Yousef Bozorgnia, PEER

1
Ground Motion Selection, Modification and
Simulation for Seismic Analysis of Tall
Buildings Status Report

• Yousef Bozorgnia, PEER
• Nicolas Luco, USGS
• Farzad Naeim, JAMA
• Paul Somerville, URS

LATBSDC 2007 Meeting May 4, 2007
2
Outline

• Various tasks and groups, related to ground
  motion tasks for tall buildings
• Current status
• On-going and future tasks

3
Latest Developments in Next Generation
Attenuation (NGA) of Ground Motion

• With Emphasis on Long Period Components

4
NGA Attenuation Models

• Ground motion parameters
• Horizontal components
• PGA, PGV, PGD
• Pseudo spectral acceleration at 5 damping
• Period 0 - 10 sec

5
Example of Predicted Acceleration
Spectra(Campbell-Bozorgnia) Strike Slip, RRUP
10 km, VS30 760 m/s
6
Effect of DistanceCB06 Predicted Acceleration
SpectraStrike Slip, M 7.0, VS30 760 m/s
7
Behavior at Long Periods CB06 Predicted
Spectral DisplacementStrike Slip, RRUP 10 km,
VS30 760 m/s
8
Standard Deviation vs. Magnitude
9
NGA Models Give Prediction on Geometric Mean of
Two Horizontal Components
10
NGA Reports Papers

• Reports are available at PEER web site

• Including Excel files of the models
• Journal publications will be printed in special
  issue of EERI Spectra, March 2008

11
USGS Is Adopting NGA Models for the 2007 US
National Seismic Hazard Maps
12
USGS is Adopting

• NGA relations with documentation
• Boore and Atkinson
• Campbell and Bozorgnia
• Chiou and Youngs
• With equal weighting

Courtesy Art Frankel
13
Ratio of 0.2 s SA 2007 and 2002 national seismic
hazard maps for WUS at 2 probability of
exceedance in 50 years.
14
Ratio of 1.0 s SA 2007 and 2002 national seismic
hazard maps for WUS At 2 probability of
exceedance in 50 years
15
In Summary

• NGA models provide much more robust estimates of
  ground motions than the old models
• Especially at long periods (e.g., for tall
  buildings)
• The authors of NGA and old models, strongly
  recommend using NGA vs the 1997 models

16
Tall Buildings Initiative, Task 5

• Synthetically generated ground motions
• Paul Somerville (Task Leader)
• Brad Aagaard, Rob Graves

17
PEER Tall Buildings Initiative, Task 5

• The goal is to add simulated ground motions to
  the NGA recorded motions to have a combined
  (hybrid) database

18
TBI Task 5

• Use validated broadband ground motion simulation
  procedures
• Simulate ground motions at San Francisco and Los
  Angeles sites

19
Simulations for Puente Hills Blind Trust Fault

• Simulated motions at 648 sites were generated

20
Various scenarios were considered

• Next scenarios
• San Andreas Fault (north and south)
• Hayward Fault

21
Tall Buildings Initiative, Task 6

• Review and validation of synthetically generated
  ground motions
• Farzad Naeim, Yousef Bozorgnia
• Yoshi Moriwaki, CB Crouse, Doug Dreger, Norm
  Abrahamson, Yuehua Zeng, Joe Maffei

22
TBI Task 6

• Goal is to establish a validated reference set of
  synthetically generated ground motions
• Review the procedures and results of simulations
• The review will be carried out by both
  seismologists and engineers

23
TBI Task 6

• Documentations of Puente Hills simulated ground
  motions have been sent to various seismologists
  for review
• Questions and comments are being collected from
  the reviewers

24
TBI Task 6

• Elastic and inelastic response spectra of
  simulated ground motions are being computed
• Resulted spectra will be compared with the
  results of attenuation of NGA records (both
  elastic and inelastic spectra)

25
Inelastic spectra of NGA database

• Various versions of inelastic spectra were
  computed for more than 3200 horizontal records
• For periods
• 0.02 to 10 sec

• Overall gt 1,240,000 nonlinear time history
  analyses were performed

26
Attenuation of Inelastic Spectra (NGA), For
Different Ductility Ratios
27
Attenuation of Inelastic Spectra (NGA)
28
TBI Task 6

• Spectral comparison (NGA vs
  simulated motions) will be submitted to
  engineers for evaluation and comments

29
PEER Ground Motion Selection Modification
(GMSM)
30
PEER GMSM Working Group

• A multi-disciplinary team of
• Seismologists
• Geotechs
• Structural engineers

31
Objectives

• (1) Estimate true response of many structural
  models
• Buildings
• Bridges
• Dams
• OpenSees models
• Perform models
• (2) Examine existing and upcoming GMSM methods
  with reference to the true response

32
Estimate the true response of structural model

• Excite the structural model with hundreds of
  input GMs
• Suppose you have no limitations in number of runs
• Compute statistics of various structural responses

33
True structural response can be computed

• Using the input motions, un-scaled and scaled,
  obtain correlation between structural response
  and elastic response spectra of input motions
• Running for a large set of input motions and
  correlate the structural response directly to
  basic parameters

34
Example of a work-in-progress (Zareian
Bozorgnia)

• Run a nonlinear frame model for thousands of NGA
  records (un-scaled)
• Correlate structural response to magnitude,
  distance, site,
• Estimate also uncertainty

35
Estimate Structural Response
36
Once we know the true response

• Evaluate various ground motion selection and
  modification methods
• We have contacted all authors of GMSM methods
• Requested them to provide records
• Will run the same structural models with the
  provided input motions
• Will compare results with true response

37
Examine various GMSM methods
38
Thank You !