Professional Documents
Culture Documents
STEELBUILDINGDAMAGEFROMTHECHRISTCHURCHEARTHQUAKESERIESOF2010and2011
CharlesClifton1,MichelBruneau2,GregMacRae3,RobertoLeon4,andAlistairFussell5
SUMMARY
Thispaperpresentspreliminaryfieldobservationsontheperformanceofselectedsteelstructuresin
Christchurchduringtheearthquakeseriesof2010to2011.Thiscomprises6damagingearthquakes,on
4Septemberand26December2010,February22,June6andtwoonJune13,2011.Mostnotableof
thesewasthe4Septemberevent,atMs7.1andMM7(MMasobservedintheChristchurchCBD)and
mostintensewasthe22FebruaryeventatMs6.3andMM910withintheCBD.Focusisonperformance
ofconcentricallybracedframes,eccentricallybracedframes,momentresistingframes,andindustrial
storageracks.Withafewnotableexceptions,steelstructuresperformedwellduringthisearthquake
series,totheextentthatinelasticdeformationswerelessthanwhatwouldhavebeenexpectedgiven
theseverityoftherecordedstrongmotions.Somehypothesesareformulatedtoexplainthis
satisfactoryperformance.
INTRODUCTION
Widespreadfailuresofunreinforcedmasonrybuildings,thecollapseofafewreinforcedconcrete
buildings,andseveresoilliquefactionacrossthecityofChristchurchcontributedtomaketheFebruary
22,2011,earthquakeatragicnationaldisaster.Thescaleofhumancasualtiesandpropertydamage
fromtheFebruary22eventisinsharpcontrasttothe4Septembereventandotherearthquakesinthe
serieswhichdidnotcauselossoflife.The5kmshallowdepthofthatearthquakeshypocentre,atan
horizontaldistanceofroughly10kmfromthecitysCentralBusinessDistrict(CBD)resultedinground
excitationsbetween3and6timeshigherthanthoserecordedduringthe4Septemberfirsteventinthe
series.Detailedanalysesofthecomprehensivesetofstrongmotiondatarecordedshowsthatthe4
Septembereventwasapproximately0.7timestheUltimateLimitState(ULS)designlevelspecifiedby
theNewZealandseismicloadingstandardovertheperiodrangeof0.5to4seconds,the22February
eventwas1.5to2timestheULSandthelargest13Juneearthquake0.9timesULS.
Whilethedurationofstrongshakingofeachearthquakewasshort(around10to15seconds)the
cumulativedurationofstrongshakingwasover60seconds.Cautionwasexpressedfollowingthe
SeptemberandFebruaryearthquakesthattheshortdurationofstrongshakingineacheventmeant
thatdurationrelateddamagemighthavebeensuppressedcomparedwithwhatonecouldhaveseen
fromasingleearthquakeoflongerduration.However,thiscautionislesswarrantedwhenconsidering
thedurationofthetotalearthquakeseries.Furthermore,therewerereportsofdurationdamagesuch
AssociateProfessorinStructuralEngineering,Dept.ofCivilEngineering,UniversityofAuckland,Auckland,New
Zealand
2
Professor,Dept.ofCivil,Structural,andEnvironmentalEngineering,UniversityatBuffalo,Buffalo,NY,USA
3
AssociateProfessorinStructuralEngineering,Dept.ofCivilandNaturalResourcesEngineering,Universityof
Canterbury,Christchurch,NewZealand
4
Professor,SchoolofCivilandEnvironmentalEngineering,GeorgiaTech,Atlanta,GA,USA
5
SeniorStructuralEngineer,SteelConstructionNewZealand,ManukauCity,NewZealand
ENG.UOA.0003.2
aslowcyclefatiguefractureofreinforcingbarandattachmentdetailstocladdingpanelsfollowingthe
June13events.Metallurgically,theextendedperiodofthisearthquakeseriesislikelytohavebeen
moreseverethanasingleeventofcomparableduration,duetostrainageingofthesteelfromthemost
intense22Februaryearthquakeraisingtheyieldstrengthanddecreasingtheductilityofyielded
componentsbeforethesecondstrongesteventof13June.Forthesereasons,theperformanceofsteel
structuresisinstructive,providingauniqueopportunitytogagetheadequacyofthecurrentNew
Zealandseismicdesignprovisionsforsteelstructures.Thisistheobjectiveofthepaper.
SEISMICDEMAND
ThissectionfocusesontheFebruary11eventdemand,whichwasthemostsevereoftheseries.Figure1
showstheCBDultimatelimitstate(ULS)designspectrumandmaximumconsideredevent(MCE)
spectrumforbuildingsofnormalimportance(basedona2500yearreturnperiod),thelargerhorizontal
componentsfromthefourstrongmotionrecordersintheCBDandtheaverageofthesecomponents.
TheaverageisabovetheMCEforperiodsof0.3secondsandabove,exceptfortheperiodrangeof1.8
to2.7seconds,whereitstillremainssubstantiallyabovetheULSlevel.Thecorrespondingearthquake
excitationsfromoneofthestrongmotionrecordingstationsintheCBD,giveninFigure2,show
substantiallygreateraccelerationsrecordedduringtheaftershockcomparedtothemainshock,andalso
highlighttherelativelyshortdurationofstrongmotion,typicallyontheorderof10seconds.
CENTRAL CITY AND NZS1170 SPECTRA
CLASS D DEEP OR SOFT SOIL
Larger Horizontal Components
1.8
1.6
1.2
SA(T) (g)
CHHC_MaxH_FEB
1
CCCC_MaxH_FEB
0.8
0.6
CBGS_MaxH_FEB
0.4
REHS_MaxH_FEB
0.2
GM_Larger_FEB
0
0
0.5
1.5
2.5
3.5
4.5
Period T(s)
Figure
1:NZS1170.5SpectraandLargestHorizontalDirectionRecordedfromtheCBDStrongMotion
Records
NotestoFigure1:
1. ThedottedlineistheULSdesignspectrumfornormalimportancebuildingsforthesoftsoil
type,ClassD,generallyconsideredintheCBD
2. ThesolidblacklineistheMaximumConsideredEventdesignspectrumfornormal
importancebuildingsforClassDsoilintheCBD
ENG.UOA.0003.3
3. Thesolidredlineistheaveragefromthe5recordingstations
Figure2HorizontalandVerticalSpectrafromtheCanterburyCollegeStrongMotionRecorder
NoteCCCC=ChristchurchCathedralCollege
STEELSTRUCTURESINTHECHRISTCHURCHAREA
ThenumberofsteelstructuresisrelativelylowintheChristchurcharea.Thisisattributedtoboththe
historicalavailabilityofcheapconcreteaggregatesdepositedinriverbedsfloodedbytheseasonal
meltinginthemountainrangeandglacierswestofChristchurch(leavingtheriverbedmostlydryand
accessibletherestoftheyear),andlabourdisputesinthe1970sthatcrippledthesteelindustryinNew
Zealanduntilthe1990s.ConstructionofmodernsteelbuildingsinChristchurchstartedtoreceivedue
considerationfollowingtheendoftheearly1990srecession.Hence,mostofthesteelbuildingsinthe
Christchurchareaarerecentanddesignedtothelatestseismicprovisions.Themarketshareforsteel
framedstructuresnationallyhasincreasedconsiderablyinthelastfewyearstobeclosetothatof
reinforced/precastconcretestructures.Inparticular,afewnotablebuildingshavingsteelframes
openedlessthanthreeyearspriortotheFebruary2011earthquake.Table1providesalistingofthe
multistoreysteelframedbuildingsintheCBDandsomeinthesuburbs.Thereareasimilarnumberof
lowerrisemodernsteelframedbuildingsinthesuburbsthatarenotlistedinthistable.Inaddition,a
numberofprincipallyconcreteframedbuildingsbuiltinthelastdecadeincludepartgravitysteelframes
and/orpartseismicresistingsystems.Mostoftheselaterstructuresarenotlistedinthistable.
ENG.UOA.0003.4
Table1.MultiStoreySteelFramedBuildingsofSignificanceinChristchurchCBDandSuburbs
Noof
Storeys
SeismicResisting
System
FloorSystem
Year
Completed
22
EBFsandMRFs
CompositeDeckandSteelBeams
2010
12
EBFsandMRFs
CompositeDeckandSteelBeams
2009
ShearWallsandCBFs
CompositeDeckandSteelBeams
1985
PerimeterMRFs
CompositeDeckandSteelBeams
1989
MRFs
CompositeDeckandSteelBeams
2010
EBFs
CompositeDeckandSteelBeams
2008
3+Note1
EBFs
Precastcolumnsandhollowcoreunitswithtopping
2003
EBFs
Precastcolumnsandhollowcoreunitswithtopping
2010
Notes:
1. Currently3storeys;withprovisionforadditional1storey
SEISMICPERFORMANCEOFMULTISTORYECCENTRICALLYBRACEDFRAMES
TworecentlydesignedandbuiltmultistorybuildingsintheCBDhadeccentricallybracedframesaspart
oftheirlateralloadresistingsystem.The22storeyPacificResidentialTowerinChristchurchs
CBD,completedin2010,andtheClubTowerbuilding,completedin2009.Bothweregreentagged
followingtheearthquake,indicatingthattheyweresafetooccupybutwouldrequiresomeminor
repairs.
TheClubTowerBuilding(Figure3a)haseccentricallybracedframeslocatedonthreesidesofanelevator
coreeccentricallylocatedclosertothewestsideofthebuilding,andaductilemomentresistingframe
(DMRF)alongtheeastfaade.Thesteelframeissupportedonaconcretepedestalfromthebasement
tothe1ststory,andfoundationsconsistofa1.6mthickraftslab.OnlytheEBFsontheeastsideofthat
corecouldbevisuallyinspectedwithoutremovalofthearchitecturalfinishes(Figure3d),howevermore
detailedinvestigationwasmadeoftheSouthsideactivelinksthroughremovalofceilingtilesto
ascertainthemostsignificantlyyieldedbraces.Figure3cshowsalinkatlevel3ontheSouthsidewhich
hasthegreatestobservedinelasticdemand.Estimatesofthepeakinelasticdemandinthatbracewere
madebytwoindependentmeans.Firstwasthroughassessmentofthevisiblestateofthemetalinthe
yieldedweboftheactivelinkandsecondlythroughestimateofthepeakinterstoreydrift.Bothmethods
gaveapeakshearstrainofbetween3%and4%.Thelinkswerefreeofvisibleresidualdistortions.
Assessmentofdamageaccumulationinthesteelatapeakshearstrainof4%overanestimatedtwo
completecyclesofloadingusingthedamagecriteriondevelopedby(Seal,2009)andchangein
ENG.UOA.0003.5
transitiontemperaturebasedontheworkof(Hyland,2006)showedthattheyieldedactivelinkshave
sufficientlyrobustmetallurgicalpropertiestobeleftinplace.Previouslyreportedslabcracking(Bruneau
etal.2010)couldnotbedetectedastheconcretefloorslabwascoveredbyfloorcarpeting,exceptat
onelocationatthefixedendofasegmentofthefloorcantileveringononesideofthebuilding(a
featurepresentonlyovertwostoriesforarchitecturaleffect).CrackwidthsafterFebruary22,2011
appearedsimilartowhathadbeenobservedafterSeptember4,2010,beinglocalisedonly.Substantial
shearcrackingofthegypsumplasterboard(sheetrock)finishontheexteriorwallofthatcantilevering
partofthefloorwasalsoobserved(Figure3b);onlyhairlinecrackingofgypsumplasterboardfinishes
wasobservedelsewherethroughoutthebuilding,supportingpostearthquakesurveymeasurements
showingthatthebuildinghasapostearthquakeresidualdriftofonly0.1%.Onenonstructuralmasonry
blockwallinstalledforsoundproofingpurposesadjacenttomechanicalunitsonthepedestalroof
sufferedminorshearcracking,whereithadbeenplacedhardagainstacantileveringfloorbeam.
Giventhemagnitudeoftheearthquakeexcitations,withdemandsabovetheULSdesignlevel,
substantialyieldingoftheEBFlinkswouldhavebeenexpected.EBFsdesignedincompliancewiththe
NZS3404(SNZ,1997/2001/2007)provisionsaretypicallysizedconsideringaductilityfactor(,
equivalenttoRinUSpractice)ofupto4,correspondingtoaleveloflinkdeformationsthatwould
correspondtosignificantsheardistortionsofthelinks.Yet,yieldingwasbelowthatdetermined
necessaryinsubsequentdetailedassessmenttorequirestructuralreplacementoftheEBFactivelinks.
Beyondtheusualfactorscontributingtooverstrengthinsteelframes(e.g.expectedyieldstrength
exceedingnominalvalues,modellingassumptions,etc.),anumberofadditionalfactorscanexplain
behaviourinthisparticularcase,includingstrengthofthecompositefloorslabaction(neglectedin
design),mobilizationofthesolidnonstructuralwallconcretecladdingadjacenttothestaircase,elastic
stiffnessofthegravityframeespeciallythecolumnsandtherelativelyshortdurationofearthquake
excitation.
TheductileMRFalongtheeastwalldidnotshowanyevidenceofyielding.Itsdesignhadbeen
governedbytheneedtolimitdrift,particularlyundertorsionalresponseduetotheeccentricityofthe
core,anditscorrespondingeffectiveductilityfactor()waslowat1.25.
Overall,thebuildingwasdesignedforaslightlylowerlevelofstructuralductilitydemandthanistypical
foranEBF,duetoitsheightandplandimensions,andperformedwellduringtheearthquake.No
structuralrepairswererequired;nonstructuralremedialworkconsistedofminordrywallcrackrepair
andrealignmentoftheliftguiderails.ThebuildingwasopenandfullyreoccupiedinJuly2011,
becomingthefirstnormalimportancehighrisebuildinginChristchurchtobereturnedtousefollowing
theearthquakeseries.
ENG.UOA.0003.6
Figure3:ClubTower[PhotosbyM.BruneauandCClifton]:(a)Globalview;(b)crackingofpartitionin
cantileveringportionofstory;(c)paintflakingofpartiallyhiddenEBFlink;(d)globalviewofEBFbraces
obstructedbyvariousutilityruns.
ENG.UOA.0003.7
AsanewlandmarkandthetallestbuildingontheChristchurchskyline6,the22storeyPacificTower
consistsofperimeterEBFsuptothesixthflooronthewesternsideanduptotheeleventhflooronthe
northsideofthebuilding,shiftingtojointheotherEBFsaroundtheelevatorcoreabovethoselevels.,
withtransferslabsdesignedtohorizontallydistributetheseismicloadsatthosetransitionpoints.
SeveralsectionsoftheEBFsatlevelsbelowthelevel6transferslabwerevisible,apartfromatthetopof
theperimetersystem,astheselevelshousedamechanicalmultilevelparkingelevatorsystem.The
separatebracingsystemofthatmechanicaldeviceconsistedofflatplatesconnectedwithturnbuckles
andhooks.Someofthosedetailsfailedasthebarsunhookedwhenreturningintocompressionafter
tensionyieldingexcursionsthatelongatedthebraces.TheEBFsatintermediatelocations(ontheNW
frame)werenotintegralwiththefloorslabandsodidnotbenefitfromthestrengthincreaseprovided
bythatintegralactionthroughouttherestofthebuilding.Arangeofviewsforthisstructurearegivenin
Figure4.
PaintflakingandresiduallinksheardeformationswereobservedintheEBFlinksatthoselevels.Design
oftheEBFsinthatbuildingwasgovernedbytheneedtolimitdrift,withacorrespondingresulting
designductilityfactor()of1.5(eventhoughupto4.0ispermittedforEBFsystems,asmentioned
earlier).ThisistypicalofEBFsintallbuildingsinNewZealandsmoderatetolowseismiczones;
Christchurchismoderateinaccordancewiththeearthquakeloadingsstandard,NZS1170.5andamore
typicaldesignductilitylevelfactorrangeforsuchbuildingsis2to3.Whentheinitialinternalinspections
wereundertaken,therewasanabsenceofsignificantdamagetoarchitecturalandothernonstructural
finishingsexceptatlevel6whereafewofthehotelroomdoorsalongthecorridorcouldnotbeclosed,
suggestinggreaterresidualdeformationsatthatlevel.Thislevelwasthefirstinwhichadetailed
evaluationwasundertaken.OnefracturedEBFactivelinkwasdiscovered(Figure4e)inthetoplevel
(undersideofLevel6)oftheEBFsystemattheNorthWesterncornerofthebuilding.Theframesits
behindthelouvresystemnearestthecamerainFigure4a.Thislinkhadundergoneatleastonefullcycle
ofwebpanelyieldingpriortoafracturepropagatingfromonetopcorneracrosstheactivelinkregion
andresultinginsignificantresidualdeformation.Temporarystrapcrossbracingwasweldedtothe
activelinkframetoprovidelateralloadresistancewhilearepairstrategywasimplemented,which
comprisedcuttingoutthedamagedlink,weldingonanendplatesystemtoeachcollectorbeam/brace
faceandreplacingwithasiteboltedendplateactivelink.Thereplacementisscheduledforearly
October2011andistheonlyrepairtothestructuralframerequiredforthisbuilding.Adetailed
evaluationwasundertakenofallactivelinksintheadjacentstoreysandthroughoutthebuilding,with
thefrequencyofinspectionreducedasnofurtherexamplesrequiringreplacementwerefound.This
inspectionrequiredremovalofarchitecturalfinishes.
ThistypeoffailurehasnotbeenreportedineitherEBFstestedinthelaboratoryorfromdamagereports
fromotherearthquakes;thereasonsforthislinkfracturearenotcurrentlyclearanditistobethe
subjectofadetailedmetallurgicalandstructuralevaluationonceremoved..
TheGrandChancellorHotelis85metres,thePriceWaterhouseCooperbuildingis76.3metres,andtheC1
Building(a.k.a.thePacificTower)standsat73metres,istoppedbya13metresspire,foratotalof86metres.
ENG.UOA.0003.8
AswithClubTower,somerepairofdrywallcrackingandrealignmentofliftshaftguiderailsistheonly
otherworkrequiredandtheintentionistohavethiscompletedintimeforthebuildingtobefully
openedwhenpublicaccessisrestoredintothisarea.Itisalsoworthnotingthatthisislikelytobethe
onlyoneofthesixhighrisebuildingsinChristchurchthatwillbereturnedtoservice.
Itisnotedthathavingthelateralloadresistingsystemhiddenbyarchitecturalelementsisahindrance
topostearthquakeinspection,makingitoftenonlypossibletoinferthepresenceofstructuraldamage
fromthecrackingofnonstructuralfinishesandotherevidenceoflargeinterstorydriftsuntilthelinings
areremoved.Whilethismayworkwellinmanycases,experiencefollowingtheNorthridgeearthquake
suggeststhatmajorfracturesofstructuralelementsmayremainhiddenforyearsifonlynonstructural
damageisrelieduponasanindicatorofpossibleproblemswiththelateralloadresistingstructure.
Futurebuildingcodecommitteesmayconsiderthemeritofrequiringthatbuildingsbedesigntoprovide
easyinspectionofkeystructuralelementsandcriticalnonstructuralelementsfollowingsevere
earthquakes.
ENG.UOA.0003.9
Figure4PacificTower[PhotosbyM.BruneauandCClifton]:(a)Globalview;(b)FlakedpaintonEBF
activelink;(c)and(d)Multistorymechanicalgaragestackerfailedbraces;(e)FracturedEBFactivelink
intoplevelofEBFsysteminfrontfaceofatrium;(e)ResidualsheardeformationsofEBFlinkincar
stackertower.
SEISMICPERFORMANCEOFECCENTRICALLYBRACEDFRAMESINPARKINGGARAGES
ThetwolowriseparkinggarageshavingeccentricallybracedframesdescribedinBruneauetal.(2010)
wereinspectedfollowingtheFebruary22event.
TheEBFsinathreelevelparkinggarageofashoppingmallwestoftheCBDdidnotexhibitinelastic
deformations(Figure5a),howevertherewasevidenceofveryminormovementoftheboltedsplice
connectionsinthebraces.ThebasicallyelasticresponseoftheEBFsisnotsurprisinginthiscase,given
thattheseframeshadbeendesignedtoaccommodatethreeadditionalparkinglevelstobeaddedata
latertimeandtheintensityofshakingwaslowerthanintheCBD.Liveloadpresentatthetimeofthe
earthquakemayalsohavebeenlessthanconsideredindesign,althoughitwashigherthaninthe
Septemberearthquakewhentheshoppingmallwasnotoccupied.Movementofprecastunits
ENG.UOA.0003.10
previouslyreportedwasobservedtohaveintensified.Thisresultedinfractureofthespandrel
panelsbesidetheepoxymasticconnectionbetweenpanels,presumablyindicatingthattheepoxymastic
wasstrongerthantheprecastpanelsintension(Figure5b).Thesefracturesoccurredinallpanelsover
theheightofthestructure.Thesespandrelpanelswerealsodesignedtocarrygravityloadsinthe
parkingstructuresotheirfracturecompromisedtheserviceabilityofthebuilding.Nofurtherdamageis
reportedfromthethreeJuneearthquakes.
Figure5:ShoppingmallonDilworthStandClarenceSt,Christchurch[PhotosbyG.MacRae];(a)View
fromtheEast,(b)FractureofPrecastSpandrelBeamsonSouthside
TheEBFsinahospitalparkinggarageclosertotheepicentre(Bruneauetal.2010)alsoperformedwell,
althoughsomelinkfractureswereobservedintwobracedbays(Figure6).NotethatatleastsixEBF
frameswereusedateachlevelineachofthebuildingsprincipaldirections,andthatthissignificant
redundancycontributedtomaintainsatisfactoryseismicperformanceofthebuildinginspiteofthose
significantfailures.Residualdriftsoftheparkingstructureordamagetothegravityloadcarryingsystem
werenotvisuallynoticeable,whichsuggeststhatthesefractureswouldhavenothavebeendiscovered
ifhiddenbynonstructuralfinishes.
Notethatthisparkingstructurewasalsodesignedtoaccommodatetwoadditionalfloors.Yet,someof
thelinksatthefirststoryshowedpaintflakingasevidenceofinelasticdeformations.Evidenceofsoil
liquefactionwasalsoobservedoverpartsoftheslabongrade.Dependingonthefoundationtype,
liquefiedsoilscanactasasortofbaseisolationorasamethodtolengthentheperiod.Thisgenerally
resultsinaloweryieldaccelerationandlowerstructuraldemands.Assuch,itispossiblethatthis
parkinggaragewasnotsubjectedtogroundmotionsassevereasthoseshowninFigure1,inspiteof
beingonly1.5kmsawayfromstationCCCCinFigure2.However,becausetheseEBFswerenotdrift
dominatedtheyweredesignedforthemaximum=4ductilitydemand.Alsotheseactivelinkswere
addedasfinishedcomponentsintothelargelyprecastconcretestructureandsowerenottiedintothe
floorslabwithshearstudsastheywereforthetallerbuildingspreviouslydiscussed.Thismeantthat
ENG.UOA.0003.11
theydidnothavethesamestrengthenhancementduetoresistancetooutofplanedeformationofthe
floorslabasthetallerbuildingshad.
Thefractures,asshownincloseupinFigure6(c),wereofparticularconcernasthesewerethefirst
fracturesrecordedinEBFsworldwide(thePacificTowerfractureasmentionedabovewasdiscovered
later).Furtherpuzzlementwasaddedbythefactthatthefractureplane,showninFigure6(c),indicated
aductileoverloadfailureratherthanabrittlefracture.However,thelikelyexplanationliesintheoffset
ofthebraceflangefromthestiffener.ThisoffsetisshowninFigure6(c)andmeansthat,whenthebrace
wasloadedintension,theaxialtensionforceinthebracefedintotheactivelink/collectorbeampanel
zonethroughaflexiblebeamflangeratherthandirectlyintothestiffener.Thismeantthatthejunction
betweentheunstiffenedbeamflangeandthebeamwebwasseverelyoverloaded,leadingtofracture
betweenthesetwosurfacesandthisfracturespreadingacrossthebeamflangeandthroughtheweb.
Evidenceinsupportofthisisfromthefollowing:
wheretheflangesofthebracelineupwiththestiffeners,asintherighthandsideoftheactive
linkshowninFigure6(b)orthepanelzoneshowninFigure6(e),therewasnodamagetothis
panelzoneregion
thedamagetothepanelzoneregionisdirectlyproportionaltotheeccentricitybetweenthe
braceflangeandtheactivelinkendstiffener
Thisshowsthatloadpaththroughtheasconstructeddetailisparticularlyimportantwheninelastic
demandisrequiredfromthesystem.
Also,therampatthetoplevel,builtinanticipationoffutureadditionalstories,suffereddamageasthe
onlyEBFontheuppersegmentoftherampwaslocatedattheeastendofthatramp,inducingtorsional
responseandshearfailureofthecolumnsinmomentframeactionatthewestendoftherampthese
shearfailureshadnotbeenrepairedbythetimeoftheaftershockandexhibitedmoresignificant
damage(temporarylateralbracingwereinstalledtopreventfurtherswaymotions).Steelangles,
originallyaddedattheexpansionjointomeetthedesignrequirementforsupportlengthofhollowcore
slabpreventedunseatingoftheramp.TheEBFlinkattheramplevelitselfexhibitedsubstantialinelastic
distortions.
ThelateralbracingoftheactivelinksinthebuildingshowninFigure6wasonlyintheformofa
confiningangleeachsideofthetopflange,asshowninFigure6(d)and6(e).Nolateralmovementor
twistingoftheendsoftheactivelinkswasobserved,indicatingthatthelateralrestraintprovisionshad
beenadequateinpractice,despiteonlybeingappliedtothetopflangeandforEBFsnotintegralwith
theslababovealsobeingnoncompliantwithNZS3404.
ENG.UOA.0003.12
ENG.UOA.0003.13
Figure6:ParkinggarageonStAsaphStandAntiguaSt,Christchurch[PhotosbyM.Bruneau];(a)
RedundancyprovidedbymultipleEBFbays;(b)EvidenceofEBFlinkyielding;(c)Fracturedlinkatlower
levelEBF;(d)EvidenceofinelasticdeformationsattoplevelEBF;(e)Closeupviewofsame;(f)
Displacementatexpansionjoint,topramp.
Asofmid2011,thefracturedactivelinkshavebeencutoutandarebeingreplacedtobringthebuilding
backintoserviceinadvanceofwhenpublicaccessisrestoredtothisarea.
CONNECTIONS
Connectionsinmodernsteelframesperformedverywellandasexpected.Figure7(a)showsa
brace/beam/columnconnectioninwhichthegussetplateisweldedtothebeamandboltedtothe
columnwithaflexibleendplateconnection,whichisdesignedanddetailedtoberigidforverticalload
transferandflexibleinthehorizontaldirection,toaccommodatechangeintheanglebetweenbeam
andcolumnduringtheearthquake.Thisflexibleendplatehasundergonelimitedoutofplaneyielding,
protectingthegussetplatefrominelasticdemand.Figure7(b)showsaflushendplatespliceinaMRF
beamthathasperformedwell.
Inamomentendplateconnectioninaportalframebuildinginastronglyshakenregiononsoftground
nearthefault,tensilefailureofarowofboltswasobserved.Theconnectionhadnotopenedupduring
theearthquakeandwasrapidlyrepaired.
Nodamagewasevidentfromabriefinternalinspectiontotheweldedbeamtocolumnconnectionsin
the7storeyperimetermomentframebuildingmentionedbelowandshowninFigure10.These
connectionswillrequiredetailedevaluationaspartoftheassessmentofthisbuildingasanylocal
fractures,e.g.betweenthebeamflangeandcolumnflangeinasimilarmannerasoccurredinthe1994
Northridgeearthquake,wouldbehiddenbythepassivefireprotectionandonlythreejointswere
lookedatindetailduringthevisitinwhichthosepicturesweretaken.
ENG.UOA.0003.14
Figure7:ConnectionsinClubTowerBuilding,Christchurch[PhotosbyGCClifton];(a)
Brace/beam/columnconnectionshowingoutofplaneyieldinginendplatebutnoinelasticdemandin
gussetplate;(b)Flushmomentendplatespliceconnection
CBFBUILDINGS
Asinglesuspendedlevelparkinggaragewithconcentricallybracedframe(CBF)wasfoundtohave
performedpoorly(Figure8).Thegaragehadsolidprecastpanelwallsonthreesides,andtwo
individualCBFbaysalongthefourthside(onebayoneachsideofthegaragedoor).Whilethecolumns
ofthewesternmostCBFtiedtoasteelbeamattheirtop,theeasternmostCBFwasnotsimilarlyaligned
withasteelbeam.Anonductilereinforcedconcreteextensionframingintoaconcretebeamatthetop
performedpoorly.Theotherbraceofthatframefailedattheweldsundertensionloads;thesewelds
didnotappeartobedesignedtodevelopthetensioncapacityofthebraceaccordingtothecapacity
designprinciplesofNZS3404.ThewesternmostCBFperformedbetter,withoutfracturesandwith
properattachmenttothesupportedflooratbothcolumns,withvisiblepostearthquakeresidual
bucklingasaconsequenceofbraceelongation.
ENG.UOA.0003.15
ENG.UOA.0003.16
Figure8:LowriseCBFparkinggarage[PhotosbyM.Bruneau]:(a)Poorcolumnconnectiondetail;(b)
Buckledbrace;(c)and(d)Fracturednonductilebracetocolumnconnection.
AsevenstoreysteelframedhotelbuildingwithacombinationshearwallsinonedirectionandCBFsin
theotherdirection,couldnotbeinspectedbecauseofitsimmediateproximitytothe22storeyGrand
ChancellorHotelwhichwasconsideredtobeinastateofimminentcollapsefollowingthe22February
earthquake.Itishopedtovisitthisbuilding,ifitisstillintact,oncetheGrandChancellorhasbeen
demolished.Thereisnoindicationofdamagefromthestreet.
MULTISTOREYMRFBUILDINGS
Anewparkinggarage(constructioncompletedaftertheSeptember2010earthquake)appearedtohave
performedverywell,withnovisiblesignofinelasticdeformationatthebeamtocolumnconnections
(Figure9)orinanyotherpartofthestructure.However,thisassessmentcouldonlybedonefromthe
groundbelowasacollapsedconcretecarparkingbuildingnextdoorprecludedaccessintothebuilding.
AlowriseMRFbuildingintheCBD,whichhousedagymnasium,wasinspectedindetailinternallyand
externallyandhadnostructuraldamage.
Finallya7storeybuildinglocatedintheregionoftheCBDthatexhibitedsignificantgroundinstability
wasinspectedinsideandout.Thestructurecomprisesaperimetermomentresistingframealongall4
sides,withanonstructuralstairandservicescoreandcompositefloor.Inspectionofthesteelframe
andfloorshowednovisibledamage,howevertheperimeterframehadsunkanoticeableamountin
relationtothecore(Figure10b)andhadactedaspinnedbase,causingsignificantinterstoreydriftwhich
hassubsequentlysignificantlydamagedstairs(Figure10d)andnonstructuralcomponents.Theextent
ofgroundmovementaroundthebuildingwasconsiderableanditislikelythatsignificantfoundation
movementhasoccurred.Thequestionofwhattodowiththisbuildingwillrestonwhathashappened
belowground.
ENG.UOA.0003.17
Figure9:LowriseMRFparkinggarage[PhotosbyM.Bruneau]:(a)Globalview;(b)and(c)Typical
momentconnections.
ENG.UOA.0003.18
Figure10:SevenstoreyPMRFbuilding[PhotosbyCClifton]:(a)Internalviewofpartiallystrippedfloor;
(b)Crackingroundfloorslabwhereperimeter(totherightofthecrack)hasmoveddownwardsrelative
tothecore;(c)Typicalbeamtocolumnconnectionwithfireprotectionpartiallyremovedforinspection;
(d)Gapopeninginstairatlandingduetolateralmovement(stairshadindependentstructuralsupport).
HISTORICALBUILDINGS
ENG.UOA.0003.19
PartialoutofplanefailurearoundthedomeatthetopoftheRegentTheatre)Buildingrevealedthata
bracedsteelframehadbeenusedthere(Figure11).Althoughsubsequentinspectionwillberequiredto
verifytheintegrityoftheconnections,itappearedtobeingoodconditionfromadistance.Thebuilding
wasbuiltbefore1910andthescenewasreminiscentofpicturesofsimilarbuildingsfollowingthe1906
SanFranciscoearthquake.However,theCBFsappearedtobeweldedconstruction(tobeverified)which
meanstheyarelikelytobenewerthantherestofthebuildingandhadbeenaddedinasubsequent
retrofit.
Figure11:BraceddomeattopofRegentonWorchesterBuilding[PhotosbyM.Bruneau]:(a)Global
view;(b)Closeupview.
Steelbracedframesweresometimesusedtoretrofitunreinforcedmasonrystructures(e.g.Figure12).
Driftlimitstopreventfailureoftheunreinforcedmasonrytypicallygoverndesigninthoseinstances,
whichexplainthesignificantmembersizesoftheseframesproportionaltothereactivemass,andtheir
elasticresponse.
ENG.UOA.0003.20
Figure12:Bracedframeasaretrofittounreinforcedmasonrybuilding[PhotosbyM.BruneauandC
Clifton]:(a)Closeupview;(b)Globalview.
BuildingsintheCBDthathadbeenstrengthenedpriortotheSeptember2010earthquaketypically
sufferedminimaltonodamageinthatevent.TheywerenotsofortunateinthemuchstrongerFebruary
2011event.Figure13showsonegroupofthreebuildings,with(a)showingthesefollowingthe
September2010eventand(b)showing(fromadifferentvantagepoint)thethreefollowingthe
February2010event.Noteespeciallythestrengthenedbuildingonthecornerhascollapsed
Figure13:StrengthenedURMbuildings[PhotosbyCClifton]:(a)isfollowingtheSeptember2010event
and(b)followingtheFebruary2011event,takenfromaslightlydifferentviewpoint.
ENG.UOA.0003.21
Finally,notethattheheritagestructuredescribedinBruneauetal.(2010)atthecornerofManchester
andHerefordstreets,severelydamagedbytheSeptember2010earthquake,hadbeendemolishedby
itsownerpriortotheFebruaryaftershock.
INDUSTRIALandEDUCATIONALFACILITIES
Manywarehousesclosetotheepicentresufferedlimiteddamage.Theseindustrialfacilitiestypically
havelightroofsandaredesignedtoresisthighwindforces;lightrodbracesaretypicallyusedforthis
purpose.Followingtheearthquakes,steelfabricatorsinspectedmultiplewarehouses,andretightened
saggingbracesthathadstretchedduetoyieldingduringtheearthquake.
AswasthecasefollowingtheSeptember2010Darfieldearthquake,aproprietarysystemoftenusedin
thesewarehouses(soldasakit)whichusedaparticularbananaendfitting,sufferedsomebrittlefailures
ofthecaststeelconnectors(asshowninFigure14).Theseoccurredinanewwarehousewhenthe
fittingfelltothegroundfollowingtheshearingofthepinretainingclip.Giventhattheseconnectorsare
ratedforearthquakeloadingbasedonstatictestsconductedbythemanufacturer,inlightofthefew
fracturesreportedfollowingthetwoearthquakes,someengineershaveexpressedconcernsregarding
theirpotentialbrittleness.Performanceofthisandsimilarsystemsneedstobevalidatedundera
dynamictestregimemorerepresentativeoftheirexpectedseismicdemands,particularlysimulatingthe
impactforcesappliedwhenpreviouslybuckledbracesretightenduringearthquakeexcitations.
Figure14:Exampleoffracturedbananaendofproprietarybraceconnectorintheroofplaneofalong
spansteelportalframebuilding[PhotosbyM.Bruneau]:(a)Globalview;(b)Closeupview.
Extensivefailureofsteelstoragerackswasobservedinindustrialfacilities,insomecasesinspiteof
additionalmeasurestakenfollowingtheSeptemberearthquake.Forexample,onefacilityownerwho
hadracksstacked6palletlevelshighthatcollapsedduringtheSeptember2010earthquake,purchased
newracksdesignedtoresistMagnitude7earthquakesofthetypeexpectedin[themostactiveseismic
ENG.UOA.0003.22
zoneof]Wellingtonandrestructuredhisoperationstolimitstackingtothreelevels.Inspiteofthose
measures,allracksexperiencedtotalcollapse,asshowninFigure15.Whileracksthatfailedinthe
transversedirectioncouldhavebeenpushedduetospillingofthepalletsandpilingupoftheproducts
intotheaisles,thiswasnotafactorinthelongitudinalrackfailuresthatexhibitedacombinationof
overloadedandfracturedbeamtocolumnconnections,andcolumnlocalbuckling.Itappearedthatthe
semirigidbeamtocolumnconnectionsinthelongitudinaldirectionweretooweakfortheintensityof
shakinganddesigngravityloads.
ENG.UOA.0003.23
Figure15:Exampleofcollapsedindustrialstorageracks[PhotosbyM.BruneauandGCClifton]:
Anecdotally,inanotherfacility,existingrackshadbeenretrofittedbycouplingtworacksbacktoback
withflatbarbraces(Figure16).Thesebarsshowedevidenceofelongationandresidualbuckling,but
didnotcollapse,inspiteoffloormovementsduetoliquefaction,whereastheonlyrackthatwasnot
retrofitted(foritwasnotadjacenttoasecondracktowhichitcouldhavebeentied)collapsed.The
rackshasalsobeenallegedlytiedtotherafterstopreventlongitudinalfailures,butsuchtiescouldnot
beidentified.
Theseaboveselectedexampleshighlightthefactthatperformanceofindustrialstorageracksisamajor
issuethatremainstobesatisfactorilyaddressed;howevertheirperformancehastobeconsideredin
lightoftheveryhighintensityofshaking.
ENG.UOA.0003.24
Figure16:Industrialstorageracksthatsurvived,withevidenceofsoilliquefaction[PhotosbyM.
Bruneau];(a)Globalview;(b)Closeupofbuckledbrace.
Multipleexamplesoftiltuppanelmovementsduetogroundliquefactionwereobserved,sometimes
leadingtofractureofnonductilebracesunabletoaccommodatetheimposeddeformations.Onesuch
exampleisshowninFigure17,showingafracturedbraceanditscounterpartbuckledbrace.
Figure17:Industrialfacilityroofbracing[PhotosbyM.Bruneau];(a)Globalview,showingbuckledbrace
ENG.UOA.0003.25
andfracturedbrace;(b)Closeupviewoffracturesweldoftensionbrace.
Anchorageoftiltupwallstosteelstructuresalsofailedinafewinstances.Figure18showsroofbeams
buckledincompressionbytheinwardmovementofthetiltuppanels,andfailureoftheanchorsdueto
theiroutwardmovement(i.e.awayfromthebuilding).Giventhatthishappenedinmodern
construction,andbecausetiltupwallsofgreaterslendernesshaveprogressivelybeenimplementedin
NewZealand,acarefulreassessmentoftheirseismicdesignprovisionsmaybedesirable.
Figure18:Failureoftiltuppanelconnections[PhotosbyM.Bruneau];(a)Globalview;(b)Closeup
viewoffracturesconnection;(c)Globalviewofbuckledbeams;(d)Localviewofonesuchbeam.
Figure19showsthesteelstructurestandingwhentheroofinghascollapsed.Thisshowsverygood
performanceofthesteelmembers,butpoorperformanceoftheroofing/connections.
ENG.UOA.0003.26
Figure19:FailureofroofandwallsinolderindustrialfacilityonSalisburyStreet[PhotosbyMacRae];
AtHeathcoteValleyPrimarySchoolsomeofthemostextremeshakingduringtheeventwasrecorded.
TherewasonenewsinglestoreybuildingwithasteelmomentframeandblockwallsasshowninFigure
20a.Aftertheearthquakethewallwasleaningtotheeastatthesouthernend,andtothewestatthe
northernend.Theconcretebaseplatewasblownoutonthesoutheastsideofthebuildingasshownin
Figure20b.
(a) OverallViewfromtheSouth
(b)BaseplateboltatSEcornerofthebuilding
Figure20.HeathcoteValleyPrimarySchoolSteelMomentFramebuilding(Photos:MacRae)
AsteelframedwallwithabrickfaadewaserectedinasmallparkasshowninFigure21,inapartof
townthatwheresignificantoveralstructuraldamageoccurred.Thewallwasplacedthereafterthe
September2010earthquakeasstatesRebuild,BrickbyBrick.Thewallsufferednodamageduringthe
ENG.UOA.0003.27
subsequentearthquakesandthebricktiesbetweenthesteelframingandthebricksshowdnosignsof
distress.
(a) OverallViewofWall
(b)BackViewofWall
Figure21.September2010RebuildingStandConsistingofBricksSupportedbySteelFrame(Photos:
MacRae)
LIGHTSTEELFRAMEDHOUSES
Thereisasmallnumberoflightsteelframedhousesintheaffectedarea.Preliminaryreportsarethat
damagetoframing,brickworkandliningswaslessthanfromtheSeptemberearthquake,discounting
damageresultingfromsoilliquefactionandlateralspreading.
Inonehousewithbrickveneer,afewbricksonthetopcourseandadjacenttowindowopeningshad
beenloosened,butnotdislodged.
Thisbehaviourisconsistentwiththeverygoodperformanceofbrickveneeronsteelframinginaseries
ofshakingtabletestsattheUniversityofMelbournein2009(PatonColeetal,2009)
BRIDGES
TherearerelativelyfewsteelbridgesintheChristchurcharea.ApedestrianarchbridgeattheAntigua
BoatshedsandoneatVictoriaSquareshowednovisibledamage(Figure22).
ENG.UOA.0003.28
Figure22UndamagedoldersteelpedestrianbridgesontheAvonRoverneartheCBD(Leon)
AlthoughsubstantialliquefactionoccurredalongtheAvonRiverneartheCBD,theonlyoldersteel
bridgeinthisareaonlyshowedspectacularbucklingofitsfasciaarches;theactualbridge,supportedon
straightrivetedgirdersappearedundamagedeventhoughlargesettlementshadoccurredatthe
abutments(Figure23).TheoldrailbridgeovertheWaimakariririverbehavedwelleventhoughitwas
clearthatthepierhadmovedover100mmtowardtheriverandbackduringthisshake(Figure24a).The
oldroadbridgesufferedsomelongitudinalbucklingofthelowerflangeofonebeam(Figure24b)aswell
assomespallingofconcreteonthewestsideoftheabutment.Theonlymajormodernsteelbridgeat
thePortofLyttleton,athreespancontinuousplategirder,hadonlyminordamageattheabutment
(Figure25).
ENG.UOA.0003.29
Figure23ColomboStreetbridge(a)Slumpingofriverbankclosetobridge;(b)Bucklingoffascia
arches;(c)Slumpingofabutmentsatendofbridge;(d)Undamagedstraightrivetedgirders.(Leon)
Figure24WaimakaririBridges,Southend(a)OldRailBridge,(b)Oldroadbridge(MacRae)
Figure25LyttletonPortBridge(a)PlateGirder(b)AbutmentSpalling(MacRae)
ENG.UOA.0003.30
ThefootbridgesshowninFigure26weredamagedintheSeptember2010earthquakeandhadnotbeen
repairedatthetimeoftheFebruary22event.Duetofurtherlateralspreadingandslumpingof
abutments,theywereevenmoredamagedinthisshaking.
Figure26Footbridges(a)TrussbridgeoverAvonRiver,(b)SuspensionbridgewithtimberdeckOver
Kaiapoiriver,(c)SuspensionbridgeatGroynes(MacRae)
CONCLUSIONS
SteelstructuresgenerallyperformedwellduringtheChristchurchearthquakeseries,comprising6
damagingeventsfrom4September2010to13June2011,withintensityupto2xULSdesignleveland
cumulativedurationofstronggroundshakinginexcessof60seconds.However,afeweccentrically
bracedframesdevelopedlinkfractures,CBFbracefractureswereobservedinconnectionsunableto
developthebracegrosssectionyieldstrength,andmultipleindustrialsteelstoragerackscollapsed.
Thediscoveryofafracturedactivelinkina22storeybuilding,inwhichallotherlinksperformedwell,is
unexplainedatthetimeofwritingthispaper,anditwillbeaprioritytodeterminethecauseofthat
fracturewhenthedamagedlinkisremovedandaccessibleforcloseinspection.
ACKNOWLEDGMENTS
ThisworkwasfundedinpartbytheFoundationforResearchinScienceandTechnologythroughthe
EngineeringthemeoftheNaturalHazardsPlatformofNewZealand,theUniversityofAuckland,the
UniversityofCanterbury,theErskineVisitingFellowshipprogramattheUniversityofCanterbury.
ParticipationofMichelBruneautothisearthquakereconnaissancestudywasfundedbyMCEER
(UniversityatBuffalo).RonDeVall(ReadJonesChristoffersenLtd,Vancouver,Canada)isalsothanked
forsharinginsightsonthebehaviourofEBFs.However,anyopinions,findings,conclusions,and
recommendationspresentedinthispaperarethoseofthewritersanddonotnecessarilyreflectthe
viewsofthesponsors.
REFERENCES
ENG.UOA.0003.31
Bruneau,M.,Anagnostopoulou,M.,MacRae,G.,Clifton,C.,Fussell,A.,PRELIMINARYREPORTONSTEEL
BUILDINGDAMAGEFROMTHEDARFIELDEARTHQUAKEOFSEPTEMBER4,2010,BulletinoftheNew
ZealandSocietyforEarthquakeEngineering,Vol.43,No.4,pp.351359.
HayesGetal,THE09/03/2010DARFIELDEARTHQUAKEANDITSAFTERSHOCKS,INCLUDINGTHE
02/21/2011CHRISTCHURCHEVENT,EducationalSlides,USGeologicalSurvey,NationalEarthquake
InformationCenter,2011.
HylandCandFergussonWG,AFRACTUREMECHANICSBASEDAPPROACHTOTHEASSESSMENTOF
SESIMICRESISTINGSTEELSTRUCTURESKeyEngineeringMaterials312(2006)8994
PatonCole,V.P;Gad,E.F;Clifton,G.C;Heath,D.J;Davies,C;Hicks,S;Lam,N.SEISMICPERFORMANCE
OFABRICKVENEERSTEELFRAMEDHOUSE,AEES2009,Proceedingsofthe2009Conferenceofthe
AustralianEarthquakeEngineeringSociety,Newcastle,Australia,2009
Seal,C,PLASTICITYOFSTEELUNDERSESIMICLOADCONDITIONSPhDThesisreport,Facultyof
Engineering,TheUniversityofAuckland,2009
StandardsNewZealand,NZS1170Part5:2004'EARTHQUAKEACTIONSNEWZEALAND'partoftheJoint
AustralasianLoadingsStandardsetAS/NZS1170'StructuralDesignActions'.
StandardsNewZealand,NZS3404:1997incorporatingAmendmentNo1:2001andAmendmentNo2:
2007,STEELSTRUCTURESSTANDARD,Wellington,NewZealand
GNS,CHRISTCHURCHCENTRALBUSINESSDISTRICTSPECTRA,25/02/2011