analysis procedures to estimate seismic demands for ... · outline •background introduction...
TRANSCRIPT
![Page 1: Analysis Procedures to Estimate Seismic Demands for ... · Outline •Background Introduction •Classification of Analysis Procedures NSP •Nonlinear Static Pushover NDT •Nonlinear](https://reader033.vdocuments.site/reader033/viewer/2022052918/5ade68c27f8b9afd1a8b5a87/html5/thumbnails/1.jpg)
www.sgh.com
Analysis Procedures to Estimate
Seismic Demands of Structures
Seismic Lessons Learned Panel Meeting
27 May 2015
Said Bolourchi
![Page 2: Analysis Procedures to Estimate Seismic Demands for ... · Outline •Background Introduction •Classification of Analysis Procedures NSP •Nonlinear Static Pushover NDT •Nonlinear](https://reader033.vdocuments.site/reader033/viewer/2022052918/5ade68c27f8b9afd1a8b5a87/html5/thumbnails/2.jpg)
Outline
• Background
• Classification of Analysis Procedures Introduction
• Nonlinear Static Pushover NSP
• Nonlinear Dynamic Time-history NDT
• Multi-Mode Pushover MMP
• Conclusions Conclusions
27 May 2015 Seismic Lessons learned Panel Meeting 2
![Page 3: Analysis Procedures to Estimate Seismic Demands for ... · Outline •Background Introduction •Classification of Analysis Procedures NSP •Nonlinear Static Pushover NDT •Nonlinear](https://reader033.vdocuments.site/reader033/viewer/2022052918/5ade68c27f8b9afd1a8b5a87/html5/thumbnails/3.jpg)
Background
“The future effective use of performance-based engineering
depends on the continued development of reliable and
credible inelastic analysis procedures… ongoing research
promises important modifications, improvements, and
alternatives to current NSPs”
[FEMA 440 (2005), Improvement of Nonlinear Static Seismic Analysis Procedures]
27 May 2015 Seismic Lessons learned Panel Meeting 3
Introduction
![Page 4: Analysis Procedures to Estimate Seismic Demands for ... · Outline •Background Introduction •Classification of Analysis Procedures NSP •Nonlinear Static Pushover NDT •Nonlinear](https://reader033.vdocuments.site/reader033/viewer/2022052918/5ade68c27f8b9afd1a8b5a87/html5/thumbnails/4.jpg)
Classification of Analysis Procedures
LSP – Linear Static Pushover
LDT – Linear Dynamic Time-history
NSP – Nonlinear Static Pushover
NDT – Nonlinear Dynamic Time-history
MMP – Multi-Mode Pushover
27 May 2015 Seismic Lessons learned Panel Meeting 4
Introduction
![Page 5: Analysis Procedures to Estimate Seismic Demands for ... · Outline •Background Introduction •Classification of Analysis Procedures NSP •Nonlinear Static Pushover NDT •Nonlinear](https://reader033.vdocuments.site/reader033/viewer/2022052918/5ade68c27f8b9afd1a8b5a87/html5/thumbnails/5.jpg)
LSP and LDT
• LSP results can be very inaccurate for irregular
structures, (FEMA 356)
• LSP can be applied to simple structures using
1.5GPeak(Sa), (ASCE 4)
• LDT is more accurate than LSP, but can result in
responses significantly different than nonlinear structural
behavior.
• Nonlinear responses can approximately be accounted by
F, and damping.
27 May 2015 Seismic Lessons learned Panel Meeting 5
Introduction
![Page 6: Analysis Procedures to Estimate Seismic Demands for ... · Outline •Background Introduction •Classification of Analysis Procedures NSP •Nonlinear Static Pushover NDT •Nonlinear](https://reader033.vdocuments.site/reader033/viewer/2022052918/5ade68c27f8b9afd1a8b5a87/html5/thumbnails/6.jpg)
NSP - Nonlinear Static Pushover
Advantages
• Simple analysis and has been widely used in industry.
• Can account for stiffness degradation.
• Simulates first mode behavior.
• Most validations are done for lateral behavior of tall buildings.
• Uses simplified displacement amplification factors.
• Computationally efficient
• Can use acceleration loading from prior linear dynamic analysis (Two step method).
Disadvantages
• Approximate method.
• Multiple directional runs are needed.
• Inaccurate responses for irregular
structures.
• Few studies are done on vertical
responses.
• Doesn’t account for higher mode
responses.
• Inaccurate local responses such as
joint rotations.
• Doesn’t typically account for changes
in dynamic responses due to structural
degradation.
• Simplified displacement amplification
factors with limiting values can result in
inaccurate responses.
27 May 2015 Seismic Lessons learned Panel Meeting 6
NSP
![Page 7: Analysis Procedures to Estimate Seismic Demands for ... · Outline •Background Introduction •Classification of Analysis Procedures NSP •Nonlinear Static Pushover NDT •Nonlinear](https://reader033.vdocuments.site/reader033/viewer/2022052918/5ade68c27f8b9afd1a8b5a87/html5/thumbnails/7.jpg)
NDT - Nonlinear Dynamic Time-history
Advantages
• Widely recognized as the best
predictive procedure to simulate
nonlinear response.
• 3 directions of motions can be
analyzed simultaneously
Disadvantages
• Computationally expensive.
• Requires experience to perform nonlinear analysis.
• Complexity with modeling details; material hysteresis, partial collapse.
• Can be impractical for complex structural systems.
• Limited ability to capture elastic system damping.
• Response depends on selected time history sets.
• Probabilistic or averaging of multiple sets may need to be used to address the variability.
• Evaluations may require evaluations at multiple demand levels.
27 May 2015 Seismic Lessons learned Panel Meeting 7
NDT E
art
hquake M
agnitude
Deflection
![Page 8: Analysis Procedures to Estimate Seismic Demands for ... · Outline •Background Introduction •Classification of Analysis Procedures NSP •Nonlinear Static Pushover NDT •Nonlinear](https://reader033.vdocuments.site/reader033/viewer/2022052918/5ade68c27f8b9afd1a8b5a87/html5/thumbnails/8.jpg)
Moehle, Ghannoum, and Bozorgnia (2004)
• 3-Bay, 3-story RC frame model is subjected to seven ground accelerations recorded
during the 1994 Northridge.
• Chosen records from a single earthquake to exclude the earthquake-to-earthquake
variability.
• Selected record sites in the same general area to reduce spatial variability.
• Response is sensitive to the ground motions, with responses varying from almost no
yielding of longitudinal steel to total collapse.
27 May 2015 Seismic Lessons learned Panel Meeting 8
NDT
Variability of Collapse as a Function of Input ground Motions
![Page 9: Analysis Procedures to Estimate Seismic Demands for ... · Outline •Background Introduction •Classification of Analysis Procedures NSP •Nonlinear Static Pushover NDT •Nonlinear](https://reader033.vdocuments.site/reader033/viewer/2022052918/5ade68c27f8b9afd1a8b5a87/html5/thumbnails/9.jpg)
MMP – Multi-Mode Pushover
Advantages
• Accounts for higher mode responses and local behaviors.
• Can simulate adaptive structural nonlinear ductile behavior.
• Can simulate site-specific displacement amplification factors.
• More accurate than NSP, and more efficient and practical than NDT for large models.
• Validations are done primarily for lateral responses of tall structures.
Disadvantages
• No single method has been
accepted for general use.
• Requires validation for general
use.
• Approximate method.
• Multiple directional runs are
needed.
• Not many studies are done on
vertical responses.
• Computationally more
demanding than NSP for
complex structures.
27 May 2015 Seismic Lessons learned Panel Meeting 9
MMP
![Page 10: Analysis Procedures to Estimate Seismic Demands for ... · Outline •Background Introduction •Classification of Analysis Procedures NSP •Nonlinear Static Pushover NDT •Nonlinear](https://reader033.vdocuments.site/reader033/viewer/2022052918/5ade68c27f8b9afd1a8b5a87/html5/thumbnails/10.jpg)
• Elnashai (2002), Antoniou et al. (2002), FEMA 273 (1997), ATC-40
(1996) proposed a simple approach which considers a single
seismic load pattern intended to capture the effect of multi-mode
response by selecting the loading pattern based on a combination of
multiple modes.
• Chopra and Goel (2000) proposed an equivalent SDOF time domain
analyses for determination of modal demands (also acknowledge a
spectral based estimation of demand).
• Gupta and Kunnath (2000), Chopra (2001), and many others
proposed rigorous approaches involve calculating the response of
each modal pushover separately then combining the effects.
27 May 2015 Seismic Lessons learned Panel Meeting 10
MMP
Representative Proposed MMP methods
![Page 11: Analysis Procedures to Estimate Seismic Demands for ... · Outline •Background Introduction •Classification of Analysis Procedures NSP •Nonlinear Static Pushover NDT •Nonlinear](https://reader033.vdocuments.site/reader033/viewer/2022052918/5ade68c27f8b9afd1a8b5a87/html5/thumbnails/11.jpg)
• Kalkan and Kunnath (2006) proposed an adaptive modal pushover analysis based on incremental loading and combination (adaptive coupled). Displacement is based on pushover with step size based on constant energy. Demand is based on matching modal pushover curves to constant ductility spectra at the pushover ductility.
• Aydinoglu (2003) proposed displacement based pushover analysis based on equal displacement / equal energy approximation (also used in ATC 40 and FEMA 440) for use in modal scaling for pushover increments
– Modal pushover based on elastic spectral displacement scaled by mass participation (equal displacement approximation) considering amplification factors
– Incremental loading
– Evaluation stress state at each increment
– Inelastic seismic demand is evaluated at the estimated spectral displacement
– Does not rely on time history methods
27 May 2015 Seismic Lessons learned Panel Meeting 11
MMP
Representative Proposed MMP methods
![Page 12: Analysis Procedures to Estimate Seismic Demands for ... · Outline •Background Introduction •Classification of Analysis Procedures NSP •Nonlinear Static Pushover NDT •Nonlinear](https://reader033.vdocuments.site/reader033/viewer/2022052918/5ade68c27f8b9afd1a8b5a87/html5/thumbnails/12.jpg)
Kalkan and Kunnath (2006)
• Figure on left shows pushover curve for a mode
• Figure on right shows matching the pushover curve to inelastic
demand spectra defined in acceleration-displacement space
27 May 2015 Seismic Lessons learned Panel Meeting 12
MMP
Representative Proposed MMP methods
![Page 13: Analysis Procedures to Estimate Seismic Demands for ... · Outline •Background Introduction •Classification of Analysis Procedures NSP •Nonlinear Static Pushover NDT •Nonlinear](https://reader033.vdocuments.site/reader033/viewer/2022052918/5ade68c27f8b9afd1a8b5a87/html5/thumbnails/13.jpg)
Increment Adaptive Modal Superposition
• The approach is theoretically based on incremental piecewise
approach to nonlinear time history analysis.
• Assuming that increments are small and within an increment
the stiffness and mass matrices are constant, the incremental
motion can be resolved into incremental modal components.
• The approach relies on the assumption that the increment
step size is taken small enough such that the modal matrix
can be assumed to not change within the increment.
• The response is resolved similarly to a response spectrum
analysis within each increment. Each increment include
loading the modal load shapes for a static pushover step and
recombination using the complete quadratic procedure.
27 May 2015 Seismic Lessons learned Panel Meeting 13
MMP
![Page 14: Analysis Procedures to Estimate Seismic Demands for ... · Outline •Background Introduction •Classification of Analysis Procedures NSP •Nonlinear Static Pushover NDT •Nonlinear](https://reader033.vdocuments.site/reader033/viewer/2022052918/5ade68c27f8b9afd1a8b5a87/html5/thumbnails/14.jpg)
27 May 2015 Seismic Lessons learned Panel Meeting 14
Update component stiffnesses based on
latest strain state
Extract natural frequencies and
mode shapes and perform mode
tracking
Calculate displacement
amplification factor and acceleration to
be applied to each mode
Perform response spectrum analysis,
modal CQC, directional
combination, & compute structural
responses
Combined non-seismic and seismic load effects
including seismic soil pressure.
Apply non-seismic loads to unstressed
structure
Select pushover direction and mode shape for seismic sign to update stiffness
𝑅 = 1.0 𝑅1 + 0.4 𝑅2 + 0.4 𝑅3
Start
Incremental MMP Pushover Methodology Implemented for PF-4
MMP
![Page 15: Analysis Procedures to Estimate Seismic Demands for ... · Outline •Background Introduction •Classification of Analysis Procedures NSP •Nonlinear Static Pushover NDT •Nonlinear](https://reader033.vdocuments.site/reader033/viewer/2022052918/5ade68c27f8b9afd1a8b5a87/html5/thumbnails/15.jpg)
Component Backbone Curve
27 May 2015 Seismic Lessons learned Panel Meeting 15
(Figure taken from Visual Catalog of Reinforced Concrete Bridge Damage, California Department of Transportation Structure
Maintenance and Investigations 2006)
Brittle Behavior Ductile Behavior
MMP
![Page 16: Analysis Procedures to Estimate Seismic Demands for ... · Outline •Background Introduction •Classification of Analysis Procedures NSP •Nonlinear Static Pushover NDT •Nonlinear](https://reader033.vdocuments.site/reader033/viewer/2022052918/5ade68c27f8b9afd1a8b5a87/html5/thumbnails/16.jpg)
IMK Validation against test result
27 May 2015 Seismic Lessons learned Panel Meeting 16
MMP
![Page 17: Analysis Procedures to Estimate Seismic Demands for ... · Outline •Background Introduction •Classification of Analysis Procedures NSP •Nonlinear Static Pushover NDT •Nonlinear](https://reader033.vdocuments.site/reader033/viewer/2022052918/5ade68c27f8b9afd1a8b5a87/html5/thumbnails/17.jpg)
Displacement Amplification Factor
• Displacement amplification factors based on FEMA 356,
FEMA 440, or alternate methods are intended to be code
friendly and are based on typical design spectra.
• A site-specific and structure specific displacement
amplification factors may be calculated
– Performing time response history analyses for an elastic single degree
of freedom (SDOF) model and for a nonlinear SDOF model with
damaged hysteresis having the same initial stiffness but with a bilinear
non-linear representation.
27 May 2015 Seismic Lessons learned Panel Meeting 17
MMP
![Page 18: Analysis Procedures to Estimate Seismic Demands for ... · Outline •Background Introduction •Classification of Analysis Procedures NSP •Nonlinear Static Pushover NDT •Nonlinear](https://reader033.vdocuments.site/reader033/viewer/2022052918/5ade68c27f8b9afd1a8b5a87/html5/thumbnails/18.jpg)
Cμ Comparisons to Ramirez
27 May 2015 Seismic Lessons learned Panel Meeting 18
MMP
Horizontal Vertical
• Ramirez C factors are derived from many time histories based on a spectra
with peak accelerations between 1 and 10 Hz.
• PF-4 horizontal matches relatively well because the spectra also has a peak
between 1 and 10 Hz.
• PF-4 vertical differs because the peak is around 15 Hz and no plateau.
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1
2
3
4
5
6
7
8
9
10
0 1 10 100
Sp
ectr
al A
cce
lera
tio
n, g
Cμ
(am
plif
ica
tio
n fa
cto
r)
Frequency, Hz
Ramirez, HR = 0.05, R = 1.50 Ramirez, HR = 0.05, R = 6.00
HR = 0.05, R = 1.50, Full Loop HR = 0.05, R = 6.00, Full Loop
Vertical Design Spectra
Fault-Normal ExcitationFault-Normal ExcitationFault-Normal Excitation
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1
2
3
4
5
6
7
8
9
10
0 1 10 100
Sp
ectr
al A
cce
lera
tio
n, g
Cμ
(am
plif
ica
tio
n fa
cto
r)
Frequency, Hz
Ramirez, HR = 0.05, R = 1.50 Ramirez, HR = 0.05, R = 6.00
HR = 0.05, R = 1.50, Full Loop HR = 0.05, R = 6.00, Full Loop
Horziontal Design Spectra
Fault-Normal ExcitationFault-Normal ExcitationFault-Normal Excitation
![Page 19: Analysis Procedures to Estimate Seismic Demands for ... · Outline •Background Introduction •Classification of Analysis Procedures NSP •Nonlinear Static Pushover NDT •Nonlinear](https://reader033.vdocuments.site/reader033/viewer/2022052918/5ade68c27f8b9afd1a8b5a87/html5/thumbnails/19.jpg)
Cμ Enveloping and Smoothing
• Displacement amplification factors (Cμ) can have large variation between
two relatively close frequencies, leading to sensitivity in response.
• Cμ factors have been recalculated using enveloping and smoothing.
27 May 2015 Seismic Lessons learned Panel Meeting 19
MMP
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
1 10
Cμ
(am
plif
ica
tio
n fa
cto
r)
Frequency, Hz
HR = 0.10, R = 1.25 HR = 0.10, R = 2.00 HR = 0.10, R = 5.00
Fault-Parallel ExcitationFault-Parallel Excitation
Solid lines = Raw Cμ DataDashed lines = Enveloped and
Smoothed Cμ Data
![Page 20: Analysis Procedures to Estimate Seismic Demands for ... · Outline •Background Introduction •Classification of Analysis Procedures NSP •Nonlinear Static Pushover NDT •Nonlinear](https://reader033.vdocuments.site/reader033/viewer/2022052918/5ade68c27f8b9afd1a8b5a87/html5/thumbnails/20.jpg)
27 May 2015 Seismic Lessons learned Panel Meeting 20
• Nonlinearity in shear wall
• 14 Increments (first is 0.5x remainder are 0.25x, total of 3.75x)
• 5 Modes per increment
0
100
200
300
400
500
600
700
0 0.002 0.004 0.006 0.008
Stre
ss (
psi
)
Strain
Input Shear Wall Backbone Curve
MMP
Shifting of Structural Frequencies
![Page 21: Analysis Procedures to Estimate Seismic Demands for ... · Outline •Background Introduction •Classification of Analysis Procedures NSP •Nonlinear Static Pushover NDT •Nonlinear](https://reader033.vdocuments.site/reader033/viewer/2022052918/5ade68c27f8b9afd1a8b5a87/html5/thumbnails/21.jpg)
Shifting of Structural Frequencies
27 May 2015 Seismic Lessons learned Panel Meeting 21
0
2
4
6
8
10
12
14
0 1 2 3 4
Freq
uen
cy (
Hz)
DBE
mode 1
mode 2
mode 3
mode 4
mode 5
MMP
![Page 22: Analysis Procedures to Estimate Seismic Demands for ... · Outline •Background Introduction •Classification of Analysis Procedures NSP •Nonlinear Static Pushover NDT •Nonlinear](https://reader033.vdocuments.site/reader033/viewer/2022052918/5ade68c27f8b9afd1a8b5a87/html5/thumbnails/22.jpg)
3 DOF Validation Example
27 May 2015 Seismic Lessons learned Panel Meeting
Deformed shape
comparisons at
different scale
factors
Story drift comparisons for bottom, middle, and
top story at different scale factors.
MMP
22
![Page 23: Analysis Procedures to Estimate Seismic Demands for ... · Outline •Background Introduction •Classification of Analysis Procedures NSP •Nonlinear Static Pushover NDT •Nonlinear](https://reader033.vdocuments.site/reader033/viewer/2022052918/5ade68c27f8b9afd1a8b5a87/html5/thumbnails/23.jpg)
Portal Frame Validation Example
• Displacements at the mid-height of the column (Node 7) and top of
the column (Node 13) are compared to time history results
27 May 2015 Seismic Lessons learned Panel Meeting
MMP
23
![Page 24: Analysis Procedures to Estimate Seismic Demands for ... · Outline •Background Introduction •Classification of Analysis Procedures NSP •Nonlinear Static Pushover NDT •Nonlinear](https://reader033.vdocuments.site/reader033/viewer/2022052918/5ade68c27f8b9afd1a8b5a87/html5/thumbnails/24.jpg)
2D Model Validation Example
• The pushover matches
time history story drifts
with a maximum of
percent difference of 7%
and 14% for the first and
second stories,
respectively.
27 May 2015 Seismic Lessons learned Panel Meeting
14%
7%
6%
7%
MMP
24
![Page 25: Analysis Procedures to Estimate Seismic Demands for ... · Outline •Background Introduction •Classification of Analysis Procedures NSP •Nonlinear Static Pushover NDT •Nonlinear](https://reader033.vdocuments.site/reader033/viewer/2022052918/5ade68c27f8b9afd1a8b5a87/html5/thumbnails/25.jpg)
2D Model Validation Example
• In the linear range, the pushover matches generally matches the average of the 11 time histories.
• When the structure is nonlinear, the pushover generally meets or exceeds the mean + one standard deviation (84th percentile) of the 11 time histories.
27 May 2015 Seismic Lessons learned Panel Meeting 25
MMP
![Page 26: Analysis Procedures to Estimate Seismic Demands for ... · Outline •Background Introduction •Classification of Analysis Procedures NSP •Nonlinear Static Pushover NDT •Nonlinear](https://reader033.vdocuments.site/reader033/viewer/2022052918/5ade68c27f8b9afd1a8b5a87/html5/thumbnails/26.jpg)
Conclusions
• LSP and LDT are useful design tools for new regular
structures with sufficient ductile details.
• NDT can be impractical for fragility analyses for
nonlinear complex structures without ductile details,
even with today’s high-performance computers.
• MMP is more accurate than NSP for irregular structures
with higher mode effects.
• MMP with proper validation is a reliable and credible
inelastic analysis procedure, and should be considered
for response calculations.
27 May 2015 Seismic Lessons learned Panel Meeting 26
Summary