Performance Based Seismic Design Guidelines For Tall .

Performance Based SeismicDesign Guidelines for TallBuildings and their ApplicationsFarzad Naeim1

What is a Tall Building? Overall height as a measureü Some codes such as ASCE 7 impose limits onlateral systems to be used based on height Aspect ratio as a measure Vibration period as a measure Prevalence of higher modes in responseas a measure No universally accepted definition existsbut you know one when you see one!2

Should tall buildings be treatedlike other buildings? Tall buildings are occupied by hundreds if notthousands of people The consequence of failure of tall buildings is muchmore severe than an ordinary building Codes provide a “one size fits all” approach toseismic design. Tall buildings as small class of specializedstructures will perform better during earthquakes ifspecial attention is afforded to their individualcharacteristics. Prescriptive codes are not equipped with means todistinguish these differences.3

Why prescriptive codes are notsuitable? Because they simply cannot give you what you need.Linear analysis is incapable of accurately predictingcollapse and failure which are inherently nonlinearThe overwhelming majority of construction in UnitedStates and worldwide consists of low-rise buildings1 to 3 Stories (93%)14 Stories and Taller (1%)4 to 13 Stories (6%) Prescriptive provisions are not generally written with tallbuildings in mind.

We will examine two guidelines.

ASCE 41and Tall Building Design Guidelines ASCE41 is officially intended for seismicrehabilitation of existing structures However, its component-based performancelimits for NDP are routinely referenced byguidelines for performance based design oftall buildings Engineers who believe ASCE 41limits are tooconservative, or are not applicable to theirproject, conduct tests to establish appropriatelimits Peer review approval is always necessary forany deviation from ASCE 416

Common Performance Objectives SEAOC-99 ASCE 41–Similar objectives permitted. Emphasis on two events: 475 years (10% in 50 years), and 2,475 years (2% in 50 years) Tall Building Design Guidelines– Serviceability: 43 years– Collapse Prevention: 2,475 years7

Analytical Procedures ASCE-41 permits four types of analyses:1. Linear elastic static procedure (LSP)2. Linear dynamic procedure (LDP) or responsespectrum analysis3. Non-linear static procedure (NSP) commonlyreferred to as the push-over analysis, and4. Dynamic nonlinear response analysis (NDP). Tall Building Design Guidelines permit onlytwo:1. 3D LDP or NDP for serviceability check2. 3D NDP for all other checks8

PEER-TBI & LATBSDC Performance Objectives1. Serviceable behavior under events having a 50%probability of being exceeded in 30 years (43 yearreturn period) building structural and nonstructural components retaintheir general functionality during and after earthquakeRepairs, if necessary, are expected to be minor andcould be performed without substantially affecting thenormal use and functionality of the building2. A low probability of collapse under events havinga 2% probability of being exceeded in 50 years(2,475 year return period) Demands are checked for all structural members (lateralas well as gravity system)Claddings and their connections to the structure mustaccommodate MCE displacements without failure

PEER-TBI & LATBSDC Provisions1. Use 2.5% damping instead of 5% damping butpermit DCR 1.5 for deformation controlledmembers for serviceability.2. 2011 LATBSDC limits DCR to 0.70 for forcecontrolled members in serviceability check.3. 2010 PEER requirements for collapse preventionare more elaborate and detailed than 2011LATBSDC4. No minimum base shear capacity requirement

Design Procedures None of the guidelines tell you how to design For example, 2011 LATBSDC states:ü Use Capacity Design Techniquesü Develop Project-specific Design Criteria, andü Clearly define where nonlinearity can occur and make sure itdoes not occur elsewhereü Recommends preferred zones of nonlinearity But they do not explain how the engineer issupposed to achieve this design.11

ROSE School 2013Source: 2011 LATBSDC12

Evaluation Procedures All guidelines require a threedimensional detailed mathematicalmodel of the physical structure Realistic estimates of stiffness anddamping Expected material properties forductile elements Specified material properties forbrittle elements13

Source: 2011 LATBSDCPerformance Based Seismic Assessment of Tall Buildings – I14

Effective Stiffness Values forLinear AnalysisROSE School 2013Source: 2008 LATBSDC, 2010 PEER15

Effective Stiffness Values for Linear AnalysisROSE School 2013Source: 2011 LATBSDCPerformance Based Seismic Assessment of Tall Buildings – I16

Analysis Methods Serviceability:ü Can use either1. Linear Response Spectrum Analyses CQC mode combination90% mass participation2. Nonlinear Response History Analyses For MCE (ultimate state) evaluation:ü Must use Nonlinear Response History AnalysesInherent torsional properties of thestructural system should always beconsidered.17

P-Δ Inclusion P-Δ effects must beincluded in allROOF DRIFT ANGLE vs. NORMALIZED BASE SHEARanalysesPushover (NEHRP '94 k 2 pattern); LA 20-StoryNormalized Base Shear (V/W)0.14P-Delta effect includedP-Delta effect excluded0.120.10.080.060.040.02000.010.020.03Roof Drift AngleFigure courtesy of Prof. Helmut Krawinkler0.040.05

Modeling Nonlinear BehaviorFigure courtesy of Prof. Prof. Greg Deierlein19

Modeling Nonlinear Behavior Concentrated plasticity model for beams andcolumns and fiber elements for walls aremost commonAll other elements and components that incombination significantly contribute to oraffect the total or local stiffness of thebuilding should be included in themathematical model.Axial deformation of gravity columnsin a core-wall system is one exampleof effects that should beconsidered in the structural modelof the buildingFigure courtesy of MKA20

Accidental Eccentricity (AE) 2011 LATBSDCü Consider implications during serviceability evaluationü Address if significant during MCE evaluation 2010 PEER TBIü Do not need to consider Consideration of AE in nonlinear analysesrequires multiple evaluations and little is gainedby such time-consuming exercises.21

Modeling Strength / Stiffness Degradation 2010 PEER TBIü Provides detailed guidelines on fourapproved methods for modelingdegradation 2011 LATBSDCü Adopts the first two of the detailedprocedures contained in 2010 PEER.22

2010 PEER TBI Degradation Modeling OptionsFigure courtesy of Prof. Helmut Krawinkler23

Upper Limit on Column Axial Forces Large axial forces reduce availablecolumn ductility 2011 LATBSDCü MCE: P 0.4f’cAg 2010 PEER TBIü MCE: P balanced load 0.3f’cAg24

Soil-Foundation-Structure-Interaction (SFSI) Naeim & Stewart (2008)demonstrated the difficulties ofrealistic modeling of SFSI in a designenvironment. 2010 PEER TBI has tworecommended modeling techniques 2011 LATBSDC recommends a singleapproach for this.25

2010 PEER TBI SuggestedModeling Techniques for SFSI2011 LATBSDC26

Damping A particularly thorny issueü In nonlinear analyses most of the damping isrepresented by hysteretic behavior of the elementsü Some small additional viscous damping may bejustified for: Energy dissipation provided by componentsand systems not explicitly modeled As necessary to avoid numerical instability 2011 LATBSDCü Limits viscous damping to 2.5% for bothserviceability and MCE. 2010 PEER TBIü 2.5% for linear serviceability evaluationü Refers to ATC-72 for nonlinear evaluation27

Ground Motion Selection and Scaling A minimum of 7 pairs is usually required 2011 LATBSDCü Adopts by reference Chapter 21 of ASCE 7 2010 PEER TBIü More flexibleü Permits scaling, matching or CMSü Multiple CMS required if CMS is used, makingthis impractical for tall buildings Most practicing engineers prefer matchingü One must be careful as, matched motioncontains less record to record dispersion28

Acceptance Criteria -- Maximum Drift Absolute Maximum Transient DriftLimitü Serviceability: 2011 LATBSDC & 2010 PEER TBI:0.005 overallü MCE: 2011 LATBSDC & 2010 PEER TBI:0.030 max average at any story0.045 max. interstory drift at any story under anyrecord29

Acceptance Criteria -- Maximum Drift Absolute Maximum Residual DriftLimitü Serviceability: 2011 LATBSDC 0.005 overallü MCE: 2011 LATBSDC and 2010 PEER:0.010 average max. of time histories0.015 maximum from any30

Acceptance Criteria -- Serviceability 2011 LATBSDCü Brittle Actions:Strength Demand 0.7*Capacityü Ductile Actions: Linear AnalysisStrength Demand 1.50 Capacity Nonlinear AnalysisCan use up to IO limit of ASCE 4131

Acceptance CriteriaMCE 2010 PEER and 2011 LATBSDCü Ductile Actions: Deformation Demand ASCE 41-06 CPDeformation Capacity Continuous Load Path Capacity exhausted when it drops below80% of maximum strength32

Acceptance Criteria -- MCE 2010 PEERü Brittle Actions: Two Groups: Critical Actions failure mode pose severe consequences tostructural stability under gravity and/or lateralloads Design for mean 1.3 to 1.5 times SD Noncritical Actions Design for mean values Use ϕ 0.75 for shear 2011 LATBSDCü Essentially the same, except uses 1.5 timesmean and ϕ 1.033

R/C Specific Requirements None in 2010 PEER Several in 2011LATBSDCü Detailing The spacing limit of 12inches of ACI 318§21.5.3.2 (d) is reducedto 6 inches.ü High-StrengthConcrete34

Peer Review Requirements Each project needs a Seismic Peer Review Panel(SPRP) SPRP is to provide an independent, objective,technical review of design Paid by the owner but reports to Building Official Responsibility for the structural design remainssolely with the EOR SPRP is not a plan checking entity Minimum of three members with recognizedexpertise in relevant fields such as:ü ü ü ü ü ü structural engineeringearthquake engineering researchperformance-based earthquake engineeringnonlinear response history analysistall building designearthquake ground motions, geotechnical engineering,geological engineering35

Instrumentation Requirements 2010 PEER TBIü No requirements 2011 LATBSDCü Detailed requirementsü Consistent with CGS / CSMIP36

A typical tall buildinginstrumented by CSMIPROSE School 2013Performance Based Seismic Assessment of Tall Buildings – I37

SeismicInstrumentationROSE School 2013CSMIP sensor layout38

Applications Many tall buildings have been designed usingthese guidelines in Los Angeles, San Francisco,San Diego, and elsewhere Here are some examplesü Los Angeles: ü San Diego ü 888 Olive1133 Olive1212 Flower TowersWilshire & GrandMetropolis Tower7th & AshSan Francisco Transbay Tower39

888 Olive Streetin downtownLos Angelesü ü ü ü ü ü ü 34 storiesCore wall constructionPodiumSubterranean levelsBasement wallsFlat platesGravity columnsIllustrations and drawings courtesy of Onni Group and Glotman-Simpson

Illustrations and drawings courtesy of Onni Group and Glotman-Simpson