Professional Documents
Culture Documents
Chapter 15
Key words: Seismic Performance; Performance Based Design; Seismic Demand; Capacity; ADRS Spectrum; Target
Displacement; Performance Objectives; Push-over Analysis; Capacity Spectrum; Static Analysis; Nonlinear
Analysis; Damage Control; Life safety, Collapse Prevention; Immediate Occupancy
Abstract: Performance based seismic engineering is the modern approach to earthquake resistant design. Rather than
being based on prescriptive mostly empirical code formulations, performance based design is an attempt to
predict buildings with predictable seismic performance. Therefore, performance objectives such as life-
safety, collapse prevention, or immediate occupancy are used to define the state of the building following a
design earthquake. In one sense, performance based seismic design is limit-states design extended to cover
the complex range of issues faced by earthquake engineers. This chapter provides a basic understanding of
the promises and limitations of performance based seismic engineering. The state-of-the-art methodologies
and techniques embodied in the two leading guidelines on this subject (ATC-40 and FEMA 273/274) are
introduced and discussed. Numerical examples are provided to illustrate the practical applications of the
methods discussed.
758 Chapter 15
15. Performance Based Seismic Engineering 759
Heavy
Moderate None
2%
15% 19%
Insignificant
64%
Figure 15-1. Damage State in 530 Buildings within 15 km of epicenter Surveyed After the 1994 Northridge Earthquake
outside the academic research settings is fairly homes were surveyed and although more than
limited and not up to the standards needed for a 90 percent of the homes in the sample were old
reliable prediction of seismic performance. and built prior to the 1971 San Fernando Valley
For example, following the Northridge earthquake the cases of moderate to high
earthquake, the Applied Technology Council damage were infrequent (less than 2-percent).
conducted a survey of 530 buildings which Most occurrences of serious damage were
were located within 300 meters of strong- located in foundation systems and were
motion recording sites(15-4). From the total of associated with localized site conditions such as
530 buildings which were located in the areas liquefaction, fissuring, and hillside slope
of strong shaking (San Fernando Valley, Santa failures. Here again, design analysis would
Monica, and West Los Angeles) with peak have predicted much larger damage percentage
ground acceleration in their vicinity ranging than the 2-percent number reported by Crandell.
from 0.15g to 1.78g, only 10 (less than two- Large uncertainties also exist in our
percent) showed heavy damage, a total of 78 estimates of design ground motion. For
buildings (about 15-percent) showed moderate example, median estimates of spectral
damage and 340 (64-percent) were marked by accelerations for a magnitude 7.0 event at
insignificant damage (Figure 15-1). If response rupture distance of 10 km obtained from
of these buildings were predicted by standard various attenuation relations can vary by as
design analysis techniques, a far worse picture much as 50 percent(15-6). If the uncertainties
would have been predicted. associated with other source and regional
Crandell(15-5) performed a similar variables are also considered, the variance
statistically-based study of the seismic could be significantly larger. Most attenuation
performance of residential buildings located relations are updated every few years (Figure
within a 10-mile radius of the Northridge 15-3), indicating that there are still many things
earthquake epicenter (Figure 15-2). Three to be learned about the generation and
hundred forty one of the 375 randomly selected propagation of earthquake ground motion.
15. Performance Based Seismic Engineering 761
350
300
250
200
150
100 Roof
Walls
50
Foundation to Walls
0
Foundation
Sample Size
No Damage
Low Damage
Moderate Damage
High Damage
Figure 15-2. Description of Damage During the 1994 Northridge Earthquake to Single Family Dwellings Within a 10 Miles
Radius of the Epicenter (data from Crandell, 1997)
4.00
3.50
PSEUDO-RELATIVE VELOCITY (FT/SEC)
3.00
2.50
2.00
CAMPBELL (1991R)
CAMPBELL (1989)
0.50
0.00
0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00
UNDAMPED NATURAL PERIOD (SEC)
Figure 15-3. Evolution of a Typical Attenuation Relation (Spectral velocity estimates are shown for a magnitude 7.0 event
at 5.0 km for a strike-slip fault)
762 Chapter 15
Another source of uncertainty is critical single event (Mexico, 1985). Clearly, this is
shortage of recorded earthquake ground motion one of the areas where more information is
where they are needed most. Despite the needed for performance based design
tremendous growth in the number of earthquake Since PBSE is inherently multi-disciplinary
records during the past decade, the number of in nature, further educational efforts are also of
recordings from large earthquakes close by. vital importance in bringing PBSE to fruition
Figure 15-3(15-7) shows a bivariate histogram of by developing a common understanding of
horizontal components recorded in north and issues and a common PBSE language and
central America categorized by magnitude and vocabulary. Only a broad multi- disciplinary
epicentral distance, indicating practically no approach can succeed in reduction of
record of M >7.5 at distances less than 20 km. uncertainties, knowledge gaps, and common
All of the data for M>8 records come from a misunderstandings.
Figure 15-4. Distribution of Magnitude and Distance among Available Earthquake Records for North and Central America,
1933-1994 (M>5.5; PGA>0.05g)
15. Performance Based Seismic Engineering 763
Table 15-1. Combinations of Structural and Non-structural Levels to form Building Performance Levels (15-1)
Building Performance Levels
Structural Performance Levels
Non- SP-1 SP-2 SP-3 SP-4 SP-5 SP-6
structural Immediate Damage Life Safety Limited Structural Not
Performanc Occupancy Control Safety Stability Considered
e Levels (Range) (Range)
NP-A 1-A 2-A NR NR NR NR
Operational Operational
NP-B 1-B 2-B 3-B NR NR NR
Immediate Immediate
Occupancy Occupancy
NP-C 1-C 2-C 3-C 4-C 5-C 6-C
Life Safety Life Safety
NP-D NR 2-D 3-D 4-D 5-D 6-D
Reduced
Hazards
NP-E NR NR 3-E 4-E 5-E Not
Not Structural Applicable
Considered Stability
Legend
Commonly referenced Building Performance Levels (SP-NP)
Other possible combinations of SP-NP
NR Not recommended combinations of SP-NP
Thus, for the Nonlinear Static Procedure, a elements, which should be completely modeled
static pushover analysis is performed using a in the non-linear analysis. Secondary elements,
nonlinear analysis program for an increasing which do not significantly contribute to the
monotonic lateral load pattern. An alternative is buildings lateral force resisting system, do not
to perform a step by step analysis using a linear need to be included in the analysis.
program. The base shear at each step is plotted
again roof displacement. The performance point Possible
Performance
is found using the Capacity Spectrum Global
Point
Procedure[15-8,15-9,15-10] described in subsequent Base Capacity
Curve
Shear
sections. The individual structural components V
are checked against acceptability limits that
depend on the global performance goals. The
nature of the acceptability limits depends on
specific components. Inelastic rotation is
typically one of acceptability parameters for
beam and column hinges. The limits on Immediate Life Structural
inelastic rotation are based on observation from Occupancy Safety Stability
tests and the collective judgement of the Roof Displacement, R
development team.
The key step for the entire analysis is In concrete buildings, the effects of
identification of the primary structural earthquake shaking are resisted by vertical
frame elements or wall elements that are
766 Chapter 15
connected to horizontal elements (diaphragms) and consistent with the general methodology
at the roof and floor levels. The structural should be used for the effects of the
elements may themselves comprise of an foundations.
assembly of elements such as columns, beam, The response parameters of foundation
wall piers, wall spandrels etc. It is important to elements are dependent on structural as well as
identify the failure mechanism for these geotechnical components. Spread footings
primary structural elements and define their elements, for example, might consist of a rigid
non-linear properties accordingly. The structural plate component model of the
properties of interest of such elements are concrete footing bearing on soil represented by
relationships between the forces (axial, bending geotechnical components with appropriate
and shear) and the corresponding inelastic force-displacement properties. Some generic
displacements (displacements, rotations, drifts). models for typical foundation elements and
Earthquakes usually load these elements in a acceptance criterion for structural components
cyclic manner as shown in Figure 15-6a. For of the foundations are provided in ATC-40.
modeling and analysis purposes, these There is a large degree of uncertainty
relationship can be idealized as shown in Figure associated with both strength and stiffness of
15-6b using a combination of empirical data, the geotechnical components. Thus, ATC-40
theoretical strength and strain compatibility. recommends enveloping analysis to determine
Using the component load-deformation data the sensitivity of seismic performance to
and the geometric relationships among foundation behavior (See Figure 15-8).
components and elements, a global model of Guidance in provided for representative
the structure relates the total seismic forces on a properties of normally encountered soil
building to it overall lateral displacement to materials that are based on limited initial
generate the capacity curve. During the investigations in ATC-40. If the analysis shows
pushover process of developing the capacity sensitivity to foundation behavior than more
curve as brittle elements degrade, ductile detailed investigations and tests of geotechnical
elements take over the resistance and the result properties may be warranted.
is a saw tooth shape that helps visualize the Geotechnical properties are very ductile and
performance. Once the global displacement failure is rarely encountered. Thus, deformation
demand is estimated for a specific seismic limits of geotechnical components are not
hazard, the model is used to predict the explicitly defined. However, deformation of
resulting deformation in each component. The geotechnical components may affect the
ATC 40 document provides acceptability limits deformation and acceptability of components in
for component deformations depending on the the superstructure. It should also be noted that
specified performance level. geotechnical components tend to accumulate
residual displacements. This tendency may
15.3.5 Geotechnical effects affect the acceptability of a structure for higher
performance objectives such as Immediate
The deformation and movement of the Occupancy. Soil structure interaction also has
foundations of a building can significantly beneficial affects such as lower demands on
affect the seismic response and performance of structural members due to base rotation, lower
structures. As the structural components are forces due to uplift and damping effects that
represented by non-linear load-displacement reduce demand on the superstructure.
relationships, analogous techniques compatible
15. Performance Based Seismic Engineering 767
250
50
150
30
Lateral Load (kips)
-10 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 -50
Displacement Ductility Factor
-30
-150
-50
-250
-70
-350
-90
-450
-110
-18 -14 -10 -6 -2 2 6 10 14 18
Top Displacement (inches)
Backbone
F curve F C,D
C
Idealized
B component B
behavior
A D A E D
Ductile Semi-ductile
(deformation controlled)
F B,C,D
A E D
Brittle
(force controlled)
(15-1)
(b) Idealized component behavior from backbone curves
Column/Wall
Grade Beams
Spread Footing
Soil
Structural Components
Force Q
Envelope
Upper Qc
Lower Qc
Kstiff Kflexible
Actual Behavior
Displacement, d
Ti 2
Sd i = Sa i g (15-1)
4 2 i
Sd i = (15-4)
( PF1 1,roof )
Standard demand response spectra contain a
range of constant spectral acceleration and a Where 1 and PF1 are the modal mass
second range of constant spectral velocity, Sv. coefficient and participation factors for the first
Spectral acceleration and displacement at natural mode of the structure respectively. 1,roof
period Ti are given by: is the roof level amplitude of the first mode.
The modal participation factors and modal
2 T
Sai g = Sv , Sd i = i Sv (15-2) coefficient are calculated as:
Ti 2
n
The capacity spectrum can be developed ( wi i1 ) / g
from the pushover curve by a point by point PF1 = i =n1 (15-5)
(w 2 ) / g
conversion to the first mode spectral i i1
coordinates. Any point Vi (Base Shear), i (Roof i =1
n
2 the structure. The -factor (called -factor in
( wi i1 ) / g ATC-40) is a modification factor to account for
1 = n i =1 (15-6) the extent to which the actual building
n
wi / g ( wi i1 ) / g
2 hysteresis is well represented by the bilinear
i =1 i =1 representation of the capacity spectrum (See
Table 15-3 and Figure 15-11).
Where wi is the weight at any level i. The term o can be calculated using:
As displacement increase, the period of the
structure lengthens. This is reflected directly in 1 ED
o = (15-8)
the capacity spectrum. Inelastic displacements 4 E So
increase damping and reduce demand. The
Capacity Spectrum Method reduces the demand Where ED is the energy dissipated by damping
to find an intersection with the capacity and ESo is the maximum strain energy. The
spectrum, where the displacement is consistent physical significance is explained in Figure 15-
with the implied damping. 11.
Base Bilinear representation of
Shear
Capacity Spectrum
-V
Spectral Capacity Spectrum
Vi, i,Roof Acceleration
Sapi Keffective
Kinitial
Say
Roof Displacement -
Capacity Curve
Spectral
Acceleration
- Sa Sdy Sdpi
Spectral Displacement
Sai,Sdi
the spectral reduction factors are arbitrary and Reduced Response Spectrum
need farther study.
Spectral Displacement
Spectral
Acceleration
2.5CA
Elastic Response
Spectrum 5% Damped
SRA x 2.5CA
Intersection point of reduced demand
spectrum and capacity spectrum
Capacity Spectrum
Sapi
CV/T
Say
Bilinear SRV x CV/T
representation of
capacity Reduced Response Spectrum
Sdy Sdpi Sdp
Spectral Displacement
6. If the reduced demand spectrum intersects 1. For the global building response, verify
the capacity spectrum at Sapi, Sdpi or if the a. The lateral force resistance has not
intersection point Sdp is within 5% of Sdpi, degraded by more than 20% of the peak
then this point represents the performance resistance.
point. b. The lateral drift limits satisfy the limits
7. If the intersection point does not lie within given in the Table 15-5.
acceptable tolerance (5% of Sdpi or other) 2. Identify and classify the different elements
then select another point and repeat Steps 4 in the building in the following types:
to 7. The intersection point obtained in Step beam-column frames, slab-column frames,
6 can used as the starting point for the next solid walls, coupled walls, perforated walls,
iteration. punched walls, floor diaphragms and
Procedure B is also an iterative method to foundations.
find the performance point, which uses the 3. Identify all primary and secondary
assumption that the yield point and the post elements.
yield slope of the bilinear representation, 4. For each element type, identify the critical
remains constant. This is adequate for most components and actions to check as
cases, however, in some cases this assumption detailed in Chapter 11 of ATC-40.
may not be valid. Procedure C is graphical 5. The strength and deformation demands at
method that is convenient for hand analysis. the performance point should be equal to or
less than the capacities detailed in Chapter
15.3.7 Checking Performance at Expected 11 of ATC-40.
Maximum Displacement 6. The performance of secondary elements
(such as gravity load carrying members not
Once the performance point Sap, Sdp (which part of the lateral load resisting system) are
are in spectral ordinates) is found, the base reviewed for acceptability for the specified
shear (Vp) and roof displacement (p) at the performance level.
performance point are found using Equation 15- 7. Non-structural elements are checked for the
3 and 15-4. The following steps should be used specified performance level.
in the performance check:
15. Performance Based Seismic Engineering 773
1.2
T = 0.5 secs T = 1.0 secs
1
Spectral Acceleration (g)
0.8
T = 1.5 secs
Equal Displacement Approx
T = 2.0 secs
0.4
0.2
Capacity Spectrum Curve
0 Sd y Sd p1
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Spectral Displacement (inches)
1.2
T = 0.5 secs T = 1.0 secs
1
Spectral Acceleration (g)
0.2
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Spectral Displacement (inches)
Figure 15-15. Determination of Intersection Point and Comparison with the First Trial Performance Point
1.2
T = 0.5 secs T = 1.0 secs
1
Spectral Acceleration (g)
0.8
T = 1.5 secs
T = 2.0 secs
0.4
0.2
Capacity Spectrum Curve
0 Sd y Sd p2
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Spectral Displacement (inches)
1.2
T = 0.5 secs T = 1.0 secs
1
Spectral Acceleration (g)
0.8
T = 1.5 secs
New Reduced Demand Spectrum
0.2
Capacity Spectrum Curve
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Spectral Displacement (inches)
deformation at the performance point is given performance levels are similar to those defined
by the one intersection point where the ductility in ATC 40 (See Section 15.3.2).
factor calculated from the capacity spectrum FEMA-273 defines ground motion hazard
matches the value associated with the levels in a probabilistic basis. Four ground
intersected demand spectrum. motion hazard levels are defined
Another method for determining the
performance point is suggested by Fajfar(15-19). Earthquake Probability Mean Return
Here the ductility demand is determined using of Exceedence Period (years)
the equal displacement rule and the inelastic 50% in 50 years 72
design spectra. Another variant of the Capacity 20% in 50 years 225
Spectrum method called the Yield Point BSE-1 10% in 50 years 474
Spectra(15-20) has recently been suggested. Here BSE-2 2% in 50 years 2,475
the yield displacement is plotted on the abscissa
instead of the spectral displacement and Ry--T Where BSE is the Basic Safety Earthquake. The
relations or exact computations are used instead broad range of performance objectives
of equivalent viscous damping. recommended for a given earthquake hazard
levels are shown in Table 15-7
15.4 FEMA 273 and 274
Table 15-7. Rehabilitation Objectives(15-2)
Building Performance Levels
Performance Level
Performance Level
15.4.1 Introduction
Operational Level
Occupancy Level
NEHRP Guidelines for the Seismic
Prevention
Life Safety
Immediate
Rehabilitation of Buildings (FEMA-273)(15-2)
Collapse
and the associated commentary (FEMA-274)(15-
(1-A)
(3-C)
(1-B)
3)
was developed by the Building Seismic
Safety Council (BSSC) with subcontractors 50%/50
Earthquake Hazard Level
a b c d
American Society of Civil Engineering (ASCE) yrs
and the Applied Technology Council (ATC) 20%/50
e f g h
with the funding provided by the Federal yrs
BSE-1
Emergency Management Agency (FEMA). The
10%/50 i j k l
primary purpose of FEMA-273 was to provide yrs
technically sound and nationally acceptable BSE-2
guidelines for the seismic rehabilitation of 2%/50 m n o p
buildings. Although the document was written yrs
with the objective of performance based retrofit k+p = Basic Safety Objective
k+p+any of a, e, i or m; or b, f, j, or n = Enhanced
of existing structures, the procedures described
Objectives
therein are equally applicable for new design. o = Enhanced Objectives
Unlike the ATC-40 document, these k alone or p alone = Limited Objective
recommendations are applicable to all building c, g, d, h = Limited Objectives
materials and define acceptability limits for
linear as well as non-linear analysis. From Table 15-7, it is clear that FEMA-273
The basic procedure is similar to that specifies a two-level design to achieve the
recommended in ATC-40. The owner decides Basic Safety Objective (BSO), Life Safety
the performance object that needs to be Performance Level for BSE-1 demands and
achieved. The engineer then designs the retrofit Collapse Prevention Level for BSE-2 demands.
or new structure to achieve the performance However, for new structures it is possible to
objective. The definitions of the basic control ductility and configuration of the design
778 Chapter 15
to an extent that will permit those structures increase in the relative deformation between the
designed to achieve Life Safety Performance connected components.
Level for a BSE-1 level earthquake to also
avoid collapse for much larger events. 15.4.2.2 Horizontal Torsion
Two sets of earthquake hazard maps are In addition to the actual eccentricities
distributed with FEMA-273 and 274. One set between the centers of mass and centers of
provide key response acceleration for the rigidity, a additional accidental torsional
Maximum Considered Earthquake (MCE) moment should be included which may be
which in most areas represents a 2%/50 years produced by including a horizontal offset in the
exceedence level. The other uses 10%/50 years centers of mass equal to a minimum of 5% of
exceedence probability. Thus, it is possible to the horizontal dimension at a given floor level.
obtain a BSE-1 and BSE-2 level spectra from For buildings with rigid diaphragms, the
these maps. effects of torsion must be included when the
maximum displacement at any point in a
diaphragm exceeds the average displacement in
15.4.2 Mathematical Modeling that diaphragm by more than 10%. For linear
analysis, the effect of accidental torsion is
FEMA-273 provides four analysis amplified by a factor Ax:
procedures for systematic design and
2
rehabilitation of buildings. The Linear Static
(LSP) and Linear Dynamic Procedures (LDP) Ax = max (15-11)
are linearly elastic analysis, which may include 1.2 avg
geometric non-linearity. Also some material
non-linearity is also introduced by use of Where max and avg are the maximum and
cracked properties for concrete and masonry average displacements in a diaphragm. Ax is
components even though the analysis is linear. greater than 1 and not greater than 3.
In the Nonlinear Static (NSP) and Nonlinear If =max/avg is greater than 1.5, then a
Dynamic Procedures (NDP) material non- three-dimensional analysis is required. For two-
linearity is included in the analysis. dimensional analysis subject to this limitation,
the effect of torsion can included for LSP and
15.4.2.1 Basic Assumptions LDP by increasing the design forces and
In general, a three dimensional analysis displacement by . For NSP, the target
consisting of an assembly of elements and displacement is increased by and for NDP the
components is recommended. Three- amplitude of the ground acceleration record is
dimensional analysis is required when the increased by .
building has plan irregularities and when
torsional effects cannot be ignored or indirectly 15.4.2.3 Primary and Secondary Elements
captured. Primary elements are key parts of the
For buildings with flexible diaphragms, the seismic framing system required in the design
diaphragms may be individually modeled and to resist earthquake effects. These must be
analyzed as two-dimensional assemblies of evaluated to resist earthquake forces as well as
components and elements or three-dimensional gravity loads if required. Secondary elements
models with flexible elements. are not designed to be part of the lateral force
Explicit modeling of connections is not resisting system but must be evaluated to ensure
required if the connection is stronger than the they can simultaneously sustain earthquake
connected components or when the deflection induced deformation and gravity loads.
of the connection does not cause a significant For linear analysis procedures, the
secondary elements must not constitute more
15. Performance Based Seismic Engineering 779
than 25% of the total stiffness of the primary Table 15-8. Calculation of Element Capacities(15-2)
elements at any level and may not be included Parameter Deformation Force
in the analysis. For nonlinear procedures, the Controlled Controlled
Linear Procedures
stiffness of the primary as well as the secondary
Existing Material Mean value with Lower bound
elements must be included in the model. Strength allowance for (Mean Std
Additionally, the stiffness of non-structural strain hardening Dev)
elements must not exceed 10% of the total Existing Capacity m QCE QCE
lateral stiffness of any story. If this is exceeded, New Material Mean value Specified value
then the non-structural elements must be Strength
included in the model. New Capacity QCE QCE
Nonlinear Procedures
Deformation x N/A
15.4.2.4 Deformation and Force Capacity Deformation
Controlled Elements Existing Element limit
Elements can be classified as either Deformation Deformation N/A
deformation controlled or force controlled. A Capacity New limit
deformation controlled element is one that has Element
an associated deformation that is allowed to Strength Capacity N/A QCL
Existing
exceed yield value, that is, the maximum
Element
associated deformation of the element is limited Strength Capacity N/A QCL
by the ductility of the element. A force New Element
controlled element is one where the maximum = Knowledge factor
associated displacement is not allowed to m = Demand Modifier for expected ductility
exceed yield value. Elements with limited QCE = Expected Strength
QCL = Lower Bound Estimate of Strength
ductility shall be considered to be force
controlled. See Table 15-8 for calculation of
element capacities used to compare with
demands. 15.4.2.7 Diaphragms
Diaphragms transfer earthquake induced
15.4.2.5 Stiffness and Strength inertial loads to the vertical elements of the
Assumptions seismic framing system. Connection between
Element and component stiffness properties the diaphragms and the vertical elements of the
and strength assumptions for most material lateral load resisting system must have
types are provided in FEMA-273. Guidelines sufficient strength to transfer the maximum
for structural and foundation elements are also calculated inertial loads. Diaphragms may be
provided. These are similar to those provided in flexible, stiff or rigid. Flexible diaphragms are
ATC-40. those where the maximum lateral deformation
of the diaphragm is more than twice the average
15.4.2.6 Foundation Modeling inter-story drift of the story below the
Foundation modeling assumptions are diaphragm. Rigid diaphragms are those where
similar to ATC-40 (See Section 15.3.5). The the maximum lateral deformation of the
foundation system may be included in the diaphragm is less than half the average inter-
model for analysis with stiffness and damping story drift of the associated story. Diaphragms
properties as defined in Chapter 4 of FEMA- that are neither rigid nor flexible can be
273. Otherwise, unless specifically prohibited, considered to be stiff.
the foundation may be assumed to rigid and not Mathematical models of buildings with stiff
included in the model. or flexible diaphragms must consider the effect
of diaphragm flexibility. For buildings with
flexible diaphragms at each floor level, the
780 Chapter 15
combinations and therefore only the critical Where w and d are in-plane wall and
load combination (by inspection) may be used. diaphragm displacements in inches due to a
lateral loads in the direction under
15.4.3 Analysis Procedures consideration equal to the weight tributary
to the diaphragm. For multiple span
diaphragms, a lateral load equal to the
15.4.3.1 Linear Static Procedure gravity weight tributary to the span under
In this procedure a linear elastic model is consideration can be applied to each span to
used in the analysis with an equivalent damping calculate a separate period for each
that approximates values expected for loading diaphragm span. The period so calculated
near the yield point. A pseudo-lateral load is that maximizes the pseudo lateral load is to
computed as shown in the following section and be used for the design of all walls and
applied to the model. The resulting forces and diaphragm spans in the building.
displacements in the elements are then checked The total pseudo lateral load, V in a given
against capacities modified to account for horizontal direction is determined as
inelastic response demands.
V = C1C 2 C 3 S aW (15-17)
15.4.3.1.1 Pseudo Lateral Load
To compute the pseudo lateral load, the Where
fundamental period must be first determined. C1 = Modification factor to relate expected
The period may be determined by one of the maximum inelastic displacements to
following methods: displacements calculated for the linear elastic
1. Eigenvalue value analysis of the building. response. C1 can be calculated as in Section
For buildings with flexible diaphragms, the 15.4.3.3.4 with the elastic base shear substituted
model must consider representation of for Vy. Alternatively C1 may be calculated as
diaphragm flexibility unless it can be follows
shown that the effects of the omission will C1=1.5 for T < 0.10 secs
not be significant. C1=1.0 for T T0 secs
2. Use of the following equation Linear interpolation can be used to calculate C1
for intermediate value of T.
T = Ct hn
3/ 4
(15-15) T = Fundamental period of the building in the
direction under consideration. For SSI, the
Where T is the fundamental period in effective fundamental period should be used.
seconds under the direction under T0 = Characteristic period of the response
consideration and hn is the height above the spectrum, defined as the period associated with
base to the roof. the transition from the constant acceleration
Ct = 0.035 for steel moment resisting segment of the spectrum to the constant
frames. velocity segment of the spectrum
Ct = 0.030 for moment resisting frame C2 = Modification factor to represent the effect
system of concrete and eccentrically braced of stiffness degradation and strength
steel frames. deterioration on the maximum displacement
Ct = 0.020 for all other framing systems. response. Values for different framing for
Ct = 0.060 for wood buildings. different performance levels are listed in Table
3. For one-story buildings with flexible 15-9. Linear interpolation can be used to
diaphragms: calculate C2 for intermediate value of T.
C3 = Modification factor to represent the
T = (0.1 w + 0.078 d ) 0.5 (15-16) increased displacement due to dynamic P-Delta
effect. This effect is in addition to P-Delta
782 Chapter 15
n
1 wx
Fx = C vxV (15-18) Fpx =
C1C 2 C 3
F i n
(15-20)
i= x
w
i= x
i
mode should be combined using SRSS (square displacement is reached or the building
root sum of squares) or CQC (complete collapses. The target displacement is intended
quadratic combination). It should also be noted to represent the maximum displacement likely
that the directivity of the forces is lost in the to be experienced during the design earthquake.
response spectrum analysis and therefore the The nonlinear effects are directly included in
combination of forces must reflect this loss. the model and therefore the calculated inertial
Multidirectional effects should also be forces are reasonable approximations of those
investigated when using the response spectrum expected during the design earthquake.
analysis. The target displacement can be calculated
by any procedure that accounts for nonlinear
15.4.3.2.3 Time History Method response on displacement amplitude as well as
All the requirements for response spectrum damping effects at the performance point. One
analysis are also identical for the time history such procedure called the Displacement
analysis. Response parameters are computed for Coefficient Method is described in FEMA 273.
each time history analysis. If 3 pairs of time ATC-40 also includes this method as an
histories are used, the maximum response of the alternative method of finding the performance
parameter of interest shall be used for the point. The advantage of this method over the
design. If seven or more pairs of time histories Capacity Spectrum procedure is it simplicity.
are used, the average response (of the maximum The modeling requirements for NSP are
of each analysis) of the parameter of interest is similar to those described in ATC-40. The
to be used. pushover analysis is performed and a curve
Multidirectional effects can be accounted by relating the base shear force and the lateral
using a three dimensional mathematical model displacement of the control node are established
and using simultaneously imposed pairs of between 0 and 150% of the target displacement,
earthquake ground motions along each of the t. Acceptance criterion is based on the forces
horizontal axes of the building. and deformation corresponding to the
displacement of the control node equal to t.
15.4.3.2.4 Acceptance Criteria to satisfy The analysis model must be sufficiently
Performance Point requirements discretized to represent the load-deformation
The acceptance criterion for LDP is similar response of each element or component.
to that described for LSP. However, all Particular attention needs to be paid to
deformations and force demands obtained from identifying locations of inelastic action along
either the response spectrum or the time history the length of element or component. Thus, local
analysis must be multiplied by the product of models of elements or assemblages of elements
the modification factors C1, C2 and C3. Force need to be studied before embarking on the
demands on elements of the floor diaphragm global models.
need not be increased by these factors. The
seismic forces on the diaphragm obtained in the 15.4.3.3.1 Control Node
analysis must not be less than 85% than those The control node is usually the center of
obtained in LSP (See Equation 15-20). mass of the roof of the building. The top of the
penthouse should not be considered to be the
15.4.3.3 Nonlinear Static Procedure roof. As the displacement of the control node is
In the Nonlinear Static Procedure (NSP) the compared with the target displacement, the
nonlinear load-deformation characteristics of choice of the control node is very important.
individual elements and components are
modeled directly. The mathematical model of 15.4.3.3.2 Lateral Load Patterns
the building is subjected to monotonically The lateral load should be applied to
increasing lateral load until a target building in profiles that approximately bound
15. Performance Based Seismic Engineering 785
modeling approaches and acceptance criterion some areas, including a 17-inch thick
for the NDP are similar to those of the NSP. "sonovoid" slab, a cast-in-place concrete slab
With the NDP the design displacements are not with voids. The sonovoid slabs are located at
established using a target displacement, but the ground and first floor. The slabs, beams,
determined directly through the dynamic time and girders are supported by tied and spirally
history analysis. As the analysis can be very reinforced concrete columns and concrete
sensitive to characteristics of individual ground bearing walls. The columns rest on spread
motions, it is advisable to perform the analysis footings, with continuous footings under the
with more than one ground motion. Ground perimeter and interior walls.
motions used for the analysis and the analysis There are some unusual features in the
procedure should be similar to those used in vertical load-carrying system. Along the north
LDP (See Section 15.4.3.2). and south exterior walls and the Central Wing,
It should be noted that the volume of data vertical loads are carried by concrete columns
generated in NDP is enormous and it is difficult outside the building envelope. At the second
to condense the data to useful performance level, columns are discontinuous and are
based design information. Sensitivity analysis supported by transfer girders. At the First Floor,
to various parameters is also a prerequisite for the Central Wing relies on massive concrete
NDP analysis. Thus, NDP must only be used frames to resist vertical loads.
with caution for very important, irregular and
unusual structures.
East Wing
15.4.4 Example
S X 1 BS 0.975
T0 = = = 0.65 seconds
S XS B1 1.5
Sa 1 1.5 1
Base Shear (kips)
R= =
6000
5000
0.6V Y
V y / W C0 7,200 / 38,064 1.3
4000
3000
= 6.1
2000
1000
T 1
C1 = 1 + ( R 1) 0
0
0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50 5.00
9000
This value is reduced to the maximum value of
8000
C1 in Section 15.4.3.1.1, which is 1.28
7000 (interpolated for Te = 0.41 seconds). Thus:
VY
Base Shear (kips)
6000
Te2
t = C 0 C1C 2 C3 S a
5000
g
4000
0.6V Y 4 2
3000
0.412
2000 = 1.3 1.28 1 1 1.5 g
1000
4 2
0
= 4.11 inches
0.00 1.00 2.00 3.00 4.00 5.00 6.00
acceptance criteria for BSE-1 in the East-West One very useful characteristic of the ATC-
and North-South directions. 40 and FEMA 273/274 documents is that they
The maximum displacement reached and the provide a step-by-step approach for PBSE.
type and number of hinges formed for the This is an important first step towards a
various pushover analyses performed was building code implementations of performance
recovered. From the results of the pushover based design.
analyses, it was seen that the Modal pattern is There are some weaknesses that need to be
more detrimental to this building as more addressed with additional research. Three
number of hinges were formed for a given broad areas need work:
displacement level compared to the Uniform 1. A more reliable and conservative
pattern. This also goes to show that the lower methodology, which is widely accepted,
floors of this building are relatively stronger needs to be developed for establishing the
than the upper floors. However this building in performance point. More accurate
its existing configuration was unable to achieve equations need to be developed to find the
its target displacement. The building could only effective damping or equivalent ductility
be pushed to a displacement of 2.8 in the used to reduce the design response spectra
negative East-West direction and 4.34 in the to levels consistent with observed structural
negative North-South direction. behavior.
The analyses also revealed a number of 2. More sophisticated computer analysis
columns supporting walls above to have programs are needed which can do
rotations beyond collapse. Many of the walls nonlinear analysis of concrete/masonry/
and beams also had plastic rotations beyond the plywood shear walls, concrete and steel
Life Safety requirement at the target joints, confined concrete sections, etc.
displacement. Some of the columns in the There is also a need to reduce the data to a
central wing had shear failures under the finite number of parameters than can be
uniform pattern for push in the East-West used for design.
direction. Clearly, this building does not meet 3. The element capacities and deformations
the acceptance criteria of the basic safety limits for various performance levels are
objective, and therefore needs retrofit. currently based on engineering judgment or
relatively small number of experiments.
15.5 Conclusions More experimental and theoretical work is
needed to establish reliable element
The principal advantage of PBSE is that the capacities and deformation limits for given
choice of performance goals lies with the owner performance objectives.
who can decide the acceptable damage state.
The engineer can also convey to the owner a
better understanding of the expected damage REFERENCES
state. PBSE does not eliminate the risks
associated with uncertainties in ground 15-1 Applied Technology Council (1996), Seismic
motions, material properties, element behavior Evaluation and Retrofit of Concrete Buildings,
ATC-40, Volume 1 and 2, Report No. SSC 96-01,
or geotechnical properties. However, it Seismic Safety Commission, Redwood City, CA.
provides a new technique to remove 15-2 Federal Emergency Management Agency (1997),
unnecessary conservatism for some parameters NEHRP Guidelines for the Seismic Rehabilitation
and discover unidentified deficiencies for of Buildings, FEMA-273, Washington, D.C.
others. If implemented correctly and 15-3 Federal Emergency Management Agency (1997),
NEHRP Commentary on the Guidelines for the
competently, PBSE can produce a design that is
Seismic Rehabilitation of Buildings, FEMA-274,
more reliable than traditional procedures. Washington, D.C.
792 Chapter 15
15-4 King, S.A. and Rojhan, C. (1997), "ATC-38 Seismic Analysis and Design of Reinforced
Database on the Performance of Buildings Near Concrete Buildings, P. Fajfar and J. Krawinkler,
Strong-Motion Recordings," Proceedings of Editors., Elsevier Applied Science, New York.
Northridge Earthquake Research Conference, 15-18 Vidic, T., Fajfar, P. and Fischinger, M., 1994,
CUREe, Los Angeles, August. Consistent Inelastic Design Spectra: Strength and
15-5 Crandell, J.H. (1997), "Statistical assessment of Displacement, Earthquake Engineering and
Residential Construction Damage by the Structural Dynamics 23(5).
Northridge Earthquake," Proceedings of Northridge 15-19 Fajfar, P., 2000, A Nonlinear Analysis Method for
Earthquake Research Conference, CUREe, Los Performance Based Seismic Design, Accepted for
Angeles, August. Publication in Earthquake Spectra, EERI, Oakland,
15-6 Naeim, F. and Kelly, J.M. (1999), Design of California.
Seismic Isolated Structures From Theory to 15-20 Aschheim M., Black, E.F., 2000, Yield Point
Practice, John Wiley & Sons, New York. Spectra for Seismic Design and Rehabilitation,
15-7 Naeim, F. (1998), Earthquake Ground Motions Earthquake Spectra, Volume 16, Number 2, EERI,
and Performance Based Design, Performance Oakland, California.
Based Seismic Engineering Invitational Workshop, 15-21 Cormartin, C.D., Niewiarowski, Freeman, S.A. and
Earthquake Engineering Research Institute, San Turner, F.M., 2000, Seismic Evaluation and
Diego, California. Retrofit of Concrete Buildings; A Practical
15-8 Freeman, S.A., Nicoletti, J.P. and Tyrell, J.V., Overview of the ATC-40 Document, Earthquake
1975, Evaluation of Existing Buildings for Spectra, Volume 16, Number 1, EERI, Oakland,
Seismic Risk: A Case Study of Pudget Sound California.
Naval Shipyard, Bremerton, Washington, 15-22 Chai, W. and Guh, J., 1999, Performance-Based
Proceedings of U.S. National Conference of Design of Concrete Shear Wall Buildings,
Earthquake Engineers, Berkeley, Earthquake Proceedings of 1999 SEAOC Convention,
Engineering Research Institute. Structural Engineers Association of California,
15-9 Freeman, S.A., 1998, Development and use of Santa Barbara, California.
Capacity Spectrum Method, Paper No. 269,
Proceedings of the 6th U.S. National Conference of
Earthquake Engineering, Seattle, Washington.
15-10 U.S. Army, 1986, Seismic Design Guidelines for
Essential Buildings, Departments of the Army
(TM5-809-10-1), Navy (NAVFAC P355.1), and
the Air Force (AFM88-3), Washington, DC.
15-11 Structural Engineers Association of California
(SEAOC), 1995, Vision 2000: Performance-Based
Seismic Engineering of Buildings, Sacramento,
California.
15-12 International Code Council, 2000, International
Building Code 2000.
15-13 International Conference of Building Officials,
1997, Uniform Building Code, Whittier, CA.
15-14 Chopra, A.K. and Goel R.K., 1999, Capacity-
Demand-Diagram Methods for Estimating Seismic
Deformation of Inelastic Structures: SDF Systems,
Pacific Earthquake Engineering Research Center,
PEER-1999/02, University of California, Berkeley,
California.
15-15 Chopra, A.K. and Goel R.K., 2000, Capacity-
Demand-Diagram Methods Based on Inelastic
Design Spectrum, Earthquake Spectra, Volume
15, Number 4, EERI, Oakland, California.
15-16 Newmark, N.M., and Hall,W.J., 1982, Earthquake
Spectra and Design, Earthquake Engineering
Research Institute, Berkeley, California.
15-17 Krawinkler, H. and Nassar, A.A., 1992, Seismic
Design based on Ductilities and Cumulative
Damage Demands and Capacities, in Nonlinear