Tall Building Initiative
Response EvaluationHelmut KrawinklerProfessor EmeritusStanford
University
On behalf of the Guidelines writers:Y. Bozorgnia, C.B. Crouse,
R.O. Hamburger, R. Klemencic, H. Krawinkler, J.O Malley, J.P.
Moehle, F. Naeim, J.P. Stewart Quake Summit 2010, October 8,
2010
Quake Summit 2010, October 8, 2010
Performance ObjectivesDemonstrate that structure will be capable
of essentially elastic response and limited damage under
Service-level Earthquake shaking (mean RP = 43 years = 50/30)
Demonstrate, with high confidence, that structure will respond
to Maximum Considered Earthquake (MCE) shaking without loss of
gravity-load-carrying capacity without inelastic straining of
important lateral-force resisting elements to a level that will
severely degrade their strength; and without experiencing excessive
permanent lateral drift or development of global structural
instability.
Quake Summit 2010, October 8, 2010
MCE Level EvaluationObjective: provide, implicitly, adequate
life safety protection Protection against collapseProtection
against life threatening falling hazardsProtection against
aftershocks & condemnationUse 3-D nonlinear response history
analysis for at least 7 ground motion pairsUse a realistic model of
the structural systemFollow capacity design principles (enforced in
acceptance criteria)Minimum base shear not required
Quake Summit 2010, October 8, 2010
Acceptance Criteria at Component LevelForce-controlled actions
with severe consequences:
Fu fFn,e
Fu = smaller of1.5 times meanMean + 1.3s but 1.2 times meanFn,e
= nominal strength based on expected material propertiesf =
resistance factor
Quake Summit 2010, October 8, 2010
Acceptance Criteria at Component LevelDeformation controlled
actions:No specific limitations, but use realistic model of
component behavior, including deterioration, or limit maximum
deformation to a conservative (low) value du.
If d > du in any one analysis:Strength in this action should
drop to zeroEffect on related strength properties should be
evaluated
Quake Summit 2010, October 8, 2010
Mean of max. transient drift in every story 3.0%Max. transient
drift in every story 4.5%Mean of max. residual drift in every story
1.0%Max. residual drift in every story 1.5%Loss in story strength
at max. drift should not be more than 20%Acceptance Criteria at
System Level
Quake Summit 2010, October 8, 2010
System Modeling IssuesIncorporate all components and all
behavior modes (e.g., shear in RC) that significantly affect
prediction of seismic responseMight require post-analysis review
and re-analysisFlexibility of floor diaphragms should be modeled if
deemed importantAnalysis should provide information needed to
quantify diaphragm forcesPodium and backstay effects must be
represented realisticallyP-Delta effects must be includedInclude
real torsion, but no requirement for accidental torsion
Quake Summit 2010, October 8, 2010
Wall Hinging at the BaseLoadingStory Sheargy=Vy/WStory OTM
Quake Summit 2010, October 8, 2010
NRHA force demands may be very different from elastic
expectationsMaximum moment in shear wall
Quake Summit 2010, October 8, 2010
NRHA force demands may be very different from elastic
expectationsMaximum shear force in shear wall
Quake Summit 2010, October 8, 2010
Component ModelingDeterioration in strength and stiffness must
be considered if it significantly affects the response of the
structure to the MCE ground motions Or conservative estimates must
be made of strength and deformation capacities
Quake Summit 2010, October 8, 2010
Modes of Deterioration
Quake Summit 2010, October 8, 2010
Basic ObservationThe cyclic envelope curve is different from the
monotonic backbone curve
Quake Summit 2010, October 8, 2010
Resource DocumentATC-72-1Interim Guidelines on Modeling and
Acceptance Criteria for Seismic Design and Analysis of Tall
Buildings
Quake Summit 2010, October 8, 2010
GENERAL MODELING ISSUESTypes of ModelsDeteriorationP-Delta
effectsDampingUncertainties
PROPERTIES OF NONLINEAR STRUCTURAL COMPONENTSSteel beams and
columnsSteel panel zonesAxially loaded steel bracesRC beams,
columns, and joints
PLANAR AND CORE WALL SYSTEMS AND COMPONENTSPlanar walls, flanged
walls, core wallsCoupling beamsSlab-columns and connections
FLOOR DIAPHRAGMS, COLLECTORS, AND PODIUM AND BACKSTAY
EFFECTSRigid, semi-rigid, and flexible diaphragmsPodium and
backstay effects Resource DocumentATC-72-1
Quake Summit 2010, October 8, 2010
Source: G. Deierlein
Quake Summit 2010, October 8, 2010
Use of Strain-based Models (Fiber & Curvature
Models)Argument for their use:
whenever lumped plasticity models are not available
Columns subjected to biaxial bending and large axial forceShear
walls with (and without?) openingsSpandrel beams?
Quake Summit 2010, October 8, 2010
Use of Strain-based Models (Fiber & Curvature
Models)Arguments against their use:RC:Rebar buckling?Rebar
fracture?Bond slippage and pullout?Shear?Steel:Local
instabilities?Fracture?Joint panel zones?
Need to account for cyclic deterioration
Quake Summit 2010, October 8, 2010
Use of Concentrated Plasticity (Spring) ModelsRotational spring
models if inelastic behavior mode is bending
Translational spring models if inelastic behavior mode is
shear
Arguments for their useCan capture deterioration characteristics
if good calibrations are available from experimental dataAre
relatively simple
Arguments against their useAre approximate Not available for
many important failure modes
Quake Summit 2010, October 8, 2010
ASCE 41 models may be used if deemed appropriateThey were
intended to be used in conjunction with pushover analysisThey were
not intended to be used for hysteresis modelingThe sharp drop from
C to D is not representative of reality except for brittle failure
modesThey may not be applicable to many new components
Quake Summit 2010, October 8, 2010
Component Models with Deterioration (see ATC-72) Q-HYST
Degrading Stiffness Flag-Shaped Bi-Linear Hysteresis1. Monotonic
(initial) backbone curve:Cyclic deterioration parameterDescription
of hysteresis loopsFd
Quake Summit 2010, October 8, 2010
Modeling Option #1 ATC-72Use of monotonic backbone curve and
explicit incorporation of cyclic deterioration
Quake Summit 2010, October 8, 2010
Modeling Option #2 ATC-72Use of cyclic envelope curve as
modified backbone curve, and no incorporation of cyclic
deterioration limit du to max. observed in test
Quake Summit 2010, October 8, 2010
Modeling Option #3 ATC-72Use of factors to generate modified
backbone curve from monotonic backbone curve, and no incorporation
of cyclic deterioration
- capping strength Fc* = 0.9 Fc- plastic deformation capacity
dp* = 0.7dp- post-capping deformation capacity dpc* = 0.5dp-
residual strength Fr* = 0.7Fr- ultimate deformation capacity du* =
1.2dc
Quake Summit 2010, October 8, 2010
Modeling Option #3
Quake Summit 2010, October 8, 2010
Modeling Option #4 ATC-72 No deterioration at all in analytical
model
ultimate deformation capacity du* corresponding to 80% of
capping strength on descending branch of Options 2 or 3
Quake Summit 2010, October 8, 2010
Comparison of ATC-72 Modeling Options
Quake Summit 2010, October 8, 2010
qyqcqp0.5qpcqpcqcqp=0.7qp1.5qcMc0.8McInitial backbone
curveModified backbone curve, Option 3Ultimate rotation, Option
4qpuUltimate rotation, Option 3Penalties for Options 3 and 4
Quake Summit 2010, October 8, 2010
What is new?No radical changesExplicit formulation of
performance objective and acceptance criteria at two levels of
ground motions (SLE & MCE)Consideration of deterioration in
component properties if it is importantOr acceptance of penalty in
component modelingConsistent design and performance evaluation
process
Quake Summit 2010, October 8, 2010
I think we are making progress
*********