Report - PEER - University of California, Berkeley

4. ANALYSIS METHODS FOR PERFORMANCE-BASED DESIGN,ASSESSMENT OR RETROFITTING OF BUILDINGSPart 1 of Eurocode 8 (CEN 2004a) includes the following analysis options for thedesign of buildings or the assessment of their performance:• Linear static (termed “lateral force” method).• Linear modal response spectrum analysis.• Nonlinear static analysis (“pushover”).• Nonlinear dynamic (response time-history).Linear modal response spectrum analysis is the standard procedure for design,applicable to all types of new buildings. Countries are allowed to limit the use ofnonlinear analysis methods for the design of new buildings via their National Annex.Pushover analyses should be performed under two lateral load patterns: one foruniform lateral accelerations and another similar to the lateral forces used in linearstatic (lateral force) analysis, if applicable, or are derived from a modal responsespectrum one, if it isn’t. The target displacement is derived via the well-known N2procedure (Fajfar, 2000), which is given in an informative annex and employs Eq. (1).Part 3 of Eurocode 8 (CEN 2004c) adopts a fully displacement-based approach.So the main objective of the seismic analysis is the estimation of deformationdemands in structural members. The four analysis options of Part 1 are available. Inall of them the seismic action is given by the 5%-damped elastic spectrum or thequantities of interest derived from it: the target displacement for non-linear staticanalysis, the acceleration time-histories for nonlinear dynamic analysis. The spectrumis anchored to the PGA corresponding to the hazard level chosen for the Limit Statefor which analysis results will be used, with multiplication by the “importance factor”of the building (the same as in Part 1). All analyses, except nonlinear dynamic ones,essentially use the equal-displacement rule: at the level of member deformations, e.g.,chord-rotation demands, for the first two types of analysis, or of the displacement ofan equivalent SDOF system for pushover analysis (as modified by Eq.(1)).For linear analysis to be applicable, ductility demands should be fairly uniformlydistributed in the entire structure. Taking the ratio D/C of bending moment at amember end from elastic analysis, D, to the corresponding capacity, C, as a measureof the local ductility demand (D/C is roughly equal to the demand value of the chordrotationductility ratio), the maximum of the D/C-ratio in all primary elements shouldnot exceed its minimum value over all primary elements with D/C>1 by more than aNDP-factor between 2 to 3 with a recommended value of 2.5. The D/C ratio is takenequal to 1 (elastic) at sections around beam-column joints where plastic hinges willnot form on the basis of the sum of beam flexural capacities compared to that ofcolumns. No limit is set to the ductility demands on “ductile” elements.If linear analysis is applied, internal forces in “brittle” mechanisms of behaviorare estimated as in capacity-design: from equilibrium, assuming that “ductile”locations delivering force to the “brittle” mechanisms develop their force capacity, or11