Professional Documents
Culture Documents
Structures Congress Lignos DKK 2011
Structures Congress Lignos DKK 2011
dimitrios.lignos@mcgill.ca
2
Degengolb Engineers; 300 Frank H. Ogawa Plaza, Suite 450, Oakland, CA 94612;
PH (510) 250-1242; email: cputman@degenkolb.com
3
University of California at Irvine, Department of Civil and Environment Engineering
EG E4141; CA 92697-2175; PH (949) 824-9866; FAX (949) 824-2117; email:
zareian@uci.edu
4
Stanford University, Department of Civil and Environmental Engineering, Room
231; 473 Via Ortega Stanford, CA, 94305; PH (650) 723-4129; FAX (650) 723-7514;
email: krawinkler@stanford.edu
ABSTRACT
INTRODUCTION
freedom (MDOF) response, and that improved nonlinear analysis techniques to more
reliably address MDOF effects were needed. This paper summarizes part of a study
that was initiated by the National Institute of Standards and Technology (NIST) to
address this need. Using FEMA 440 as a starting point, the goal is to improve
nonlinear MDOF modeling for structural design practice by providing guidance on:
(1) the minimum level of MDOF model sophistication necessary to make
performance-based engineering decisions and (2) selection of appropriate nonlinear
Downloaded from ascelibrary.org by GEORGE MASON UNIVERSITY on 07/21/13. Copyright ASCE. For personal use only; all rights reserved.
analysis methods.
Steel Moment Resisting Frames: The structures comprise 3-bay moment resisting
frames that serve as the lateral load resisting system of steel buildings. The final
report of this project (NIST, 2010) describes the SSMFs in detail. The subset consists
of three structures that have been designed based on response spectrum analysis
(RSA). In summary, these buildings are,
• 2-story SSMF (Archetype ID 2-Dmax-RSA)
• 4-story SSMF (Archetype ID 4-Dmax-RSA)
• 8-story SSMF (Archetype ID 8-Dmax-RSA)
Reinforced Concrete Shear Walls: The salient features of the subset of RCSWs
utilized in this study are summarized below. The subset consists of three structures,
designated here as,
• 2-story ATC-76-1 RCSW (Archetype ID 12)
• 4-story ATC-76-1 RCSW (Archetype ID 13)
• 8-story ATC-76-1 RCSW (Archetype ID 14)
Note that the structures have been designed with a low 0.075Agfc’ axial stress level.
Steel Moment Resisting Frames: The bare frame structure is analyzed with the
latest version of either the Drain-2DX program (Prakash et al., 1993) or the Open
System for Earthquake Engineering Simulation (OpenSees) platform
results and provide different values for different steel W-sections as a function of
geometric section properties and material properties that control deterioration in
strength and stiffness due to local and lateral torsional buckling (Lignos and
Krawinkler, 2011). This component model is referred to as the “modified IK model,”
and analysis performed based on this component model is referred to as Analyt.M1.
Results obtained with Analyt.M1 account for cyclic deterioration in NRHA but they
do not account for cyclic deterioration in a pushover analysis, which is based on the
initial backbone curve. An example of global pushover curves for the 4- and 8-story
SSMFs is shown in Figure 1 after utilizing the Analyt.M1 component model.
NORMALIZED BASE SHEAR vs ROOF DRIFT RATIO-Analyt.M1 NORMALIZED BASE SHEAR vs ROOF DRIFT RATIO-Analyt.M1
ATC76-1 Steel SMF, 4-Story-RSA-Dmax,T1=1.56sec ATC76-1 Steel SMF, 8-Story-RSA-Dmax,T1=2.14sec
1st Mode Lateral Load Pattern 1st Mode Lateral Load Pattern
0.2 0.2
1-Bay 1-Bay
3-Bay 3-Bay
Norm. Base Shear V/W
Norm. Base Shear V/W
0.15 0.15
0.1 0.1
0.05 0.05
0 0
0 0.01 0.02 0.03 0.04 0 0.005 0.01 0.015 0.02 0.025 0.03
Roof Drift θ r=δ r/H [rad] Roof Drift θ r=δ r/H [rad]
Figure 1. Comparison of global pushover curves (VI/W –θr) of 1- and 3-bay steel SMF
analytical models, 4-, and 8-story steel SMFs.
as 20% of the peak confined concrete stress). Confined concrete model parameters
were varied over the wall height at locations where transverse reinforcement
changed. Concrete tensile strength (ft), concrete tensile modulus (Et), and unloading
parameter (λ), which defines the unloading slope in terms of the initial concrete
modulus (unloading slope = λE0), were selected based on the information provided
by Orakcal et al. (2006). Reinforcement was modeled as a “hysteretic material” in
OpenSees. Expected values of yield and ultimate strength of reinforcement were
taken as 68 ksi and 100 ksi, respectively. A tensile strain value of 0.05 was selected
to correspond to failure associated with rebar buckling and subsequently rebar
fracture. After reaching a strain of 0.05, the stress capacity of the reinforcing
bar drops to near zero. Flexural behavior was modeled by 6”x6” fiber elements.
Shear behavior was modeled by a translational shear spring. Such a spring was
inserted in every story. The properties of the shear spring were based on the NIST
(2010) study.
Given a prescribed lateral load pattern, there are different options for modeling a
structure for pushover analysis and for selecting the method for target displacement
prediction. In general, the latter is based on predicting the displacement demand for
an equivalent single-degree-of-freedom (SDOF) system that represents the first mode
characteristics of the multi-degree-of-freedom (MDOF) structure and mapping this
demand back to the global pushover curve to find the point at which the structure
should be evaluated (sometimes referred to as performance point). In the ASCE/SEI
41-06 coefficient method (ASCE, 2007) this process is greatly simplified and the
equivalent SDOF system does not become an explicit part of the target displacement
estimation. Figures 2 and 3 show an example of the global pushover analysis of a 2-
story SSMF and 4-story RCSW structure, respectively, together with an equivalent
single degree of freedom system.
Pushover Analysis Options for SSMFs: The following two options are explored for
models of structural components, which then are assembled in the OpenSees analysis
platform:
• ASCE41: ASCE/SEI 41-06 component models are used, but assuming a post-
capping stiffness obtained by linearly connecting peak point C and point E of the
generic ASCE/SEI 41-06 model. This modification is made in order to avoid
numerical analysis stability problems and to conform better to data and analysis
models developed over the last decade.
• Analyt.M1: Modified IK component model with monotonic backbone curve is
used, i.e., ATC-72-1 analysis option 1 (PEER/ATC, 2010); cyclic deterioration is
not reflected in pushover analysis.
Pushover Analysis Options for Reinforced Concrete Shear Walls: In this study
Downloaded from ascelibrary.org by GEORGE MASON UNIVERSITY on 07/21/13. Copyright ASCE. For personal use only; all rights reserved.
the following options are explored for models of structural components, which then
are assembled in the OpenSees analysis platform:
• ASCE41: ASCE/SEI 41-06 component models are used, but assuming a post-
capping stiffness obtained by linearly connecting peak point C and point E of the
generic ASCE/SEI 41-06 model.
• FM: Fiber model described in NIST (2010) is utilized to represent flexural
behavior, and a story shear spring is used to represent shear behavior.
The FM analysis option was executed for all three RCSWs, whereas the
ASCE41 option was explored only for the 4-story RCSW.
NORMALIZED BASE SHEAR vs ROOF DRIFT RATIO –Analyt.M1 EQUIVALENT SDOF SYSTEM-Analyt.M1
ATC76-1 Steel SMF, 2-Story-RSA-Dmax,1-bay,T =0.95sec ATC76-1 Steel SMF, 2-Story-RSA-Dmax,1-bay,T 1=0.95sec
1
st
1 Mode Lateral Load Pattern 0.6
0.6
0.5
0.5
Norm. Base Shear V*/W*
Norm. Base Shear V/W
0.4
0.4
0.3 0.3
0.2 0.2
V(I) 0.1
0.1
V(I+P-Δ)
Trilinear
0 0
0 0.01 0.02 0.03 0.04 0.05 0.06 0 0.01 0.02 0.03 0.04 0.05 0.06
Roof Drift Ratio θ r=δ r/H [rad] Drift Ratio θ *=δ */H* [rad]
Target Displacement Options for Steel Moment Resisting Frames and Reinforced
Concrete Shear Walls
• ASCE41: Target displacement obtained from ASCE/SEI 41-06 coefficient
method. For a shear wall system the target displacement is obtained with an
idealized trilinear force-displacement curve recommended in ASCE 41-06 as
shown in Figure 3a. This curve utilizes an effective elastic stiffness obtained by
placing a line through the displacement at 0.6Vy.
• EqSDOF: Target displacement based on median displacement obtained from a
first mode equivalent SDOF system and NRHA using the 44 ground motions of
the FEMA P-695 set and the analysis tool IIIDAP (Lignos, 2009). Equivalent
SDOF properties are obtained from the base shear VI – roof displacement
pushover curve (not the VI+P- – roof displacement pushover curve), which
implies that P-delta effects are accounted for approximately in the development of
the equivalent SDOF system (see Figure 2). In the case of shear wall structures in
which the global pushover curve shows a clear kink around cracking, the
idealized force-displacement curve is better represented by a multi-linear curve
that has at least 4 stiffnesses (pre-cracking, post-cracking, post-yielding, and post-
capping) as shown in Figure 3.
correspond to three scale factors (SF) of median ground motion intensity (SF=0.5, 1.0
and 2.0)
NORMALIZED BASE SHEAR vs ROOF DRIFT RATIO –FM NORMALIZED BASE SHEAR vs ROOF DRIFT RATIO -FM-Eq.SDOF
ATC76-1 RCSW, 4-Story-RSA-Dmax-Low ,T =0.66sec
1 ATC76-1 RCSW, 4-Story-RSA-Dmax-Low,0.66sec
1st Mode Lateral Load Pattern 1st Mode Lateral Load Pattern
0.5 0.5
0.4 0.4
Norm. Base Shear V/W
0.2 0.2
(a) (b)
Figure 3. Utilization of global pushover of 4-story RCSW to develop an equivalent
SDOF system; (a) ASCE 41 procedure; (b) Eq.SDOF system
Findings for Special Steel Moment Frames: Only a summary of results from all the
studies is presented herein due to brevity. For illustration, peak story drift ratios, peak
story shear forces and peak floor overturning moments for a scale factor of 2 are
compared between NSP and NRHA for the 4- and 8-story SSMFs. A summary of
findings for the 2-story frame structure is also included. It should be noted that story
shear forces are obtained as the summation of the inertia and P-Delta forces noted as
VI+P-Δ. Story shear forces are normalized with respect to the seismic weight W of the
frame structure. Similarly, floor overturning moments are based on VI+P-Δ multiplied
by the individual story heights and are normalized with respect to the seismic weight
W multiplied by the total height H of the frame structure.
2-story Steel Moment Resisting Frame: In the Analyt.M1 component model the
post-yield stiffness is caused by a 10% increase in My regardless of the amount of
pre-capping plastic hinge rotation capacity.
The ASCE/SEI 41-06 criteria for post-capping behavior result in a steep post-
capping tangent stiffness leading to rapid deterioration. This does not correspond to
statistical data from component tests on which the Analyt.M1 model is based (Lignos
and Krawinkler, 2009, 2011). The ASCE/SEI 41-06 coefficient method pays no
attention to the aforementioned differences, i.e., it results in the same target
displacement prediction for both the ASCE41 and Analyt.M1 pushovers.
All target displacement predictions are within 10% of the median roof drift
obtained from NRHA. NSP results compare well with NRHA results, except for the
story drift predictions for SF = 2.0 based on the ASCE/SEI 41-06 pushover. The
reason is the negative tangent stiffness in the post-yield range of the ASCE41
pushover.
4- and 8-story Steel Moment Resisting Frames: In theses structures the ASCE/SEI
Downloaded from ascelibrary.org by GEORGE MASON UNIVERSITY on 07/21/13. Copyright ASCE. For personal use only; all rights reserved.
Floor
3 5
2 3
1 1
0 0.01 0.02 0.03 0.04 0.05 0.06 0 0.01 0.02 0.03 0.04
Peak Story Drift Ratio (θ si=(δ fi-δ fi-1)/hi) Peak Story Drift Ratio (θ si=(δ fi-δ fi-1)/hi)
Figure 4. Peak story drift ratios for SF=2 for 4- and 8-story SSMFs
NSP story shear predictions are not good representations of NRHA results in
the inelastic range (SF = 2.0). Story shears are consistently underestimated,
particularly in the upper stories. But also the maximum NSP base shear, which
corresponds to the peak of the VI+P-delta pushover curves is about 25% below the
NRHA median base shear. The reason is dynamic redistribution, which amplifies
story shear forces compared to those obtained from a predetermined lateral load
pattern. If story shears are an important performance consideration, then the validity
of quantitative values obtained from a pushover analysis diminished for the 4- and 8-
story steel SMFs (see Figures 5a and 5b). Similar observations apply to floor
Floor
3 5
2 3
1 1
0 0.1 0.2 0.3 0.4 0 0.1 0.2 0.3 0.4
Peak Normalized Story Shear (V(I+P-Δ )/W) Peak Normalized Story Shear (V(I+P-Δ )/W)
(a) (b)
PEAK NORM. FLOOR OVERTURNING MOMENTS, OTM(I+P-Δ ), SF=2 PEAK NORM. FLOOR OVERTURNING MOMENTS, OTM(I+P-Δ ), SF=2
ATC76-1 Steel SMF, 4-Story-RSA-Dmax, 1-Bay,1.56sec, 44 Records ATC76-1 Steel SMF, 8-Story-RSA-Dmax, 1-Bay,2.14sec, 44 Records
roof roof
NRHA Median NRHA Median
ASCE41-ASCE41 8 ASCE41-ASCE41
Analyt.M1-ASCE41 ASCE41-EqSDOF
4 Analyt.M1-EqSDOF 7 Analyt.M1-ASCE41
Analyt.M1-EqSDOF
6
Floor
Floor
3 5
2 3
1 1
0 0.05 0.1 0.15 0.2 0 0.05 0.1 0.15 0.2
Peak Normalized Floor Overturning Moment (Σ [V(I+P-Δ )*h ]/(W*H)) Peak Normalized Floor Overturning Moment (Σ [V(I+P-Δ)*h ]/(W*H))
i i
(c) (d)
Figure 5. Peak normalized story shear forces and floor overturning moments for the
4- and 8-story SSMFs for SF=2
Findings for Reinforced Concrete Shear Walls: For all RCSW structures (2, 4, and
8 stories) the FM-ASCE41 NSP option (as recommended presently in ASCE/SEI 41-
06) leads to rather poor predictions of story drift ratios. An example can be seen from
Figure 6 for the 4-story RCSW. Therefore, emphasis was placed on a better
representation of the global pushover curve in the equivalent SDOF system in order
to obtain a better estimate of the target displacement. For this purpose the global
pushover curve was represented in a multi-linear manner rather than the trilinear
manner recommended in ASCE/SEI 41-06. An example of this improved SDOF
system can be seen in Figure 3b. The target displacement of the corresponding
equivalent SDOF system was predicted with the program IIIDAP (Lignos, 2009), by
subjecting the equivalent SDOF system to the set of 44 FEMA P-695 ground motions
and computing the median displacement, which was then mapped back to the MDOF
domain. This mapped value was used as target displacement to predict the EDPs
Downloaded from ascelibrary.org by GEORGE MASON UNIVERSITY on 07/21/13. Copyright ASCE. For personal use only; all rights reserved.
from the global pushover. The EDP predictions obtained in this manner (denoted as
FM-Eq.SDOF) show a large improvement over the predictions obtained from use of
the trilinear ASCE/SEI 41-06 equivalent SDOF system for SF=0.5 and 1.0. Figures 6
and 7 summarize the peak story drift ratios, normalized peak story shear forces and
peak normalized overturning moments along the height of the 4-story RCSW
structure for the two scale factors of interest.
PEAK STORY DRIFT RATIOS, SF=0.5 PEAK STORY DRIFT RATIOS, SF=1.0
ATC76-1 RCSW, 4-Story-RSA-Dmax-Low,0.66sec, 44 Records ATC76-1 RCSW, 4-Story-RSA-Dmax-Low,0.66sec, 44 Records
roof roof
NRHA Median NRHA Median
FM-ASCE41 FM-ASCE41
FM-EqSDOF FM-EqSDOF
4 ASCE41-ASCE41 4 ASCE41-ASCE41
Floor
Floor
3 3
2 2
1 1
0 0.005 0.01 0.015 0.02 0 0.005 0.01 0.015 0.02
Peak Story Drift Ratio (θ si=(δ fi-δ fi-1)/hi) Peak Story Drift Ratio (θ si=(δ fi-δ fi-1)/hi)
Figure 6. Peak story drift ratios for the 4- story RCSW for SF=0.5 and 1.0
As an alternative, EDP predictions for the 4-story RCSW were also obtained from
using ASCE/SEI 41-06 component models to perform the pushover analysis together
with the ASCE/SEI 41-06 target displacement procedure (ASCE41-ASCE41 option).
The challenge here was to determine moment and shear force strength properties
based on ASCE/SEI 41-06 recommendations. The values for bending strength
obtained from the FM model and computed from ACI criteria together with
ASCE/SEI 41-06 criteria for expected material properties differ by a significant
amount. Due to brevity, details about this model can be found in NIST (2010) report.
The estimated strength values (Mc) are significantly larger than would be
calculated from basic engineering models and using the reinforcement layout
provided in the NIST (2010) report. The FM peak moment value Mc can be obtained
only with a much larger axial force that was used in design of this structure. The
bending capacity of the wall is very sensitive to the axial force in the wall. A small
difference in axial force changes the bending capacity by a large amount, and
consequently the failure mode in the wall may change from bending to shear or vice
versa. Nonlinear dynamic response in a wall that deforms primarily in shear may be
very different from that deforming primarily in bending. This is the reason for the
large differences between NSP predictions (ASCE41-ASCE41) compared to those
obtained from FM-ASCE41 and FM-Eq.SDOF in story drift ratios (see Figure 6) and
peak story shear forces and overturning moments (see Figure 7). ASCE41-ASCE41
predictions were possible only for SF = 0.5 and 1.0, because for larger ground motion
scale factors the predicted roof drift exceeded the deformation capacity indicated by
the ASCE/SEI 41-06 pushover.
PEAK NORMALIZED STORY SHEAR FORCES, V(I+P-Δ ), SF=0.5 PEAK NORMALIZED STORY SHEAR FORCES, V(I+P-Δ ), SF=1.0
Downloaded from ascelibrary.org by GEORGE MASON UNIVERSITY on 07/21/13. Copyright ASCE. For personal use only; all rights reserved.
Floor
3 3
2 2
1 1
0 0.1 0.2 0.3 0.4 0 0.1 0.2 0.3 0.4
Peak Normalized Story Shear (V(I+P-Δ )/W) Peak Normalized Story Shear (V(I+P-Δ )/W)
PEAK NORM. FLOOR OVERT. MOMENTS, OTM(I+P-Δ ), SF=0.5 PEAK NORM. FLOOR OVERT. MOMENTS, OTM(I+P-Δ ), SF=1.0
ATC76-1 RCSW, 4-Story-RSA-Dmax-Low,0.66sec, 44 Records ATC76-1 RCSW, 4-Story-RSA-Dmax-Low,0.66sec, 44 Records
roof roof
NRHA Median NRHA Median
FM-ASCE41 FM-ASCE41
FM-EqSDOF FM-EqSDOF
4 ASCE41-ASCE41 4 ASCE41-ASCE41
Floor
Floor
3 3
2 2
1 1
0 0.05 0.1 0.15 0.2 0 0.05 0.1 0.15 0.2
Peak Normalized Floor Overturning Moment (Σ [V(I+P-Δ )*h ]/(W*H)) Peak Normalized Floor Overturning Moment (Σ [V(I+P-Δ )*h ]/(W*H))
i i
Figure 7. NSP to NRHA comparison, 4-story ATC-76-1 RCSW, three different NSP
options, SF = 0.5 and 1.0.
SUMMARY
This paper summarizes part of a study that was conducted in order to provide
guidance on the selection of nonlinear static procedures (NSPs) for steel special
moment frame and reinforced concrete shear wall structures. It also summarizes
limitations of use for NSPs to predict engineering demand parameters. Rather than
developing new, complex NSP methods, improvements on current NSP methods that
are utilized in various guideline documents are suggested.
ACKNOWLEDGEMENTS
This paper relies on results obtained under Task Order 6 of the NEHRP Consultants
Joint Venture (a partnership of the Applied Technology Council and Consortium of
Universities for Research in Earthquake Engineering), under Contract
SB134107CQ0019, Earthquake Structural and Engineering Research, issued by the
National Institute of Standards and Technology. The views expressed do not
necessarily represent those of the organizations identified above.
REFERENCES