S O L U T I O N S F O R I M P R O V E D S E I S M I C P E R F O R M A N C E O F N E W A N D

E X I S T I N G S T R U C T U R E S

Habilitation thesis

Aurel Stratan

2015

i

ACKNOWLEDGEMENTS

Writingthishabilitationthesisgavemetheopportunitytolookbackoverthelast12yearsofmyprofessionalexperience,sincedefendingthePhDthesisin2003atthePolitehnicaUniversityofTimisoara.Isawasteadyenlargementofthespectrumofactivities,increaseofthecomplexityofthework,andmoreliabilities.Italsocompelledmelookingintothefutureandplanningitinastructuredway.

The scientific, professional and academic accomplishments described in this thesis would nothavebeenpossiblewithoutthecontributionofalonglistofpeople,ofwhichonlyafewwillbementionedhere.Thereforethisworkisnotonlymine,butalsoofallofthem.

FirstofallIwouldliketodeeplythankProfessorDanDubina,whoconstantlysupportedmeinthescientificandprofessionalcareer,openingnewopportunities,providingusefuladvicesandmotivatingmetogoonindifficultsituations.Thisthesiswouldnothavebeenwrittenwithouthisincentives.

I would like to thank my fellows from the department of Steel Structures and StructuralMechanics, Adrian Ciutina, Daniel Grecea, Florin Dinu, Raul Zaharia, Viorel Ungureanu, MirceaGeorgescu, AdrianDogariu, Edward Petzek and RamonaSzabo for providing a motivating andsupportiveworkingenvironment.

Thanks to all younger colleagues, former or present PhD students, Cristian Vulcu, IonelMărginean, Adriana Ioan, Norin Filip-Văcărescu, Călin Neagu, Nicu Muntean, Sorin Bordea,Andrei Crișan, Ioan Both and Cosmin Mariș for their dedicated work in different researchprojects.

Thankstoallpartnersininternationalandnationalresearchprojectsfortheircollaborationandmotivation.

Lastbutnotleast, thankstomyfamilyforsharingmewithmyjobandpatientlytoleratingmyworkschedule.

ii

SUMMARY

Thishabilitationthesispresentsthemainscientific,professionalandacademicachievementsofAurel Stratan following the defence of the PhD thesis in 2003 at the Politehnica University ofTimisoara,aswellasthefuturedevelopmentplan.

The main research area of the author fits into the broad and multi-disciplinary area ofearthquake engineering, with particular emphasis on seismic performance of steel structuresand rehabilitation of existing buildings using metal-based solutions. The most important andrelevant research directions pursued by the author are: "Re-centring eccentrically bracedframes","Cold-formedsteelpitched-roofportalframeswithboltedjoints","Highstrengthsteelin seismic resistant structures", "Seismic rehabilitation of existing reinforced concrete andmasonry buildings with steel-based solutions", "Validation of the technical solution for braceswith truepinconnections", "Seismicperformanceofmulti-storeysteelstructures with frictiondampers"and"Prequalificationofboltedbeamtocolumnjointswithhaunches".Experimentalinvestigationmethods represent themaintoolof theresearch,supportedat thesametimebyadvanced numerical simulations and analytical tools. The habilitation thesis summarises theevolution of the research performed by theauthor following the defence of the PhD thesis, aswellasthemainoutcomes,outliningalsothecontextinwhichtheresearchwasperformed,i.e.funding scheme, dissemination of results, and associated PhD theses. There were 12 grantssupportingtheresearch:4nationalgrants,7internationalgrantsand1researchcontractwithindustry.Theresultsweredisseminatedin92publications(journalandconferencepapers,andbookchapters).SixPhDstudentswereinvolvedintheresearch(5PhDthesesweresuccessfullydefendedand1iscurrentlyunderdevelopment).AurelStratanhadanactiveroleinguidingthePhDcandidates.

Professional development of Aurel Stratan followed a wide pallet of activities, includingparticipationto trainingcourses, structuraldesign, industry-orientedresearch, involvement inprofessionalorganisationsandtechnicalcommittees,codedrafting,developmentoftheresearchinfrastructure, organisation of scientific events, short-term scientific missions, involvement inadministrative duties and peer-review of scientific publications. Though the author had arelatively limited involvement in structural design activity, it lead to important achievements,suchasparticipationinthestructuraldesignoftheBucharestTowerCenter,anofficebuildingwithtotalheightof106.3m,currentlythethirdtalleststructure inBucharest.AurelStratanismemberinseveralnationalprofessionalorganisations:AICPS,APCMR,AGIR-SBIS.Heisalsoanactivememberinseveralnationalandinternationaltechnicalcommittees:TechnicalCommitteeTC13 "Seismic Design" of the European Convention for Constructional Steelwork (ECCS),CEN/TC250/SC8"Eurocode8:Earthquakeresistancedesignofstructures",CEN/TC340/WG5"Revision of EN 15129 – Anti-seismic devices", ASRO CT 343 "Basis of design and structuraleurocodes", CTS4 " Actions on structures", Ministry of Regional Development and PublicAdministration (MDRAP). The main work undertaken within these structures concernsimprovement of national and European design guidelines in the field of earthquake-resistantdesign. Aurel Stratan is actively involved in maintaining and upgrading of the existing testingfacilities in the laboratory of Steel Structures from the department of Steel Structures andStructuralMechanics.Hewasmemberofthescientificcommitteeofthreeconferences,memberin theorganisingcommitteeof twoconferencesandchaired twosessionswithininternationalconferences.

Inwhatconcernstheacademicarea,AurelStratanteachesthecoursesof"Structuraldynamicsand earthquake engineering" and "Basis of structural design", as well as the project of "Steelstructures" at bachelor level, and "Performance based design" and "Redesign of existing

iii

structures" at the master levels. All of them are supported by teaching aids developed by theauthor, which includes a book and electronic course notes posted on faculty website. Severaldiplomaworksandmasterdissertationsarecoordinatedeachyear, the latterbeingcorrelatedwiththeon-goingresearchprojects.Theauthorwasrecently involvedinteachingactivitiesofthe European Erasmus Mundus Master Course "Sustainable Constructions under NaturalHazards and Catastrophic Events". He coordinated the module 2C09 "Design for seismic andclimatechanges"thatweredeliveredinatthePolitehnicaUniversityofTimisoarainthespringof2014.PartofthelectureswasdeliveredbyAurelStratanattheUniversityofNaplesFedericoII in March 2015, as part of the 2014-2016 edition of the European Erasmus Mundus MasterCourse.

Inwhatconcerns the futuredevelopmentplan,severalnewresearchdirectionsare identified:extensionoftheconceptofre-centringstructurestodualBuckling-RestrainedFrames(BRBFs)and Steel Plate Shear Walls (SPSWs), pseudo-dynamic and shaking table testing techniques,BucklingRestrainedBracedFrames(BRBFs),seismicprotectionofbuildingsusingpassiveandsemi-active control, improved seismic design criteria for steel structures in case of groundmotionswithhighfrequencycontentinthelong-periodrange,designcriteriaforconnectionsinconcentrically braced frames and prequalification of typical European connections, seismicvulnerability and risk assessment. Moreover, development of the international, national andindustry-orientedcooperation is envisaged,as wellasamoreactiverole ingrantapplicationsandimprovementoftheresearchinfrastructure.Lastbutnotleast,theteachingmethodswillbeimproved by involving the students through interactive learning techniques. Furtherdevelopment of the international cooperation at the academic level and continuing educationarealsotargeted.

iv

REZUMAT

Aceastătezădeabilitareprezintăprincipalelerealizăriștiințifice,profesionaleșiacademicealelui Aurel Stratan după susținerea tezei de doctorat in 2003 la Universitatea PolitehnicaTimișoara,precumșidirecțiileviitoarededezvoltarealeacestora.

Principaluldomeniudecercetareaautoruluiseîncadreazăîndomeniulvastșimulti-disciplinaral ingineriei seismice, cu un accent special pe performanță seismică a structurilor din oțel șireabilitarea clădirilor existente folosind soluții bazate pe metale. Cele mai importante șirelevante direcțiile de cercetare urmărite de autor sunt: "cadre contravântuite excentric cucapabilitatedere-centrare","cadreportaldinprofile formate larececuîmbinăricușuruburi","utilizarea oțelului de înaltă rezistență la structurile anti-seismice", "reabilitarea seismică aclădirilor existente din beton armat și zidărie folosind soluții bazate pe metale", "validareasoluțieitehnicepentrucontravântuiricuîmbinarecubolț","performanțăseismicăastructurilormetalicemultietajatecuamortizoricufrecare"și"precalificareaîmbinărilorgrindă-stâlpvutatecu șuruburi". Metoda de investigare experimentală reprezintă principalul instrument decercetare, susținut în același timp de simulări numerice avansate și metode analitice. Teza deabilitare prezintă evoluția cercetărilor efectuate de autor după susținerea tezei de doctorat,precum și principalele rezultate obținute, subliniind, de asemenea, contextul în care a fostefectuatăcercetarea,adicăschemadefinanțare,moduldediseminarearezultatelorșitezelededoctoratasociate.Activitateadecercetaredescrisăafostsusținutăde12degranturi:4granturinaționale, 7 granturi internaționale și 1 contract de cercetare cu industria.Rezultatele au fostdiseminateîn92publicații(articoleînrevisteșilaconferințeștiințifice,precumșisicapitoledecarte). În activitatea de cercetare au fost implicați șase doctoranzi (5 teze de doctorat au fostsusținut cu succes, iar 1 se află în curs de elaborare). Aurel Stratan a avut un rol activ înîndrumareadoctoranzilor.

Dezvoltarea profesională a lui Aurel Stratan a inclus o paletă largă de activități, precumparticiparea la cursuri de formare, proiectare, cercetare aplicativă, implicarea în organizațiiprofesionale și comitetele tehnice, elaborarea de coduri de proiectare, dezvoltareainfrastructuriidecercetare,organizareadeevenimenteștiințifice,stagiidecercetare,implicareaînactivitățiadministrativeșirecenzieapublicațiilorștiințifice.Deșiautorulaavuto implicarerelativ limitată în activitatea de proiectare, aceasta a condus la realizări importante, cum ar fiparticiparea la proiectarea structurală a Bucharest Tower Center, o clădire de birouri cuînălțimea totala de 106,3 m, în prezent pe locul trei în top-ul structurilor cele mai înalte dinBucurești. Aurel Stratan este membru în mai multe organizații profesionale naționale: AICPS,APCMR, AGIR-SBIS. El este de asemenea un membru activ în mai multe comitete tehnicenaționaleșiinternaționale:ComitetulTehnicTC13"Proiectareaseismică"aConvențieiEuropenedeConstrucțiiMetalice(ECCS),CEN/TC250/SC8"Eurocod8:Proiectareastructurilorpentrurezistențălacutremur",CEN/TC340/WG5"RevizuireaEN15129-dispozitiveanti-seismice",ASRO CT 343 "Bazele proiectării şi eurocoduri pentru structuri", CTS4 "Acțiuni asupraconstrucțiilor", Ministerul Dezvoltării Regionale și Administrației Publice (MDRAP). Principalaactivitateîntreprinsăîncadrulacestorgrupurisereferălaîmbunătățireacodurilordeproiectarenaționaleșieuropene îndomeniulproiectăriiantiseismice.AurelStrataneste implicatactiv înîntreținereașimodernizareaechipamentelorexperimentaleexistenteînlaboratoruldeStructurial departamentului de Construcții Metalice și Mecanica Construcțiilor. El a fost membru încomitetuluiștiințificatreiconferințe,membruîncomitetuldeorganizareadouăconferințeșiaprezidatdouăsesiuniîncadrulunorconferințelorinternaționale.

În ceea ce privește domeniul academic, Aurel Stratan este titular de curs la disciplinele"Dinamica structurilor și inginerie seismică" și "Bazele proiectării structurilor", precum și

v

proiectul de "Structuri metalice", la nivel de studii de licență, și "Proiectarea anti-seismică pecriteriideperformanță"și"Procedeepentrureproiectareaconstrucțiilor"lanivelulstudiilordemasterat.Toateacesteasuntsusținutedematerialedidacticeelaboratedeautor,careincludocarte și note de curs electronice postate pe site-ul facultății. Autorul coordonează anualelaborareacâtorvalucrăridediplomășidedisertație,subiectulcelordinurmăfiindcorelatcuproiectele de cercetare în curs de desfășurare. Autorul a fost implicat recent în activitățile depredare la cursul de master European Erasmus Mundus "Sustainable Constructions underNaturalHazardsandCatastrophicEvents".Elacoordonatmodulul2C09"Designforseismicandclimatechanges"predatlaUniversitateaPolitehnicadinTimișoaraînprimăvaraanului2014.Oparte din prelegeri au fost susținute de către Aurel Stratan la Universitatea "Federico II" dinNapoli în martie 2015, în cadrul ediției 2014-2016 a cursului de master European ErasmusMundus.

Înceeacepriveșteplanuldedezvoltareacariereiștiințifice,suntidentificatemaimultedirecțiide cercetare noi: extinderea conceptului dere-centrare lastructuri dualecucontravântuiri cuflambaj împiedicat (BRBFs) și pereți de forfecare din oțel (SPSWs), tehnici experimentalepseudo-dinamiceșidinamice,cadrecucontravântuiricuflambajîmpiedicat(BRBFs),protecțiaseismicăaclădirilorfolosindcontrolulpasivșisemi-activ,îmbunătățireacriteriiledeproiectareseismicăastructurilormetaliceîncazulmișcărilorseismicecuunconținutridicatdefrecvențeîndomeniulperioadelorlungi,criteriiledeproiectarepentruîmbinărilacadrelecontravântuitecentric și precalificarea unor îmbinări europene tipice, evaluarea vulnerabilității și a risculuiseismic.Înplus,seareînvederedezvoltareacooperăriiinternaționale,naționaleșicuindustria,precumșiasumareaunuirolmaiactivînaplicațiiledefinanțareacercetăriișiainfrastructuriide cercetare. Nu în ultimul rând, metodele de predare vor fi îmbunătățite prin implicareastudenților folosind tehnici interactive de predare. Sunt vizate, de asemenea, dezvoltarea încontinuareacooperăriiinternaționalelanivelacademicșiaeducațieicontinue.

vi

CONTENTS

ACKNOWLEDGEMENTS ...................................................................................................................................................... I

SUMMARY ............................................................................................................................................................................. II

REZUMAT ............................................................................................................................................................................. IV

CONTENTS ............................................................................................................................................................................ VI

1 SCIENTIFIC, PROFESSIONAL AND ACADEMIC ACHIEVEMENTS ................................................................. 1

1.1 SCIENTIFICACHIEVEMENTS....................................................................................................................................................1

1.2 LISTOFRELEVANTPUBLICATIONS......................................................................................................................................10

1.3 PROFESSIONALACHIEVEMENTS..........................................................................................................................................11

1.4 ACADEMICACHIEVEMENTS..................................................................................................................................................14

2 RESEARCH OUTCOMES ......................................................................................................................................... 16

2.1 RE-CENTRINGECCENTRICALLYBRACEDFRAMES.............................................................................................................16

2.1.1 The concept of re-centring structure ................................................................................................................... 16

2.1.2 Design criteria for dual frames with re-centring capability ....................................................................... 17

2.1.3 Experimental tests on one-storey eccentrically braced frame ................................................................... 23

2.1.4 Large-scale tests on re-centring dual eccentrically braced frame ........................................................... 30

2.1.5 link replacement order in higher-rise frames ................................................................................................... 37

2.1.6 Outcomes ......................................................................................................................................................................... 41

2.1.7 Publications .................................................................................................................................................................... 42

2.2 COLD-FORMEDSTEELPITCHED-ROOFPORTALFRAMESWITHBOLTEDJOINTS...........................................................45

2.2.1 Tests on joint specimens ............................................................................................................................................ 45

2.2.2 Tests on full-scale pitched-roof portal frames .................................................................................................. 56

2.2.3 Outcomes ......................................................................................................................................................................... 61

2.2.4 Publications .................................................................................................................................................................... 62

2.3 HIGHSTRENGTHSTEELINSEISMICRESISTANTSTRUCTURES.........................................................................................64

2.3.1 Steel-concrete connection in concrete-filled rectangular hollow section columns ........................... 64

2.3.2 Experimental tests on welded beam to concrete-filled RHS column joints ........................................... 69

2.3.3 Numerical Investigation of welded beam to concrete-filled RHS column joints ................................. 78

2.3.4 Outcomes ......................................................................................................................................................................... 85

2.3.5 Publications .................................................................................................................................................................... 86

2.4 SEISMICREHABILITATIONOFEXISTINGREINFORCEDCONCRETEANDMASONRYBUILDINGSWITHMETAL-BASED

SOLUTIONS..............................................................................................................................................................................................89

vii

2.4.1 Strengthening of r.c. frames with buckling restrained braces ................................................................... 89

2.4.2 Strengthening of masonry walls with metal-based techniques ................................................................. 96

2.4.3 Publications .................................................................................................................................................................. 103

2.5 VALIDATIONOFTHETECHNICALSOLUTIONFORBRACESWITHTRUEPINCONNECTIONS......................................105

2.5.1 Pre-test finite element analyses ........................................................................................................................... 106

2.5.2 Experimental investigation .................................................................................................................................... 110

2.5.3 Publications .................................................................................................................................................................. 113

2.6 SEISMICPERFORMANCEOFMULTI-STOREYSTEELSTRUCTURESWITHFRICTIONDAMPERS..................................114

2.6.1 Experimental program ............................................................................................................................................ 116

2.6.2 Seismic performance of multistorey frames with SERB dampers ........................................................... 119

2.6.3 Publications .................................................................................................................................................................. 126

2.7 PREQUALIFICATIONOFBOLTEDBEAMTOCOLUMNJOINTSWITHHAUNCHES..........................................................127

2.7.1 Experimental program ............................................................................................................................................ 127

2.7.2 Pre-test numerical simulations ............................................................................................................................ 128

2.7.3 Publications .................................................................................................................................................................. 133

3 PROFESSIONAL DEVELOPMENT PLAN ..........................................................................................................134

3.1 SCIENTIFICDEVELOPMENTPLAN.....................................................................................................................................134

3.2 PROFESSIONALDEVELOPMENTPLAN..............................................................................................................................136

3.3 ACADEMICDEVELOPMENTPLAN......................................................................................................................................137

4 REFERENCES ...........................................................................................................................................................138

1

1 SCIENTIFIC,PROFESSIONALANDACADEMICACHIEVEMENTS

Thishabilitationthesispresentsthemainscientific,professionalandacademicachievementsoftheauthorfollowingthedefenceofthePhDthesisin2003atthePolitehnicaUniversityofTimisoara.

1.1 SCIENTIFICACHIEVEMENTS

Themainresearchareaoftheauthorfitsintothebroadandmulti-disciplinaryareaofearthquakeengineering,withparticularemphasisofseismicperformanceofsteelstructuresandrehabilitationof existing buildings using metal-based solutions. The most important and relevant researchdirectionspursuedbytheauthorare: Re-centringeccentricallybracedframes. Cold-formedsteelpitched-roofportalframeswithboltedjoints. Highstrengthsteelinseismicresistantstructures. Seismicrehabilitationofexistingreinforcedconcreteandmasonrybuildingswithsteel-based

solutions. Validationofthetechnicalsolutionforbraceswithtruepinconnections. Seismicperformanceofmulti-storeysteelstructureswithfrictiondampers. Prequalificationofboltedbeamtocolumnjointswithhaunches.

"Re-centring eccentrically braced frames"representsthebackboneoftheresearchinterestsoftheauthor.Moststructuresdesignedaccordingtomodernseismicdesigncodesfollowtheconceptofdissipativebehaviour,whichreliesonplasticdeformationsinordertoabsorbtheseismicenergy.Importantstructuralandnon-structuraldamageisthereforeoccurringinbuildingsunderlargeandeven moderate earthquakes. The most radical solutions to avoid such damage are base isolationandvariousimplementationsofactiveandsemi-activestructuralcontrol.Otherstrategiesrelyonsupplemental damping conferred to the structure through various devices based on viscous,friction, or yielding dampers. Self-centring systems, a relatively new concept that addresses thedrawbacksoftheconventionalyieldingsystems,havereceivedmuchattentionrecently.

An alternative solution is to provide re-centring capability (as opposed to self-centring), byremovable dissipative members and dual (rigid-flexible) structural configuration. Application ofthisapproachtodualeccentricallybracedframeswithreplaceablelinkswasfirstsuggestedbytheauthor in his PhD thesis (Stratan, 2003). Though experimental investigations on isolatedreplaceable links and numerical assessment of seismic performance of dual eccentrically bracedframeswithreplaceablelinkswereperformedaspartofthePhDthesis,thesolutionwasstillinitsconceptualstage.

In the following years this idea was steadily developed. First a series of experimental tests wereperformedononestorey,one-bayeccentricallybracedframewithreplaceablelinks(Figure1a),inorder to investigate the connection design parameters and feasibility of link replacement at thesystemlevel.Additionally,numericalsimulationsonboltedlinksanddualstructuralsystemswereperformed. These research activities were conducted at the Politehnica University of Timisoara,and were supported by a national research grant coordinated by the author: grant no. 1434/27.04.2006 type CEEX-ET module 2 (2006-2008). "Structuri metalice duale cu elemente disipative demontabile pentru construcţii amplasate în zone seismice (Dual metal structures with removable dissipatie elements for constructions in seismic areas)", financed by the Ministry ofEducationandResearch.

Startingwith2010,theresearchreachedanewlevel,aftertheteamfromthePolitehnicaUniversityof Timisoara, CEMSIG research center, was awarded a FP7 SERIES project: JRC N° 31817 / 2010 (2010-2014). "Full-scale experimental validation of dual eccentrically braced frame with removable

2

links (DUAREM)", Transnational Access within the framework of Grant Agreement No. 227887financed by the European Commission. Project objectives were to (1) validate the re-centeringcapability of dual structures with removable dissipative members; (2) assess overall seismicperformanceofdualeccentricallybracedframes;(3)obtaininformationontheinteractionbetweenthesteel frameandthereinforcedconcreteslab in the linkregion.TheresearchconsortiumwasconstitutedbyPolitehnicaUniversityofTimisoara(theLeadUser),UniversityofNaples"FedericoII", University of Liege, University of Ljubljana, and University of Coimbra. Within this fundingscheme,aseriesoffull-scalepseudo-dynamictestswereperformedattheEuropeanLaboratoryforStructural Assessment (ELSA) at Joint Research Centre (JRC) in Ispra, on a three-storey dualeccentricallybracedstructurewithreplaceablelinks(Figure1b).Acomprehensivesetofnumericalsimulations were also performed in order to assess the seismic performance and optimal linkreplacementorderofdualeccentricallybracedframeswithreplaceablelinks.Theoutcomesofthisresearch validated the concept of re-centring structure, proving the technical feasibility of linkremoval and replacement, as well as of the design approach. It cleared the route towardimplementationintopracticeofthisinnovativestructuraldesignconcept.Anin-depthcoverageofthisresearchsubjectcanbefollowedinsection2.1.

(a)

(b)

Figure1.Experimentaltestonalmostfull-scaleframewithboltedlinksintheStructuralLaboratoryofthePolitehnicaUniversityofTimisoara,CEMSIGresearchcentre(a)andtheexperimentalmock-upinfrontofthereactionwalloftheEuropeanLaboratoryforStructuralAssessment(ELSA)atJRC

inIspra(b).

Obtainedresultsweredisseminatedin25publications(journalpapers,conferenceproceedingsanda report); for a complete list see section 2.1.7. A PhD thesis entitled "Seismic performance of re-centringdualeccentricallybraced frames withremovable links"waselaboratedanddefendedaspartoftheresearchbyAdrianaIoanunderthesupervisionofProf.dr.ing.DanDubina.TheauthorofthishabilitationthesishadanactiveroleinguidanceofthePhDcandidate.

Within the second research direction, "Cold-formed steel pitched-roof portal frames with bolted joints",anexperimentalprogramwascarriedoutinordertoevaluatetheperformanceofpitched roof cold-formed steel portal frames of back-to-back channel sections and bolted joints.Threedifferentconfigurationsofridgeandeavesjointsweretested(Figure2a).Thebehaviourandfailure mechanisms of joints were observed in order to evaluate their stiffness, strength andductility.Jointsbetweencold-formedmemberswithboltsinthewebonlyresultinareductionofjointmomentcapacityandprematurewebbuckling.Thecomponentmethodwasappliedinordertocharacterizethejointstiffnessandmomentcapacityforthepurposeofframeanalysisanddesign.The influence of jointscharacteristics on the global frame response under lateral (seismic) loadswas analysed by considering three connection models. Full-scale tests were performed on cold-formedpitched-roofportalframes(Figure2b).

3

(a)

(b)

Figure2.Componenttests(a)andsystem-leveltests(b)oncold-formedsteelportalframesofback-to-backchannelsectionsandboltedjoints.

The research was funded through the national grant no. 32940/2004 (2004-2006) type A, CNCSIS code 164 "Incercari experimentale pe cadre portal din profile din otel formate la rece pentru cladiri civile si industriale in zone seismice – EXPOFORS (Experimental tests on portal frames from cold-formed steel profiles for civil and industrial buildings located in seismic areas)", financed by theministryofresearchandeducation.Theauthorofthehabilitationthesispartoftheresearchteam.

The study suggested that the classical calculation model for connections, assuming the centre ofrotationtobelocatedatthecentroidoftheboltgroupandalineardistributionoftheforcesoneachbolt,isinappropriateforframesfeaturingcold-formedmembers.Theforcedistributionisunequalduetotheflexibilityoftheconnectedmembers.Infact,theforceisanorderofmagnitudebiggerintheouterboltrowscomparedtomostinnerones.Aconnectionwithboltsonlyonthewebcausesconcentrated forces in the web of the connected member and leads to premature web buckling,reducingthejointmomentcapacity.Thistypeofconnectionsisalwayspartialstrength.Iftheloadbearingcapacityoftheconnectedbeamistobematchedbytheconnectionstrength,boltsontheflanges become necessary. The ductility of the connection is reduced both under monotonic andcyclicloadsandthedesign,includingthedesignforearthquakeloads,shouldtakeintoaccountonlytheconventionalelasticcapacity.Becausethereisnosignificantpost-elasticstrength,therearenosignificant differences in ductility and capacity of cyclically tested specimens compared with themonotonicones.Amoredetailedcoverageofthisresearchsubjectcanbefollowedinsection0.

Obtained results were disseminated in 12 publications (journal and conference papers); for acomplete list see section 2.2.4. It has to be mentioned that the paper "Monotonic and cyclicperformance of joints of cold formed steel portal frames" authored by D. Dubina, A. Stratan, A.Ciutina,L.Fulop&Zs.Nagyreceivedthe"BestPaperAward"attheFourthInternationalConferenceon Thin-Walled Structures, Loughborough, UK, 22-24 June 2004. A PhD thesis entitled "Study ofconstructional solutions and structural performance of light-gauge halls from cold-formed steelprofiles(Studiulsoluţiilorconstructiveşiperformanţelorstructuralealehaleloruşoarecustructurarealizatădinprofiledeoţelformatelarece)"waselaboratedanddefendedaspartoftheresearchby Zsolt NAGY under the supervision of Prof.dr.ing. Dan Dubina. The author of this habilitationthesiswasalsoactivelyinvolvedinguidanceofthePhDcandidate.

The third research direction concerns "High strength steel in seismic resistant structures".Seismic resistant building frames designed as dissipative structures must allow for plasticdeformations inspecificmembers,whosebehaviorhas tobepredictedbyproperdesign. InDual

4

Frames (e.g. MRF + CBF or EBF) members designed to remain predominantly elastic duringearthquakes,suchascolumnsforinstance,arecharacterizedbyhighstrengthdemands.Dual-steelstructuralsystems,optimizedaccordingtoaPerformanceBasedDesignPhilosophy,inwhichHighStrengthSteelisusedin"elastic"membersandconnectioncomponents,whileMildCarbonSteelindissipativemembers,canbeveryreliableandcosteffective.Basedonthisidea,aresearchprogramwascarriedout inordertoevaluate theperformanceofmomentresisting jointsofHSSandMCScomponents,undermonotonicandcyclicloading.

Theresearchwasfundedthroughthenationalgrant no. 29/2005 (2005-2008) CEEX MATNANTECH "Sisteme constructive si tehnologii avansate pentru structuri din oteluri cu perfomante ridicate destinate cladirilor amplasate in zone cu risc seismic – STOPRISC (Structural systems and advanced technologies for structures from high-performance steels for buildings located in high-seismicity areas)" financed by the ministry of education and research. The project was coordinated by DanDubina,whileAurelStratanwasthescientificresponsible.

Tests on welded details indicated that welds between components of different steel gradesperformedadequatelyunderbothmonotonicandcyclicloading,foralltypesofwelds(fillet,singlebevel and double bevel). The most important factor affecting the ductility of T-stub componentsunder monotonic loading was the failure mode. Most ductile response was observed forcomponents failing by end-plate bending (mode 1), while failure modes involving bolts (mode 2and3)were lessductile.Thedegree towhichcyclic loadingaffected theductilityofT-stubswas,verymuchdependentonthefailuremode.Specimensfailingbyend-platebending(mode1)werecharacterizedbyanimportantdecreaseofductilitywithrespecttomonotonicloading,duetolow-cyclefatigue.Ontheotherhand,ductilityofspecimensinvolvingboltfailure(modes2and3)wasnot much affected by cyclic loading. Stiffening of Y-stubs increased their strength, but reducedslightlytheductility.T-stubswithend-platesrealizedfromhighstrengthsteelshowedcomparablestrengthwiththoserealizedfrommildcarbonsteel.However,oneremarksthatthinnerendplatesrealized from highstrengthsteel,at thesamestrength,areprovideequaloreven largerductility(duetofailureinmode1or2)thanthickermildcarbonsteel,evenifelongationatruptureofhighstrengthsteelwaslowerthantheoneofmildcarbonsteel.

The research was continued at the European level, through a research project coordinated byPolitehnicaUniversityofTimisoara:RFSR-CT-2009-00024 HSS-SERF 2009-2013 "High Strength Steel in Seismic Resistant Building Frames - HSS-SERF",financedbytheEuropeanCommission-ResearchFundforCoalandSteel.Otherpartnersinvolvedintheprojectwere:Univ.ofStuttgart(Germany),Univ. of Liege (Belgium), Univ. of Ljubljana (Slovenia), Univ. “Federico II” of Naples (Italy), VTTTechnicalResearchCentre(Finland),GIPACLtd.DesignOffice(Portugal),RIVAAcciaioS.p.AwithUniv.ofPisaassubcontractor(Italy),andRuukkiConstructionOy(Finland).

Alargenumericalandexperimentalprogramwasundertaken,addressingthefollowingobjectives:(1) to find reliable structural typologies and joint/connection detailing for dual-steel buildingframes,andtovalidate themby tests andadvancednumericalsimulations; (2) todevelopdesigncriteria and PBD methodology for dual-steel structures using HSS; (3) to recommend relevantdesign parameters (i.e. behaviour factor q, overstrength factor Ω) to be implemented in furtherversionsoftheseismicdesigncode,suchasEN1998-1,inordertoapplycapacitydesignapproachfor dual-steel framing typologies; (4) to evaluate technical and economic benefit of dual-steelapproachinvolvingHSS.

Themainoutcomesandcontributionsoftheprojectwerethefollowing:(1)principlesanddesignrecommendationsfordual-steelframes(guidelines);(2)theinvestigatedframetypologiesbasedonthedual-steelapproachwithcompositecolumns,aresolutionswithahighinnovativecharacterintheEuropeancontext;(3)characterisationintermsofglobalductilityandover-strengthdemandsof dual-steel frames; (4) proposal of a series of innovative beam-to-column joint typologies withcompositecolumns(partiallyencased–PE,fullyencased–FE,andconcretefilledtubes–CFT),for

5

whichthestructuralperformancewasconfirmedbyexperimentalandnumericalinvestigations;(5)recommendationsforwelddetailsandappropriatecomponentmethoddesignapproaches.

TheresearchactivitiesspecificallyconductedatthePolitehnicaUniversityofTimisoaraincludedanexperimentalprogram (Figure3)ontwotypesofmomentresisting joints indual-steel framesofconcrete filled highstrength steel rectangular hollow section (CF-RHS) columns and mild carbonsteel beams. The parameters considered in the configuration of the joints are given by two jointtypologies(reducedbeamsectionRBS,coverplatesCP),twosteelgradesfortheRHStubes(S460,S700), and two failure modes (beam, joint components). Besides, two loading conditions(monotonic,cyclic)wereconsidered,leadingtoatotalnumberof16jointassemblies.

(a) (b)

(c)

Figure3.Moment-rotationcharacteristic(a),failuremode(b)andFEMsimulations(c)oftheS700-RBS-Cjoint.

Asageneralconclusion,theexperimentalinvestigationsevidencedagoodconceptionanddesignofthe joints (RBS and CP), justified by: elastic response of the connection zone, formation of theplastichinge in thebeam,satisfactoryresponseofconnectioncomponents. Inaddition, thestudyprovedthefeasibilityofusinghighersteelgradesinnon-dissipativemembers(columns)andjointcomponents.Furthermore,theseismicperformanceoftheweldedbeam-to-CFTcolumnjointswasevaluated. Corresponding to the Significant Damage performance level, the RBS and CP jointconfigurations evidenced rotation capacities larger than the 40 mrad, and therefore the seismicperformanceofthejointswasconsideredasacceptable.Amoredetailedcoverageofthisresearchsubjectcanbefollowedinsection0.

Obtainedresultsweredisseminatedin28publications(journalandconferencepapers,aswellasaresearchreport);foracompletelistseesection2.3.5.APhDthesisentitled"Seismicperformanceofdual steel frames of CFRHS and welded beam-to-column joints" was elaborated and defended aspartoftheresearchbyCristianVULCUunderthesupervisionofProf.dr.ing.DanDubina.Theauthorof this habilitation thesis was actively involved in guidance of the PhD candidate. Moreover, aninternational workshop on "Application of High Strength Steels in Seismic Resistant Structures"wasorganisedbytheUniversityofNaples"FedericoII"andPolitehnicaUniversityofTimisoara,incooperation with European Convention for Constructional Steelwork - Committee TC13 "SeismicDesign"on June28-292013, inNaples, Italy.Workshopproceedingswereedited byDanDubina,RaffaeleLandolfo,AurelStratan,andCristianVulcuattheOrizonturiUniversitarepublishinghouse(ISBN:978-973-638-552-0).

The fourth research direction is "Seismic rehabilitation of existing reinforced concrete and masonry buildings with steel-based solutions". Two innovative strengthening solutions formasonry walls were investigated. First one consists in sheeting some steel or aluminium plateseither on both sides or on one side of the masonry wall. Metallic plates are fixed either withprestressed steel ties, or using chemical anchors. The second one is derived from the FRPtechnique, but applies a steel wire mesh bonded with epoxy resin to the masonry wall. The

Rotation [mrad]

Mo

me

nt

[kN

m]

S700-RBS-CEnvelope

-90 -60 -30 0 30 60 90

DLSD

NC

-1000

-1000

-500

500

0

6

proposedstrengtheningsolutionsareanalternativetoFRPtechnologyenablingtoobtainaductileincrease of strength, but without increasing the stiffness of the wall. It was concluded that steelplates increases mainly the ductility, while wire mesh increases the resistance. Both techniqueswere more efficient when applied on both sides. The prestressed tie connections were moreappropriateandthespecimenssheetedwithaluminiumplateshaveshownabetterbehaviourthanonessheetedwithsteel.

(a)

(b)

Figure4.Testingofamasonrywall(a)andabucklingrestrainedbrace(b).

A related subject concerned gravity-only designed reinforced concrete (RC) buildings located inseismic zones that need seismic upgrade in order to comply with modern seismic designrequirements. The effectiveness of the Buckling Restrained Braces (BRB) in strengthening andincreasingductilityinnon-seismicreinforceconcreteframewasexamined.TheBRBmembershavebeen designed to respond to the strengthening demand resulted from non-linear analysis andtested in order to observe their functionality. Numerical analysis showed that the use of BRBsystemhastobeassociatedwithlocalFRPconfinementofcolumnsatleast(confinementofbeamswouldbebeneficial, too).BRBspecimens havebeendesignedaccordingto thedemandsresultedfromtheanalysisandaccordingtoFEMA356criteria.Usingthesamesteelcore,threetypesofBRBhave been prepared, using three different unbonding materials i.e. polyethylene film, asphalticbitumenandrubber.Bothmonotonicandcyclictestshavebeenperformed.CyclictestshavebeencarriedoutaccordingtobothAISC2005andECCSRecommendations.AllthethreeBRBsolutionssatisfied thedemand,howevertheoneusingpolyethylene film provedabetterbehaviour. ItwasconcludedthattheBRBsystemscanbedesignedandappliedasaneffectivestrengtheningsolutionfor poor seismic resistant RC Frames. A more detailed coverage of this research subject can befollowedinsection0.

Theresearchwassupportedbyfourprojects,oneofwhichwascoordinatorbytheauthor: RFCS-CT-2007-00050 STEELRETRO (2007-2010) "Steel solutions for seismic retrofit and upgrade

of existing constructions",financedbytheEuropeanCommission-ResearchFundforCoalandSteel.

FP6 INCO-CT-2004-509119/ 2003 (2003-2008) "Earthquake Protection of Historical Buildings by reversible Mixed Technologies - PROHITECH",financedbytheEuropeanCommission.

C18873/28.12.2005 (2006-2008) bilateral Romanian-Greek research program "Strengthening and rehabilitation of historical buildings by reversible technologies",financedbytheministryofeducationandresearch(programcoordinator).

RFS2-CT-2014-00022 (2014-2015) "Steel based applications in earthquake-prone areas (STEELEARTH)",financedbytheEuropeanCommission-ResearchFundforCoalandSteel.

Obtainedresultsweredisseminatedin11publications(journalandconferencepapers,aswellastwo books); for a complete list see section 2.4.3. A PhD thesis entitled "Dual frame systems of

7

bucklingrestrainedbraces"waselaboratedanddefendedaspartoftheresearchbySorinBORDEAunder the supervision of Prof.dr.ing. Dan Dubina. The author of this habilitation thesis was alsoactivelyinvolvedinguidanceofthePhDcandidate.

Thefifthsubjectdescribedhereinconcerns"Validation of the technical solution for braces with true pin connections".Theresearchwastriggeredbyapracticalproblemthatcameupduringthedesignofamulti-storeybuildingwithtwoundergroundand29levelsabovegroundinBucharest,Romania by SC Popp & Asociatii SRL. The building is characterised by in-plan dimensions of atypical floor of 52.0x25.6 m, and a total height of 117.6 m. The structure uses steel framing forresisting gravity forces. In the transversal direction the main lateral force resisting system iscomposed of two reinforced concrete cores, while in the longitudinal one the cores aresupplemented by steel braces located in the facade of the building. The braces are placed in Xconfigurationdevelopedovertwostoreys.Thisreducesthenumberofbraceconnectionsandhelpsincomplyingwithcode limitationsonslenderness.Bracesarerealisedfromhot-finishedCircularHollowSections(CHS)andhaveconnectionswithpins."Truepin"connectionswithgussetplatesand pin were adopted for brace connections. One of the brace connections has an eccentric pin,allowing forvariationof thepin-to-pin length,which facilitates erectionononehand,andallowscompensationforaxialforcesinbracesduetogravityloadsontheotherhand.Highstrengthsteelwasusedforgussetsandpin,inordertokeepconnectiondimensionstoaminimum.

(a)

S,Mises[N/mm2]

(b)

(c)

Figure5.Buildingunderdesign(a);FEManalysesonthetrue-pinconnection(b)andexperimentaltestingofthebrace(c).

Numericalsimulationsandexperimentaltestingwereperformedinordertovalidatethedesignofpinned brace for a seismic resistant multi-storey building (Figure 5). Based on finite elementnumericalsimulations,amorecompactsolutionwaspro-posedandadopted,usinghigh-strengthsteel components. Due to connection detailing and tubular shape of the cross-section, the braceassemblywasshowntobesensitivetooutofplanebuckling,leadingtofailureoftheconnectioninabrittleway.Firmlyfixingthewasherstothepinhelpspreventingbrittlefailureoftheconnection,evenwhenbucklingtakesplaceoutoftheplaneoftheconnection.Themaincausesofoutofplanebucklingare(1)thefreeoutofplanerotationsoftheconnectionatsmalldeformationsduetothetolerancebetweenthepinandtheholeingussets,(2)frictionthatrestrainstosomeextentin-planerotations of the connection and (3) initial member and connection imperfections. In-plane buck-lingofthebraceassemblyisfavouredbythefollowing:(1)designin-planeconnectioneccentricity,(2) reduction of out of plane rotation of the connection through smaller tolerances at the pin orlargerspacingbetweengussets,(3)lowerfrictionatthepin–gussetsinterface,(4)slenderbraces,and (5) brace cross section with different moments of inertia about the two principal axes(elliptical,RHS,wideflange).Amoredetailedcoverageofthisresearchsubjectcanbefollowedinsection2.5.

8

Obtained results were disseminated in 3 publications (journal and conference papers); for acompletelistseesection2.5.3.Researchwassupportedbythecontract no. BC79/04.07.2011 (2011-2012) "Testarea şi validarea contravântuirilor din ţeavă rotundă şi a îmbinărilor acestora pentru proiectul Smart Park din Bucureşti, Clădire de birouri clasa A (clădire Turn, alte clădiri şi parcaj) Şoseaua Străuleşti nr. 37, sector 1, Bucureşti",financedbySCPopp&AsociațiiSRL.

The sixth subject investigated "Seismic performance of multi-storey steel structures with friction dampers". The general aim of the research program was to establish the seismicperformanceofmultistoreysteelconcentricallybracedstructuresequippedwithstrainhardeningfriction dampers. For this purpose experimental and numerical analyses have been conducted.Based on the experimental data numerical models were calibrated and applied to evaluate theperformanceofconcentricallybracedframesequippedwithsuchdevicesinthebraces.Theaimofthe experimental program was to evaluate and characterise the damper, in a first step, and thebehaviourofthebrace-damperassembly,inthesecondstep(Figure6).

(a)

(b)

Figure6.Experimentaltestsetupfordampertests(a)andbracewithdampertests(b).

TheparticulardampertypologyinvestigatedwithinthisresearchprogramshowedtobeefficientinreducingtheseismicresponseofabuildingforearthquakescharacterizedbyshortcornerperiodTC=0.5s(stiffsoil)bypreventingtheformationofplastichingesatSLSandreducingthepermanentdisplacementof thestructure.Forearthquakes characterizedby longcornerperiodTC=1.6s (softsoil) this type of damper is not effective in improving the behaviour of the structure. Under thistype of seismic motions the structures with dampers form plastic hinges in non-dissipativeelementswithvaluesthatexceedtheacceptancecriteriaforthecorrespondingperformancelevels.Amoredetailedcoverageofthisresearchsubjectcanbefollowedinsection0.

The research was funded through the national grant no. 31042/2007 (2007-2011) type PN II Partnerships "Sisteme structurale şi soluţii tehnologice inovative pentru protecţia clădirilor la acţiuni extreme în contextul cerinţelor pentru dezvoltare durabilă – PROACTEX (Structural systems and innovative technologies for protection of buildings under extreme actions taking into account sustainable design criteria)",financedbytheministryofeducationandresearch.

Obtained results were disseminated in 10 publications (journal and conference papers); for acomplete list see section 2.6.3. A PhD thesis entitled "Seismic performance of multistorey steelconcentricallybracedframesequippedwithfrictiondampers"waselaboratedanddefendedaspartoftheresearchbyNorinFILIP-VĂCĂRESCUunderthesupervisionofProf.dr.ing.DanDubina.TheauthorofthishabilitationthesiswasalsoactivelyinvolvedinguidanceofthePhDcandidate.

The last research direction covered herein deals with "Prequalification of bolted beam to column joints with haunches"andrepresentsanongoingresearch.Modernseismicdesigncodesrequire that the seismic performance of beam-to-column connections in steel moment resisting

9

framestobedemonstratedthroughexperimentalinvestigations,whichoftenprovetobeexpensiveandtime-consuming.Asolutionto thisproblem is thepre-qualificationof typicalconnections forthedesignpractice.TheresearchprojectEQUALJOINTSiscurrentlyunderwayandaimsatseismicpre-qualification of several beam-to-column connection typologies common in the Europeanpractice. As a result, the current investigation outlines the experimental program on boltedextendedend-plateconnectionswithhaunchesthatwillbecarriedoutatthePolitehnicaUniversityof Timisoara in the framework of the mentioned project. The design objective was to get full-strengthandrigidjoints.AnumericalmodelwascalibratedbasedonthepastexperimentalresultscarriedoutonT-stubelementsandwasusedinordertovalidatethedesignprocedure(Figure7).

(a)

(b)

(c)

Figure7.CyclicresponseforEH2-TS-35:(a)moment-rotationcurve(monotonicandcyclicloadingconditions),(b)vonMissesstressdistribution,(c)equivalentplasticstrain.

Thepre-testnumericalsimulationscarriedoutforthejointassembliesdesignedbasedonEN1993-1-8confirmedtheintendedfailuremode,i.e.formationoftheplastichingeinthebeamclosetothehaunch ending. However, other assumptions considered in design were not confirmed, inparticular:forcedistributionintheboltrows,andpositionofthecompressioncentre(middleofthecompressedhaunchflange).Incontrast,theactiveboltrowswerethosesituatedneartheflangeintension,andthecompressioncentrewas locatedatadistanceequal to60%of thehaunchdepthmeasuredfromthebottomflangeofthebeam.Asaresult,thedesignprocedurewasadjustedandthe beam-to-column joint configurations were re-designed and analysed. The parametric studyallowed investigating the influence of: member size, haunch geometry, web panel strength, andcyclicloading.Consequently,thestudyevidencedhigherstraininmembers(beams)oflargercross-section, for the same joint rotation. In addition, the capacity and the initial stiffness of theconnection increased for higher values of haunch angle and depth, but for economic andarchitectural reasons, haunches with minimum depth and a smooth slope are recommended. Amoredetailedcoverageofthisresearchsubjectcanbefollowedinsection0.

This research is supported by the European grant no. RFSR-CT-2013 – 00021 (2013-2016). "European pre-QUALified steel JOINTS (EQUALJOINTS)", financed by the European Commission -Research Fund for Coal and Steel. The project is implemented by the following partnership:UniversityofNaplesFedericoII(coordinator),ArcelormittalBelval&DifferdangeSA,UniversitedeLiege,UniversitateaPolitehnicaTimisoara, ImperialCollegeofScience,TechnologyandMedicine,UniversidadedeCoimbra,EuropeanConventionforConstructionalSteelworkVereniging,Cordioli&C.SPA.

Obtained results were disseminated in 3 publications (journal and conference papers); for acomplete list see section 2.7.3. A PhD thesis entitled "Experimental study of haunched beam-to-column connections of bolted extended end plate under cyclic loading (Studiul experimental alîmbinărilor riglă-stâlp, cu placă de capăt extinsă, vută și șuruburi, solicitate în regim ciclic)" isunderdevelopmentbyCosminMARISunderthesupervisionofProf.dr.ing.DanDubina.TheauthorofthishabilitationthesisisactivelyinvolvedinguidanceofthePhDcandidate.

-1250-1000

-750-500-250

0250500750

10001250

-0.08-0.06 -0.04 -0.02 0 0.02 0.04 0.06 0.08

Mo

men

t [k

Nm

]

Rotation [rad]

Cyclic FEM

Monotonic Sag. FEM

Monotonic Hog. FEM

10

1.2 LISTOFRELEVANTPUBLICATIONS

Thefollowing10paperswereselectedasthemostrepresentativefortheresearchactivitiesoftheauthordescribedinthishabilitationthesis.

[1] Dubina,D.,Stratan,A.,Dinu,F.(2008)."Dualhigh-strengthsteeleccentricallybracedframeswithremovablelinks".EarthquakeEngineering&StructuralDynamics,Vol.37,issue15,pp.1703-1720,(OnlineISSN:1096-9845,PrintISSN:0098-8847).

[2] Stratan, A., Ioan, A., Dubina, D., Poljanšek, M., Molina, J., Pegon, P., Taucer, F. (2014). "Dualeccentrically braced frames with removable links: Experimental validation of technicalsolution through large-scale pseudo-dynamic testing". Proceedings of the Fifth NationalConference on Earthquake Engineering and First National Conference on EarthquakeEngineeringandSeismology–5CNIS&1CNISS,Bucharest,Romania,June19-20,2014.RaduVăcăreanu, Constantin Ionescu (Eds.), Bucharest: Conspress, 2014, ISBN (print) 978-973-100-342-9,ISBN(CD)978-973-100-341-2,pp.323-330,PaperNo.31(onCD-ROM)

[3] Dubina,D.,Stratan,A.,Nagy,Zs.(2009)."Full–scaletestsoncold-formedsteelpitched-roofportalframeswithboltedjoints".AdvancedSteelConstructionVol.5,No.2,pp.175-194.ISSN1816-112X.

[4] Vulcu,C.,Stratan,A.,Dubina,D.(2012)."Numericalsimulationofthecyclicloadingforweldedbeam-to-CFTcolumn joints ofdual-steel frames",PollackPeriodica,vol.7,no.2,pp.35–46.DOI:10.1556/Pollack.7.2012.2.3.PaperISSN:1788-1994,OnlineISSN:1788-3911.

[5] VulcuC.,StratanA.,CiutinaA.,DubinaD.(2011)."Beam-to-columnjointsforseismicresistantdual-steelstructures",PollackPeriodica6(2),pp.49-60.PaperISSN:1788-1994,OnlineISSN:1788-3911.

[6] Dubina, D., Bordea, S., Stratan, A. (2009). "Performance Based Evaluation of a RC FramestrengthenedwithBRBSteelBraces".Proc.oftheInternationalConferenceonProtectionofHistorical Buildings PROHITECH 09, Rome, Italy, 21-24 June 2009. Ed. F.M. Mazzolani. CRCPress,ISBN978-0-415-55803-7,p.1741-1746.

[7] Dubina,D.,Dogariu,A.,Stratan,A.,Stoian,V.,Nagy-Gyorgy,T.,Dan,D.andDaescu,C.(2007)."Masonrywallsstrengtheningwithinnovativemetalbasedtechniques".Proc.ofthe3rdIntl.Conf.onSteelandCompositeStructures(ICSCS07),Manchester,UK,30July-1August2007,Eds.WangandChoi.Taylor&Francis,ISBN:978-0-415-45141-3,pp.1071-1077.

[8] Stratan, D. Dubina, R. Gabor, C. Vulcu, I. Marginean (2012). "Experimental validation of abrace with true pin connections", Proceedings of the 7th International Workshop onConnections in Steel Structures, May 30 - June 2, 2012, Timisoara, Romania, ECCS, DanDubinaandDanielGrecea(Eds),ISBN978-92-9147-114-0,pp.535-546.

[9] Filip-Vacarescu, N., Stratan, A., and Dubina, D. (2014). "Experimental validation of a strainhardening friction damper". Proceedings of the Romanian Academy Series a - MathematicsPhysicsTechnicalSciencesInformationScience,15(1),60-67.

[10] Maris,C.,Vulcu,C.,Stratan,A.,andDubina,D.(2014).“NumericalSimulationsofBoltedBeamtoColumnConnectionswithHaunchesinSteelMomentFrames.”ProceedingsoftheSecondInternational Conference for PhD Students in Civil Engineering and Architecture, UTPRESS,Cluj-Napoca,Romania,109–116.

11

1.3 PROFESSIONALACHIEVEMENTS

Professional development of Aurel Stratan followed a wide pallet of activities, includingparticipation to training courses, structural design, industry-oriented research, involvement inprofessional organisations and technical committees, code drafting, development of the researchinfrastructure, organisation of scientific events, short-term scientific missions, involvement inadministrativedutiesandpeer-reviewofscientificpublications.

Recognizing the fact that an academic andscientificcareer is in needofcontinuous learning andimprovement of personal skills, Aurel Stratan followed two training courses related toexperimentaltechniques,oneofhismajorfieldsofinterest: "NewApproachestoAnalysisandTestingofMechanicalandStructuralSystems".International

CentreforMechanicalSciences,Udine,Italy,2007. "PreparatoryCourseonPseudodynamicExperimentalTesting",EuropeanLaboratoryfor

StructuralAssessment(ELSA),JRCIspra,Italia,2010.

He also fulfilled a short-term scientific mission as invited researcher at VTT – the TechnicalResearchCentreofFinlandinEspoo(September1-8,2006).Newresearchconnectionsweresought,including through the participation to the COST C26 action "Urban Habitat Constructions underCatastrophicEvents"(2006-2010).

Taking into account the highly applicative character of the domain of civil engineering, AurelStratantriedtokeepclosecontactwithstructuraldesignandindustry-orientedresearch.Thus,hehadalimitedinvolvementindesignactivity,themostimportantofwhichistheparticipationinthestructuraldesignoftheBucharestTowerCenter(Figure8).Thebuildinghasthreebasementlevels,26floorsandatotalheightof106.3m.Asof2015,itisthethirdtalleststructureinBucharest,aftertheFloreascaCityCenterandBasarabTower(ListoftallestbuildingsinBucharest,2015).Designofthe Bucharest Tower Center received two awards: "ECCS European Award for Steel Structures2007" and 1st prize of the Romanian Association of Structural Engineering (AICPS) in 2007(structuraldesignteam:D.Dubina,F.Dinu,A.Stratan,A.Ciutina).

(a)

(b)

Figure8.BucharestTowerCenterinBucharest:erectionphase(a)andcompletedbuilding(b).

At the same time, close cooperation with industry was sought in the area of research andconsultancy. Participation in the development of "Requirements for multi-storey buildings inseismic areas" for the Rautaruukki Corporation from Finland within the RUUKKI/2009 contract(coordinator D. Dubina) can be mentioned. Moreover, Aurel Stratan was the coordinator in the

12

contractBC79/04.07.2011(2011-2012)thatinvestigatedanoptimalsolutionforatrue-pinbraceconnectionusedinthedesignofahigh-risebuildinginBucharestbyPopp&Asociatii,describedindetailinsection2.5.

Aurel Stratan is member in several national professional organisations: APCMR - AsociaţiaProducătorilor de Construcţii Metalice din România (Romanian Association of SteelworkProducers); AGIR-SBIS - Asociaţia Generală a Inginerilor din România (General Association ofEngineersfromRomania),SocietateaBănăţeanădeInginerieSeismicăSBIS(RomanianAssociationof Engineers, Banat Seismic Society); AICPS - Asociaţia Inginerilor Constructori Proiectanţi deStructuri (RomanianAssociationofStructuralEngineers).He isalsoanactivemember inseveralnationalandinternationaltechnicalcommittees: TechnicalsecretaryofTechnicalCommitteeTC13"SeismicDesign"oftheEuropeanConvention

forConstructionalSteelwork(ECCS) MemberofCEN/TC250/SC8"Eurocode8:Earthquakeresistancedesignofstructures",

EuropeanCommitteeforStandardization(CEN) MemberofCEN/TC340/WG5"RevisionofEN15129–Anti-seismicdevices",European

CommitteeforStandardization(CEN) MemberofASROCT343"Bazeleproiectăriişieurocoduripentrustructure(Basisofdesignand

structuraleurocodes)",RomanianStandardsAssociation-AsociaţiadestandardizaredinRomânia(ASRO).

MemberofCTS4"Acțiuniasupraconstrucțiilor(Actionsonstructures)",MinisterulDezvoltăriiRegionalesiAdministrațieiPublice(MDRAP)–MinistryofRegionalDevelopmentandPublicAdministartion.

The main work undertaken within these structures concerns improvement of national andEuropean design guidelines in the field of earthquake-resistant design. As an example, theoutcomesoftheactivitiesperformedwithintheTechnicalCommitteeTC13"SeismicDesign"oftheEuropeanConventionforConstructionalSteelwork(ECCS)wererecentlypublishedinthefollowingbook: Jean-MarieAribert,DarkoBeg,JoséMiguelCastro,HerveDegee,FloreaDinu,DanDubina,

AhmedElghazouli,AmrElnashai,VighL.Gergely,VictorGioncu,MohammedHjiaj,BennoHoffmeister,RaffaeleLandolfo,Pierre-OlivierMartin,FedericoM.Mazzolani,PaoloNegro,VincenzoPiluso,AndrePlumier,CarlosRebelo,GerhardSedlaceck,AurelStratan(2013)."AssessmentofEC8provisionsforseismicdesignofsteelstructures",RaffaeleLandolfo(Ed.),EuropeanConventionforConstructionalSteelwork.

AurelStratanhadanimportantactivityincode-drafting.Itconcernedelaborationofnationalcodes,especiallytheonesrelatedtoseismicdesignofsteelstructures,andalsodevelopmentofRomanianversionsandNationalAnnexesofstructuralEurocodes.ThefollowingisalistofnationalcodesinwhichAurelStratanwasamemberinthedraftingteam: 2013revisionoftheRomanianseismicdesigncodeP100-1/2013."Coddeproiectareseismică-

ParteaI-Prevederideproiectarepentruclădiri".VolumulIșiII.MonitoruloficialalRomâniei,nr.558bis/2013.

2006revisionoftheRomanianseismicdesigncodeP100-1/2006."Coddeproiectareseismică-ParteaI-Prevederideproiectarepentruclădiri".BuletinulConstrucţiilor,Vol.12-13,2006.

GP078-03"Ghidprivindproiectareahalelorusoarecustructurametalica(Designguideforlight-gaugeindustrialsteelhalls)",Bul.ConstructiilorVol16,2004,p59-273.

GP082-03."Ghiddeproiectareaimbinarilorductilelastructurimetaliceinzoneseismice(Designguideforductileconnectionsofsteelstructuresinseismicareas)".Bul.Constr.Vol.16,2004,p.3-58.

Below is a list of standards for which the author elaborated the Romanian versions and/or theNationalAnnexes:

13

SREN1998-1:2004/NA:2008"Eurocode8:Designofstructuresforearthquakeresistance–Part1:Generalrules,seismicactionsandrulesforbuilding.NationalAnnex".RomanianStandardsAssociation-AsociaţiadestandardizaredinRomânia(ASRO).

EN15129:2010"Anti-seismicdevices".RomanianStandardsAssociation-AsociaţiadestandardizaredinRomânia(ASRO).

SREN1993-1-12:2007"Eurocode3-Designofsteelstructures-Part1-12:AdditionalrulesfortheextensionofEN1993uptosteelgradesS700".RomanianStandardsAssociation-AsociaţiadestandardizaredinRomânia(ASRO).

SREN1993-1-7:2007"Eurocode3-Designofsteelstructures-Part1-7:Platedstructuressubjecttooutofplaneloading".RomanianStandardsAssociation-AsociaţiadestandardizaredinRomânia(ASRO).

Most of the research performed at the CEMSIG research centre has a strong experimental andnumericalcomponent,whichisrequiresasolidinfrastructure.AurelStratanisactivelyinvolvedinmaintaining and upgrading of the existing testing facilities. He participated in the followingcontractsthatspecificallyaddressedthedevelopmentoftheresearchinfrastructureoftheresearchcentre: Contractnr.662/2014(2014-2015)POSCCEID1827/SMIS48741"Platformaintegratăde

cercetare-dezvoltarepentrucomportareaconstrucţiilorlaacţiuniextreme(ACTEX)"–memberinthemanagementteam.

90CP/I/2007(2007-2010)GranttipPNIIModulI,Capacităţi:"Dezvoltarelaboratorpentruîncercăripestructurilascarămare-INSTRUCT"–scientificresponsible.

42/2006(2006-2008).GranttipPlatforme/Laboratoaredeformaresicercetareinterdisciplinara:"CentrudeStudiiAvansatesiCercetareînIngineriaMaterialelorsiStructurilor–CESCIMS"-collaborator.

32940/2004(2004-2005).GrantE,Tema3,codCNCSIS31"Standexperimentalpentruîncercăriciclice"–collaborator.

(a)

(b)

(c)

Figure9.Instronuniversaltestingmachinefordynamictestsof1000kNcapacity,withhydraulicgrips(a);Reactionwall5.0x6.2mandstrongconcreteslab5.0x9.5mforstaticandpseudo-dynamic

tests(b);HydraulicactuatorsMTS650/1015kNandhydraulicpowerunit.

Among the infrastructure / equipment that has been developed or upgraded with the directcontributionofAurelStratan, the followingonescouldbementioned:reactionstructure(5.0x6.2m) and strong concrete slab (5.0x9.5 m) for quasi-static and pseudo-dynamic tests; hydraulicactuatorsMTS650/1015kN,withhydraulicandcontrolsystems;Instronuniversaltestingmachinefordynamictestsof1000kNcapacity,withhydraulicgrips;UTSuniversaltestingmachineof250kNcapacity,withhydraulicgrips;Zwickuniversaltestingmachineof10kNcapacity;Charpyimpacttestingmachine(Figure9).Amajordevelopmentoftheexistingresearchinfrastructureiscurrently

14

underwaywithinthePOSCCEID1827/SMIS48741ACTEXproject.Aspartofthemanagementteamandtechnicalandscientificresponsibleoftheproject,AurelStrataniscoordinatingthepurchasingof a uniaxial shaking table, two dynamic and two static actuators, optical metallographicmicroscope,high-resolutionvideodataacquisitionsystem,dynamicdataacquisitionsystem,high-performance computing (HPC) system, workstations and advanced structural analysis software.AurelStratanwas/ismemberofthescientificcommitteeofthefollowingconferences: A12-aConferinţăNaţionalădeConstrucţiiMetalice,Timişoara,26-27noiembrie2010. A13-aConferinţăNaţionalădeConstrucţiiMetalice,București,21-22noiembrie2013. A14-aConferinţăNaţionalădeConstrucţiiMetalice,Cluj-Napoca,22-24noiembrie2015.

Hewasalso involvedas member in theorganisingcommitteeof the"Internationalconference inmetal structures", Poiana Brasov, Romania, September 20-22, 2006 and the XIII edition of the“ZileloracademiceTimişene(TimisaoraacademicDays)”,24may2013.AurelStratanwasamongthe editors of the international workshop on "Application of High Strength Steels in SeismicResistant Structures" was organised by the University of Naples "Federico II" and PolitehnicaUniversityofTimisoara, incooperationwithEuropeanConventionforConstructionalSteelwork-Committee TC13 "Seismic Design" on June 28-29 2013, in Naples, Italy. Workshop proceedingswereeditedbyDanDubina,RaffaeleLandolfo,AurelStratan,andCristianVulcuattheOrizonturiUniversitarepublishinghouse(ISBN:978-973-638-552-0).

AurelStratanchairedsessionswithinthefollowinginternationalconferences: Session23“Earthquake&Vibration”,SecondInternationalWorkshoponPerformance,

Protection&StrengtheningofStructuresunderExtremeLoading,ShonanVillageCenter,Hayama,Japan,August19-21,2009

Sessionon"Seismic-resistantstructures5"withinthe7thEuropeanConferenceonSteelandCompositeStructuresEUROSTEEL2014,10-12September2014,Napoli,Italy

Aurel Stratan seeks a more active role at the administrative level by acting as member of theCouncil of the Faculty of Civil Engineering (2013-present), member in the council and scientificsecretary of the Department of Steel structures and Structural Mechanics (2004-present), andscientific coordinator of the Research Center for Mechanics of Materials and Structural Safety –CEMSIG(2011-present).

Aurel Stratan is a reviewer in the following ISI journals: "Earthquake Engineering and StructuralDynamics", John Wiley & Sons, ISSN: 1096-9845;"Steel and Composite Structures", Techno-Press,ISSN:1598-6233;"JournalofEarthquakeEngineering",Taylor&Francis,ISSN1559-808X.

1.4 ACADEMICACHIEVEMENTS

CurrentlyAurelStratanteachesthecoursesof"Structuraldynamicsandearthquakeengineering"and"Basisofstructuraldesign",aswellas theprojectof "Steel structures"atbachelor level,and"Performancebaseddesign"and"Redesignofexistingstructures"atthemasterlevels.Thefirstofthese courses is supported by a book published by the author: Stratan, A. (2007). "Dinamicastructurilor şi inginerie seismică (Dynamics of Structures and Earthquake Engineering)", Ed.OrizonturiUniversitare,Timişoara,ISBN978-973-638-388-0,223p.Itwasreceivedthe1stawardatthe"CarteaTehnică2008"contestbythe"AsociaţiaGeneralăa InginerilordinRomânia,FilialaTimiş(RomanianAssociationofEngineers)".

Electronic supporting material is provided for other courses, some of which are posted on thefacultywebsiteandconstantlyupdated: Stratan,A."DinamicaStructurilorsiInginerieSeismica".

http://www.ct.upt.ro/users/AurelStratan/index.htm,Suportdecurspentruanul3CCIAși3CFDP,155p.

15

Stratan,A."StructuralDynamicsandEarthquakeEngineering".http://www.ct.upt.ro/users/AurelStratan/index.htm,Suportdecurspentruanul3PI,153p.

Stratan,A."BasisofStructuralDesign".http://www.ct.upt.ro/users/AurelStratan/index.htm,Suportdecurspentruanul2PI,229p.

Anotherworkpublishedbytheauthorandthatservesassupportformasterstudentsis: DanDubina,FloreaDinu,AurelStratan,NorinFilip-Văcărescu(2014)."Calcululstructural

globalalstructurilormetalice.Recomandări,comentariișiexempledeaplicareînconformitatecuSREN1993-1-1siSREN1998-1".BuletinulConstrucțiilor,nr.7/2014.ISSN1222-1295,201p.

Aurel Stratan coordinates several diploma works and master dissertations each year, the latterbeingcorrelatedwiththeon-goingresearchprojects.

Theauthorwas recently involved in teachingactivities of theEuropeanErasmus Mundus MasterCourse "Sustainable Constructions under Natural Hazards and Catastrophic Events". Hecoordinated themodule2C09"Design forseismicandclimate changes" thatweredelivered inatthePolitehnicaUniversityofTimisoarainthespringof2014.Lecturenotesareavailableat: Stratan,A.,Grecea,D.,D'Aniello,M.,Dubina,D.(2014)."Designforseismicandclimatechanges:

Lecturenotes".http://www.ct.upt.ro/suscos/2C09.htm,EuropeanErasmusMundusMasterCourse-520121-1-2011-1-CZ-ERAMUNDUS-EMMC,2013-2015edition,859p.

PartofthelectureswasdeliveredbyAurelStratanattheUniversityofNaplesFedericoIIinMarch2015,aspartofthe2014-2016editionoftheEuropeanErasmusMundusMasterCourse.

Theauthorsupportedextra-curricularactivitiesofthestudentsfromthefacultyofcivilengineeringbydeliveringlectureswithineventsorganisedbytheInternationalAssociationofCivilEngineeringStudents(IACES): Lecture"Earthquakeengineering:anintroduction"withinthesummercourseIACES

"TIMELESS",Timisoara,Romania,18-25July2010. Lecture"EarthquakeAction"withinthesummercourseIACES"ExchangeBuildingChallenges",

Timisoara,Romania,11-17July2005.

16

2 RESEARCHOUTCOMES

2.1 RE-CENTRINGECCENTRICALLYBRACEDFRAMES

2.1.1 THECONCEPTOFRE-CENTRINGSTRUCTURE

Most structures designed to modern codes would experience inelastic deformations even undermoderateseismicaction,andpermanent(residual)displacementsafteramajorearthquake.Insuchcases,therepairisdifficultandtherearedifferentstrategiestoreducestructuraldamage.Themostradical solutions to avoid such damage are base isolation and various implementations of activeandsemi-activestructuralcontrol.Otherstrategiesrelyonsupplementaldampingconferredtothestructure through various devices based on viscous, friction, or yielding dampers. While thesesolutions are efficient, they all require specialised knowledge at the design and erection stage,carefulmaintenance,andpresentahighinitialcost.

However,thestructuraldamagecanbecontrolledbyamoreconventionaldesignapproach.VargasandBruneau(2006)investigatedadesignapproachaimingatconcentratingdamageonremovableand easy to repair "structural fuses", with the main structure designed to remain elastic or withminor inelastic deformations. With emphasis on buckling restrained braced frames, Kiggins andUang (2006) showed that even if such structures exhibit a favourable energy-dissipatingmechanism,thelowpost-yieldstiffnessofthebracesleavethesystemvulnerabletounfavourableresponsesuchaslargepermanentdrift.Thepotentialbenefitofusingbuckling-restrainedbracesinadualsystemtominimizepermanentdeformationswasshowntoreducesignificantlytheresidualstorydrifts.

Self-centringsystems,arelativelynewconcept thataddresses thedrawbacksof theconventionalyieldingsystems,havereceivedmuchattentionrecently.Amongmanypractical implementations,the following are most relevant to our study: self-centring moment-resisting frames with post-tensionedbeamtocolumnjoints(Herningetal.2011)andcolumnbases(ChiandLiu2012),self-centringconcentricallybracedframes(RokeandJeffers2011).

An alternative solution is to provide re-centring capability (as opposed to self-centring), byremovabledissipativemembersanddual(rigid-flexible)structuralconfiguration.

Adualeccentricallybraced frame(EBF)withremovabledissipativemembers (links) is shown inFigure10(StratanandDubina2004).Thelinktothebeamconnectionisdonebyaflushend-plateand high-strength friction grip bolts. The main advantage over other dissipative devices is thatremovablelinkscanbedesignedusingmethodsalreadyavailabletostructuralengineersandcanbefabricatedandinstalledusingstandardprocedures.

Figure10.Boltedlinkconcept(StratanandDubina2004).

e

17

The re-centring of the system is attained by designing the structure as a combination ofeccentricallybraced frames(EBFs)andmoment-resisting frames (MRFs).Theelasticresponseofthe flexible subsystem (MRF) provides the restoring forces, once the links damaged during anearthquakeareremoved.Forthisprincipletobeefficienttheflexiblesubsystemshouldremainintheelasticrange.Standardcapacitydesignprinciplescanbeusedtomeetthisobjective.However,nonlinear structural analysis is advised. To ensure the elastic response of the flexible subsystemsome members could be made in high-strength steel (Dubina et al. 2008). Additionally, thedamagedlinksshouldbereasonablyeasytoremove.Ifthelinkdeformationsaresmall,linkscouldbesimplyunboltedandthenremoved. Incaseof largeresidual linkdeformations,unboltingmayprovedifficult.Insuchacasetheresidualdeformationsshouldfirstbereleasedbyflamecuttingthelink.

2.1.2 DESIGNCRITERIAFORDUALFRAMESWITHRE-CENTRINGCAPABILITY

Dual steel frames are obtained by combination of moment resisting frames (MRFs) andconcentrically braced frames (CBFs) or eccentrically braced frames (EBFs). There are threerequirementsfordesignofdualframesinEN1998-1(2004):(1)asinglebehaviourfactorisusedfor determining the design seismic action, (2) seismic forces should be distributed between thedifferentframesaccordingtotheirelasticstiffnessand(3)eachstructuralsubsystem(MRFs,CBFs,EBFs)shouldbedesignedaccordingtotheirspecificprovisions.Anadditionalrequirementcanbefoundinothercodes(e.g.NEHRP,2003,P100-1,2006):theMRFsshouldbeableofresistingatleast25% of the total seismic force. This last requirement is based on judgement (NEHRP, 2003).According to code rules, the main benefit of dual configuration is the possibility to adopt largervalues of behaviour factor (EN 1998-1, 2004; NEHRP 2003) in comparison with simple bracedframesorstructuralwalls,duetolargerredundancyofthedualsystem.

Severalresearchers(Akiyama,1999;IyamaandKuwamura,1999;Astaneh-Asl,2001)investigatedseismicperformanceofdualsystems,consistingofrigidandflexiblesubsystems.Akiyama(1999)showed that "flexible-stiff" structural types belong to the most efficient earthquake resistantstructures. It is believedthat the flexiblesubsystem willpreventexcessivedevelopmentofdrifts,while the rigid subsystem will dissipate seismic energy by plastic deformations. Iyama andKuwamura(1999)studiedtheprobabilisticaspectofdualsystems.ThedualsystemwasobtainedbycombiningagenericCBFandMRF,withdifferentnaturalperiodsofvibration.Theobtaineddualstructurewastermed"fail-safe",becauseitprovidesanalternativeloadpathtoearthquakeloading(MRF)incaseoffailureoftheprimarysystem(CBF).Analysesonsingledegreeoffreedomsystemsshowedthatadualsystemhasahighersafetyfactorthanahomogeneoussystem,consideringtheunknown characteristics of future earthquakes. It was also showed that the benefits of the dualsystemisenhancedwhenthedifferencebetweenthenaturalperiodsofthesubsystemsislarge,andwhen the ductility of the subsystems is small. Astaneh-Asl (2001) studied steel shear wallstructures, emphasizing the importance of multiple lines of resistance, naturally achieved bycombiningasteelshearwallandaMRF.

Adifferentuseofdualframesistheoneinwhichremovabledissipativemembersareemployedintherigidsubsystem.Providedtheflexiblesubsystemdoesnotyield,andtheinelasticdeformationsare constrained to the removable members alone, the flexible subsystem would provide therestoring force necessary to re-centre the structure once the damaged dissipative members areremoved. It would be possible then to install new dissipative members, regaining the originalstrengthofthestructure.

Vargas and Bruneau (2006) investigated a design approach aiming to concentrate damage onremovable and easy to repair structural elements ("structural fuses"), with the main structuredesigned to remain elastic or with minor inelastic deformations. A systematic procedure wasproposedtodesignbuildingswithmetallicstructuralfuses.KigginsandUang(2006)showedthat

18

evenifbucklingrestrainedbracedframesexhibitafavourableenergy-dissipatingmechanism,thelow post-yield stiffness of the braces leave the system vulnerable to unfavourable behaviouralcharacteristics such as large permanent drift. The potential benefit of using buckling-restrainedbraces in a dual system to minimize permanent deformations was investigated and shown toreducesignificantlyresidualstorydrifts.

In order to analyse the factors controlling the two requirements for structures with removabledissipative elements (isolation of damage and re-centring capability), it is useful to consider asimple dual system consisting of two inelastic springs connected in parallel (see Figure 11a).Providedthattheflexiblesubsystemisnotveryweak,plasticdeformationsappearfirstintherigidsubsystem. In order to maximize system performance, plastic deformations in the flexible

subsystemshouldbeavoided.Atthe limit,whentheyieldforceFyfandyielddisplacementdyfareattainedintheflexiblesubsystem,therigidsubsystemexperiencestheyieldforceFyrandthetotal

displacementdyr+dplr(seeFigure11b).Equatingthetwodisplacements:

yf yr plrd d d (1)

andconsideringtherelationshipbetweenforceanddeformation( F k d ),itcanbeshownthat:

yr plr yf yf r

D

yr yr yr f

F K

F K

d d d

d d

(2)

(a)

(b)

Figure11.Simplifiedmodelofageneralizeddualsystem.

The notation D represents the "useful" ductility of the rigid subsystem, for which the flexiblesubsystemstillrespondsintheelasticrange.Itcanbeobservedthattherearetwofactorsthatneedtobeconsideredinordertoobtainaductiledualsystemwithplasticdeformationsisolatedintherigidsubsystemalone.Thefirstoneistheratiobetweentheyieldstrengthoftheflexibleandrigidsubsystems(Fyf/Fyr),whilethesecondoneistheratiobetweenthestiffnessoftherigidsubsystemand the one of the flexible subsystem (Kyf/Kyr). The larger are these two ratios, the larger is the

"useful"ductilityDofthedualsystem.

Re-centring capability of the dual system is not easily attainable. Though the dual configuration

results in smaller permanent driftsdpD in comparison with permanent deformations of the rigid

system alone dpr (see Figure 12a), they are not eliminated completely after unloading. Afterunloadingof thedualstructure, there is aresidual force in therigidsubsystem (Fpr)equal to theopposing force in the flexiblesubsystem(Fpf), seeFigure12b.However,permanentdeformationscanbeeliminatediftherigidsubsystemisrealisedtoberemovable.Onceremovableelementsaredismantled,stiffnessandstrengthofthesystemisprovidedbytheflexiblesubsystemalone.Ifthe

Kf, Fyf

Kr, Fyr

F

d

d

F

dyr dyf

Fyf

Fyr

Fyf

+Fyr

KfKr

K

Kf

Fyr+Kfxdyrd

plr

flexible subsystem

rigid subsystem

dual system

19

flexible subsystem is still in the elastic range, it will return the system to the initial position,implyingzeropermanentdeformations.

When designing a dual system, the size of structural members will be governed by thecharacteristicsoftherigidsubsystem,whichprovidesmostofthestrengthandstiffnesstolateralforces. If removable dissipative members are used, then the characteristics of the flexiblesubsystem should be chosen such that it would remain in the elastic range up to displacementscorresponding to attainment of ultimate deformations in the rigid subsystem. This requirementcouldbeexpressedas:

yf ur yr plrd d d d (3)

wheredurultimatedisplacementoftherigidsubsystem.

Thus,theabilityoftheflexiblesubsystemofadualstructurewithremovabledissipativemembersintherigidsubsystemtoprovidethere-centringcapability(seeequations(2)and(3))dependson:

thedemandofinelasticdeformationsintherigidsubsystemdur=dyr+dplr,and

theyielddisplacementdyf=Fyf/Kfintheflexiblesubsystem.

Theyielddisplacementintheflexiblesubsystemiscontrolledbybothstrengthandstiffness(dyf=Fyf/Kf). Therefore, the code requirement for the flexible subsystem in dual frames in terms ofstrengthonlymaynotassurethedesiredbehaviourfordualstructureswithremovabledissipativemembers.

(a)

(b)

Figure12.Permanentdeformationsinaconventionaldualsystem(a)andinadualsystemwithremovabledissipativemembers(b).

The basic design requirement for dual frames with removable dissipativemembers is to preventyieldinginMRFsuptotheattainmentofultimatedeformationcapacityintheEBFwithremovablelinks, checking the equation (3). Simple procedures suitable for design are required in order toestimatetheultimatedisplacementinEBFs,aswellasdisplacementatyieldintheMRFs.Therefore,strength (Fy) and stiffness (K) need to be known for basic components of the system: aneccentricallybracedframe(seeFigure13a)andamoment-resistingframe(Figure14).

Fromthefree-bodydiagramofthebasicEBF(seeFigure13a),andconsideringthefactthatplasticdeformations should be constrained to the link, and sketching the free-body diagram of half thebasicEBF(seeFigure13b),theyieldstrengthoftheEBF(FyEBF)couldbedeterminedas:

,EBFy pl link

LF V

H (4)

d

F

dprdpD

Fyf

Fyr

Fyf

+Fyr

d

F

Fyf

Fyr

Fyf

+Fyr

=F

pf

Fpr

20

whereVpl,linkistheshearstrengthofthelink.

The most important components contributing to the stiffness of an EBF are the link sheardeformationandbraceaxialdeformation.Itcanbeshownthatthestiffnesstolateralforcesofthesetocomponentscanbeexpressedas:

2

EBF slink

G ALK L e

H e

(5)

22 cosEBFbr

br

E AK

l

(6)

whereGistheshearmodulus,Asisthelinksheararea,EistheYoung'smodulus,Aisbracecross-sectionarea,lbristhebracelength,whiletheothernotationsareshowninFigure13.

(a)

(b)

Figure13.Basicone-storeyEBFcomponent.

ThetotalstiffnessoftheEBFcanbeobtainedbycombiningthecontributionduetoshearofthelinkwiththeoneduetoaxialdeformationofbraces:

EBF EBF

EBF link br

EBF EBFlink br

K KK

K K

(7)

Oncethestrength(eq.(4))andstiffness(eq.(7))oftheone-storeycomponentsofaEBFareknown,theultimatedisplacementoftheEBFcanbedeterminedasthesumoftheyielddisplacementandtheplasticdisplacement(seeeq.(3)).Thelatteriscontrolledbytheplasticdeformationcapacityof

the link (pl,u) and the geometry of the EBF. Thus, the ultimate displacement of the EBF can bedeterminedas:

,

EBFyEBF EBF EBF

u y pl pl uEBF

F eH

K L ed d d

(8)

ThebasicMRFcomponentcanbeidealisedasanassemblyofabeamandtwocolumnsshowninFigure14a,assuminginflexionpointsathalfthestoreyheight.Fromthefree-bodydiagramoftheMRF(seeFigure14a),andconsideringthe factthatplasticdeformationsshoulddevelopatbeamends,andsketchingthefree-bodydiagramsofthebeamandthecolumn(seeFigure14b),theyieldstrengthoftheMRFcouldbedeterminedas:

,2 pl bMRF

y

MF

H (9)

whereMpl,bisthebeamplasticmoment.

L

H

eF/2

R

d

F/2R

F/2

F/2

R

Vpl,link

F/2

F/2

21

(a)

(b)

Figure14.Basicone-storeyMRFcomponent.

Considering the flexural deformations only, the stiffness of the basic MRF component can bedeterminedas:

2

2

6 12

MRF

b c

KL H

HE I E I

(10)

whereIbisthemomentofinertiaofthebeamandIcisthemomentofinertiaofthecolumn.

Oncethestrength(seeeq.(9))andstiffness(seeeq.(10))oftheone-storeycomponentsofaMRFareknown,theyielddisplacementoftheMRFcanbedeterminedas:

MRFyMRF

y MRF

F

Kd (11)

Inordertoassurethere-centringcapabilitytothedualeccentricallybracedframes,itisnecessarythatyieldinginthemoment-resistingframesispreventeduptolargeenoughplasticdeformationsin the dissipative components (links) of the eccentrically braced frames. The design verificationthatneedstobecheckedisthus:

EBF MRFu yd d (12)

(a)

(b)

(c)

Figure15.Structuralconfiguration:a)3Dview;b)planview;c)transversalendwallframe

Amediumriseframestructurewasconsideredtochecknumericallytheanalyticalprocedure.Thebuildinghas3spansof6metersand5baysof6meters.Buildinghas3storeysof3.5meterseach.Ineachdirection,themainlateralloadresistingsystemismadebytwobracedframes,locatedonthe perimeter (Figure 15a). Additionally, there are 4 moment resisting frames on transversaldirectionsand10momentresistingframesonlongitudinaldirection(representedwiththicklines),toassuretherestoringforcesafteranearthquake(Figure15b).Allotherbeamshavepinnedends.Thecolumnsarepinnedatthebase.ThedesignwascarriedoutaccordingtoEN1993,EN1998and

L

H

F/2

RR

F/2

F/2 F/2R

Mpl,b

V

Mpl,b

VV

R

V

F/2

F/2

F/2

F/2

22

RomanianseismicdesigncodeP100-1/2006.A5.0kN/m2deadloadand3.0kN/m2liveloadwereconsidered. The building is located in Bucharest, characterised by the following seismicparameters:0.24gdesignpeakgroundacceleration,softsoilconditionswithTC=1.6s,abehaviourfactor q=6 and interstorey drift limitation of 0.008 of the storey height. Columns are made ofHEA240profiles.BeamsinthemomentresistingframesaremadeofIPE240.Bracesandbeamsinthe braced frames are made of HEA240 profiles while links are made of welded H section230x140x10x6mm(Figure15.c).

In order toconcentrate the plastic deformations in links and keep the MRFs in the elastic range,different steel grades were used for members (Dubina et al., 2008). Thus, the beams in momentresistingframesandthecolumnsaremadefromHSSsteel(gradeS460),whileallothermembers,including braces and links are made from mild carbon steel (grade S355) (Figure 15.c). Aremovableboltedlinkwasused.

The analysis was performed with SAP2000 computer program. Inelastic behaviour of beams,columns and braces were modelled with concentrated plastic hinges. The inelastic shear linkelement model was based on the Ricles and Popov (1994) proposal, but adapted to the bilinearenvelope curve available in SAP2000. The stiffness of bolted links was reduced to an equivalentstiffness of 0.25 of the theoretical shear stiffness of continuous links (Dubina et al., 2008). The

ultimate link deformations atULS was consideredpl,u=0.10 rad. Inorder to assess the structuralperformanceofthestructure,nonlinearstaticanalyseswereperformed.Analysesweredoneonaplanar end wall transversal frame, extracted from the 3D structure, with seismic masscorresponding to half of the total floor masses, in order to take into account that, in the 3Dstructure,theendwallbracedframescarryhorizontalseismicforcesfrominternalframesalso.Thelateral force pattern, used in the push-over analysis, has a "uniform" distribution and isproportional to mass, regardless of elevation (uniform response acceleration). One mater ofinterest incaseofdual frames is thecontributionof themomentresistingframestothestiffnessandstrengthoftheentirestructure.Therefore,inordertoevaluatetheparticipationofeachpart,thestructurewasanalyzedin3differentsituations,whichare(Figure16): EBF:outerframesarepinned,nocontributiontothelateralresistingsystem MRF:outerframesarerigid,bracesareremoved DUA(EBF+MRF):bothbracesandmomentframeareactive

(a)

(b)

(c)

Figure16.Contributiontothedualstructure:a)EBFconfiguration;b)MRFconfiguration;c)DUAconfiguration.

Table1.Comparisonofanalyticalandnumericalpredictionsofstoreydisplacements.

duEBF,mm dyMRF,mm dyMRF/duEBF

Analytical 20.6 82.8 4.0

Numerical 22.6 80.5 3.6

23

Figure17.StoreypushovercurvesforEBF,MRFandDUAconfigurations

The analytical procedure involves checking that the displacement corresponding to ultimatedeformationof theEBF is smaller thanthedisplacementcorresponding to firstyield in theMRF.StoreypushovercurvesfortheEBF,MRFandDUAstructuresareshowninFigure17.Firstplastichinge inbeams ismarkedonthecurves, togetherwith theattainmentofultimaterotationin thelinks. Table 1 presents a comparison of yield displacements in the MRFs and ultimatedisplacementsintheEBFforthefirststorey,where largestdemandsarepresent.Fairagreementcanbeobservedbetweenanalyticalandnumericalresults.

2.1.3 EXPERIMENTALTESTSONONE-STOREYECCENTRICALLYBRACEDFRAME

ThreeseriesofexperimentaltestsonremovablelinkassemblieswerecarriedoutatthePolitehnicaUniversityofTimisoarainordertodeterminecyclicperformanceofboltedlinksandtocheckthefeasibility of the removable link solution (Stratan and Dubina 2004, Dubina et al. 2008, Danku,2011).Thefirstseriesoftestswasrealisedonisolatedlinks(seeFigure18a),whiletheothertwoon almost full-scale model of a single bay and single storey eccentrically braced frame withremovablelink(seeFigure18b).

The first series of tests on links showed an important influence of the connection on the totalresponseoftheboltedlink,intermsofstiffness,strengthandoverallhystereticresponse.Shorterlinkswerefoundtobesuitablefortheboltedsolution,asplasticdeformationswereconstrainedtothelink,whiletheconnectionresponsewasalmostelastic,allowingforaneasyreplacementofthedamagedlink.

Thesecondseriesoftestsconsistedin8almostfull-scaleframespecimens,representingaone-bayone-storey eccentrically braced frame sub-assembly (see Figure 18). The main parameter of thetest was the thickness of the end-plates of link-beam connections. The links were designed andconstructedwiththreethicknessesoftheendplate:t=10mm,t=15mmandT=20mm.Twotestswereperformedforeachtypeoflink-beamconnection:onemonotonicandonecyclic.Cyclictestswere performed according to the ECCS procedure. Monotonic tests were performed at positivedisplacement(compressioninthebeam).Additionally,forspecimenswith15mmand20mmthicklinkend-plates,monotonictestsundernegativedisplacement(tensioninthebeam)wereperfomedas well. Table 2 summarizes synthetically the experimental program. For testing the eightspecimensasingleframewasused,changingaftereachtestthedamagedremovablelinks.

Toensureapredominantlyelasticresponseof the link-beamconnection, thecapacitydesignwasused.Thus,connectionactioneffects,shearforceVj,EdandflexuralmomentMj,Edweredeterminedaccordingtothefollowingrelationships:

, ,j Ed ov pl linkV V (13)

0

100

200

300

400

500

600

700

800

900

1000

0.00 0.02 0.04 0.06 0.08 0.10 0.12

1st storey displacement, [m]

1st s

tore

y sh

ear,

[kN

]

MRF

EBF

Dual

dy,MRF

du,EBF

24

,

2

j

j Ed

V eM

(14)

whereVp,linkistheshearresistanceofthelink,accordingtoEN1998-1(2004)andP100-1(2006)

ov is the ratio between the actual and the nominal values of the yield strength of dissipativeelements(consideredequalto1.25,accordingtoEN1998-1(2004)andP100-1(2006))

istheratiobetweenthemaximumandtheyieldshearforce(amountingto1.5,accordingtodatafrompreviousexperimentaltestsandliterature)

eisthelinklength.

Shearstrength(Vj,Rd)andflexuralstrength(Mj,Rd)ofthelinkconnectionweredeterminedaccordingtoEN1993-1-8.

(a)

(b)

Figure18.Experimentaltestonremovableboltedlink(a)andtestsonalmostfull-scaleframewithboltedlinks(b).

Table2.Experimentalprogram.

Specimen L10m1 L15m1 L20m1 L10c L15c L20c L15m2 L20m2End-plate

thicknesstep,mm10 15 20 10 15 20 15 20

Loadingtype monotonic(positive) cyclic monotonic(neg.)

Table3.Propertiesoflinkconnections.

tep,mm Bolts Vj,Ed,kN Vj,Rd,kN Mj,Ed,kNm Mj,Rd,kNmFailuremode

10M16gr8.8 268.3

137.840.2

33.4 Mode1-215 223.9 49.5 Mode220 268.8 52.4 Mode3

25

(a)

(b)

Figure19.Lateralforcevs.displacement(a)andfailuremode(b)ofthespecimensL10c,L15candL20c.

Table3presentsthedesignvaluesofshearforceandbendingmomentinthelinkconnection,thedesignresistanceofshearforceandbendingmoment,aswellasfailuremodeoftheendplate.Onlyoneconnectionhasthenecessarystrengthinshearandbending–theonewithathicknessof20mm.Thegoverningfailuremodeoftheendplateismode3,throughboltfracture.Theothertwoconnections (with end-plate thicknesses of 15 mm and 10 mm) were considered in order toinvestigate the possibility of relaxing the connection design criteria given by equations (13) and

(14),aswellasbythe factor .Theconnectionwithend-platethicknessof15mmisnotstrongenoughinshear,whiletheendplatefailsinmode2–combinedboltandend-plate.Theconnectionwithend-platethicknessof10mmisnotstrongenoughneither inshearnorinbending,andtheendplate fails inmode2,butclose tomode1(endplateonly).Frictiongripmechanismwas notaccounted for when determining the shear strength of the connections, though bolts have been

26

pretensioned.Itwasconsideredthatcyclicloadingandinelasticresponseofthelinkwouldrenderfrictiongripineffective.

Figure 19 presents the lateral force vs. displacement relationship (a) and failure mode (b) forspecimensL10c,L15candL20c,subjectedtocyclicloading.

Monotonictestsunderpositiveloadingshowedexcellentductility,thelink-beamconnectionhavingaquasi-elasticresponse.Thesespecimensfailedbycompleteyieldingofthelink.Specimenstestedmonotonicallyundernegativeloadingshowedlowercapacitiesandultimatedeformationsthantheonesubjectedtopositiveloading.Thisresponseoccurredduetotensileforcespresentinthelinkconnectionundernegativeloading.

Responseofspecimenssubjectedtocyclicloadingwasinfluencedtoagreatextentbythedifferentbehaviourof theconnectionunderpositiveandnegative loading.Thus, the failuremodeof thesespecimenswasgovernedbynegativesemi-cycles.

Endplatethicknessdidnothaveasignificantinfluenceontheinitialstiffnessattheframelevel,butaffected the ductility of the structure under negative displacements. Endplate thickness had aminimalinfluenceontheoverallresponseofframesformonotonictestsunderpositiveloadingduetocompressionforcesintheconnection,whichreducedtensilestressesinboltsandfavouredtheelasticresponseofthejoint.

The monotonic negative loading generated tensile forces in the link connection. The compoundeffect of tensile axial force and bending generated high demands in the end plateand bolts. As aresult, the maximum force and deformation capacity of the removable link and of the framedecreased(Figure20).Theweakestperformancewasobservedforthespecimenwiththethickerend plate, due to the fact that deformation demands concentrated mostly in the bolts, with arelativelybrittlebehaviour.

Figure20.Lateralforcevs.displacementforspecimensL15m1,L15m2andL15c.

Under cyclic loading, the effect of end-plate thickness is not so straightforward. Thus, undernegative loading, tensile axial force leads to higher demands in the bolts, which produces areduction in the maximum force and deformation capacity of the frame. The most importantreductionofmaximumforceanddeformationcapacitywasobservedforconnectionswiththickerend-plates (L15c and L20c), as most of the inelastic demands concentrate in the bolts. For thespecimenswiththinendplate(L10c),theinelasticdeformationdemandsissharedbytheendplateandbolts,whichleadstoanimprovedductilityofthesespecimens.However,thesmallthicknessoftheendplateleadstoagradualdegradationofstiffnessandstrengthduetolow-cyclefatigueeffect.

27

Figure21.Lateralforcevs.displacementforspecimensL10c,L15candL20c:influenceofendplatethickness.

Figure22.Dimensionsandlayoutofthetestedframe.

Figure23.Experimentalsetup.

At the frame level, the experimental tests showed that the removable link solution is feasible.Inelasticdeformationswereconstrainedtotheremovablelinksalone,allotherframemembersandconnections remaining in the elastic range. Additionally, it was possible to replace the damagedremovablelinkswithnewelements.

The third series of tests concerned experimental validation of link removal process. The mainobjectiveoftestwastochecktheprocedureusedtoremovethelinkfollowingplasticdeformationslikelytooccurunderadesignearthquake.Oxy-fuelcuttingwasusedinordertoreleasetheresidualstressesinthelinkafterunloading.

In order to provide the elastic restoring force necessary in order to pull back the eccentricallybraced steel frame after link removal, the columns were realised with a fixed base. The mainstructuraldimensionsandlayoutareshowninFigure22,whiletheexperimentalsetupinFigure23.Theforcewasappliedatthecolumntopbyahydraulicservo-actuatorTheloadcellintegratedintheactuatorwasusedtorecordtheappliedforce.

HE200A

HE

260

B

HE

260B

HE

180A

3.26

2.10 2.10

4.50

0.30

HE

180A

28

a) b)

c) d)

e) f)

g)

Figure24.Linkremovalprocedureforthesecondspecimen.

Displacement transducers were used to monitor different absolute or relative deformations ordisplacements between different components. They were positioned in order to recorddeformationsthatcanbedividedintothreemaincategories:(1)globaldeformationsoftheframe;(2) deformations of the dissipative element – bolted link; (3) slip and deformations of the non-dissipativeelements.

29

Thespecimenwasloadedmonotonicallyuptoalateraltopdisplacementof37.4mm(Figure24a),correspondingroughlytoalinksheardeformationof0.1rad,butwithanintermediateunloadingat0.05 radians. Afterunloading and removal of theactuator (see Figure 24b) the top displacementamounted to 23 mm. After oxy-fuel cutting of the link’s web (see Figure 24c-d), the topdisplacementdroppedtoabout11mm. Itwas concluded that removing the linkwebbyoxy-fuelcutting is notenough foreliminatingtheresidualstresses in the link,due to the fact that flangeshave an important contribution to the ultimate shear strength of the link. Therefore, the bottomflange of the link was also cut (see Figure 24e), followed by the upper one (see Figure 24f).Following complete removal of the link by oxy-fuel cutting both the web and the flanges, theresidualtopdisplacementamountedtoavalueof0.65mm(seeFigure25andFigure26).

Figure25.Load-displacementcurve.

Figure26.Topdisplacementvs.timeresponse.

The non-dissipative elements such as the braces and the columns did not suffer any unwanteddamage, nor did their connections (except from slips in the braces connections). These elementsremained in the elastic domain. The fixed column base did not slip, and the beam-to-columnconnectionsbehavedinasatisfactorymanner.

Fromtheperformedtestsitcanbeconcludedthatboththewebandflangeshavetobeflamecutinordertoalloweasyreplacementof the links.Therewerenoparticulartechnologicalproblemsorspecialsafetymeasuresnecessaryduringlinkremovalandreplacement.

-100

0

100

200

300

400

500

600

700

800

0 5 10 15 20 25 30 35 40

Top displacement [mm]

Fo

rce [

kN

]

0

5

10

15

20

25

30

35

40

0 50 100 150 200 250

Time [min]

To

p d

isp

lacem

en

t [m

m]

beginning of web cut

web elimination

beginning of bottom flange

cut

bottom flange elimination

beginning of top flange cut

maximum force

top flange elimination

unloading

30

2.1.4 LARGE-SCALETESTSONRE-CENTRINGDUALECCENTRICALLYBRACEDFRAME

Afull-scaleexperimentalinvestigationonadualeccentricallybracedframewithreplaceablelinkswasperformedattheEuropeanLaboratoryforStructuralAssessment(ELSA)oftheJointResearchCentre (JRC) in Ispra, Italy. Its objectives were to: (1) validate the re-centring capability of dualstructureswithremovabledissipativemembers(links),(2)assessoverallseismicperformanceofdualeccentricallybraced frames and(3)obtain informationonthe interactionbetweenthesteelframeandthereinforcedconcreteslabinthelinkregion.Theprototypestructurehad3spansof6metersand5baysof6meters,and3storeyswiththeheightof3.5meterseach.Themainlateralloadresistingsystemiscomposedofeccentricallybracedframes.Additionally,thereare4momentresisting frames in transversal directions and 10 moment resisting frames in longitudinal (test)direction, to assure the restoring forces after an earthquake. Considering that in the transversaldirectionthelateralforceresistingsystemislocatedontheperimeterframesonly,andinordertoreducethecostoftheexperimentalcampaign,theteststructureiscomposedofthetwoendframesonly(Figure27a).

The test structure was devised in order to study two alternative solutions of the slab – linkinteraction.Inthesouthframetheremovablelinkwasdisconnectedfromthereinforcedconcreteslab,byanadditionalsecondarybeamplacedinparallelwiththebeamcontainingthelink(Figure27b).Aconventionalsolutionwasadoptedinthenorthframe,wheretheslabwascastoverlinks,butnoshearstudswereusedinthelinkregion.

(a)

(b)

Figure27.Theexperimentalmock-upinfrontofthereactionwall(a)andplanlayout(b)oftheteststructure.

The prototype structure was designed according to EN1990, EN1991, EN1992, EN1993, EN1994and EN1998. Gravity loads of 4.9 kN/m2 (permanent load) and 3.0 kN/m2 (live load) wereconsidered.Thestructurewasassumedtobelocatedinanareacharacterisedby0.19gpeakgroundacceleration and stiff soil conditions (EC8 spectrum for soil type C). A behaviour factor q = 4(ductilityclassM)andinter-storeydriftlimitationof0.0075ofthestoreyheightwereassumed.Thestructural steel components were designed using S355 grade steel, with two exceptions. GradeS460 steel was used for the columns, in order to obtain a larger capacity without increasing thestiffness.Thisapproachhelpspromoteelasticresponseofnon-dissipativecomponents.Linksweredesigned using S235 grade steel (which was replaced during fabrication with the equivalentDOMEX240YPB).

31

Pre-testnumericalsimulations

In order to assess the structural performance and to validate the feasibility of the link removalprocedure, numerical analyses were made on a model of the test structure using SeismoStructsoftware (Seismosoft, 2011). In the pseudo dynamic (PsD) test campaign the equation of motionwill be solved for the frame 1 (with links) disconnected from the slab, and the resultingdisplacementsappliedtobothframes.Therefore,a2Dnumericalmodelwasusedinwhichforce-based plastic hinge elements for beams and columns were used. For braces, the physical theorymodel is used (two force based plastic hinge elements per brace; initial out-of-plane memberimperfection).

(a)

(b)

Figure28.Linkmodel(a);Shearforce–displacementrelationshipfortheLH5link(b).

Boltedlinksweremodelledusingaforce-basedinelasticbeamwitharotationalspringateachend.Theformerwasusedinordertomodeltheflexuralstiffnessofthelink,whilethelatterwereusedinordertoaccountforshearstiffnessofthelink,aswellasrotationaldeformationsandslipinthebolted connection (Figure 28a). The rotational springs were modelled using multi-linear linkelements that can simulate the deteriorating behaviour of strength, stiffness and slip. Materialcharacteristicsobtainedfromqualitycertificateswereused.Thenumericalmodeloftheremovablelinks was calibrated using the experimental results obtained from removable links tests atPolitehnicaUniversityofTimisoara,foralinkwiththeeffectivelengthof400mm,withpropertiessimilar to the links in the DUAREM structure. Figure 28b shows a comparison between theexperimentalandcalibratednumericalresponse.ThecalibratednumericalmodelwasadoptedfortheDUAREMlinksandintroducedinthenumericalmodeloftheteststructure.SheardeformationreportedinFigure28refertothetotaldeformationofthelink,withcontributionsduetoshearoftheelementandrotationandslipinthetwoconnectionsofthelink.

Pushoveranalyseswereperformedonthemodeldescribedaboveinordertoestimatethecapacityof actuators and check the plastic mechanism. Two lateral load patterns were considered: modal(inverted triangular) and uniform. The pushover curves are shown in Figure 29, together withsome characteristic events: first yielding in links, attainment of ultimate deformation capacity inlinks,firstyieldinginelementsotherthanlinks.Itcanbeobservedthatallelementsoutsidelinksyield after attainment of ultimate deformation capacities in links. This proves that moment–resistingframescanprovidetherestoringforceuptocompletefailureoflinks.

SevennaturalrecordswereselectedbymatchingEC8elasticresponsespectrumusedinthedesign.The records were selected from the RESORCE database (Akkar et al., 2014), the main selectioncriteria being thematchingof thecontrolperiod to theoneof the targetspectrum (TC=0.6sec).TherecordswerethenscaledtothetargetspectrumusingEN1998-1rules.StructuralperformancewasevaluatedforthelimitstatesshowninTable5.Thekeyparametersofstructuralresponseat

32

thelimitstatesconsideredareshowninTable4.AcceptancecriteriaforshearlinkswerebasedonFEMA356(0.005radatDL,0.11radatSDand0.14radatNC).

AttheDamageLimitation(DL)limitstateallstructuralcomponentsexceptlinksareintheelasticrange. Shear deformations in links exceed the FEMA 356 limits, but this is expected, since thedesigncarriedoutaccordingtoEN1998-1didnotimposeanylimitationonyieldingofstructuralmembersatDLlimitstate.Evenso,permanentinterstoreydriftsremainverylow(0.007%or0.2mminaverage,andamaximumvalueof0.028%or0.98mm).Intheseconditions,theadvantageofthe proposed system is obvious, since the damaged dissipative members (links) can be easilyreplaced due to negligible permanent drifts. Even if somestructural damage is present, it can berepairedeasilybyreplacingtheboltedlinks.

Figure29.Pushovercurves.

Table4.Keyparametersofresponse.

Peaktransientinterstoreydrift,%

Permanentinterstoreydrift,%

Sheardeformationinlinks,rad

Record DL SD NC DL SD NC DL SD NC00385_H1 0.45 0.86 2.82 0.001 0.001 0.155 0.054 0.108 0.43214336_H1 0.46 0.76 1.68 0.003 0.005 0.016 0.056 0.093 0.25615613_H2 0.43 0.76 2.87 0.003 0.003 0.154 0.051 0.101 0.44115683_H2 0.40 0.91 3.06 0.028 0.011 0.128 0.046 0.123 0.16916035_H2 0.51 0.86 2.14 0.004 0.002 0.125 0.058 0.108 0.32716889_H1 0.39 0.72 3.69 0.006 0.025 0.238 0.044 0.094 0.57017167_H1 0.56 1.17 2.82 0.006 0.033 0.199 0.069 0.155 0.429

mean 0.46 0.86 2.73 0.007 0.011 0.145 0.054 0.108 0.432

At theSignificantDamage(SD) limit statedamage isstillconstrainedto linksonly,whichexhibitplastic deformation demands very close to the acceptable 0.11 rad. Permanent drifts are onlyslightlylargerthanattheDLlimitstate,theaveragevaluesbeingabout0.011%or0.4mm,andthemaximum ones of 0.033 % or 1.2 mm. Due to low permanent drifts, the structure is easilyrepairableatthislimitstateaswell.

AttheNearCollapse(NC) limitstatethestructuraldamageiswidespread.Sheardeformations inlinks are well over the acceptable values (0.432 rad compared to 0.14 rad). However, due tomomentresistingframes,theoverallperformanceofthestructurecanbeconsideredacceptableforthislimitstate,averagetransientdriftsbeingoftheorderof2.73%.Plasticdeformationdemandsare present in moment resisting frames (beams and columns) and braces. Even so, permanentinterstorey drifts are not very large (average of 0.15% or 5.1 mm). However, repairing of thestructureisconsiderednottobefeasibleattheselargelevelsofseismicinput.

33

Experimentalprogram

Theinstrumentationoftheexperimentalmock-upconsistedof38localdisplacementtransducers(tomonitorthelinkdeformationsandtheslipinthespliceconnectionofeveryEBFbraceandtheEBF beams from the first storey, in the south frame), nine global displacement transducers (tomonitortheglobal longitudinalandtransversaldisplacementsof thestructure),22 inclinometers(tomonitortherotationsofthebeamtocolumnandcolumnbase-jointzones)and32straingages(tomonitortheyieldatthemiddleoftheEBFbracesandattheendoftheMRFbeamsofthefirststorey).Potentialdissipativeareaswerewhite-washedtovisuallyobserveyielding.

Table5.Limitstatesandcorrespondingscalingfactorsforseismicinput

Limitstate Meanreturnperiod,years

Probabilityofexceedance

ag/agr ag,g

Fulloperation(FO) - - 0.062 0.020DamageLimitation(DL) 95 10%/10years 0.59 0.191SignificantDamage(SD) 475 10%/50years 1.00 0.324

NearCollapse(NC) 2475 2%/50years 1.72 0.557

Table6.SequenceofPsDandlinkremovaltests.

Links Test Scope

Set1

Fulloperation(FO1) Assesselasticresponseofthestructure.DamageLimitation(DL) Observestructuralresponsetoamoderateearthquake.

Replacementofset1links(LR1)

Investigateremovaloflinksbyunboltingandtheirreplacement.

Set2

Fulloperation(FO2) Assesselasticresponseofthestructure.SignificantDamage(SD) Observestructuralresponsetoadesign-levelearthquake.

Pushover(PO1) Inducelargepermanentdrifts.Replacementofset2links

(LR2)Investigateremovaloflinksbyflamecuttingandtheir

replacement.

Set3Fulloperation(FO3) Assesselasticresponseofthestructure.Nearcollapse(NC) Observestructuralresponsetoasevere-levelearthquake.

Pushover(PO2andPO3) Investigateultimatecapacityofthestructure.

SevennaturalrecordswereselectedbymatchingEN1998-1elasticresponsespectrumusedinthedesignandwereusedtoperformpre-testnumericalsimulations(Stratanetal.2014).TherecordswereselectedfromtheRESORCEdatabase(Akkaretal.2014),themainselectioncriteriabeingthematchingof thecontrolperiodtotheoneof thetargetspectrum(TC=0.6sec).TherecordswerethenscaledtothetargetspectrumusingEN1998-1rules.Structuralperformancewasevaluatedforthe limit states shown in Table 5. Record selected for pseudo-dynamic (PsD) tests is 15613_H2(recorded at Yarimca (Eri) station during the Izmit aftershock event in Turkey on 13.09.1999).Basedonpre-testnumericalsimulations, itprovidedresponseclosesttothemeanresponsefromallrecordsintermsoftopdisplacement,interstoreydriftsandsheardeformationsinlinks.Testingsequenceonthemock-upinthereactionwallfacilityofELSAconsistedinmodalevaluation,snap-backandPsDtests.Table6summarisesthesequenceofPsDtestsandlinkreplacementsperformedontheteststructure.Linkswerereplacedtwotimes:afterDLtestandafterSDtest.

FO1,FO2andFO3tests

EachofthethreeseriesofPsDtestsshowninTable6startedwithaFullOperationtest(FO1,FO2andFO3).Seismicinputwaschosensmallenoughtoallowassessmentoftheelasticresponseofthestructure and check instrumentation. There was no visible or measured damage during the FO1

34

test. The peak top displacement amounted to 5.7 mm, while permanent drifts were negligible.SimilarresponsewasobservedduringtheFO2andFO3tests.

DLandLR1tests

During thedamage limitation(DL) test, thepeak topdisplacementwas32mm(0.3%),while theresidualoneof5mm(0.05%).Alowresidualinter-storeydriftamountingtoamaximumof3mm(lessthan0.1%)wasobserved.Relativelysmallplasticdeformationsoccurredinthelinks(Figure30), the largest peak rotation was 0.032 rad and the largest residual deformation was 0.014 rad(linkinthe3rdstoreyofthesouthframe).Noslipoccurredinthebraceorbeamspliceconnections.Nodamagewasobservedinthesteelelementsofthestructurewiththeexceptionof links.Minorcrackswereobservedintheconcreteslab.

(a)

(b)

Figure30.DLtest:linkfromstorey3ofthesouthframe(a)andhystereticresponseofthelinksfromthe3rdstorey(b).

(a)

(b)

Figure31.Topdisplacementtimehistoryforthesouthframe(a)andpeakvaluesoftopdisplacementduringDLandLR1tests(b).

AftertheDLtest,thefirstsetofdamagedlinkswasremovedandreplacedwithanewsetoflinks.ConsideringthatresidualdriftsaftertheDLtestsweresmall(maximumvaluesofabout5mmforbothframes),thelinkswereremovedafteruntighteningthebolts,startingwiththefirstfloorofthenorthframe.Afteruntighteningthebolts,ahydraulicjackwasusedforpushingapartbraceends,inordertopulloutthelinks.Theforcesthathadtobeappliedbythehydraulicjackinordertopullouteasilythelinksvariedbetween360kNand430kN.Tofacilitatemountingofthenewlinks,thereplacementlinkswerefabricatedby2mmshorterthantheoriginalones.Theinstallationofthenewlinksdidnotraiseparticularproblemsandthehydraulicjackwasnotneeded.Installationof

-300

-200

-100

0

100

200

300

400

-0,01 0 0,01 0,02 0,03 0,04

Lin

k s

hea

r fo

rce

[kN

]

Link rotation [rad]

Storey 3

N

S

-20

0

20

40

0 10 20 30 40 50

To

p d

isp

lace

men

t [m

m]

Time [s]

residual displacementafter DL test

removal of links

32

5

1 2

32

5 4

2

0

5

10

15

20

25

30

35

tran

sien

t

resi

du

al

link

rem

oval

link

repla

cem

ent

Top

dis

pla

cem

ent,

mm

South

North

35

thelinksinthesouthframewasstraightforwardbecausetheslabdidn’tobstructtheaccess.Itwasmoredifficultinthenorthframewheretheslabwascastacrossthelinkswhichhadtobeinsertedfromtheside.

As can be seen in Figure 31, the residual top displacement at the end of the DL test (5 mm)decreasedaftertheeliminationofthedamagedlinksto1mmforthesouthframeand4mmforthenorthframe.Betterre-centringwasobservedinthesouthframe,wherethelinkisnotconnectedtothe concrete slab. An additional re-centring wasobserved during thereplacement of the links (2mmfinaltopdisplacementforbothframes).Afterlinkreplacementprocedure,anegligibleresidualtopdisplacementwasobserved(H/5250forbothframes),considerablysmallerthantheerectiontolerance(H/300)accordingtoEN1090-2,thestructurebeingessentiallyre-centred.

SD,PO1andLR2tests

Thepeaktopdisplacementrecordedduringthesignificantdamage(SD)testwas50mm(0.48%),andtheresidualonewas13mm(0.12%).Thepeaktransientinterstoreydriftamountedto20mm(0.57%), while the residual interstorey drift was approximately 5 mm (0.14%). Moderatedeformationsoccurred in the links (Figure32), the largestamountingto0.061rad transientand0.022radresidual(inthelinkfromthe1ststoreyofthesouthframe).Noslipoccurredinthebraceorbeamspliceconnections.Nodamagewasobservedinthesteelelementsofthestructurewiththeexceptionoflinks.Moderatecrackswereobservedintheconcreteslab.

(a)

(b)

Figure32.SDtest:linkfromstorey1ofthesouthframe(a)andhystereticresponseofthelinksfromthe1ststorey(b).

(a)

(b)

Figure33.Flamecuttingforreleasingresiduallinkforces(a)andalinkafterflamecutting(b).

-600

-400

-200

0

200

400

600

-0.01 0.01 0.03 0.05 0.07

Lin

k s

hea

r fo

rce

[kN

]

Link rotation [rad]

Storey 1

N

S

36

(a)

(b)

Figure34.SD,PO1andLR2tests:topdisplacementforthesouthframe(a)andpeakvaluesoftopdisplacement(b).

SincetheresidualdeformationsaftertheSDtestswererelativelysmall,thepushovertestPO1wasperformed,togeneratelargeenoughresidualdisplacementsthatwouldrequireflamecuttingofthelinks. The test was controlled by the displacement of the south frame in the third floor with aninverted triangular distribution of forces. The displacements applied to the north frame wereidenticaltothoseonthesouthframe.

ThemaximumtopdisplacementrecordedduringthePO1testswasof67mmand68mmforthesouthandnorthframesrespectively.Noyieldwasobservedinthesteelelementsoutsidethelinks.Larger deformations occurred in the links (maximum of 0.075 rad peak rotation and 0.066 radresidual one). There was no slip detected in the brace or beam splice connections. More visiblecracks were observed in the concrete slab. After the completion of the PO1 test, the structureexhibited a significant residual top displacement of 45 mm (0.43%), while the correspondingresidualinter-storeydriftamountedtoamaximumof18mm(0.5%).

AfterthePO1test, thesecondsetofdamagedlinkswasremoved(LR2)andreplacedwithanew(third)setoflinks.Inthiscase,removalofthelinksrequiredthattheybeflamecut(Figure33),inordertoreleaseresidualforceslockedinthedamagedlinks,priortountighteningthebolts.Linkswere flame cut and removed starting from the top storey downwards. The residual topdisplacementattheendoflinkremoval(LR2test)was10mminthesouthframeand19mminthenorthframe(Figure34).Betterre-centringwasobservedinthesouthframe,wherethelinkswerenot connected to the concrete slab. Although it was not necessary to use a hydraulic jack forremovingthelinks,itwasneededtoplacethenewsetoflinksinthestructure.Theforcesthathadtobeappliedby thehydraulic jack inorder toeasily insert the linksvariedbetween286kNand486 kN. An additional re-centring was also observed during the replacement of the links. Finalresidualtopdisplacementsafterplacementofnewlinksamountedto2mm(H/5250)forthesouthframe and 6 mm (H/1750) for the north one. Both values are considerably smaller than theerection tolerance 11.7 mm (H/300) according to EN1090-2, the structure being essentially re-centred.

NC,PO2andPO3tests

The near collapse (NC) test was stopped automatically at 5.3 seconds of the input accelerogrambecause the difference in displacement recorded by the actuator and the external transducerexceededthelimitof10mm.InordertocontinuewiththePsDtest,thestructurewasbroughttoastate corresponding to the distribution of displacements measured before the oil pressure wasreleased.Aftertherestart,thetestwasstoppedagainafewstepslaterastheactuatorsofthefirstfloorreachedtheirmaximumcapacitywhenthestructureentered thesecondmodeofvibration.Thiscorrespondedtoasituationwheretheactuatorsfromthefirstfloorwereactinginadirection

-20

0

20

40

60

80

0 20 40 60 80

Top

dis

pla

cem

ent

[mm

]

Time [s]

residual displacementafter SD+PO1 test

removal of links

50

13

67

45

10

2

50

13

68

47

19

6

0

10

20

30

40

50

60

70

80

tran

sien

t S

D

resi

dual

SD

max

. PO

1

resi

dual

PO

1

link

rem

oval

link

repl

acem

entT

op

dis

pla

cem

ent,

mm

South

North

37

opposite to that of the actuators of the second and third floors. The maximum displacementrecordedatthetopof thestructurewas118.1mm.Theseismic linksreachedhighdeformations,exceedingthemaximumlimitofthetransducers(±25mm)placedacrossthelinksofthefirsttwofloors.

The test campaign continued with a push-over test PO2 with an inverted triangle distribution offorces.Equaldisplacementsbetweenthenorthandsouth frameswere imposedatthethirdfloorandtheactuatorsofthefirstandsecondfloorswereinforcecontrol.Althoughequaldisplacementswereimposedonthenorthandsouthframesofthethirdfloor,thetwolowerfloorsrotatedinthehorizontal plane due to the different stiffness between the two frames. The test was stopped at231.6 mm displacement at the top floor. An additional cyclic push-over test PO3 with a uniformforce distribution in order to avoid force saturation at the actuators was carried out underdisplacementcontrolconsideringaMaximumdisplacementsatthethirdfloorwere405and499.4mmonthenorthandsouthframesrespectively.

Thelastthreetestsproducedextensivedamagethroughoutthestructure.Verylargedeformationsoccurredinthelinks,0.15–0.38radforthefirsttwolevelsand0.09–0.13radforthethirdlevel.Linksfromthefirsttwostoreysfelldown,withfailureoccurringintheweldsconnectingthelinkstotheendplates.SignificantdamagewasalsoobservedinthecolumnbasezonesandattheendoftheMRFbeams,justoutsidethehaunch.Theconcreteslabwasheavilydamagedinthenorthframe.

2.1.5 LINKREPLACEMENTORDERINHIGHER-RISEFRAMES

In order to validate numerically the design methodology of current practice EBF structures withremovable links and the link replacement procedure, two dual structural configurations weredesigned: the first, like in the case of the experimental specimen (Stratan et al., 2015), having acentral EBF (the link being placed at mid-span) and two side MRFs, being further referred to as"configuration A" (see Figure 35a)and thesecond having acentralMRF and two side EBFs. (thelinks being placed marginally, connected to columns on one side) being further referred to as"configuration B" (Figure 35b). The links from EBFs were conceived as removable (bolted)dissipativeelementsbecausetheyareintendedtoprovidetheenergydissipationcapacityandtobeeasily replaceable. The more flexible moment resisting frames provide the necessary re-centringcapabilitytothestructure

Bothstructures have3spansof7.5meters and5bays of7.5meters, and6storiesof3.5meterseach(4.0matthegroundlevel).Themainlateralloadresistingsystemiscomposedofeccentricallybracedframes(2oneachhorizontaldirectionforstructureAand4oneachhorizontaldirectionforstructureB).Additionally,thereare4momentresistingframesoneachhorizontaldirectionincaseof configuration A and 2 moment resisting frames on each horizontal direction in case ofconfiguration B, to assure the restoring forces after an earthquake. All the other frames aregravitationalloadsresistingsystems(withpinnedcompositesteel-concretebeams).

(a)

(b)

Figure35.ConfigurationA(a)andconfigurationB(b)structures.

38

The capacity design of the structure was carried out according to EN1990, EN1991, EN1993,EN1994 and EN1998. A 4.9 kN/m2 dead load and 3.0 kN/m2 live load were considered. Thebuildingwasanalyzedforstiffsoilconditions(EC8type1spectrumforsoiltypeC),characterizedby0.35gpeakgroundacceleration.Abehaviourfactorq=4(ductilityclassM)andinter-storeydriftlimitationof0.0075ofthestoreyheightareused.

Theperimetercolumnsarefixedatthebaseandallthecentralones(resistingtogravityloadsonly)are pinned at the base. Short links were used, with the length e = 800 mm and with equivalentstiffnessreducedfromthetheoreticalshearstiffnessofcontinuouslinks(toaccountfortheshearstiffness of the link and rotational deformations in the bolted connection), made from welded Hsections. All thestructural elements from EBFs are made from mild carbon steelS355, while theonesfromMRFsaremadefromhigh-strengthsteelS460.

TheelementsectionspresentedinTable7andTable8wereobtained.

Table7.ElementssectionsforconfigurationA.

Storey LinksEBF

beamsBraces

EBFcolumns

MRFbeams MRFcolumns

1 640x260x22x12 HEA650 HEM300 HEM340 IPE450 HEB3402 590x260x22x10 HEA600 HEM280 HEM340 IPE400 HEB3403 540x240x22x10 HEA550 HEM260 HEB340 IPE360 HEB3004 490x230x22x9 HEA500 HEM240 HEB340 IPE360 HEB3005 440x230x20x8 HEA450 HEM220 HEB300 IPE300 HEB3006 390x210x16x6 HEA400 HEM180 HEB300 IPE300 HEB300

Table8.ElementssectionsforconfigurationB.

Storey LinksEBF

beamsBraces

EBFcolumns

MRFbeams MRFcolumns

1 540x260x20x9 HEA550 HEM300 HEM340 IPE600 HEB3402 490x230x20x8 HEA500 HEM280 HEM340 IPE550 HEB3403 440x230x20x8 HEA450 HEM260 HEM340 IPE550 HEB3004 390x230x20x8 HEA400 HEM240 HEM340 IPE500 HEB3005 350x230x18x8 HEA360 HEM240 HEB340 IPE500 HEB3006 330x190x15x5 HEA340 HEM200 HEB340 IPE400 HEB300

In order to obtain a global numerical model that can be used in further post-test numericalsimulationsoncurrentpracticedualeccentricallybracedframeswithremovablelinks,acalibrationwas performed based on experimental results of pseudo-dynamic tests of the DUAREM testingprogramme.

Atime-historyanalysiswasperformedusingSeismoStructprogram,applyingateachofthethreelevels of the DUAREM specimen model, the displacements obtained from the experimental tests.The values used for material characteristics were the ones obtained from tensile tests on steelsamples originating from experimental specimen material batch, performed at PolitehnicaUniversityTimisoara.

Force-based plastic hinge elements for beams, columns and braces were used. Bolted links weremodelled using a force-based inelastic beam with two rotational springs at the end (see Figure28a).Theformerwasusedinordertomodeltheflexuralstiffnessofthelink,whilethelatterwereused in order to account for shear stiffness of the link, as well as rotational deformations in thebolted connection. The rotational springs were modelled using the smooth hysteresis curve (asdefined by SeismoStruct) for link elements, having the initial stiffness computed as describedbefore,theratiobetweenultimateshearforce(Vu)andyieldshearforce(Vy)equalto1.7andtheultimatesheardeformationγu=0.15rad.

39

The results of numerical simulations and experimental testing were compared in terms of firststorey link rotation versus shear force and drift versus base shear force (see Figure 36). A verygoodmatchwasobserved,thenewcalibratednumericalmodelbeingadoptedforfurthernumericalsimulationsoncurrentpracticestructures.

Figure36.Firststorey-linkandglobalhysteresis.

Forastructurewithre-centringcapability, thedesignobjectiveconsists inpreventingyielding inmembers other than removable dissipative ones, up to a desired deformation. Ideally the lattershould be the ultimate deformation capacity of the removable dissipative member. From apreliminarypushoveranalysis,itwasobservedthatfollowingcode-basedcapacitydesignruleswasnotenoughtoaccomplishtheobjectivestatedabovefortheinvestigatedstructures.But,usingS690higher-strengthsteelinmomentresistingframeswasshowntobeefficientinavoidingyieldinginthetheirmembers, increasingtheirstrength,butnotthestiffness, thisbeingthematerialusedinfurtheranalyses.

The calibrated 2D numerical models of the designed structures were subjected to pushoveranalysesontransversaldirection,withamodal(invertedtriangular)distributionoflateralforces,in order to assess their structural performance. The pushover curves presented in Figure 37illustratetheseismicperformanceofthetwostructures.

Figure37.Pushovercurves.

-500

-400

-300

-200

-100

0

100

200

300

400

500

-0,02 0 0,02 0,04 0,06 0,08

Lin

k sh

ear

fo

rce

[kN

]

Link shear deformation [rad]

Link storey 1

experimentalnumeric -1000

-800

-600

-400

-200

0

200

400

600

800

-20 -15 -10 -5 0 5 10

She

ar f

orc

e [

kN]

Inter-storey drift [mm]

Storey 1

experimentalnumeric

0,005 rad

0,11 rad 0,14 rad

MRF beam

0,005 rad

0,11 rad MRF beam

0,14 rad

0

1000

2000

3000

4000

5000

6000

7000

0 0,1 0,2 0,3 0,4

Bas

e sh

ear

fo

rce

[kN

]

Top displacement [m]

PO_A

PO_B

40

TheobjectiveofhavingnoyieldingintheMRFsbeforetheattainmentoftheULSdeformationintheremovable links (0.11 rad plastic rotation according to FEMA356) of the EBFs is accomplished,representing the basic design requirement for dual frames with removable dissipative members.MRFsprovidethere-centringofthestructureevenuntilafterthelinksultimatedeformation(0.14radplasticrotationaccordingtoFEMA356)incaseofconfigurationA.

Seven natural seismic records of European earthquakes were used for seismic performanceassessment of the test structure using nonlinear time-history analysis. Individual records werescaledtothetargetspectrum(EN1998type1,soiltypeC,ag=0.35g)usingEN1998criteria.

Structural performance was evaluated for the limit states shown in Table 5, where agr is thereference peak ground acceleration (corresponding to 10% / 50 years earthquake) and agrepresentsthepeakgroundaccelerationforaspecificearthquakelevel.

AttheDamageLimitation(DL)limitstateallstructuralcomponentsexceptlinksareintheelasticrange. Shear deformations in links exceed the FEMA 356 limits of 0.005 radians (0.039 rad forstructureAand0.026radforstructureB),butthisisnormal,sincethedesigncarriedoutaccordingtoEN1998didnotimposeanylimitationonyieldingofstructuralmembersatDLlimitstate.Evenso, peak inter-storey drifts are within the limits imposed (<0.75%). In these conditions, theadvantageof theproposedsystemisobvious,sincethedamageddissipativemembers(links)canbereplacedeasilyduetonegligiblepermanentdrifts.Evenifsomestructuraldamageispresent,itcanberepairedeasilybyreplacingtheboltedlinks.

At theSignificantDamage(SD) limit statedamage isstillconstrainedto linksonly,whichexhibitplasticdeformationdemandsbelow0.11rad(0.092radforstructureAand0.078radforstructureB).PermanentdriftsareonlyslightlylargerthanattheDLlimitstate.Duetolowpermanentdrifts,thestructuresareeasilyrepairableatthislimitstateaswell.

AttheNearCollapse(NC) limitstatethestructuraldamageiswidespread.Sheardeformations inlinks are well over acceptable values of 0.14 rad (0.282 rad for structure A and 0.266 rad forstructureB).However,duetomomentresistingframes,theoverallperformanceofthestructurescan be considered acceptable for this limit state. Plastic deformation demands are present inmomentresistingframes(beamsandcolumns)andbraces.Evenso,permanentinter-storeydriftsare not very large. However, repairing of the structures is considered not to be feasible anddesirableattheselargelevelsofseismicinput.

Inordertonumericallysimulatethelinkremovalorder,thestructuresweresubjectedtoauniformdistribution of lateral forces up to the attainment of 0.11 rad plastic rotation in links on thetransversaldirection,simulatingtheseismicaction.Thenthestructureswereunloaded,simulatingthestateofthestructureafteranearthquakeandlinksareremovedlevelbylevel.

Threepossibilitiesoflinksremovingorderwithinastoreywerestudied:firstlyremovingthelinksonthelongitudinaldirectionandsecondlytheonesonthetransversaldirection(seeFigure37a.d),viceversa(seeFigure37b.e),andinacircularpattern(seeFigure37c.f).

Itwasobservedthatforthefirstversion(seeFigure37a-d)theresidualshearforcedropisabout21%(configurationA)and16%(configurationB)smallerthanforthesecondversion(seeFigure37b-e) and the redistribution of forces between the links of the same storey is also smaller. Thesituationforthethirdversion(seeFigure37c-f)isinbetweentheothertwo.

Onthefirstversionoflinkremovalorderwithinastorey(seeFigure37a-d),wasanalysedalsotheremoval of linksstarting from themost loaded to the least loadedstorey (from the lower storeytowardtheupperone).Inthiscase,wasobservedalargerinteractionbetweenstories,butvaluesofthe shear force drop with 43% (configuration A) and 39% (configuration B) smaller than in thecase of eliminating links from the upper one toward the lower one and smaller redistribution offorcesbetweenthelinksofthesamestorey.

41

(a)

(b)

(c)

(d)

(e)

(f)

Figure38.LinkremovalsolutionforconfigurationA(atoc)andforconfigurationB(dtof).

Numericalsimulationswereperformedinordertoinvestigatesolutionsforremovinglinks.Threepossibilitiesoflinksremovingorderwithinastoreywerestudied:(a)firstlyremovingthelinksonthelongitudinaldirection(theleastloadedones)andsecondlytheonesonthetransversaldirection(themostloadedones),(b)viceversa,and(c)inacircularpattern.Itwasobservedthatforthefirstversionthelink’sresidualshearforcedropissmallerthanfortheothertwostudiedsolutionsandthe redistribution of forces between the linksof the same storey is alsosmaller. This is because,whenremovinglinksinversion(a),thestructurebecomesmoreflexible(theleastloadedlinksonthe longitudinal direction are first removed and when removing the first link on the transversaldirection, only the MRFs in the other direction are still working) and the links develop smallerresidualforces.

Onthefirstversion(a)oflinkremovalorderwithinastorey,wasanalysedalsotheremovaloflinksstartingfromthemostloadedtotheleastloadedstorey(fromthelowerstoreytowardtheupperone).Inthiscase,wasobservedalargerinteractionbetweenstories,butvaluesoftheshearforcedropwithabout30-40%smallerthaninthecaseofeliminatinglinksfromtheupperstoreytowardtheloweroneandsmallerredistributionofforcesbetweenthelinksofthesamestorey.

2.1.6 OUTCOMES

Dual frames with removable dissipative members can be used to provide a structure with a re-centringcapability,andcangreatlyreducethecostsandmanpowerrequiredforpost-earthquakerepair. This work describes a large-scale experimental programme carried out at the EuropeanLaboratory for Structural Assessment (ELSA) on a three-storey dual eccentrically braced framewith replaceable links. The testing sequence on the mock-up in thereaction wall facility ofELSAconsistedofmodalevaluations,snap-back,andpseudo-dynamictests.

Thedualeccentricallybracedstructureexhibitedexcellentperformancewhensubjectedtoseismicinputwithintensitiescorrespondingto95and475yearsreturnperiod,correspondingtoDamageLimitation and Significant Damage limit states, respectively. Small residual deformations wererecorded for both seismic intensity levels, whichwere comfortably within the erection tolerancelimits.Suchsmallpermanentdeformationseffectivelymeanthatthestructureisself-centringtoadegree, which allows for an easy repair of the structure by the replacement of damaged links.Residual drifts are further reduced by removing and replacing the bolted links. Re-centring was

42

betterfortheframewithlinksdisconnectedfromtheslab.Moreover,damagetotheconcretewasavoidedinthiscase.Nevertheless,goodre-centringwasobservedevenfortheframewiththeslabcastoverthe links,whiledamagetothereinforcedconcreteslabwas insignificantat theDamageLimitation and Significant Damage tests. Provided the residual deformations after an earthquakearesmall,thelinkscanberemovedsimplybyuntightening,aswasdemonstratedaftertheDamageLimitation test. If largerresidualdrifts occur, flamecutting of links is recommended to allow forsmoothreleaseofforces,aswasdoneaftertheSignificantDamageseriesoftests.Theexperimentalinvestigation validated the re-centring capability of dual eccentrically braced frames withremovablelinks,whichwasaccomplishedwithoutmajortechnologicaldifficulties.

Forastructurewithre-centringcapability, thedesignobjectiveconsists inpreventingyielding inmembers other than removable dissipative ones, up to a desired deformation. Ideally the lattershouldbetheultimatedeformationcapacityoftheremovabledissipativemember.Followingcode-based capacity design rules was not enough to accomplish this objective for the investigatedstructures.But,usinghigher-strengthsteelinmomentresistingframeswasshowntobeefficientinavoidingyieldingintheirmembers.

2.1.7 PUBLICATIONS

[1] Dubina,D.,Stratan,A.,Dinu,F.(2008)."Dualhigh-strengthsteeleccentricallybracedframeswithremovablelinks".EarthquakeEngineering&StructuralDynamics,Vol.37,issue15,pp.1703-1720,(OnlineISSN:1096-9845,PrintISSN:0098-8847).

[2] Ioan,A.,A.Stratan,andD.Dubina.(2013)."NumericalSimulationofBoltedLinksRemovalinEccentrically Braced Frames." Pollack Periodica 8 (1): 15-26, DOI:10.1556/Pollack.8.2013.1.2.PaperISSN:1788-1994,OnlineISSN:1788-3911.

[3] Dubina, D., Stratan, A., Dinu, F. (2011). "Re-centring capacity of dual-steel frames", SteelConstruction:DesignandResearch,Vol.4,No.2,pp.73-84.

[4] Ioan, A., Stratan, A., Dubina, D., Taucer, F. (2013), "Dual-Steel Eccentrically Braced Frameswith Bolted Links – Simulation of Safe Removal Process", Acta Technica Napocensis: CivilEngineering&Architecture,vol.56,No.2,UTPRESS,ISSN1221-5848,pp.111-118.

[5] Dubina, D., Stratan, A., Florea, D., Ioan, A. (2013). "Sisteme structurale cu componentedisipativedesigurantaseismica".AICPSReview,nr.1-2/2013,ISSN:2067-4546,pp.41-50.

[6] Dubina,D.,Stratan,A.,Dinu,F.Bordea,S.,Neagu,C., Ioan,A.(2010)."Sistemestructuralecuelementemetalicedisipativedemontabilepentruclădirimultietajateînzoneseismice".AICPSReview,nr.2-3/2010,p.106-112.ISSN:2067-4546

[7] Stratan, A., Ioan, A., Dubina, D. (2012). "Re-centring capability of dual eccentrically bracedframes with removable bolted links". Proceedings of the 7th International Conference onBehaviorofSteelStructures inSeismicAreas(STESSA),9-11 January2012,Santiago,Chile,ISBN978-0-415-62105-2,pp.723-728.

[8] Dinu, F., Dubina, D., Stratan, A. (2010). "Evaluation of re-centring capability of dual frameswithremovabledissipativemembers:casestudyforeccentricallybracedframeswithboltedlinks". COST Action C26 "Urban Habitat Constructions under Catastrophic Events"Proceedings, Naples, 16-18 September 2010, Mazzolani (Ed.), Taylor & Francis Group,London,p.821-828.ISBN:978-0-415-60685-1.

[9] Dubina,D.,Stratan,A.andDinu,F.(2007)."HighStrengthSteelEBFrameswithLowStrengthBolted Links". Proc. of the 5th International Conference on Advances in Steel Structures,Singapore,5–7December2007,ResearchPublishingServices, JYRichardLiew&YSChoo(Eds.),ISBN978-981-05-9365-0,Vol.III.,pp.249-254.

43

[10] Stratan,A.,Dogariu,A.andDubina,D.(2007)."Bolted linksforeccentricallybracedframes:influence of link stiffness". Proc. of the 3rd Intl. Conf. on Steel and Composite Structures(ICSCS07), Manchester, UK, 30 July - 1 August 2007, Eds. Wang and Choi. Taylor&Francis,ISBN:978-0-415-45141-3,pp.847-853.

[11] Stratan,A.andDubina,D.(2004)."Boltedlinksforeccentricallybracedsteelframes".Proc.oftheFifthAISC/ECCSInternationalWorkshop"ConnectionsinSteelStructuresV.Behaviour,Strength&Design",June3-5,2004.Ed.F.S.K.Bijlaard,A.M.Gresnigt,G.J.vanderVegte.DelftUniversityofTechnology,TheNetherlands.ISBN:978-90-9019809-5,pp.223-232

[12] Stratan, A., Ioan, A., Dubina, D., Taucer, F., Poljanšek, M., Molina, J., Pegon, P., D’Aniello, M.,Landolfo, L.(2014). "Experimental program for large-scale tests on a re-centring dualeccentrically braced frame", 7th European Conference on Steel and Composite StructuresEUROSTEEL 2014, 10-12 September 2014, Napoli, Italy, Landolfo R, Mazzolani FM, editors,EuropeanConventionforConstructionalSteelwork,ECCS,paperno.37-300,8p.ISBN:978-92-9147-121-8.

[13] Ioan,A.,Stratan,A.,Dubina,D.(2014)."Linkreplacementorderinhigh-risere-centringdualeccentrically braced frames", 7th European Conference on Steel and Composite StructuresEUROSTEEL 2014, 10-12 September 2014, Napoli, Italy, Landolfo R, Mazzolani FM, editors,EuropeanConventionforConstructionalSteelwork,ECCS,paperno.49-302,6p.ISBN:978-92-9147-121-8.

[14] Stratan, A., Ioan, A., Dubina, D., Poljanšek, M., Molina, J., Pegon, P., Taucer, F. (2014). "Dualeccentrically braced frames with removable links: Experimental validation of technicalsolution through large-scale pseudo-dynamic testing". Proceedings of the Fifth NationalConference on Earthquake Engineering and First National Conference on EarthquakeEngineeringandSeismology–5CNIS&1CNISS,Bucharest,Romania,June19-20,2014.RaduVăcăreanu, Constantin Ionescu (Eds.), Bucharest: Conspress, 2014, ISBN (print) 978-973-100-342-9,ISBN(CD)978-973-100-341-2,pp.323-330,PaperNo.31(onCD-ROM)

[15] Stratan,A.,Ioan,A.,Dubina,D.,D'Aniello,M.,LaMannaAmbrosino,G.,Landolfo,R.,Taucer,F.,Poljansek, M. (2013). "Pre-test numerical simulation and experimental program on a dualeccentricallybracedframewithreplaceablelinks".XXIVGiornateitalianedellaCostruzioneinAcciaio,Torino,30September–2October2013,ISBN978-88-905870-0-9,pp.867-876.

[16] Stratan, A., Ioan, A., Dubina, A., Taucer, F., Poljansek, M. (2014). "Pre-test numericalsimulations and experimental program on dual eccentrically braced frame with removablelinks". Proceedings of the International Workshop "Application of High Strength Steels inSeismic Resistant Structures". Orizonturi Universitare, Timisoara, Romania, ISBN: 978-973-638-552-0,pp.137-148.

[17] Stratan, A., Dinu, F., Dubina, D. (2010). "Replacement of bolted links in dual eccentricallybraced frames".Proc.of the14thEuropeanConferenceonEarthquakeEngineering,August30–September03,2010,Ohrid,RepublicofMacedonia.MAEE,PaperNo.1581.ISBN:978-608-65185-0-9.

[18] Dubina, D., Dinu, F., Stratan, A. (2009). "Post-earthquake intervention procedure on dualeccentrically braced frames with removable links". A 4-a Conferinţă Naţională de InginerieSeismică, Bucureşti, 18 decembrie 2009. EdituraConspress. Vol. II. pp. 273-282. ISBN 978-973-100-096-1

[19] Stratan, A., Dubina, D. (2008). "Removable bolted links for eccentrically braced frames",Proceedings of the International Symposium "Urban Habitat Constructions underCatastrophic Events", Malta, 22-23 October 2008, COST Action C26, Editors: Mazzolani,

44

Mistakidis,Borg,Byfield,DeMatteis,Dubina,Indirli,Mandara,Muzeau,Wald,Wang,p.181-186.ISBN978-99909-44-42-6

[20] Dinu, F., Dubina, D., Stratan, A. (2005). "Performance Criteria for Seismic Design of SteelFrameswithEccentricBracings".Proceedingsof the4thEuropeanConferenceonSteelandComposite Structures - Eurosteel 2005. Maastricht, The Netherlands, June 8-10, 2005,VolumeC,Part.5.2,65-73.

[21] Stratan, A. and Dubina, D. (2004). "Eccentrically braced dual steel frames with removablelink". Improvement of Building’s Structural Quality by New technologies, Outcome of theCooperative Action, Final scientific report, September 2004. A.A. Balkema Publishers, ISBN04-1536-610-0.p.111-116.

[22] SabauG.A.,Poljansek,M.,Taucer,F.,Pegon,P.,MolinaRuiz,F.J.,Tirelli,D.,Viaccoz,B.,Stratan,A.,Dubina,D.,andIoanChesoan,A.(2014)."Seismicengineeringresearchinfrastructuresforeuropeansynergies.Full-ScaleExperimentalValidationofaDualEccentricallyBracedFramewith Removable Links (DUAREM)". EUR - Scientific and Technical Research Reports.Publications Office of the European Union. JRC93136.http://publications.jrc.ec.europa.eu/repository/handle/111111111/34024.147p.

[23] Stratan, A., Ciutina, A. and Dogariu, A. (2006). "Seismic-resistant steel structures withremovabledissipativeelements".Proc.of the10thNat. and4th Int.conf. "Planning,design,constructionandtheconstructionindustry".EditorsR.Folic,V.Radonjanin,M.Trivunic.NoviSad,November22-24,2006.pp:559-566.ISBN86-7892-016-5

[24] Stratan, A., Ioan, A., Dubina, D., Poljanšek, M., Molina, J., Pegon, P., Taucer, F., Sabău, G. (inprint)."Large-scaletestsonare-centringdualeccentricallybracedframe",8thInternationalConferenceonBehaviorofSteelStructures inSeismicAreasSTESSA2015,Shanghai,China,July1-3,2015.Paperno.111.

[25] Ioan,A.,Stratan,A.,Dubina,D.,D’Aniello,M.,Landolfo,R.(inprint)."Seismicperformanceandre-centring capability of dual eccentrically braced frames with replaceable links", 8thInternational Conference on Behavior of Steel Structures in Seismic Areas STESSA 2015,Shanghai,China,July1-3,2015.Paperno.113.

45

2.2 COLD-FORMEDSTEELPITCHED-ROOFPORTALFRAMESWITHBOLTEDJOINTS

The objective of the research carried out was to evaluate the performance of pitched roof cold-formed steel portal frames of back-to-back channel sections and bolted joints. Three differentconfigurations of ridge and eaves joints were tested. The behaviour and failure mechanisms ofjointswereobservedinordertoevaluatetheirstiffness,strengthandductility.Jointsbetweencold-formed members with bolts in the web only result in a reduction of joint moment capacity andpremature web buckling. The component method was applied in order to characterise the jointstiffnessandmomentcapacityforthepurposeofframeanalysisanddesign.Theinfluenceofjointscharacteristics on the global frame response under lateral (seismic) loads was analysed byconsideringthreeconnectionmodels.Full-scaletestswereperformedoncold-formedpitched-roofportal frames. Experimental observations and comparison to numerical predictions of frameresponsearepresented.

PreviousstudiesbyLimandNethercot(2004)andChungandLau(1999)showedthatboltedjointsin cold formed steel portal frames have a semi-rigid behaviour. Also, these types of joints arepartially resistant (Lim and Nethercot, 2003). When bolts are installed only on the web of cold-formed section, the local buckling is made more critical by stress concentrations, shear lag andbearingdeformationsaroundboltholes(DunduandKemp,2006),reducingthemomentresistancewellbelowthemomentresistanceoftheeffectivecross-section.Incaseofusualcold-formedsteelsections,bothtestsandnumericalsimulationsshowthatelastic-plasticelongationofbolt-holesisby far the most important component controlling the stiffness and capacity of such type ofconnections(LimandNethercot,2004,Yuetal.,2005).Thecontributionofothercomponents,suchas flanges in tension and compression due to bending action, and the web in shear due totransverseactionissignificantlylower.

Theglobalbehaviourofcold-formedsteelportalframesofboltedjointswasstudiedexperimentallyby Lim (2001), Dundu and Kemp (2006), and Kwon et al. (2006). All these studies providedevidenceofthecrucialimportanceofjointperformanceontheglobalresponseofframes.

In present research, the influence of joint characteristics on the global behaviour of cold-formedpitched-roofportalframesisinvestigated.Anexperimentalprogramonridgeandeavesjointswascarriedout.Detailedresultsonjointbehaviourarereportedelsewhere(Dubinaetal.,2004).Basedon experimental results, a calculation procedure based on the component method (EN1993-1-8)was adapted to cold-formed joints. Joint stiffness and moment capacity obtained using thecomponent method isused to develop a jointmodel for global structuralanalysis.Two full-scaletests on cold-formed pitched-roof portal frames with bolted joints were performed, with theprimaryobjectivetoassesstheirperformanceunderhorizontal(seismic)loading.Theresultsoftheexperimental investigation are presented and experimental response is compared to analyticalpredictionsofframeresponse.

2.2.1 TESTSONJOINTSPECIMENS

Inordertobeabletodefinerealisticspecimenconfigurationsasimplepitchedroofportal framewasfirstdesignedwiththefollowingconfiguration:span12m;bay5m;eavesheight4mandroofangle10°.ThisframewassubjectedtoloadscommonintheRomaniandesignpractice:selfweight

0.35kN/m2(withapartialsafetyfactorofULS=1.1fortheultimatelimitstate),technologicalload

0.15 kN/m2 (ULS=1.1) and snow load 0.72 kN/m2 (ULS=2.0). These loads were totallingapproximately 10 kN/m uniformly distributed load on the frame. The frame was analysed anddesignedaccordingtoEN1993-1-3(2001)rules.Thesizeofkneeandridgespecimensandtestingsetupwerechosentoobtainintheconnectedmembersadistributionofbendingmomentsimilartotheoneobservedinthedesignedstructure.

46

Theelementsof theportal frameweremadefromback-to-backbuiltupsectionsmadeofLindabC350/3.0 profiles (yield strength fy=350 N/mm2). Using these cross section dimensions, threealternativejointconfigurationsweredesigned(seeFigure39andFigure40),usingweldedbracketelements(S235:fy=235N/mm2)

Figure39.ConfigurationsofRidgeJoints

Figure40.ConfigurationsofKneeJoints.

Theconnectingboltsaresubjectedtoshearandtheirdesignwascarriedoutassumingtherotationof the jointaroundthecentroidof theboltgroupanda lineardistributionof forces ineachbolt,proportional to their distance from the centre of rotation. The joints weredesigned to resist thebendingmomentinthebeam,atthelocationofthecentroidoftheboltgroup.

One group of specimens (KSG and RSG) used spaced built-up gussets. In this case, bolts wereprovidedonlyontheweboftheC350profile.Intheothercases,wheretwodifferentdetailswereused for the connecting bracket – i.e. welded I sections only (KIS and RIS), and welded I sectionwithplatebisector(KIPandRIP),respectively-boltswereprovidedonthewebonly,orbothonthewebandtheflanges.JointswhereboltswereprovidedonthewebandonflangesweredenotedbyFBletters(seeTable9).

-RIS-

1460

1-1

1 (FB-only) (FB-only)

1460

2

31

3

2

-RIP-

3-3

-RSG-

1460

2-2

(FB-only)

235

4

1747

235

4

17471747

235

4

-KIP-

(FB-only)

-KSG--KIS-

47

Table9.TestsonJointspecimens.

Elementtype Code LoadingtypeRIS(RidgeconnectionwithISimpleprofile)

RIS-FB-M MonotonicRIS-FB-C1* Cyclic:modifiedECCSRIS-FB-C2* Cyclic:lowcyclefatigue

RSG(RidgeconnectionwithSpacedGusset)

RSG-M MonotonicRSG-C1 Cyclic:ECCSprocedureRSG-C2 Cyclic:ModifiedECCS

RIP(RidgeconnectionwithIprofileandend

Plate)

RIP-M MonotonicRIP-M MonotonicRIP-C1 Cyclic-ECCSproc.

KSG(KneeconnectionwithSpacedGusset)

KSG-M MonotonicKSG-C1 Cyclic-ModifiedECCSKSG-C2 Cyclic-Lowcyclefatigue

KIS(KneeconnectionwithISimpleprofile)

KIS-M MonotonicKIS-FB-M* MonotonicKIS-FB-C* Cyclic-ModifiedECCS

KIP(KneeconnectionwithIprofileandend

Plate)

KIP-M MonotonicKIP-FB-M* MonotonicKIP-FB-C* Cyclic-ModifiedECCS

*FBSpecimens(RIS,RIP,KIS,KIP)withsupplementaryboltsontheflange

Monotonicandcyclicexperimentswereperformedforeachspecimentypology,allspecimensbeingtested statically. Figure 41 shows the test setup and specimen instrumentation. In the kneeconnection tests, a short tie was used to prevent vertical displacements of the joint. Formonotonically loaded specimens the loading velocity was approximately 3.33 mm/min, and the"yield" displacement (vy) was determined according to the ECCS (1985) procedure, as thedisplacementcorrespondingtotheintersectionoftheinitialstiffnesslineandanotherlinewithaslopeof10%oftheinitialstiffness(seeFigure42a).Forthecyclictestsseveralalternativeloadingprocedures were used: (1) the standard ECCS cyclic procedure (see Figure 42b), (2) a modifiedcyclicprocedure,suggestedbytheauthors,whichisbasedontheECCSproposal(seeFigure42c)and(3)acyclicprocedureforlowcyclefatigue.TheECCSloadingprocedureconsistsoffourinitial

cyclesintheelasticrange,followedbygroupsofthreeinelasticcyclesat2vy,4vy,6vy,etc.Theinelastic demand imposed on cold-formed specimens following this procedure is too severe, the

specimen failing during the first inelastic cycle at 2vy. To overcome this problem, a modifiedprocedurewasused,thatconsistedoffourinitialcyclesintheelasticrange,followedbygroupsof

threeinelasticcyclesat1.2vy,1.4vy,1.6vy,etc.

Figure41.Loadingschemeandinstrumentation.

I1

I2I3

Drel2

I4

Drel4

Drel3

Drel1

Dgl ft

Dlat jos

Dlatsus

Dgl sp

Fact

Dact

48

(a)

(b)

(c)

Figure42.ECCSprocedurefordeterminingtheyielddisplacement(a);ECCSloadingprocedure(b)andmodifiedloadingprocedure(c)

Themonotonictestsidentifiedfailuremodesofthedifferentjointtypologies.Allspecimenshadafailure due to local buckling of the cold formed profiles; however two distinctive modes wereidentifiedforspecimenswithflangeboltsandthosewithout(Figure43;Figure44).

(a)

(b)

Figure43.FailureofridgespecimensRIP-M(a)andRIS-FB-M(b).

Ifnoboltsareprovidedontheflangeofprofiles,initiallyminorbearingelongationoftheboltholeswereobserved,thefailurebeingduetostressconcentrationinthevicinityofouterboltrow.Theresulting concentration of compressive stress in the web of the C profile causes in the ultimatestagelocalbucklingfollowedsuddenlybyweb-inducedflangebuckling.Thisphenomenonoccurredin asimilar way in thecaseof RSG and KSG specimens.No important differences were observedbetweenspecimenswherenoboltswereprovidedontheflanges.Inthecaseofthespecimenswithflangebolts,thestressesconcentratedinthevicinityoftheouterboltrowontheflange.Inthiscasenoinitialelongationoftheboltholeswereobserved;thebucklingwasfirstlyinitiatedintheflange,andonlylaterwasextendedintotheweb.

To account for the flexibility of the bolted connection in structural analysis, two models arepossible:one whichconsiders both connections independently (see Figure45a), and asimplifiedone, which considers the characteristics of the connection concentrated in one joint only (seeFigure45b).Theformerisbelievedtorepresentmoreexactlytherealbehaviouroftheassembly,while the latter has the advantage of simplicity. Similar models can be used for knee jointconfigurations. Moment-rotation relationships characterising the connection response werederivedforboththe leftandrightridgeconnections(beamandcolumnconnectioninthecaseofkneejoins).Momentswerecomputedattheendofthebracket.Thecorrespondingrelativerotation

betweenthebracketandtheconnectedelementqC*was determined fromacquireddata,soas to

y

y/10

vmax

uv

PmaxP

vvy

Py

-4

-2

0

2

4v/v

y

time

-4

-2

0

2

4v/v

y

time

49

represent both the flexibility of the connection (due to bolt bearing) and post-bucklingdeformations in the element (Dubina et al., 2004). For the simplified joint representation (as inFigure 45b), both moment (Mj) and rotations were considered at the intersection of the elementcentrelines.

(a)

(b)

Figure44.FailureofkneespecimensKIS-M(a)andKIS-FB-M(b).

(a)

(b)

Figure45.Twopossiblemodelsforridgejoints:detailed(a)andsimplified(b).

ComparativeexperimentalcurvesforridgeandkneeconnectionsarepresentedinFigure46aandFigure47.Therearenosignificantdifferencesamongthespecimenswithoutflangebolts(RSG-M,RIP-M, and KSG-M, KIS-M). This could be explained by the higher stiffness and capacity of theconnecting bolts compared to the other components of the joint. On the other hand, there is animportantgaininloadbearingcapacityandtheinitial jointstiffnesswhenboltsareinstalledalsoontheflanges,althoughthisjointtypeismoredifficulttofabricate(RIS-FB-MandKIS-FB-M).

InTable10theyieldandultimaterotation(qC,y*;qC,u*),theinitialstiffness(KiniC),andthemaximum

bendingmoment(MC,max)arepresentedandcomparedforallmonotonicallytestedspecimens,forthefailedconnection.Theinitialstiffnesswasdeterminedbyalinearfitofmoment-rotationvaluesbetween0.25and0.9ofthemaximummoment.Thelower-boundlimit(0.25)waschosendifferentfromzeroinordertoeliminatetheeffectofinitialslipduetotoleranceofbolt-holes.Thisinitialslipis believed to be ineffective in the real structure, due to loading-unloading cycles under serviceloads.Theupper-boundlimit(0.9)wasconsideredempiricallyasalimitofelasticresponseoftheconnection.Yieldrotationwasdeterminedasthepointontheinitialstiffnesslinecorrespondingtomaximum moment. Ultimate rotation was defined as the one corresponding to a 10% drop of

bracket

connection

element element

joint

bracket

50

momentcapacityrelativetothemaximummoment.Figure46bpresentsgraphicallydeterminationtheinitialstiffness,yieldrotationandultimaterotationfortheRIS-FB-Mspecimen.

Obviously, the specimens with unbolted flanges that failed prematurely by web buckling due tostress concentration around the outer bolt rows, would be the weakest part of portal frames.Consequently,thisjointtypologyisnotrecommendedtobeusedinpractice.

Table10.Monotonicresults:parametersofconnectionmoment-rotationcurves.

Specimen KiniC kNm/rad

qC,y* Rad

qC,u* rad

MCmax kNm

RSG-M 4891.3 0.021 0.034 1.6 77.1RIS-FB-M 6011.1 0.017 0.025 1.4 108.0

RIP-M 5806.8 0.018 0.028 1.6 74.3RIP-M2 6541.2 0.012 0.013 1.1 72.9KSG-M 6031.6 0.009 0.023 2.5 53.3KIS-M 4115.0 0.020 0.033 1.6 78.4

KIS-FB-M 6432.3 0.016 0.029 1.8 102.9KIP-M 7863.9 0.010 0.019 2.0 90.0

KIP-FB-M 6956.5 0.015 0.025 1.6 116.7

(a)

(b)

Figure46.Comparativeresultsfrommonotonictestsforridgejoints(a)andgraphicalrepresentationofinitialstiffness,yieldandultimaterotationsfortheRIS-FB-Mspecimen(b).

Inthecaseofcyclicloading,failureofspecimensstartedbyelongationofboltholes.Comparedtomonotonicloading,inthecaseofcyclicloadingthephenomenonwasamplifiedduetotherepeatedandreverseloading.However,thefailureoccurredalsobylocalbuckling,asinmonotonictests,butat the repeated reversals, the buckling occurred alternately on opposite sides of the profile. Thisrepeated loading caused the initiation of a crack at the corner of the C profile, in 2-3 cyclesfollowingthebuckling,closedtothepointwherethefirstbucklingwavewasobservedintheflange.

Thecrackgraduallypropagatedthroughtheflangeandweb,causinganimportantdecreaseoftheloadbearingcapacityineachconsecutivecycle.

51

Figure47.Comparativeresultsfrommonotonictestsforkneejoints.

Figure48.Comparativeresultsfromcyclictestsonridgespecimens.

The hysteretic M-q curves show a stable behaviour up to the yield limit (qC,y*) with a suddendecrease of the load bearing capacity afterwards (Figure 48 and Figure 49). Therefore the lowductilityofthespecimensmustbeunderlinedagain.Further,thecyclesshowtheeffectofslippageinthejoint(i.e.pinching)andstrengthdegradationinrepeatedcycles.Strengthdegradationismostsignificantinthefirstcycle,whileintheconsequentonesthebehaviourismorestable.

In order to obtain connection characteristics under cyclic loading, an unstabilised envelope wasfirst determined, by considering the points corresponding to maximum moment in each cycle.Based on envelope curves, connection strength, rotation capacity and ductility have beendetermined following a procedure identical to the one used in the case of monotonic specimens,andarereportedinTable11.Again,jointswithoutflangeboltswereweaker.

Thecomponentmethodisageneralprocedureforthedesignofthestrengthandstiffnessofjointsinbuildingframes,andisimplementedinEN1993-1-8.2003.Theprocedureisprimarilyintendedfor heavy-gauged construction. Its application to joints connecting light-gauge members isinvestigatedherein.

52

Figure49.Comparativeresultsfromcyclictestsonkneespecimens.

Table11.Cyclicresults:parametersofconnectionmoment-rotationcurves.

Specimen KiniC kNm/rad

qC,y* rad

qC,u* rad

MCmax kNm

RSG-C1 5060.0 0.017 0.028 1.7 78.8-5400.3 -0.019 * * -76.8

RSG-C2 4502.9 0.018 0.028 1.6 76.0-2792.5 -0.029 -0.036 1.2 -78.1

RIS-FB-C1 * * * * 106.9* * * * -108.6

RIS-FB-C2 * * * * 100.2* * * * -111.5

RIP-C1 6642.1 0.014 * * 73.8-6585.1 -0.013 * * -74.7

KSG-C1 5395.5 0.013 0.022 1.7 82.5-6672.4 -0.015 -0.028 1.9 -90.9

KSG-C2 5067.6 0.014 0.022 1.5 84.5-4684.1 -0.014 -0.017 1.2 -76.8

KIS-FB-C 6914.7 0.014 0.021 1.5 102.3-9201.5 -0.012 -0.023 2.0 -114.4

KIP-FB-C 10051.8 0.012 0.026 2.1 102.2-8193.5 -0.011 -0.021 1.9 -105.1

*resultsnotavailableduetofaultydataacquisition

Applicationofthecomponentmethodrequiresthefollowingsteps(Jaspartetal.,1998): identificationoftheactivecomponentswithinthejoint evaluationofthestiffnessandstrengthofindividualcomponents assemblyofthecomponentsinordertoevaluatestiffnessandstrengthofthewholejoint

Basedontheconclusionsoftheexperimentalprogramme,thepresentstudyinvestigatesonlyjointswith both web and flange bolts (RIS-FB-M, KIS-FB-M, andKIP-FB-M).Qualitative FEMsimulation(seeFigure50)showedthatinthecaseofspecimenswithboltsonthewebonlythereisastressconcentrationintheweb,whichcausesprematurelocalbucklingfailure.TheFEMsimulationalsodemonstratedthatloaddistributionintheboltsisnotlinear.Infact,duetomemberflexibilityand

53

localbuckling,theconnectedmembersdonotbehaveasrigidbodies,andthecentreofrotationofwebboltsdoesnotcoincidewiththecentroidofwebbolts.Thecentreofrotationoftheconnectionisshifted towards theouter bolt rows (see Figure 51), whosecorresponding force is an orderofmagnitudehigherthantheforceintheinnerbolts.Consideringthisobservation,onlytheouterboltgroup was considered for determination of connection characteristics using the componentmethod.Thisassumptionsignificantlydifferincomparisonwiththebehaviourmodelsconsideredin the papers of the list of reference, which, all, consider the centroid of the bolt group as therotationcentre.

The centre of compression of the connection was considered at the exterior flange of the cold-formed member (see Figure 51), similarly with the model used for design of bolted connectionswithangleflangecleatsinEN1993-1-8(2003).Thereareatotaloffourboltrows,ofwhichthreeboltrowsareinthe"tension"zone.Thefollowingcomponentswereidentifiedandusedtomodeltheconnectionstiffnessandstrength: Cold-formedmemberflangeandwebincompression.Onlythestrengthofthiscomponentwas

considered,whilestiffnesswasconsideredinfinite(similarlywithLimandNethercot,2004) Boltsinshear Boltsinbearingonthecold-formedmember Boltsinbearingonthebracket

(a)

(b)

Figure50.Stressconcentrationinthecaseofspecimenswithwebboltsonly(a),andbothwebandflangebolts(b).

Figure51.BoltGroupsconsideredinanalysis.

ThestiffnessandstrengthofallthesecomponentsarereadilyavailableinEN1993-1-8,onlyminoradjustmentsbeingrequiredforthecaseoftheparticularcaseconsideredhere.Inordertofacilitatecomparisonwiththeexperimentalresults, themeasuredgeometricalcharacteristicsandstrength(yieldstressfy=452N/mm2,andtensilestrengthfu=520N/mm2)wereconsideredinthecaseofthe cold-formed member. Nominal characteristics were used for the bracket and bolt

outerbolt group

innerbolt group

h3 h2 h1

centre ofcompression

54

characteristics, as experimental data was not available. Partial safety factors equal to unity wereconsideredinallcases.

Only three components were considered to contribute to thestiffness of the connection: bolts inshear(denotedkv,fforflangeboltsandkv,wforwebbolts),boltsinbearingoncold-formedmember(denotedkb,cffforflangeboltsandkb,cfwforwebbolts),andboltsinbearingonthebracket(denotedkb,bfforflangeboltsandkb,bwforwebbolts),seeFigure52a.FormulasfordeterminationofstiffnesscoefficientsareavailableinEN1993-1-8.Foreachoftheboltrowsr,aneffectivestiffnesscoefficientkeff,r is determined, by combining the individual stiffness coefficients using the followingrelationship(EN1993-1-8,seeFigure52b):

i ri

reff

k

k

,

, 1

1 (15)

Theeffectivestiffnesscoefficientsoftheboltrowsin"tension"zonearereplacedbyanequivalentspringkeq(EN1993-1-8,seeFigure52c):

eq

rrreff

eqz

hk

k

,

(16)

wherehristhedistancebetweenboltrowrandthecentreofcompression;zeqisdeterminedusingEq.(17).

rrreff

rrreff

eqhk

hk

z,

,2

(17)

(a) (b)

(c) (d)

Figure52.Mainstepsintheassemblyofcomponentsfordeterminationofconnectionstiffness.

Finally,theinitialconnectionstiffnessisdeterminedas(seeFigure52d):

,4effk

,3effk

,2effk

,1effk

Bolt row 4

Bolt row 3

Bolt row 2

Bolt row 1

fvk ,

cffbk ,bfbk ,

wvk ,

,b cfwkbwbk ,

fvk ,

cffbk ,bfbk ,

cfwbk ,bwbk ,

jS

,4effk

eqk

z

55

2

,

1

eq

j ini

ii

E zS

k

(18)

Themomentresistanceoftheboltedconnectionwasdeterminedusingatwo-stepprocedure.Inthefirst step, only components related to bolt resistance were included in order to determine themoment resistance of the bolted connection MbC,Rd. In a second step, the connection momentresistancewasobtainedastheminimumofthemomentresistancesoftheboltedconnectionMbC,Rdandtheconnectedcold-formedmemberMbeam,Rd:

, , ,min ,bC Rd C Rd beam RdM M M (19)

The adopted approach for determination of connection moment resistance allows to easilydetermineiftheconnectionisfull-strengthorpartialstrength.

Themomentresistanceoftheboltedconnectionwasdeterminedas(EN1993-1-8):

, ,

bC Rd tr Rd r

r

M F h (20)

whereFtr,Rdistheeffectivetensionresistanceofboltrowr(minimumvalueofcomponentsrelatedtoboltrowr);hristhedistancebetweenboltrowrandthecentreofcompression.

The moment resistance of the cold-formed member Mbeam,Rd was determined using measuredgeometricalandmechanicalcharacteristics,usingeffectivecross-sectionmodulus.

It was considered appropriate to use a linear distribution of forces on bolts in the case of aconnection to light-gauge members. Therefore, the effective tension resistance of bolt rows waslimitedaccordingtothefollowingrelationship:

, 1,

1

rtr Rd t Rd

hF F

h (21)

where Ft1,Rd is the effective tension resistance of bolt row 1 (farthest from the centre ofcompression);h1isthedistancebetweenboltrow1andthecentreofcompression.

Table4andTable5presentresistanceandstiffnessofboltrows.Theweakestcomponentofflangeboltsisbearingoncoldformedmember,whileinthecaseofwebboltsitisbearingonbracket(seeTable4).Thedifference isduetothe factthatboltsare insimpleshearonflangesandindoubleshearonweb,aswellasduetodifferentnumberofboltsonflanges(4boltsperrow)andweb(2bolts per row). The main contribution to the flexibility of the connection is bearing on the cold-formedmember,aswellasbearingonbracketinthecaseofwebbolts(seeTable5).

Theconfigurationoftheoutergroupofboltsbeingthesameinthecaseofallthreespecimenswithweb and flange bolts (RIS-FB-M, KIS-FB-M, KIP-FB-M), a single set of analytical connectionproperties were determined. A comparison of experimental vs. analytical characteristics ofconnections(stiffnessandmomentresistance)ispresentedinTable6.Generallyafairagreementbetween experimental and analytical stiffness of the connection can be observed. Largerexperimentalvaluesofstiffnesscanbeexplainedbythefactthatthecontributionoftheinnerboltgroup was ignored in the analytical model. The stiffness of the connection is considerably lowerthantheEN1993-1-8limitsforclassificationofjointsasrigid(25EIb/Lb),whichamountsto25256kN/m(consideringthebeamspanLbequal toframespanandusinggrossmomentof inertiaIb).

56

Therefore,thesetypesofconnectionsaresemi-rigid,andtheircharacteristicsneedtobetakenintoaccountintheglobaldesignofframe.

Table12.Resistanceofconnectioncomponents.

Boltrow

Component Bolt-rowresistance

Ftr,Rd,kN

Boltsinshear,kN

Boltsinbearingonthecold-formedmember,

kN

Boltsinbearingonthebracket,

kN1 361.4 290.6 527.0 290.62 361.4 290.6 288.0 288.03 361.4 290.6 288.0 288.04 361.4 290.6 527.0 290.6

Table13.Stiffnessofconnectioncomponents.

Boltrow

Component Bolt-roweffectivestiffnesskeff,r,

mmBoltsin

shear,mmBoltsinbearingonthecold-formedmember,

mm

Boltsinbearingonthebracket,

mm1 2.286 0.7785 1.3886 0.40952 2.286 0.7785 0.7714 0.33133 2.286 0.7785 0.7714 0.33134 2.286 0.7785 1.3886 0.4095

Table14.Experimentalvs.analyticalconnectioncharacteristics.

Specimen Initialstiffness,KiniC[kNm/rad] Momentresistance,MC,[kNm]experimental analytical experimental analytical

RIS-FB-M 6011 5224 108.0 117.8KIS-FB-M 6432 5224 102.9 117.8KIP-FB-M 6957 5224 116.7 117.8

The moment resistance of the bolted connection MbC,Rd determined by the component methodamountedto193.9kNm,whichwaslargerthanthemomentresistanceofthecold-formedmemberMbeam,Rd,amounting117.8kNm.Therefore,thistypeofconnectionisafull-strengthone.Thiswasdemonstrated also by the experimental results, failure mode being local buckling of the cold-formedmember.

2.2.2 TESTSONFULL-SCALEPITCHED-ROOFPORTALFRAMES

Followingexperimentaltestsoncold-formedjoints,twofull-scaletestsonframeswereperformed.Frame dimensions were chosen identical to the ones in the initial design used to establish thedimensionsoftestedjoints.Consideringthepoorperformanceofjointswithwebboltsonly,RIS-FBand KIS-FB configurations (with both web and flange bolts) were used for frame construction.Pinnedsupportswereusedatthecolumnbases.Theobjectiveofthefull-scaletestswastoassessthe performance of pitched-roof cold-formed portal frames with moment-resisting joints underlateralloading,withparticularemphasisonearthquakeloading.

The test setup consisted of two frames in upward position, located 1.5 m apart. Tie bracing wasprovided between the two frames in order to provide out-of plane stability. The purlins wereinstalled on the girders, but no side rails were provided on the columns. The schematicrepresentationoftestsetupisshowninFigure53.Areactionframewasusedtoapplylateralload.

57

Figure53.Experimentaltestsetupforfull-scaletests.

Inthecaseofthefirsttest(C1),onlylateralloadingwasapplied.Forthesecondtest(C2),gravityloadingcorrespondingtoseismicdesignsituation(permanentanda0.3fractionofthesnowload)was applied, followed by increasing lateral load up to failure. Total gravity loading amounted to31.2kNperframe,andwasappliedusing30corrugatedsteelsheetslaidonthepurlins.Aloadcellwas used in order to measure lateral load applied through a hydraulic jack. Frames wereinstrumented with displacement transducer to measure lateral in-plane and out-of planedisplacementsattheeaves,deflectionsattheridge,aswellasconnectionrotations.

Experimental tests on ridge and eaves joints showed that bolted connections of back-to-backchannelcold-formedmembersaresemi-rigid,evenwhenboltsareprovidednotonlyontheweb,but also on the flanges of the channel section. Therefore, deformations can be underestimated ifconnections are assumed rigid for global frame analysis. In order to assess the influence ofconnection stiffness and post-buckling resistance, three frame models were analysed (see Figure54). A nonlinear static analysis under increasing lateral load was applied to the models, and theresultswerecomparedtoexperimentalones.

The first model was a conventional model, where connections were considered rigid. Nominalgeometricalcharacteristicswereusedtomodelmembers.Finitedimensionsofbracketsweretakeninto account. Local buckling of members was modelled by rigid-plastic hinges located at theextremitiesofcold-formedmembers.ananalyticallydeterminedmomentcapacity(Mc=117.8kNm)wasconsidered.

The second model (M2, see Figure 54b) was obtained from model M1 by adopting an elasticperfectly-plastic model of the connection moment-rotation response. The initial stiffness(KiniC=5224 kNm/rad) and moment capacity (Mc=117.8kNm) were the ones obtained using theanalyticalproceduredescribedabove(seeTable6).

Inthecaseofthethirdmodel(M3,seeFigure54c),theelasto-plasticmodelwasenhancedfollowingtwo directions. The first one was related to connection behaviour under small loads, whenexperimentalevidenceshowedaverystiffinitialresponse.Thisresponseisattributedtowedgingandfrictionbetweenthecold-formedprofilesandthebracket.Consequently,arigidresponsewasassumedbefore"slipping"uptomomentMs(seeFigure55a).Thevalueofthe"slipping"momentMs was estimated based on experimental results, a value of 15% from the connection momentcapacitybeingadopted.Followingthe initialrigidbehaviour, theconnectionmodelconsistsofanelastic response at the initial stiffness KiniC (determined using the component method), up to theconnectionmomentcapacityMC.

58

(a)

(b)

(c)

Figure54.Consideredstructuralmodels:rigidconnections-M1(a),elastic-perfectlyplasticconnections-M2(b),anddegradingconnections–M3(c).

The second enhancement of the model consisted in post-elastic response. The plastic rotation

(plateau)wasdeterminedassuminganultimaterotationqCuequalto1.4timestheyieldrotationqCy.Thesofteningbranchwasdeterminedbyconsideringadropofmomentcapacityto50%fromthe

maximum one, at a rotationqCr of 4.0 times the yield rotation (Figure 55a). The same moment-rotationcharacteristicswereusedforallconnections(forbothbeamsandcolumns).Theinfluenceofaxialforceonthestiffnessandmomentresistanceoftheconnectionwereignored.Acomparisonbetweentheanalyticalandtheexperimentalmoment-rotationcurvesisshowninFigure55b.

(a)

(b)

Figure55.ParametersoftheM3connectionsmodel(a),andexperimentalandanalyticalM3modelofconnectionmoment-rotationrelationship(b).

Figure56ashowsaglobalviewof theC1 frame(testedunderhorizontal loadingonly)after thetest.Theframeresponseduringthetestwascharacterisedbyanalmostlinearresponseuptothefirst local buckle of the beam at the connection 2 (see Figure 56b, Figure 57, and Figure 58a),followed by a rapid loss of global frame resistance. The final collapse mechanism consisted ofhingingofthebeamatconnections2and5(seeFigure58b)neartheeaves.

A comparison of the experimental and numerical lateral force – deformation curves for the C1frameisshowninFigure57.Theforcecorrespondstooneofthetwoframesfromtheexperimentalsetup,assumingtheforceequallydistributedbetweenthetwoframes.Itcanbeobservedthattherigidmodel(M1)providesagoodapproximationoftheinitialresponseoftheframeuptolateralforces of about 10 kN. At larger forces, model M2, with semi-rigid connections, provide a betterapproximation of the experimental response. The M3 model, incorporating both the initial rigidresponse and subsequent semi-rigid behaviour shows the best agreement to the experimentalresults.Thesamepatternofmemberhingingasintheoneobservedintheexperimentisobtainedforthenumericalmodel(seeFigure58bforthecaseoftheM3model).TheM3modelcaptureswellthe initial and post-buckling response. M2 model overestimates lateral deformations. All threemodelsslightlyunderestimatetheglobalframeresistance.

0

20

40

60

80

100

120

140

0 0.02 0.04 0.06 0.08 0.1

q c , rad

Mc,

kN

m

KIP-FB-MKIS-FB-MRIS-FB-Mmodel

59

(a)

(b)

Figure56.C1Frame:globalview(a)andlocalbucklingoftheleftbeamconnection(b).

Figure57.FrameC1:Experimentalvs.numericallateralforce-deformationcurves.

(a)

(b)

Figure58.FrameC1:positionoflocalbucklingobservedexperimentally(b)andinthenumericalmodel(c)

0

10

20

30

40

50

60

70

80

0 100 200 300 400 500

D, mm

F,

kN

expexp: conn 2 hingingexp: conn 5 hingingM1 modelM2 modelM3 modelM3: conn 2 hingingM3: conn 5 hinging

60

In the case of the C2 frame, gravity loading corresponding to seismic design situation was firstapplied, followed by increasing lateral loading up to complete failure of the frame. Figure 59ashowsaviewoftheframeduringloading.Theglobalforce-deformationresponsewasverysimilarto the frame C1 up to 10-15 kN lateral loading. For larger lateral loading, the stiffness of the C2frame was slightly larger than the one of the C1 frame. However, the global resistance underhorizontal loadingwassmaller inthecaseof theC2frame.Theultimatecapacitywasattainedatthe first local buckle in the beam near the right eaves (connection 5, see Figure 61a), when thelateralforceresistancedroppedsuddenly.Itwasfollowedbyacombinedlocalbucklingandlateral-torsionalbucklingofoneofthecolumnsatthemid-height(seeFigure59aandFigure61a).Finally,localbucklingofthebeamatthelefteaveswasobserved(atconnection2,seeFigure61a).

(a)

(b)

Figure59.C2Frame:globalview(a)andlocalbucklingoftherightbeamconnection(b).

A comparison of the experimental and numerical lateral force – deformation response of the C2frame is shown in Figure 60. As in the case of frame C1, The M1 model (with rigid connections)providesagoodapproximationoftheinitialresponseoftheframe,uptolateralforcesofabout10kN.Forlargerforces,theM2model,accountingforsemi-rigidconnectionresponse,showsabetterapproximation of experimental response. The M3 model shows the best agreement between thenumericalandexperimentalresults.Allnumericalmodelsoverestimatetheglobalframeresistanceunderlateralloading.Therearetwofactorsthatarebelievedtohavecontributedtothissituation:(1) the numerical model did not consider buckling of the column and (2) the influence of axialforces was neglected when determining connection moment resistance. Higher axial forces arepresentintherightcolumnundercombinedeffectofgravityloadingandlateralloadingduetotheeffect of overturning. While the location of the first local buckle was correctly predicted by thenumericalmodel(atconnection5,seeFigure61andFigure60b),columnhinging(duetocombinedlocal and flexural-torsional buckling) observed in the experimental test was not confirmed bynumericalmodels.Columnhingingcanbeexplainedbyneglectedinfluenceofaxialforce,combinedwiththeeffectofno lateralrestrainingatcolumnflangesbysiderails.Bothof theseeffectswerepresentintheexperimentalsetup,butnotinthenumericalmodel.

61

Figure60.FrameC2:comparisonofexperimentalandnumericallateralforce–deformationcurves.

Figure61.FrameC2:positionoflocalbucklingobservedexperimentally(b)andinthenumericalmodel(c).

ItcanbeconcludedthattheM3modelseemstoprovidethebestagreementwiththeexperimentalresults, if initialstiffness, lateralresistance,andpost-bucklingresponseareenvisaged.Theglobalframeresistanceunderlateralloadsdropsquicklyafterthefirstlocalbuckling,whenthemaximumforce is reached. This behaviour is attributed to the similar rapid drop in moment resistance ofcold-formedcross-sections,aswellastothelowredundancyoftheframe.Therefore,forpracticalcases,theresponsetothefirstlocalbuckleinmembersisimportant,whichcanbeestimatedusingasimplerframemodel,incorporatingonlythesemi-rigidconnectionresponse,eventuallyanelasticperfectlyplasticmodel.TheGlobalframestiffnessdeterminedusingthebilinearmoment-rotationcharacteristics obtained analytically by the component method is smaller than the experimentalstiffness. The real initial stiffness of the connection may be higher at low moments, due torestraining provided by flanges of the bracket element and/or by the inner bolt group. The M3connection model is capable of representing this higher stiffness and provides the closest matchbetweenexperimentalandnumericalframestiffnessvalues.

2.2.3 OUTCOMES

This study suggests that the classical calculation model for connections, assuming the centre ofrotationtobelocatedatthecentroidoftheboltgroupandalineardistributionoftheforcesoneachbolt,isinappropriateforframesfeaturingcold-formedmembers.Theforcedistributionisunequalduetotheflexibilityoftheconnectedmembers.Infact,theforceisanorderofmagnitudebiggerintheouterboltrowscomparedtomostinnerones.Aconnectionwithboltsonlyonthewebcauses

0

10

20

30

40

50

60

70

80

0 100 200 300 400 500

D, mm

F,

kN

expexp: conn 5 hingingexp: right column hingingM1 modelM2 modelM3 modelM3: conn 5 hingingM3: conn 2 hinging

(a)

(b)

62

concentrated forces in the web of the connected member and leads to premature web buckling,reducing the jointmomentcapacity.These typeofconnections arealwayspartial strength. If theloadbearingcapacityoftheconnectedbeamistobematchedbytheconnectionstrength,boltsontheflangesbecomenecessary.Theductilityoftheconnectionisreducedbothundermonotonicandcyclicloadsandthedesign,includingthedesignforearthquakeloads,shouldtakeintoaccountonlytheconventionalelasticcapacity.Becausethereisnosignificantpost-elasticstrength,therearenosignificant differences in ductility and capacity of cyclically tested specimens compared with themonotonicones.

The application of the component method implemented in EN1993-1-8 for determination ofconnectioncharacteristicsinthecaseofcold-formedmembersispossiblewithaminimumnumberofadjustments.Fortheparticularcaseofconnectionstudiedhere(withbothflangeandwebbolts),connection characteristics can be determined with a reasonable accuracy if only the outer boltgroup of bolts is considered. The components contributing to the stiffness and strength of theconnection are: cold-formed member flange and web in compression, bolts in shear, bolts inbearing on the cold-formed member, and bolts in bearing on the bracket. It is consideredappropriatetousealineardistributionofforcesonboltsinthecaseofaconnectiontolight-gaugemembers.

Theconnectionwithbothflangeandwebboltsissemi-rigidbutfull-strength.Thereforethedesignoflight-gaugeportalframeswiththeconsideredtypeofconnectionneedstoaccountforconnectionflexibility.Connectioncharacteristicsobtainedusingthecomponentmethod(EN1993-1-8)canbeeasily incorporated in the structural model, in order to obtain a realistic response under lateralforces.Aconnectionmodelisdevelopedthatcaptureswellboththebehaviourintheelasticrangeandinthepost-elasticrange.

Thoughadetailedmoment-rotationresponserepresentingtheinitialstiffness,momentresistanceandpost-bucklingresponseprovidesthemostrealisticglobalresponse,asimpleelasticstructuralanalysis modelling connection stiffness alone can be sufficient for design purpose. Cold-formedsteel pitched-roof portal frames of back-to-back channel sections and bolted joints arecharacterisedbyarapiddegradationofstrengthafterthefirstlocalbucklinginitsmembers.Thisbehaviour is attributed to the similar rapid drop in moment resistance of cold-formed cross-sections,as wellas to the lowredundancyof the frameanalysed in thisstudy.Therefore, for theconsideredframeconfiguration,globalstrengthmaybeestimatedattheattainmentofthemomentcapacity in the most stressed cross-section using an elastic structural analysis. Axial forces canreducemomentresistanceofcold-formedmembersandneedbetakenintoaccount.

2.2.4 PUBLICATIONS

[1] Dubina,D.,Stratan,A.,Nagy,Zs.(2009)."Full–scaletestsoncold-formedsteelpitched-roofportalframeswithboltedjoints".AdvancedSteelConstructionVol.5,No.2,pp.175-194.ISSN1816-112X.

[2] Dubina,D.,Ungureanu,V.,StratanA.,NagyZs.(2008)."Structuralperformanceofpitchedroofcold-formed steel frames of bolted joints". Acta Technica Napocensis. Section: CivilEngineering–Architecture.Nr.51,vol.1,2008,ISSN1221-5848,p.385-396.ProceedingsoftheInternationalConference–CONSTRUCTIONS2008,9-10May2008,Cluj-Napoca.

[3] Nagy,Zs.,Stratan,A.,Dubina,D.(2006)."Applicationofcomponentmethodforboltedcold-formedsteeljoints".Proc.oftheinternationalconferenceinmetalstructures,PoianaBrasov,Romania,September20-22,2006.Steel–anewandTraditionalMaterialforBuilding,DubinaandUngureanu(eds.),TaylorandFrancisGroup,London,ISBN0-415-40817-2,pp:207-215

63

[4] Dubina,D.,Stratan,A.,Ciutina,A.,Fulop,L.,Nagy,Zs.(2004)."Performanceofridgeandeavesjoints in cold-formed steel portal frames". Proc. of the 17th int. Specialty conf. "Recentadvancesanddevelopmentsincold-formedsteeldesignandconstruction",Orlando,Florida,USA,04-05Nov.2004.Univ.ofMissouri-Rolla,Ed.R.A.LaBoube,W-W.Yu,p.727-742.

[5] Stratan,A.,Nagy,Zs.,andDubina,D.(2006)."Cold-formedsteelpitched-roofportalframesofback-to-back plain channel sections and bolted joints". Proc. of the 18th int. Specialty conf."Recentresearchanddevelopmentsincold-formedsteeldesignandconstruction",Orlando,Florida,USA,Oct.26-27.Univ.ofMissouri-Rolla,Ed.R.A.LaBoube,W-W.Yu,p.351-365.

[6] Dubina, D., Stratan, A., Ciutina, A., Fulop, L., Nagy, Zs. (2004). "Monotonic and cyclicperformanceofjointsofcoldformedsteelportalframes",Thin-WalledStructures:AdvancesinResearch,DesignandManufacturingTechnology,Ed.J.Loughlan,IOP,ISBN0750310065,proceedings of the Fourth International Conference on Thin-Walled Structures,Loughborough,UK,22-24June2004.pp:381-388.

[7] Dubina, D., Ungureanu, V. and Stratan, A. (2008). "Ultimate design capacity of pitch-roofportal frames made by thin-walled cold-formed members". Proceedings of the 5thInternationalConferenceonThin-WalledStructures:RecentInnovationsandDevelopments,GoldCoast,Australia,18-20June2008,Vol.1,p.387-394,ISBN:978-1-74107-239-6.

[8] Dubina,D.,Stratan,A.,Nagy,Zs.(2007)."Full–scaletestingofcold-formedsteelpitched-roofportal frames of back-to-back channel sections and bolted joints". Proc. of the Sixth Intern.Conf. on Steel and Aluminium Structures, 24-27 July 2007, Oxford, UK. Beale, R.G. (Ed.),OxfordBrookesUniversity,ISBN:978-0-9556254-0-4,pp.931-939.

[9] Dubina, D., Nagy, Zs., Stratan, A., Fulop, L. and Ungureanu, V. (2006). "Design assisted bytesting of cold-formed steel frame structures". Proc. of the 10th Nat. and 4th Int. conf."Planning,design,constructionandtheconstructionindustry".EditorsR.Folic,V.Radonjanin,M.Trivunic.NoviSad,November22-24,2006.pp:203-212.ISBN86-7892-016-5.

[10] Dubina,D.,Stratan,A.,Ciutina,A.,Fulop,L.,andNagy,Zs.(2004)."Monotonicandcyclictestson cold-formed steel joints". COST C12 Improving Buildings Structural Quality by NewTechnologies,April23-24,2004,Rzeszow,Poland.

[11] Dubina, D., Stratan, A., Ciutina, A., Nagy, Zs. (2004). "Experimental Research on Monotonicand Cyclic Performance of Joints of Cold-Formed Pitched Roof Portal Frames". Proc. "TheSecondInternationalConferenceonSteel&CompositeStructuresICSCS’04",Ed.Chang-KoonChoi, Hwan-Woo Lee, Hyo-Gyong Kwak, ISBN 89-89693-13-6-98530. 2-4 September 2004,Seoul,Korea.pp:176-190.

[12] Dubina,D.,Stratan,A.,Ciutina,A.,Fulop,L.,Nagy,Zs.(2004)."Strength,stiffnessandductilityof cold-formed steel bolted connections". Proc. of the Fifth AISC / ECCS InternationalWorkshop"ConnectionsinSteelStructuresV.Behaviour,Strength&Design",June3-5,2004.Ed. F.S.K. Bijlaard, A.M. Gresnigt, G.J. van der Vegte. Delft University of Technology, TheNetherlands.ISBN:978-90-9019809-5,pp.263-272.

64

2.3 HIGHSTRENGTHSTEELINSEISMICRESISTANTSTRUCTURES

Seismic resistant building frames designed as dissipative structures, must allow for plasticdeformations to develop in specific members, whose behaviour is expected to be predicted andcontrolled by proper calculation and detailing. Members designed to remain elastic duringearthquake, such as columns, are characterized by high strength demands. Dual-steel structuralsystems, optimized according to Performance Based Design (PBD) philosophy, in which HighStrengthSteel(HSS)isusedinpredominantly"elastic"members,whileMildCarbonSteel(MCS)isused in dissipative members, can be very reliable and cost efficient. Current European seismicdesign codes do not cover this specific configuration. An extensive research project, HSS-SERF –High Strength Steel in Seismic Resistant Building Frames, was carried out with the aim toinvestigate and evaluate the seismic performance of dual-steel building frames. Main resultsinvestigatedatthePolitehnicaUniversityofTimisoaraarepresentedinthefollowing.

2.3.1 STEEL-CONCRETECONNECTIONINCONCRETE-FILLEDRECTANGULARHOLLOWSECTIONCOLUMNS

Concretefilledtubes(CFT)havetheadvantageofimprovedrigidityandresistancetofire,aswellassimilar cross-sectional properties on principal loading directions. The combination the twomaterials(steelandconcrete)mayleadtoareducedsteelsection.However,aparticularattentionshould be given in order to distribute thestresses between thesteel tubeand theconcrete core.Within thecurrentresearch, thegoalwas tocountonbothmaterials (steelandconcrete)andtohave a composite action ensured by the use of shot fired nails, also termed as powder-actuatedfasteners.

The nailed shear connection was developed in a research project conducted at the TechnicalUniversityofInnsbruck:Fink(1997)andLarcher(1997).Asaresult,theuseofpowder–actuatedfasteners is a relatively new method for assuring shear connection in areas where loads areinducedtocompositetubularcolumns–BeckandReuter(2005).Thepowder-actuatedfastenersare driven through the tube walls from the outside and then protrude inside the tube. Theconnectionbetweenthesurrounding tubularsectionandtheconcrete inside is thenprovidedbydirectpressureoftheconcreteagainsttheshanksofthenails.Themainadvantageofthissolutionis that it is quick and easy to apply, especially for columns which are continuous over severalstoreys.

Extensive experimental investigations were performed by Beck (1999) covering a number of 30push-out tests with pipe specimens. Influences of pipe geometry, concrete strength and type offastener on the load-deflection characteristics were investigated. The load deflection behaviourexhibitedexcellentductilitycombinedwithhighload-bearingcapacityperfastener.Basedonthesefindings the nailed shear connection was introduced into practice with the Millennium Tower inVienna,afiftystoreyhighrisebuildingcompletedin1999,whereithasproventobeareliableandcost effective connection method as no welding work was required: Huber (2001). Additionalexperimentalinvestigationswereperformed,byHanswilleetal(2001)withtheaimtoinvestigatethebehaviouroftheshearconnectionsubjectedtoaserviceabilitylimitstateloadingsequence,andthelongtermbehaviourandinfluenceofconcretecreeponthenailedshearconnections.

The aim of the experimental program was to evaluate the load introduction within compositecolumns realized as concrete filled tubes of high strength steel (see Figure62). For this purposeloadintroductiontestswereperformedonanumberof6specimensvaryingparameterssuchas:loading procedure (monotonic, cyclic), connection type (steel-concrete bonding, steel-concretebondingcombinedwithconnectors),andsteelgrade(S460,S700).Theobjectiveofresearchwastoassesstheefficiencyoftheshotfirednailsinprovidingtheshearconnectionbetweenthesteeltube

65

andtheconcretecoreundermonotonicandcyclicloading.Asconnectors,anumberof24HiltiX-DSH32P10shotfirednailswereusedpercolumnstub.

Theparticularities of thecurrentexperimentalprogram,withregard to theprevious studies,arerelated to: rectangular hollow section tubes, high strength steel (S460, S700), cyclic loadingprocedure, evaluation of the friction forces between the steel tube and the concrete core. Theexperimental program is summarised in Table 15. Consequently, a number of 6 tests on columnstubs were performed under monotonic and cyclic loading taking into consideration the casesbelow: 1monotonicand1cyclictestonthecolumnstubtakingintoaccountonlytheadhesionbetween

thesteeltubeandtheencasedconcrete(S700-F-M/S700-F-C); 1monotonicand1cyclictestonthecolumnstubcontainingsteel-concretebondingcombined

withconnectors(S700-F-H-M/S700-F-H-C); 1monotonicand1cyclictestonthecolumnstubcontainingsteel-concretebondingcombined

withconnectors(S460-F-H-M/S460-F-H-C);

Figure62.Specimenconfigurationfortheloadintroductiontests.

Table15.Experimentalprogram-loadintroductiontests.

Nr. Specimen Tube Bond Load1 S700-F-M RHS250x10S700 Friction Monotonic2 S700-F-C RHS250x10S700 Friction Cyclic3 S700-F-H-M RHS250x10S700 FrictionandNails* Monotonic4 S700-F-H-C RHS250x10S700 FrictionandNails* Cyclic5 S460-F-H-M RHS300x12.5S460 FrictionandNails* Monotonic6 S460-F-H-C RHS300x12.5S460 FrictionandNails* Cyclic

*24HiltiX-DSH32P10shotfirednails

The column stub specimens were manufactured using cold formed steel hollow section tubes of800mmlength.Theconcretecorewasconsideredwithadepthof600mminsidethesteeltubes.Fortheapplicationoftheload,plateswereweldedintheshapeofadoubleTbeamontwooppositetubewalls.

For the positioning of the connectors, installation instructions were prepared by Beck (2010) atHiltiAGbasedontheparticularitiesofthecurrentexperimentalprogram(steelgradeandthicknessofthetubes).Consequently,itwasrecommendedtousetheX-DSH32P10nailsandtoapplythembeforeconcretinginordertodrivethemconsistentlystraightthroughthetubewalls.Furthermore,due to the use of high-strength steel, pre-drilling of the steel with a small pilot hole would benecessary.Theinstallationinstructionscontainedalsoinformationrelatedtothepowder-actuatedfasteningtoolequipmentandthestep-by-stepnailapplicationprocedure.

66

With theaim toassess thebehaviourandthecharacteristicsof theconnectionbetweenthesteeltubeandtheencasedconcrete,anexperimentaltestset-upwasdesigned(seeFigure63a)inordertobeusedwithinauniversaltestingmachine,andconsideringthefollowing: monotonicaswellascyclicloadingconditionsofthespecimens; topandbottomsupports(seeFigure62)incontactwithconcretecoreonly; loadtobeappliedonthesteeltube; loadpathfromthesteeltubetotheconcretecore: friction/friction+connectors;

(a)

(b)

Figure63.Conceptualscheme(frontandlateralview),andillustrationofthetestset-up(a)andConceptualschemeandillustrationofinstrumentationarrangement(b).

Theinstrumentationofthespecimens(seeFigure63b)consistedinthemeasurementoftheforceappliedbythetestingmachineandtherelativedisplacementbetweensteeltubeandconcretecore,usingtwodisplacementtransducersatthetopsideandtwoatthebottomside.

The parameters used to control the load introduction tests, were the relative displacement "d"betweenthesteeltubeandtheconcretecore,andtheforce"F".Threealternatingcycleswithapeakloadofupto25%oftheexpectedyieldloadwereappliedtothetestassemblybeforethetestitselfinordertostabilizethesystem.Duringthetestitself,theloadwasappliedinforcecontrolwithintheelasticrangeandindisplacementcontrolfortheplasticrange.Forthecyclicloading,theECCS(1986)procedurewasused.

At 28 days from concrete casting, compression tests were performed on concrete cube samples.Consequently,theaveragecharacteristicstrengthwasobtainedinamountof35.88N/mm2,avalueslightly lower than theminimum characteristicstrength (37 N/mm2) corresponding toa C30/37concreteclass.

TheX-DSH32P10nailsof4.5mmdiameterweretestedtoshearbyHiltiAG(1998)obtainingan

averageshearforceinamountof21.85kN(at25C).Consequently,thecorrespondingshearstresswasinamountof1374N/mm2andtheequivalentstresswasobtainedinamountof2380N/mm2.Inaddition,accordingtoHanswilleetal.(2001),theultimatestrengthoftheX-DSH32P10nailswascorrespondingtofu=2200N/mm2.

The test assemblies of the three column stub configurations are shown in Figure 64. Theexperimentalinvestigationsstartedwiththethreemonotonictests.Theobtainedforcevs.relativedisplacementcurvesallowedassessingtheyielddisplacements"dy",necessaryfortheECCS(1986)cyclic loading procedure. Accordingly, the yield displacements were as follows: 0.75 mm for thecolumn stub with friction bond only, and 1 mm for the two column stubs with friction andconnectors.

d d

67

(a)S700-F

(b)S700-F-H

(b)S460-F-H

Figure64.Testassemblyofthethreecolumnstubconfigurations

The monotonic and cyclic response of the friction, developed between the steel tube and theconcrete core, is shown in Figure 65 and Figure 66. It can be observed that above a relativedisplacementof2mmthemonotonicforcetransmittedthroughfrictionwasmeasuredinamountof200kN.Undercyclicloadingtheforcedevelopedthroughfrictionwaslower(approximately160kN corresponding to the maximum amplitude of a cycle and 20 kN corresponding to the initialposition,i.e.0relativedisplacement.

0

200

400

600

800

1000

1200

0 2 4 6 8 10

Relative displacement [mm]

Fo

rce

[k

N]

S700-F-H-MS700-F-M

Figure65.Contributionoftheconnectorsundermonotonicloading.

-1100-900-700-500-300-100100300500700900

1100

-24 -20 -16 -12 -8 -4 0 4 8 12 16 20 24

Relative displacement [mm]

Fo

rce

[k

N]

S700-F-H-C

S700-F-C

Figure66.Contributionoftheconnectorsundercyclicloading.

68

A graphical comparison in terms of monotonic response corresponding to the column stubswithoutnails(S700-F)andwithnails(S700-F-H)canbeseeninFigure65.Itcanbeobservedthatthecontributionofthe24connectorsissignificant.Inaddition,theresponseundercyclicloadingofthe column stubs without and with connectors can be seen in Figure 66. The connectors show asignificant contribution also under cyclic loading conditions. It can be observed that exceeding arelative displacement of approximately 5 mm, the capacity of the steel-concrete connectiondecreasedduetofractureoftheconnectorsattheinterfacebetweensteeltubeandconcretecoreaswillbefurthershown.

Theresponseundercyclic loadingof the twospecimenswithbothconnectorsandsteel-concretebondingiscomparedinFigure67.Thedifferencebetweenthetwoconfigurations isgivenbythesteeltube.TheS460tubeisasquareRHS300x12.5,whiletheS700tubeisasquareRHS250x10.Inthefirstcasethethicknessofthetubeis12.5mmandintheothercasethethicknessis10mm.Thelateralsurfaceoftheconcretecoreis660000mm2withintheS460tube,and552000mm2withintheS700tube.Itistobenotedthattheconcretedepthinbothcaseswas600mm.Aslightlyhigherfrictionforcewasthereforedevelopedwithinthecolumnstubwithlargersteeltube(S460).

-1100-900-700-500-300-100100300500700900

1100

-24 -20 -16 -12 -8 -4 0 4 8 12 16 20 24

Relative displacement [mm]

Fo

rce [

kN

]

S460-F-H-CS700-F-H-C

Figure67.Cyclicresponseofthestubswithsteel-concretebondingandconnectors.

Inordertoobservethedeformationsdevelopedwithinconnectorsandtheconcretearoundthem,two side walls of the column stubs S460-F-H-M and S460-F-H-C, were cut out with flame. TheresponseofconcreteandconnectorstomonotonicloadingcanbeobservedinFigure68.Itcanbeobservedthattheconcretewascrushedinasmallamountatthecontactwiththeshotfirednailswhichwerebent.Theresponseofconcreteandconnectorstothecyclicloadingconditionisshownin Figure 69. It can be observed thatunderalternating cycles the connectors were broken at theinterfacebetweenconcretecoreandsteeltube.Theprotrudingpartofthesteeltube,rubbedontheconcretesurface.

Figure68.Detailofconcreteandconnectoraftermonotonicloading.

69

Figure69.Detailofconcreteandconnectoraftercyclicloading.

2.3.2 EXPERIMENTALTESTSONWELDEDBEAMTOCONCRETE-FILLEDRHSCOLUMNJOINTS

An experimental program was developed and carried out with the aim to characterize thebehaviour of two types of moment resisting joints in multi-storey frames of concrete filled highstrengthsteel rectangular hollow section (RHS) columns andsteel beams.Reduced beam sectionandcoverplatebeamsareweldedtocolumnsinordertoobtainductileandover-strength joints,accordingtoEN1998-1request.

The study of the beam-to-column joints is related to the seismic performance evaluation of thedual-steel building frames within HSS-SERF research project, i.e. moment resisting frames, dualconcentricallybracedframes(D-CBF)anddualeccentricallybracedframes(D-EBF)inwhichmildcarbonsteel(MCS)isusedindissipativememberswhilehighstrengthsteel(HSS)isusedinnon-dissipative "elastic" members. Accordingly, the beams in MRF are realized of S355 steel grade,while thecolumnsareconcrete filledrectangularhollowsection tubes (CF-RHS)ofhighstrengthsteel(S460andS700).Asbasisfordefinitionoftheexperimentalprogramonbeam-columnjoints,cross-sectionsfromtheD-CBFframedesignedbyGIPAC(2011)wereused(Figure70),consideringtwocombinationsofHSS/MCS: CF-RHS300x12,5S460columnandIPE400S355beam; CF-RHS250x10S700columnandIPE400S355beam.

Beam: IPE 400 (S355)Column: RHS 300x12,5 (S460) RHS 250x10 (S700)

Beam: IPE 400 (S355)Column: RHS 400x20 (S460) RHS 350x16 (S700)

D-CBF

Beam: IPE 400 (S355)Column: RHS 250x10 (S460) RHS 220x8 (S700)

Beam: IPE 400 (S355)Column: RHS 350x12,5 (S460) RHS 300x10 (S700)

D-EBF

Figure70.DesignedframeswithCFTcolumns.

Thebeamsareweldedtothecolumnsconsideringthetwotypesofconnections:withreducedbeamsection(RBS),andwithcoverplates(CP).Duetotheflexibilityofthetubewallsundertransverseforces,theconnectionsolutionofthebeamsandcolumnswithinthecurrentresearchwasbasedonthe use of external diaphragms. The design was performed considering the development of the

70

plastic hinge in the beams. Further with the bending moment and shear force from the plastichinge, theweldedconnectionsandthecomponentsof the joint(coverplates,externaldiaphragmandcolumnwebpanel)weredesignedorcheckedsoastocompriseanequalorhighercapacityincomparison to the fully yielded and strain hardened plastic hinge. Figure 71 illustrates the twobeam-to-columnjointconfigurations,i.e.withreducedbeamsection(a)andwithcoverplates(b).

(a)

(b)

Figure71.Configurationofthedesignedbeam-to-columnjoints:withreducedbeamsection(a),andwithcoverplates(b).

The main objective of the experimental tests was to pre-qualify by tests welded connections inmoment resisting frames and dual braced frames designed using the dual-steel concept. Theexperimentalprogramonbeam-to-columnjointswithCF-RHScolumnsissummarizedinTable16.Thevariations in theconfigurationof the joints aregivenby two joint typologies (reducedbeamsection – RBS, and cover-plate – CP), two steel grades for the rectangular hollow section tubes(S460 and S700) and two intended failure modes (beam and connection zone). In addition twoloading conditions were considered for each beam-to-column joint configuration, i.e. monotonicandcyclicloadingprocedure.

Table16.Experimentalprogramonbeam-to-columnjointswithCF-RHScolumns.

Parameter Variable No.ofvariationsNo.of

specimensLoading MonotonicandCyclic 2

16Jointtype

ReducedBeamSectionandCover-Plate

2

HSSgrade S460andS690 2Failuremode Weakbeam/Weakconnection 2

Considering the two joint typologies (RBS and CP) and two steel grades for the tubes (S460 andS700),anumberoffourbeam-to-columnjointconfigurationsweredesigned.Additionally,inordertoassesstheover-strengthoftheconnectionzoneandtoobservethebasecomponentsofthejointconfigurations, tests were considered on the corresponding joints for which the beam wasstrengthenedwiththeaimtoavoidtheformationoftheplastichingeinthebeamandtoforcetheplasticdeformationsintheconnectionzone.

The intended plastic mechanism corresponding to the four designed beam-to-column joints isrelatedtotheformationoftheplastichingeinthebeamandanelasticresponseoftheconnectionand joint components. In contrast, for the beam-to-column joints with strengthened beam the

71

intended failure mode is related to the connection and joint components. Table 17 describes foreach of the 16 beam-to-column joints, the labelling, column and beam, joint typology, loadingprocedureandintendedfailuremode.

Table17.Beam-to-columnjointconfigurations.

Nr. Specimenname Column BeamJointtype

LoadingIntended

failuremode

1 S460-RBS-M RHS300x12.5S460

IPE400S355

RBSMonotonic

Beam2 S460-RBS-C Cyclic3 S700-RBS-M RHS250x10

S700IPE400

S355RBS

MonotonicBeam

4 S700-RBS-C Cyclic5 S460-CP-M RHS300x12.5

S460IPE400

S355CP

MonotonicBeam

6 S460-CP-C Cyclic7 S700-CP-M RHS250x10

S700IPE400

S355CP

MonotonicBeam

8 S700-CP-C Cyclic9 S460-RBS-R-M RHS300x12.5

S460IPE400

S355RBS

MonotonicConnection

10 S460-RBS-R-C Cyclic11 S700-RBS-R-M RHS250x10

S700IPE400

S355RBS

MonotonicConnection

12 S700-RBS-R-C Cyclic13 S460-CP-R-M RHS300x12.5

S460IPE400

S355CP

MonotonicConnection

14 S460-CP-R-C Cyclic15 S700-CP-R-M RHS250x10

S700IPE400

S355CP

MonotonicConnection

16 S700-CP-R-C Cyclic

Theconceptualschemeandanillustrationoftheexperimentaltestset-upisshowninFigure72.Ahydraulic actuator connected at the tip of the beam served as loading device. The column wassupportedatbothendsconsideringapinnedconnection.Thehorizontalandverticaldisplacementswereblockedbytherightsupport,andonlytheverticaldisplacementswererestrainedbytheleftsupport.Alateralsupportsystemwasusedtobocktheoutofplanedeformationsofthebeam.

Pinned connections

Lateral support system

Hydraulic actuator

Figure72.Conceptualschemeandillustrationofthetestassembly.

Global as well as local instrumentation of the specimens was considered using the measuringinstrumentsoftheloadingdevice,displacementtransducers,andanopticalmeasuringsystem(Vic3D).Fromtheglobalinstrumentation,informationwasobtainedrelatedtotheforceintheactuator,thedisplacementatthetipofthebeam,thehorizontalandverticaldisplacementatsupports(seeFigure73a).Themeasurementsatthesupportswereperformedatthefrontsideaswellasattheback side of the joint assembly. Local instrumentation was aimed to measure the deformationswithinthedissipativezoneandtheconnectionzone(seeFigure74),aswellasinthecolumnwebpanel (see Figure 73b). In order to identify the sequence of yielding, the connection zone waspreparedbywhitewashing(seeFigure74).TheopticalmeasuringsystemVic3Dwasusedinorder

72

to observe the initiation of yielding. Depending on each joint configuration and on the intendedfailuremode,differentcomponentswereinvestigatedusingtheopticalmeasuringsystem,i.e. thebeamwebfromthedissipativezone,theexternaldiaphragm,orthecolumnwebpanel.

(a)

(b)

Figure73.Globalinstrumentation(a)andlocaljointinstrumentation(b).

Figure74.Instrumentationofjoint,andconnectionzonepreparedbywhitewashing.

Theparameters usedtocontrol the tests were the inter-storeydrift "θ"of testassemblyandthebending moment "M" computed at column centerline. Monotonic loading was applied byprogressively increasing the displacement at the tip of the beam. One unloading-reloading phasewasconsideredforabetterestimationoftheinitialstiffness.TheANSI/AISC341loadingprotocolwasusedforthecyclicloading.

Experimental investigations wereperformedalsoonmaterialsamples withtheaimtoassess thecharacteristics of all parts of the joint assemblies. Compression tests on concrete cube sampleswereperformedat28daysfromconcretecasting,obtaininganaveragecharacteristicstrength inamountof35.88N/mm2.Inaddition,tensileandCharpyV-notchimpacttestswereperformedonthesteelsamples prepared from additional material (profiles and platesas summarized inTable18)associatedtoeachcomponentofthebeam-to-columnjoints.

73

For each component of the joint assemblies (i.e. beam, cover plates, external diaphragms, RHStubes, stiffeners, etc.), a number of three samples were prepared and tested considering bothtensileandCharpyV-notch test.Theresultsobtainedaresummarized inTable18.Consequently,for each component the average values are shown regarding the yield strength (ReH), ultimatestrength (Rm), elongation at fracture (A) and elongation at maximum force (Ag). The tableshowsalso the results obtained from the Charpy V-notch tests, i.e. the average value of the absorbedenergy,thetemperatureofthesamplesduringthetestandtheminimumrequiredenergy.Itistobenotedthatthesamplesmarkedwith"*"werepreparedwith7.5mmthickness.Theresultsobtainedon steel samples allowed assessing the properties of all joint parts and confirmed that the steelgradeswereinaccordancewiththecoderequirements.

Table18.TensileandCharpyV-notchtestresultsforsteelsamples

Steelcomponent/grade

Tensiletests CharpyV-notchtests

eHR mR

A gA T KVmin KV

N/mm2

N/mm2

% % C J J

1 IPE400flange S355JR 393.9 491.9 30.9 16.3 20 27 57.62 IPE400web* S355JR 439.5 507.5 29.1 15.8 20 27 99.63 Coverplates S355J0 432.7 489.1 30.8 16.2 0 27 266.64 Spliceplate S355J2 415.3 503.0 26.9 15.2 -20 27 221.35 RHS300x12.5 S460M 497.7 554.1 22.4 7.7 -20 40 247.36 RHS250x10* S700QL 725.4 830. 11.8 - -20 40 102.3

7External

diaphragmS460NL 463.4 618.9 24.2 13.6 -30 40 58.6

8External

diaphragmS690QL 725.8 807.1 16.3 5.5 -20 40 233.3

9 Verticalstiffener S460NL 495.1 622.4 25.9 13.6 -30 40 152.310 Verticalstiffener* S690QL 806.9 854.7 14.2 5.7 -20 40 150.6

Due to the innovative joint configurations, it was needed to have an accurate prediction for thebehaviour of the joints in order to avoid unacceptable failure during the experimental tests.Therefore,asetofnumericalsimulationshavebeenperformed.Inafirststepthematerialmodelwascalibratedbasedonresults from compression tests onconcretesamplesandtensile tests onsteelsamples.Usingthecalibratedmaterialmodel, thebehaviourofeach jointconfigurationwasobtained. The pre-test numerical simulations allowed assessing for each joint configuration, thestress distribution and plastic strain, as well as the moment-rotation curve, confirming a gooddesignofthejointsandtheintendedfailuremechanism.

Figure 75 and Figure 76 show the response and failure mechanism of RBS joints subjected tomonotonicandcyclic loading.Theyieldingwas initiatedinthebeamflangeswithintheRBSzoneand was followed by large plastic deformations – local buckling of flanges and web undercompression.

Figure 77 and Figure 78 show the response and failure mechanism of CP joints subjected tomonotonicandcyclicloading.Theyieldingwasinitiatedinthebeamflangesnearthecoverplatesand was followed by large plastic deformations – local buckling of flange and web undercompression.Nodamagewasobservedincoverplates,externaldiaphragmandcolumnwebpanel.

Figure79andFigure80showtheresponseandfailuremechanismofthejointswithstrengthenedbeam flanges (RBS-R), subjected to monotonic and cyclic loading. The yielding was initiated inbeamflanges(betweenthereinforcingplateandexternaldiaphragm)undercompression/tensionandwasfollowedbyyieldingofwebandexternaldiaphragm.Finally,thebeamflangesfracturedintheheataffectedzone(HAZ)duetotensionforces.

74

Figure81andFigure82showtheresponseandfailuremechanismof jointswithextendedcoverplate(CP-R)subjectedtomonotonicandcyclicloading.FortheS460-CP-Rjoints,theyieldingwasinitiatedintheexternaldiaphragmandwasfollowedbylocaldeformationsofcoverplatesundercompression/tensionandyieldingofthecolumnwebpanel.FortheS700-CP-Rjoints,theyieldingwasinitiatedinthecompressedcoverplateandwasfollowedbyfailureoftheweldedconnectionbetweenplates thatcomposedtheexternaldiaphragm,howeverat forces higher thanthedesigncapacity.

-1000-800-600-400-200

0200400600800

1000

-0.09 -0.06 -0.03 0 0.03 0.06 0.09

Rotation [rad]

Mo

me

nt

[kN

m]

S460-RBS-CS460-RBS-M

Figure75.S460-RBS-Cjoint:cyclicresponseandillustrationoffailuremode.

-1000-800-600-400-200

0200400600800

1000

-0.09 -0.06 -0.03 0 0.03 0.06 0.09

Rotation [rad]

Mo

men

t [k

Nm

]

S700-RBS-CS700-RBS-M

Figure76.S700-RBS-Cjoint:cyclicresponseandillustrationoffailuremode.

-1000-800-600-400-200

0200400600800

1000

-0.09 -0.06 -0.03 0 0.03 0.06 0.09

Rotation [rad]

Mo

me

nt

[kN

m]

S460-CP-CS460-CP-M

Figure77.S460-CP-Cjoint:cyclicresponseandillustrationoffailuremode.

75

-1000-800-600-400-200

0200400600800

1000

-0.09 -0.06 -0.03 0 0.03 0.06 0.09

Rotation [rad]

Mo

me

nt

[kN

m]

S700-CP-CS700-CP-M

Figure78.S700-CP-Cjoint:cyclicresponseandillustrationoffailuremode.

-1000-800-600-400-200

0200400600800

1000

-0.09 -0.06 -0.03 0 0.03 0.06 0.09

Rotation [rad]

Mo

me

nt

[kN

m]

S460-RBS-R-CS460-RBS-R-M

Figure79.S460-RBS-R-Cjoint:cyclicresponseandillustrationoffailuremode.

-1000-800-600-400-200

0200400600800

1000

-0.09 -0.06 -0.03 0 0.03 0.06 0.09

Rotation [rad]

Mo

me

nt

[kN

m]

S700-RBS-R-CS700-RBS-R-M

Figure80.S700-RBS-R-Cjoint:cyclicresponseandillustrationoffailuremode.

Theinterpretationofresultsismadeinrelationtothefollowingaspects: Overstrengthofconnectionzone; Contributionofcomponentstothejointrotation; Evaluationoftherotationcapacity.

Withtheaimtoassesstheoverstrengthofthejointandconnectionzone,acomparisonwasmadebetween the four designed joints and the corresponding joints with reinforced beam. Figure 83illustratestheoverstrengthoftheRBSjointsandFigure84illustratestheoverstrengthofCPjoints.

76

-1800

-1200

-600

0

600

1200

1800

-0.09 -0.06 -0.03 0 0.03 0.06 0.09

Rotation [rad]

Mo

me

nt

[kN

m]

S460-CP-R-CS460-CP-R-M

Figure81.S460-CP-R-Cjoint:cyclicresponseandillustrationoffailuremode.

-1800

-1200

-600

0

600

1200

1800

-0.09 -0.06 -0.03 0 0.03 0.06 0.09

Rotation [rad]

Mo

me

nt

[kN

m]

S700-CP-R-CS700-CP-R-M

Figure82.S700-CP-R-Cjoint:cyclicresponseandillustrationoffailuremode.

0

200

400

600

800

1000

0 0.04 0.08 0.12 0.16Rotation [rad]

Mo

me

nt

[kN

m]

S460-RBS-R-MS460-RBS-M

0

200

400

600

800

1000

0 0.04 0.08 0.12 0.16Rotation [rad]

Mo

me

nt

[kN

m]

S700-RBS-R-MS700-RBS-M

Figure83.OverstrengthofconnectionzoneforRBSjoints

77

0

300

600

900

1200

1500

1800

0 0.04 0.08 0.12 0.16Rotation [rad]

Mo

men

t [k

Nm

]

S460-CP-R-MS460-CP-M

0

300

600

900

1200

1500

1800

0 0.04 0.08 0.12 0.16Rotation [rad]

Mo

me

nt

[kN

m]

S700-CP-R-MS700-CP-M

Figure84.OverstrengthofconnectionzoneforCPjoints

Table19.Contributionofcomponentstotherotationcapacityofthejoints[mrad].

S460-RBS S700-RBS S460-CP S700-CPTotalassembly

rotation(EN1998-1criterion)

50 50 40 40

Columnwebpanel 1.1 3.8 1.6 2.1Connection 6.7 7.6 1.9 3.7

Plastichinge 43.1 38.2 34.1 29.1Elasticrotation 3.7 4.5 5.7 6.7

0

10

20

30

40

50

S460-RBS S700-RBS S460-CP S700-CP

Ro

tati

on

[m

rad

]

Column web panel Connection Plastic hinge Elastic rotation

Figure85.Contributionofcomponentstotherotationcapacityofthejoints.

Themoment-rotationcurveswerecomputedattheconnectiontotheexternaldiaphragm,andtheoverstrength was evaluated corresponding to the yield point and to the maximum capacity.Consequently, the overstrength of the connection zone for the RBS joints (S460 and S700) wasevaluated in the amount of 34% and 35% at yield, and respectively 53% and 60% at maximumcapacity.TheoverstrengthoftheconnectionzonefortheCPjoints(S460andS700)wascomputedinamountof55%and43%atyield,andrespectively101%and86%atmaximumcapacity.

The main plastic deformations occurred in the dissipative zone of the beam (plastic hinge). Thecontributionoftheconnectionandcolumnwebpaneltotheoveralljointrotationwasobservedto

78

below.TherotationcapacityoftheRBSandCPdesignedjointswasevaluatedconsideringtheEN1998-1(2004)criterionforwhichthereductionofstiffnessandcapacityisnotgreaterthan20%.Thetotalassemblyrotationforwhichthereductionofstiffnessandcapacitywasnotgreaterthan20%was corresponding to the50mradcycle forRBS jointsandto the40mradcycle for theCPjoints.Thecontributionofthecomponents,includingtheelasticrotation,tothetotaljointrotationissummarizedinTable19andillustratedinFigure85.

2.3.3 NUMERICALINVESTIGATIONOFWELDEDBEAMTOCONCRETE-FILLEDRHSCOLUMNJOINTS

Theobjectivesof thenumerical investigationprogramconductedontheweldedbeam-to-CF-RHScolumnjointsarerepresentedbythefollowingaspects: Calibrationofthematerialmodelbasedonmaterialsampletests; Pre-testnumericalsimulationsaimedtoofferanaccuratepredictionofthebehaviourofthe

jointsinordertoavoidunacceptablefailureduringtheexperimentaltests; Calibrationofthenumericalmodelofthejointsbasedonthemonotonicandcyclictestresults,

allowingforabetterunderstandingofthebehaviourandresponseofeachbeam-to-columnjointconfiguration;

Extensionoftheexperimentalprogramwithadditionalcases,i.e.parametricstudyfortheinvestigationof:influenceoftheconcretecore,influenceoftheaxialforceactingonthecolumn,influenceoftheexternaldiaphragm,responseofjointswithtwoandrespectivelyfourbeams;

Validationofthedesignrelationsforthejointcomponentsanddesignprocedure; Inafirststepthematerialmodelwascalibratedbasedonmaterialsampletests.Thepre-test

numericalsimulations,containingthecalibratedmaterialmodel,allowedassessingforeachjointconfiguration,thestressdistributionandplasticstrain,aswellasthemoment-rotationcurve,confirmingagooddesignofthejointsandtheintendedfailuremechanism.

The calibration of the numerical models of the joints started with the calibration of the materialmodelforeachpartofthebeam-to-columnjointassemblies.Thestress-strainrelationshipusedforconcrete, was computed using the results from the compression tests, and the analytical stress-strainmodelasreportedbyAlietal.(1990).Theconcretedamagedplasticitymodelwasusedandinadditiontheplasticityparameters,compressiondamageandtensilebehaviourfromKorotkovetal.(2004).Thecalibrationofthematerialmodelforthesteelgradeswasperformedbasedonthetensiletestresults.Consequently,Figure86illustratesfortheIPE400beamflange,thecomparisonbetweentestandsimulation,intermsofforce-displacementcurve,forwhichagoodcorrelationcanbeobserved.TheyoungmoduluswasconsideredequaltoE=210000N/mm2,andthePoissonratiowasconsideredasν=0.3.

0

200

400

600

800

0 10 20 30 40

Strain [%]

Str

ess [

N/m

m2]

S1S2S3S4

0

50

100

150

200

250

300

0 10 20 30 40 50 60Displacement [mm]

Fo

rce

[k

N]

Test

FEM

Figure86.Calibrationofthematerialmodelforsteel(IPE400beamflange).

79

Thenumerical investigationswereperformedwith the finite elementmodellingsoftwareAbaqus(2007).Allthecomponentsofthebeam-to-columnjointsweremodelledusingsolidelements.Duetothelargeamountofcontactsurfacesbetweentheconcretecoreandsurroundingsteeltube,theDynamic Explicit typeofanalysis wasconsidered.For the interactionbetweenthesteel tubeandthe concrete core, a contact law, characterised by normal as well as tangential properties, wasdefined that allowed the two parts to separate. The load was applied through a displacementcontrolat thetipof thebeamandtheboundaryconditionsconsideredattheendsof thecolumnwere consistent with the experimental test set-up, i.e. pinned and simple support. ThediscretisationoftheelementswasperformedusinglinearhexahedralelementsoftypeC3D8R.

The comparison between test results and pre-test simulations was observed to be relative close.Consequently, the calibration and refinement of the numerical models was performed using themeasuredgeometryofthespecimens.Inaddition,becausethebeamwasnotcompletelyrestrainedbytheoutofplanelateralsystem,alateralcontactwasdefined(seeFigure87)consideringasmallgap between flanges and the contact elements. It is to be noted that within the numericalsimulationsitwasnotaccountedformaterialfractureorcrackpropagation,phenomenathatwereobserved at some of the joint specimens. Because in the numerical models of these joints, nofracturedeveloped,thecapacitydidnotsufferasintheexperimentaltests.

Figure87.Illustrationofthejointmodelaccountingfortheoutofplanelateralsystem.

From thecalibration, a set of numericalmodels were obtained which were capable to reproducewith a good accuracy the response of the joints in both moment-rotation curve and failuremechanism, i.e. formation of the plastic hinge in the beam (RBS and CP designed joints) andyieldingofcomponents(jointswithstrengthenedbeam).Consequently,foreachjointconfiguration,acomparisonisshownbetweentestandsimulationintermsofmoment-rotationcurvecomputedatcolumncentreline.Theplasticstrainisshownincomparisontothefailuremodeobservedduringthe test. The results from the calibration of the monotonic tests on beam-to-column joints areshowninFigure88forthejointswithreducedbeamsection.

Thenumericalmodelscalibratedbasedonthejointssubjectedtomonotonicloadingwerefurtherused for the numerical calibration of the cyclic tests. A combined isotropic/kinematic cyclichardeningmodelwasthereforeadopted.Theinputforthematerialmodelwasrepresentedbytheyieldstrengthofthesteelpart,andinadditionthecyclichardeningparametersasgivenbyDuttaetal.(2010),i.e.C1=42096,γ1=594.45,Q∞=60,b=9.71.

Theresultsfromthecyclicanalysisareshownforeachbeam-to-columnjointconfigurationintermsmoment-rotation curve (computed at column centreline), von Misses stress distribution andequivalent plastic strain. An illustration of the failure mode, as obtained from the experimentalinvestigations,isshownaswell.ThecomparisonbetweentestandnumericalsimulationisshowninFigure89afortheRBSjoints,inFigure89bfortheCPjoints.

80

0

200

400

600

800

1000

0 0.04 0.08 0.12 0.16Rotation [rad]

Mo

me

nt

[kN

m]

S460-RBS-MFEM

0

200

400

600

800

1000

0 0.04 0.08 0.12 0.16Rotation [rad]

Mo

men

t [k

Nm

]

S700-RBS-MFEM

Figure88.ComparisonbetweentestandsimulationforRBSjoints.

Theresultsfromthenumericalinvestigationsofthecyclictestsshowagoodcorrelationwiththeexperimental results considering the moment-rotation hysteretic loops as well as the failuremechanism(plastichinge–bucklingofflangesandwebofbeaminthedissipativezone,yieldingofcomponents–failureintheheataffectedzoneofthereinforcedRBSjoints,plasticdeformationsinthe external diaphragm and column panel zone). In case of S460-CP-R-C joint, the capacitycorresponding to the negative amplitudes was obtained higher in the numerical simulationscomparedtothetestwhereacrackwasdevelopedinthecoverplate(betweenexternaldiaphragmandbeamflange)andwhichdidnotdevelopintheFEmodel.

-1000-750-500-250

0250500750

1000

-0.09 -0.06 -0.03 0 0.03 0.06 0.09

Rotation [rad]

Mo

men

t [k

Nm

]

S460-RBS-CFEM

(a)

(b)

Figure89.ComparisonbetweentestandsimulationforRBSandCPdesignedjointssubjectedtocyclicloading.

-1000-750-500-250

0250500750

1000

-0.09 -0.06 -0.03 0 0.03 0.06 0.09

Rotation [rad]

Mo

men

t [k

Nm

]

S460-CP-CFEM

81

A set of complementary testing cases were considered for the investigation through numericalsimulations with the aim to assess the influence of different parameters on the joint behaviour,suchas: Influenceoftheconcretecore–i.e.theresponseofthejointwithoutconcretecorein

comparisontothereferencejointwithCFTcolumn. Influenceoftheaxialforce–i.e.theresponseofthejointwithaxialforceinthecolumn

(N=0.5*Npl)incomparisontothereferencemodel. Responseofthejointswithtwoandfourbeamsweldedaroundtheconcretefilledtube. Influenceoftheexternaldiaphragm.

Withtheaimtoevaluatetheinfluenceoftheconcretecore–i.e.theresponseofthejointwithoutconcrete core in comparison to the reference joints with composite column (CF-RHS) – a set ofadditionalnumericalsimulationswereperformedonthecalibratednumericalmodelsfromwhichtheconcretepartwaseliminated(NC→noconcrete).Theresultsareshownintermsofmoment-rotationcurvecomputedatcolumncentreline.ItcanbeobservedinFigure90forRBSjoints,andinFigure91forCP joints, that theabsenceofconcretecoredidnotaffectsignificantlytheresponse(onlyaminorreductionofthecapacitywasobserved).

0

200

400

600

800

1000

0 0.04 0.08 0.12 0.16Rotation [rad]

Mo

men

t [k

Nm

]

S460-RBS-MS460-RBS-M NC

0

200

400

600

800

1000

0 0.04 0.08 0.12 0.16Rotation [rad]

Mo

men

t [k

Nm

]

S700-RBS-MS700-RBS-M NC

Figure90.Influenceofconcretecore–RBSjoints

0

200

400

600

800

1000

0 0.04 0.08 0.12 0.16Rotation [rad]

Mo

me

nt

[kN

m]

S460-CP-MS460-CP-M NC

0

200

400

600

800

1000

0 0.04 0.08 0.12 0.16Rotation [rad]

Mo

me

nt

[kN

m]

S700-CP-MS700-CP-M NC

Figure91.Influenceofconcretecore–CPjoints.

0

500

1000

1500

2000

0 0.04 0.08 0.12 0.16Rotation [rad]

Mo

me

nt

[kN

m]

S460-CP-R-MS460-CP-R-M NC

0

500

1000

1500

2000

0 0.04 0.08 0.12 0.16Rotation [rad]

Mo

me

nt

[kN

m]

S700-CP-R-MS700-CP-R-M NC

Figure92.Influenceofconcretecore–reinforcedCPjoint

82

In contrast, for the reinforced joints (extended CP) a significant reduction of capacity can beobserved in Figure 92. In these cases, the bending moment and implicitly the shear force in thecolumnwebpanelweremuchhigherthaninthecasesshownabove.Inotherwords,theshearforcein thecolumnwebpanelexceeded thecapacityof thesteel tube inshear,andthereforeahighercapacitywascorrespondingtothejointswithconcretecore.

Withtheaimtoassesstheinfluenceoftheaxialforceinthecolumnwithrespecttothebehaviourofthe beam-to-column joint, numerical simulations were performed on one of the joints (S460-CP)consideringthefollowingcases: Jointwithconcretecoreandanaxialforcelevelinthecolumncorrespondingto50%of

Npl,Rd,composite. Jointwithoutconcretecoreandwithanaxialforcelevelinthecolumncorrespondingto50%of

Npl,Rd,steel.

(a)

(b)

0

200

400

600

800

1000

0 0.04 0.08 0.12 0.16Rotation [rad]

Mo

me

nt

[kN

m]

FEMFEM 0.5·Npl,comp

(c)

compositeRdplN ,,5.0 Jointwithconcretecore Jointwithoutconcretecore

Figure93.Influenceof0.5·Npl,Rd,compositeaxialforce:S460-CPjoint

(a) (b)

0

200

400

600

800

1000

0 0.04 0.08 0.12 0.16Rotation [rad]

Mo

men

t [k

Nm

]

FEMFEM 0.5·Npl,steel NC

(c)

steelRdplN ,,5.0

Jointwithoutconcretecore

Figure94.Influenceof0.5·Npl,Rd,steelaxialforce:S460-CPjointwithoutconcretecore.

Foreachofthetwocases,inafirststep,thecolumnwasaxiallyloadeduptotheconsideredlevel(seeFigure93aandFigure94a),andinasecondstepthebeamwasloaded.ThestressdistributionandimplicitlythefailuremechanismofthejointwithconcretecoreareshowninFigure93b.ThecorrespondingmomentrotationcurveiscomparedinFigure93cwiththereferencemodel.Itcanbeobservedthattheaxialforceinthecolumndidnotaffecttheresponseofthejointcomparedto

83

the reference model, and that the plastic hinge formed in the beam. In addition, the stressdistributionandimplicitlythe failuremechanismof the jointwithoutconcretecoreareshowninFigure94b.Themoment-rotationcurveiscomparedinFigure94cwiththereferencemodel.Alsointhiscaseitcanbeobservedthattheaxialforceinthecolumndidnotaffecttheresponseofthejointwithout concrete core compared to the referencemodel and that the plastic hinge formed in thebeam.

The experimental tests were performed on single sided beam-to-column joints. It was thereforenecessary to assess the behaviour of the joints (including column web panel) in the situation ofloadingfromtwosideswhichcorrespondstoamoredemandingscenario.Figure95aandbshowtheplasticstraincorrespondingtotheS700-CPjointswith-andrespectivelywithoutconcretecore.Consequently, for the case with composite column and two beams (FEM 2G), a reduction of thestiffnesscanbeobservedinFigure96acomparedtothetestcurve,butthefailuremodewasnotaffected(seeinFigure95plastichingesinthebeams).Theabsenceoftheconcretecore(NC→noconcrete)leadtoreductionofthecapacity(Figure96b)andtotheyieldingofthecolumnpanel–plastichingesdidnotforminthebeamasforthejointwithconcretecore.

(a)

(b)

Figure95.BehaviourofS700-CPjointwithbeamsweldedontwosides:CFT(a)andwithoutconcrete(b).

0

200

400

600

800

1000

0 0.04 0.08 0.12 0.16

Rotation [rad]

Mo

men

t [k

Nm

]

S700-CP-MFEM 2G

(a)

0

200

400

600

800

1000

0 0.04 0.08 0.12 0.16Rotation [rad]

Mo

men

t [k

Nm

]

FEM 2GFEM 2G-NC

(b)

Figure96.ResponseofS700-Pjointswithbeamsweldedontwosides.

Asimilarinvestigationwasperformedalsoforthejointconfigurationwithreducedbeamsection.Therefore,Figure97aandbshowthestressdistributionandplasticstraincorrespondingtothejoints with- and respectively without concrete core. A reduction of the stiffness can be observedalsointhiscase(seeFigure98a)comparedtothetestcurve.Theabsenceoftheconcretecoredidnot affect the capacity or the failure mode (see Figure 98b). Consequently, the plastic hingesdevelopedinthebeams.

The weak-beam/strong-column concept was therefore confirmed for the CP and RBS joints,considering the case with beams welded on two sides, and in addition the study showed thecontributionoftheconcretecoretojointbehaviour.

84

Figure97.BehaviourofS700-RBSjointwithbeamsweldedontwosides:CFT(a)andwithoutconcrete(b).

0

200

400

600

800

1000

0 0.04 0.08 0.12 0.16Rotation [rad]

Mo

men

t [k

Nm

]

S700-RBS-MFEM 2G

(a)

0

200

400

600

800

1000

0 0.04 0.08 0.12 0.16Rotation [rad]

Mo

men

t [k

Nm

]

FEM 2GFEM 2G-NC

(b)

Figure98.ResponseofRBSjointswithbeamsweldedontwosides.

Figure99.Jointswithfourbeams–vonMissesstressdistributionandplasticstrain.

85

Figure100.Jointswithfourbeams–moment-rotationcurves.

The response of the joints was also investigated considering the case with four beams weldedaroundtheconcrete filledtube.Asreference, theS460-RBSandS460-CPjointmodelswereused.The loading was applied in displacement control at the tip of the beams. The load ratio wasconsideredinamountof100%ononedirectionandrespectively30%ontheotherdirection.AsaresultthemomentrotationcurvesareshowninFigure100comparedtothetestresults.Asitcanbeobserved,onthemainloadingdirectionthemoment-rotationcurvesufferedasmallreductionofthestiffnessbutthecapacityorthefailuremodewasnotaffected.Figure99showsthevonMissesstressdistributionandtheplasticstraininthetwojointconfigurations.Therefore,consideringtheloading of the joints from two directions it can be observed that the joint components (externaldiaphragm,columnwebpanel)didnotsufferplasticdeformations.

2.3.4 OUTCOMES

The parameters investigated within the steel-concrete connection tests consisted of loadingprocedure (monotonic, cyclic), connection (friction, friction+ connectors) and steel grade (S460,S700).Theaimwastoassesstheefficiencyofshotfirednailsinprovidingtheconnectionbetweensteeltubeandconcretecore.Theshearstrengththatdevelopedattheinterfacebetweensteeltubeandconcretecore(columnstubwithfrictiononly)wasobtainedinamountof0.4N/mm2,whichisequaltothevaluerecommendedbyEN1994-1(2004)forrectangularhollowsections.Inrelationto the specific loading (push-out tests), it was observed that the connectors have a significantcontributiontotheloadtransferfromsteeltubetotheconcretecore,inbothmonotonicandcyclicloading. From the investigation of the behaviour and failure mode, it was observed that undermonotonic loadingtheconcretewascrushedinasmallamountatthecontactwiththeshotfirednails which bent. Under alternating cycles the nails eventually broke at the interface betweenconcretecoreandsteeltube.Inadditionitwasobservedthatfromthecyclicloadingthecapacityoftheconnectorsdecreasedslightlycomparedtothemonotonic loading. Inconclusion,asobservedby Beck (1999), theX-DSH 32 P10 shot fired nails proved a significant contribution to the steel-concreteconnectionconsideringthemonotonicandcyclicloading,aswellasHSStubes(S460andS700).

Theexperimental investigationsperformedonbeam-to-column jointsunderbothmonotonicandcyclic loadingevidencedagoodconceptionanddesignof the joints(RBSandCP) justifiedbythefollowingobservations: elasticresponseoftheconnectionzone; formationoftheplastichingeinthebeam; correspondingtocyclictestsperformedaccordingtotheANSI/AISC341(2005)loading

protocol,theRBSandCPdesignedjointsevidencedrotationcapacitiesof50mrad(RBSjoints)andrespectively40mrad(CPjoints)forwhichthedegradationofstrengthandstiffnesswerenotgreaterthanthe20%limitdefinedinEN1998-1(2004);

0

200

400

600

800

1000

0 0.04 0.08 0.12 0.16Rotation [rad]

Mo

me

nt

[kN

m]

S460-RBS-MRBS - 1RBS - 0.3

0

200

400

600

800

1000

0 0.04 0.08 0.12 0.16Rotation [rad]

Mo

me

nt

[kN

m]

S460-CP-MCP - 1CP - 0.3

86

goodresponseofjointdetailingandweldedconnectionswithoneexception(S700-CP-R-M/C)whichevidencedweldfailure,howevercorrespondingtoaforcelevelhigherthenthedesigncapacitycomputedusingtherealmaterialproperties.

From the comparison between the four designed joints and the corresponding joints withreinforcedbeam,asignificantoverstrengthoftheweldedconnections,respectivelyoftheexternaldiaphragmandcolumnwebpanelwasobserved.ThereforetheoverstrengthrequirementsfromEN1998-1(2004)weresatisfied.

The contribution to the joint rotation was evaluated for the following regions: dissipative zone(plastic hinge), connection (welded connection and external diaphragm), and column web panel.For the joints with reduced beam section and with cover plates, the main plastic deformationsoccurredinthebeam(plastichinge).Thecontributionoftheconnectionandcolumnpaneltotheoverall joint rotation was observed to be low. For the joints with reinforced beam, the maincontribution to the overall joint rotation was given by the connection zone including externaldiaphragm(forstrengthenedRBS joints),andrespectivelyexternaldiaphragmandcolumnpanel(forstrengthenedCPjoint).

Asetofnumericalmodelswereobtainedthatwerecapabletoaccuratelyreproducethebehaviourof thetested jointconfigurations inbothmonotonicandcyclic loadingconditions.Thisallowedabetter understanding of the joint behaviour and validation of the analytical model based on theresultsofthecalibratedjointmodels.Inadditiontheexperimentalprogramwasextendedinordertoevaluatetheinfluenceoftheconcretecore,axialforceinthecolumn,responseofthejointswithmultiplebeamsandinfluenceoftheexternaldiaphragm.

2.3.5 PUBLICATIONS

[1] Dubina,D.,Vulcu,C.,Stratan,A.,Ciutina,A.,Grecea,D.,Ioan,A.,Tremeea,A.,Braconi,A.,Fülöp,L.,Jaspart,J.-P.,Demonceau,J.-F.,Hoang,L.,Comeliau,L.,Kuhlmann,U.,Kleiner,A.,Rasche,C.,Landolfo,R.,D’Aniello,M.,Portioli,F.,Beg,D.,Cermelj,B.,Može,P.,daSilva,L.S.,Rebelo,R.,Tenchini,A.,Kesti, J.,Salvatore,W.,Caprili,S.,andFerrini,M. (2015).Highstrengthsteel inseismicresistantbuildingframes(HSS-SERF).ResearchFundforCoalandSteel,Finalreport,EuropeanCommission,B-1049Brussels,181p.

[2] Vulcu,C.,Stratan,A.,Dubina,D.(2012)."Numericalsimulationofthecyclicloadingforweldedbeam-to-CFTcolumn joints ofdual-steel frames",PollackPeriodica,vol.7,no.2,pp.35–46.DOI:10.1556/Pollack.7.2012.2.3.PaperISSN:1788-1994,OnlineISSN:1788-3911.

[3] VulcuC.,StratanA.,CiutinaA.,DubinaD.(2011)."Beam-to-columnjointsforseismicresistantdual-steelstructures",PollackPeriodica6(2),pp.49-60.PaperISSN:1788-1994,OnlineISSN:1788-3911.

[4] Dubina, D., Stratan, A., Vulcu, C., and Ciutina, A. (2014). “High strength steel in seismicresistantbuildingframes.”SteelConstruction:DesignandResearch,7(3),173–177.

[5] Cristian Vulcu, Aurel Stratan, Adrian Ciutina, Dan Dubina (2014). "Studiul comportăriistâlpilortubulariumpluţicubetoncuconectoridinbolţuri împuşcate,solicitaţimonotonşiciclic".AICPSReview,nr.1-2/2014,ISSN:2067-4546,pp.80-85.

[6] Cosmin Maris, Cristian Vulcu, Aurel Stratan, Dan Dubina (2014). "Evaluarea eficiențeitehnico-economice a structurilor realizate în soluție «dual-steel»". AICPS Review, nr. 1-2/2014,ISSN:2067-4546,pp.191-198.

[7] Dubina, D., Grecea, D., Stratan, A. and Muntean, N. (2009). "Performance of dual-steelconnections ofhighstrengthcomponents undermonotonic andcyclic loading".Proc.of the6thInt.Conf.onBehaviourofSteelStructuresinSeismicAreas,STESSA2009,August16-20.

87

F.M.Mazzolani,J.M.Ricles&R.Sause(eds).Taylor&FrancisGroup,London,ISBN13978-0-415-56326-0.p.437-442.

[8] Dubina, D., Dinu, F., Stratan A. (2008). "Performance based analysis of high strength steelbuildingframesunderseismicactions".ActaTechnicaNapocensis.Section:CivilEngineering– Architecture. Nr. 51, vol. 1, 2008, ISSN 1221-5848, p. 361-372. Proceedings of theInternationalConference–CONSTRUCTIONS2008,9-10May2008,Cluj-Napoca.

[9] Dubina,D.,Muntean,N.,StratanA.,Grecea,D.,Zaharia,R.(2008)."Performanceofmoment-resistingjointsofhigh-strengthsteelcomponents".ActaTechnicaNapocensis.Section:CivilEngineering–Architecture.Nr.51,vol.1,2008,ISSN1221-5848,p.373-384.ProceedingsoftheInternationalConference–CONSTRUCTIONS2008,9-10May2008,Cluj-Napoca.

[10] Dubina,D.,Stratan,A.,Dinu,F.,Grecea,D.,Muntean,N.,Zaharia,R.(2008)."Structuriîncadremultietajaterealizateînsistemdual-steel:cerinţedeperformanţăşiprogramexperimental".AICPSReview,nr.1-2/2008,p.39-52.ISSN:1454-928X

[11] Muntean,N.,Stratan,A.,Dubina,D.(2008)."ExperimentalevaluationofstrengthandductilityperformanceofwelddetailsandT-stubboltedconnectionsbetweenhigh-strengthandmildcarbonsteel",ScientificBulletinofthe"Politehnica"UniversityofTimisoara,TransactionsonCivilEngineeringandArchitecture,Vol.53(67)No.1,2008,ISSN1224-6026,p.45-55.

[12] Dubina,D.,Stratan,A.,Dinu,F.,Grecea,D.,Muntean,N.,Vulcu,C.(2010)."Applicationofhighstrengthsteel toseismicresistantmulti-storeybuildings".COST ActionC26"UrbanHabitatConstructions under Catastrophic Events" Proceedings, Naples, 16-18 September 2010,Mazzolani(Ed.),Taylor&FrancisGroup,London,p.355-363.ISBN:978-0-415-60685-1.

[13] Dubina, D., Dinu, F., Stratan, A., Muntean, N., Zaharia, R., Ungureanu, V., Grecea, D. (2008)"Utilizareaoţelurilordeînaltăperformanţăînstructuraderezistenţăaclădirilormultietajateamplasate în zone cu risc seismic ridicat: studii de eficienţă şi program experimental".Structuri metalice amplasate în zone seismice - preocupări actuale. Ed. D. Dubina, V.Ungureanu.EdituraOrizonturiUniversitare,Timişoara.ISBN978-973-638-377-9,p.89-106.

[14] Dubina, D., Muntean, N., Stratan A., Grecea, D. and Zaharia R. (2008). "Testing program toevaluate behaviour of dual steel connections under monotonic and cyclic loading". Proc. ofthe 5th European Conference on Steel and Composite Structures EUROSTEEL 2008, 3-5September,Graz,Austria.R.Ofner,D.Beg,J.Fink,R.Greiner,H.Unterweger(Eds).ISBN92-0147-000-90,p.609-614.

[15] Dubina, D., Stratan, A., Muntean, N., Grecea, D. (2008). "Dual-steel T-Stub behaviour undermonotonicandcyclicloading".ProceedingsoftheSixthInternationalWorkshop"Connectionsin Steel Structures VI", June 22–25, Chicago, USA. Ed. R. Bjorhovde, F.S.K. Bijlaard, L.F.Geschwindner,AmericanInstituteofSteelConstruction,p.185-196.

[16] Dubina,D.,Stratan,A.,Muntean,N.,Dinu,F.(2008)."Experimentalprogramforevaluationofmomentbeam-to-columnjointsofhighstrengthsteelcomponents".ProceedingsoftheSixthInternationalWorkshop"ConnectionsinSteelStructuresVI",June22–25,Chicago,USA.Ed.R.Bjorhovde, F.S.K. Bijlaard, L.F. Geschwindner; American Institute of Steel Construction, p.355-366.

[17] Vulcu,C.,Stratan,A.,Dubina,D.,Bordea,S.(2012)."SeismicperformanceofdualframeswithcompositeCF-RHShighstrengthsteelcolumns",Proceedingsof the15thWorldConferenceonEarthquakeEngineering,September24-28,2012,Lisbon,Portugal,paper4937.

[18] Dubina,D.,Stratan,A.,Vulcu,C.,Ciutina,A. (2014). "Highstrengthsteel inseismicresistantbuilding frames", 7thEuropean Conferenceon Steel and CompositeStructures EUROSTEEL2014, 10-12 September 2014, Napoli, Italy, Landolfo R, Mazzolani FM, editors, European

88

Convention for Constructional Steelwork, ECCS, paper no. 10-202, 6 p. ISBN: 978-92-9147-121-8.

[19] Vulcu, C., Stratan, A., Ciutina, A., Dubina, D. (2014). "Steel-concrete connection in case ofconcrete filledrectangular hollow section columns", 7th EuropeanConference onSteelandComposite Structures EUROSTEEL 2014, 10-12 September 2014, Napoli, Italy, Landolfo R,Mazzolani FM, editors, European Convention for Constructional Steelwork, ECCS, paper no.08-409,6p.ISBN:978-92-9147-121-8.

[20] Dubină, D., Stratan, A., Vulcu, C., Ciutina, A. (2013). "Structuri în cadre multietajate cucomponetedinoțelde înaltărezistențăpentruclădiriamplasate înzoneseismice".AXIII-aConferinţă Naţională de Construcţii Metalice, Bucureşti, 21–22 noiembrie 2013, Editori:DanielBîtcășiPaulIoan,pp.165-172,ISBN978-973-100-306-1.

[21] Vulcu, C., Stratan, A., Ciutina, A., Dubina, D. (2014). "Experimental Evaluation of the Steel-Concrete Connection in Case of Concrete Filled Rectangular Hollow Section (CF-RHS)Columns".ProceedingsoftheInternationalWorkshop"ApplicationofHighStrengthSteelsinSeismic Resistant Structures". Orizonturi Universitare, Timisoara, Romania, ISBN: 978-973-638-552-0,pp.95-104.

[22] Vulcu, C., Stratan, A., Ciutina, A., Dubina, D. (2014). "Experimental Evaluation of WeldedReduced Beam Section (RBS) and Cover Plate (CP) Beam-to-CF-RHS Column Joints".Proceedings of the International Workshop "Application of High Strength Steels in SeismicResistantStructures".OrizonturiUniversitare,Timisoara,Romania,ISBN:978-973-638-552-0,pp.105-120.

[23] Vulcu,C.,Stratan,A.,Ciutina,A.,Dubina,D.(2014)."NumericalInvestigationofWeldedRBSand CP Beam-to-CF-RHS Column Joints". Proceedings of the International Workshop"ApplicationofHighStrengthSteelsinSeismicResistantStructures".OrizonturiUniversitare,Timisoara,Romania,ISBN:978-973-638-552-0,pp.121-136.

[24] Vulcu, C., Stratan, A., Dubina, D. (2012). "Seismic Performance of EB Frames of CompositeCFHS High Strength Steel Columns". Proceedings of the 10th International Conference onAdvances in Steel Concrete Composite and Hybrid Structures, ASCCS 2012, Singapore, 2–4July2012.ISBN-13:978981-07-2615-7,pp.953-960.

[25] Vulcu,C.,Stratan,A.andDubina,D.(2012)."SeismicresistantweldedconnectionsforMRFofCFTcolumnsandIbeams",Proceedingsofthe7thInternationalWorkshoponConnectionsinSteel Structures, May 30 - June 2, 2012, Timisoara, Romania, ECCS, Dan Dubina and DanielGrecea(Eds),ISBN978-92-9147-114-0,pp.275-290.

[26] Vulcu, C., Stratan, A., and Dubina, D. (2011). "Evaluation of Welded Beam-to-CFT ColumnJoints", EUROSTEEL 2011, August 31 - September 2, 2011, Budapest, Hungary, p. 489-494.ISBN:978-92-9147-103-4.

[27] Vulcu, C., Stratan, A., Ciutina, A., Dubina, D. (2010). "Simularea comportării unui modelexperimental pentru noduri de cadre etajate cu stâlpi din ţevi umplute cu beton şi grinzidubluT".Lucrărileceleide-a12-aConferinţeNaţionaledeConstrucţiiMetalice"Realizărişipreocupări actuale în ingineria construcţiilor metalice", Timişoara, 26-27 noiembrie 2010,Dubina, Ungureanu, Ciutina (Ed.), Editura Orizonturi Universitare, p. 233-244. ISBN: 978-973-638-464-6.

[28] Dubina, D., Vulcu, C., Stratan, A., Ciutina, A. (in print). "Dual frames of high strength steelRHSCF columns for seismic zones", 8th International Conference on Behavior of SteelStructuresinSeismicAreasSTESSA2015,Shanghai,China,July1-3,2015.Paperno.123.