcreteconnections.
3.6.Deformationanalysis
Everycomponentoftheconnectionsdeformswhenthesubassembliesareloadedundercyclicloading.Forexample,whenanexteriorconnectionisloaded,thedisplacementattheendofthebeam∆iscomposedofthedisplacementcausedbythebeam∆b,thedisplacementcausedbythecolumn∆c,andthedisplacementcausedbythejoint∆j,i.e.∆=∆b+∆c+∆j,asshowninFig.21.∆jiscomposedofthedisplacementcausedbythesheardeformation∆sandthedisplacementcausedbythenon-sheardeformation∆ns,i.e.∆j=∆s+∆ns[10].
ThedeformationinstrumentationofthespecimensisshowninFig.22.Duringthetest,∆ismeasuredbyDT1andDT2;∆sismeasuredbyDT10andDT11;∆jismeasuredbyDT6,DT7,DT12,andDT13;and∆cismeasuredbyRT.Thus∆nscanbeobtainedby∆ns=∆j−∆s,and∆bcanbeobtainedby∆b=∆−∆c−∆j.ThedeformationanalysisresultsareshowninFig.23.Thecurvesarecompared,andconclusionscanbedrawnasfollows:
(1)Sincethefailuremodesofthespecimensareeitheraflexuralfailuremodeorshearfailuremodeoftheconnections,thedisplacementratiocausedbythejointisthelargest.Attheendofthetest,r∆jofdifferenttypesofconnectionspecimens
1190J.Nieetal./JournalofConstructionalSteelResearch64(2008)1178–1191
Fig.22.Deformationinstrumentationofthespecimens(DT=displacementtransducer,andRT=rotationtransducer).
(a)CFRTJ-3(ID).(b)CFRTJ-4(ED).(c)CFRTJ-11(AS).
Fig.23.r∆–∆curvesofCFRTJ-3,4,11(r∆isthedisplacementratioofdifferentpartsofthespecimens,r∆b=∆b/∆,r∆c=∆c/∆,r∆j=∆j/∆,r∆s=∆s/∆,r∆ns=∆ns/∆).
areallmorethan60%,whilethedisplacementratioscausedbythebeamandthosecausedbythecolumnaresmall.
(2)Thedisplacementratiocausedbythesheardeformationr∆sgrewlargerwiththeincreaseinloadingfortheconnectionswithexteriordiaphragms.Attheendofthetest,r∆s/r∆jreachedalmost80%–90%.
(3)Thesheardeformationandnon-sheardeformationwerealmostthesamefortheconnectionswithinteriordiaphragmsatthebeginningofthetest.Duringthetestthesheardeformationdevelopedsteadily,whilethenon-sheardeformationremainedapproximatelyconstant.Whenthespecimenreachedtheultimatedisplacement,thenon-sheardeformationdevelopedquickly,andthespecimenfracturedfortheinadequateflexuralcapacityoftheconnection.
(4)Thestrengthoftheconnectionswithanchoredstudswaslow,andthesheardeformationdevelopedinsufficiently.Asthe
failuremodesofCFRTJ-11werethepunchingshearfailuremodeorfractureoftheweldsbetweenthesteeltubeflangeandweb,thenon-sheardeformationwastheprominentcauseofthedisplacementattheendofthebeam.4.Conclusions
FourteencruciformconnectionspecimenscomposedofCFSSTCsandsteel–concretecompositebeamsweretested,andthestrength,deformation,andenergydissipationcapacityoftheseconnectionswereanalyzed.Themaintestresultscanbesummarizedasfollows:
(1)Thestrengthofconnectionswithinteriordiaphragmsisadequate,buttheirductilityislow.Whilethedeformationcapacityofconnectionswithanchoredstudsisgood,theirstrengthislow.Theconnectionswithexteriordiaphragmshaveenoughstrength,goodductility,andagoodenergydissipation
J.Nieetal./JournalofConstructionalSteelResearch64(2008)1178–11911191
capacity,andasaresult,theyaremoresuitableforapplicationsinmomentresistingframesinseismicregions.
(2)Fortheconnectionswithexteriordiaphragms,whiletheirstrengthwasinsignificantlyinfluencedbytheaxialloadratioandthedimensionsoftheexteriordiaphragmsunderashearfailuremode,thedeformationcapacityandenergydissipationcapacityweremoreobviouslyinfluencedbythesetwoparameters.Withanincreaseoftheaxialloadratioorweakeningoftheexteriordiaphragms,theultimatedisplacementsandthecumulativedissipatedenergiesofthespecimensdecreased.
(3)Thestrengthandstiffnessoftheconnectionswithamiddleinteriordiaphragmarehigherthanthatoftheconnectionswithoutmiddleinteriordiaphragms.Therefore,thecontributionofthemiddleinteriordiaphragmsshouldbetakenintoaccountinthestrengthandstiffnessanalysisoftheconnections.
(4).Duringthetest,fractureoftheweldsoccurredinseveralspecimens,andasaresult,thehystereticcharacteristicswereinfluenced.Therefore,itisimportanttoensurethequalityofweldsinengineeringapplications.Acknowledgments
Thisresearchisapartoftheresearchprogramonthe“connectionscomposedofCFSSTCsandsteel–concretecompositebeams”atTsinghuaUniversity,China.ThewritersgratefullyacknowledgethefinancialsupportprovidedbytheNationalNaturalScienceFoundationofChina(GrantNo.50438020).Assistanceforexperimentalstudiesfromthestaff
atTsinghuaUniversityisalsoappreciated.ThethirdwriterappreciatesthesupportfromLouisianaStateUniversity.References
[1]MorinoS,KawaguchiJ,YasuzakiC,KanazawaS.Behaviorofconcrete-filledsteeltubularthreedimensionalsubassemblages.In:Proc.compos.constr.ofsteelandconcreteconf.NewYork:ASCE;1993.p.726–41.[2]SasakiS,TeraokaM,MoritaK,FujiwaraT.Structuralbehaviorof
concrete-filledsquaretubularcolumnwithpartial-penetrationweldcornerseamtosteelH-beamconnections.Proc.4thPacificstruct.steelconf.,vol.2.UK:Pergamon;1995.p.33–40.
[3]ShimJS,HanDJ,KimKS.Anexperimentalstudyonthestructural
behaviorsofH-shapedsteelbeam-to-concretefilledsteelsquaretubularcolumnconnections.Proc.,4thPacificstruct.steelconf.,vol.2.UK:Pergamon;1995.p.41–8.
[4]LuXL,YuY,KiyoshiT,SatoshiS.Experimentalstudyontheseismic
behaviorintheconnectionbetweenCFRTcolumnandsteelbeam.StructEngMech2000;9(4):365–74.
[5]KoesterBD.Panelzonebehaviorofmomentconnectionsbetween
rectangularconcrete-filledsteeltubesandwideflangebeams.Ph.D.dissertation.Austin(Tex):TheUniv.ofTexasatAustin;2000.
[6]RiclesJM,PengSW,LuLW.Seismicbehaviorofcompositeconcrete
filledsteeltubecolumn-wideflangebeammomentconnections.JStructEng2004;130(2):223–32.
[7]NishiyamaI,FujimotoT,FukumotoT,YoshiokaK.Inelasticforce-deformationresponseofjointshearpanelsinbeam–columnmomentconnectionstoconcrete-filledtubes.JStructEng2004;130(2):244–52.[8]TangJR.Seismicresistanceofjointsinreinforcedconcreteframes.
Nanjing(China):SoutheastUniv.Press;1989.
[9]ZhouTH,HeBK,ChenGJ,WeiCW,ShanYM.Experimentalstudieson
seismicbehaviorofconcrete-filledsteelsquaretubularcolumnandsteelbeamjointsundercyclicloading.JBuildingStruct2004;25(1):9–16.[10]FaellaC,PilusoV,RizzanoG.Structuralsteelsemirigidconnections:
Theory,designandsoftware.Fla(BocaRaton):CRCPress;2000.
因篇幅问题不能全部显示，请点此查看更多更全内容