Procedure For Evaluating Seismic Energy Demand of Framed Structures

EARTHQUAKE ENGINEERING AND STRUCTURAL DYNAMICS
Earthquake Engng Struct. Dyn. 2003; 32:229–244 (DOI: 10.1002/eqe.221)
A procedure for evaluating seismic energy demand

of framed structures
Chung-Che Chou and Chia-Ming Uang∗; †

Department of Structural Engineering; University of California; San Diego; La Jolla; CA 92093; U.S.A.
SUMMARY
Energy serves as an alternative index to response quantities like force or displacement to include the
duration-related seismic damage eect. A procedure to evaluate the absorbed energy in a multistorey
frame from energy spectra was developed. For low- to medium-rise frames, it required a static pushover
analysis of the structure to determine the modal yield force and ductility factor of an equivalent single-
degree-of-freedom system for the rst two modes. The energy spectra were then used to determine the
energy contribution of each mode. A procedure was also developed to distribute the energy along the
frame height based on energy shapes. This study showed that the second-mode response in some cases
needs to be considered to reect the energy (or damage) concentration in the upper oors. Copyright
? 2002 John Wiley & Sons, Ltd.
KEY WORDS: absorbed energy; multistorey frame; equivalent single-degree-of-freedom (SDOF) system;
ductility; pushover analysis; energy shape
1. INTRODUCTION
It is well known that seismic damage to a multistorey frame is not only caused by the maxi-
mum response such as force or lateral displacement. Inelastic excursions below the maximum
response can still cause signicant damage to structures [1]. This duration-related damage,
which can be expressed as the energy absorbed in a structure, should also be considered in
the evaluation of structural performance [2; 3]. For a single-degree-of-freedom (SDOF) sys-
tem, an attenuation-based approach has been developed for the construction of energy spectra
[4]. To extend the energy-based analysis method to multistorey frames, a procedure for the
estimation of energy demand in a multi-degree-of freedom (MDOF) system is needed.
Study on the energy demand in multistorey frames is limited. Lawson et al. [5] showed that
the hysteretic energy demand in a structure cannot be predicted by a static pushover analysis.
Fajfar and Gasplersic [6] showed that the hysteretic energy demand in an MDOF system
cannot be evaluated reliably from an equivalent SDOF system; the researchers attributed the
∗ Correspondence to: Chia-Ming Uang, Department of Structural Engineering, University of California, San Diego,
La Jolla, CA 92093, U.S.A.
† E-mail: cmu@ucsd.edu
Received 30 April 2001

Revised 13 March 2002 and 13 May 2002
Copyright ? 2002 John Wiley & Sons, Ltd. Accepted 13 May 2002
10969845, 2003, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/eqe.221 by Universidad Nacional Autonoma De Mexico, Wiley Online Library on [17/04/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
230 C.-C. CHOU AND C.-M. UANG
Table I. Earthquake data.
Record name Earthquake magnitude PGA (cm/sec2 )

1974 Tabas: LA30 7.4 972.6
1994 Northridge: Sylmar County Hospital 6.7 826.8
1989 Loma Prieta: Saratoga 6.9 493.7
problem to the higher mode eect. This eect also made it dicult to predict the energy
distribution along the height of building structures [7; 8]. A recent study by Chopra and
Goel [9] also showed that the seismic storey drift demand along the building height can be
estimated if more than one equivalent SDOF systems are considered.
OBJECTIVE
The purpose of this paper is to present a procedure that can be used to predict the seismic
energy demand at each oor of an MDOF system without performing a non-linear time-history
analysis. The rst objective is to compute the total absorbed energy in an MDOF system by
considering equivalent SDOFs for the rst two modes. The second objective is to distribute
the total energy along the frame height based on energy shapes, which are established from
a static pushover analysis. Case studies are presented to verify the accuracy of the proposed
method.
CONSTANT-DUCTILITY RESPONSE SPECTRA FOR SDOF SYSTEMS
Three earthquake ground motions (see Table I and Figure 1) were used to construct the
constant-ductility response spectra for an elastic-perfectly plastic SDOF system with 5% vis-
cous damping. The rst record shown in Table I was developed for the SAC Joint Venture
[10]; the second and third records are representative near-led and far-eld ground motions,
respectively.
For an inelastic SDOF system subjected to a ground motion, the dierential equation of
motion is
m + c˙ + f = −mg (1)
where m is the mass, c is the viscous damping coecient, and f is the restoring force. g
and are the ground displacement and relative displacement, respectively. At a given ductility
factor, , the yield force, fy , can be computed from a non-linear dynamic analysis. The yield
force can also be normalized as follows:
fy
Cy = (2)
mg
The maximum displacement, Ds , in an SDOF system is equal to times the yield displacement.
The Cy and Ds response spectra at dierent ductility levels are presented in Figure 2(a)
and 2(b).
Copyright ? 2002 John Wiley & Sons, Ltd. Earthquake Engng Struct. Dyn. 2003; 32:229–244
EVALUATING SEISMIC ENERGY DEMAND OF FRAMED STRUCTURES 231
1.0
Velocity (cm/sec)
100
Acceleration (g)
0.0
0
-100
-1.0
(a) Tabas: LA30
1.0
Velocity (cm/sec)
100
Acceleration (g)
0.0
0
-100
-1.0
(b) Northridge: Sylmar

1.0
Velocity (cm/sec)
100
Acceleration (g)
0.0
0
-100
-1.0
0 10 20 0 10 20
Time (sec) Time (sec)
(c) Loma Prieta: Saratoga
Figure 1. Ground motion acceleration and velocity time histories.
The energy equation for an inelastic SDOF system can be derived from Equation (1) as
follows [11]:
t t t
m(˙ t )2
+ (c)
˙ d + f d = (m t ) dg (3)
2 0 0 0
or
E k + E + E a = Ei (4)
where t is the total displacement. Ek , E , and Ei are the kinetic energy, viscous damping
energy and ‘absolute’ input energy, respectively. The absorbed energy, Ea , is composed of
the recoverable elastic strain energy, Es , and the irrecoverable hysteretic energy, Eh . The
absorbed energy and hysteretic energy only dier by a small amount when the ductility factor
is larger than two [12]. The equivalent velocity, Va , of Ea , expressed in Equation (5), is used
as a parameter for energy demand because it converges to the pseudo-velocity in the elastic
case [4].

Va = 2Ea =m (5)
In this study, Ea is evaluated as the maximum value obtained from the time-history analysis.
Based on the Cy values at dierent ductility levels, the absorbed energy is computed and the
equivalent velocity spectra of the absorbed energy are shown in Figure 2(c).
Tabas: LA30 Northridge: Sylmar Loma Prieta: Saratoga
4
3
Cy
2 1
0
0 1 2 3 0 1 2 3 0 1 2 3
Period (sec) Period (sec) Period (sec)
(a) Cy Response Spectra

150
100
D s (cm)
50 0
0 1 2 3 0 1 2 3 0 1 2 3
(b) Ds Response Spectra

300
Va (cm/sec)
100 0200
0 1 2 3 0 1 2 3 0 1 2 3
(c) Va Response Spectra
µ = 6; µ = 4; µ = 2; µ=1
Figure 2. Constant-ductility response spectra.
DESIGN OF MOMENT-RESISTING FRAMES
To demonstrate the development of the analysis procedure, three moment-resisting frames with
5-, 7-, and 9-stories in height were used in this study. The rst two buildings were designed
in accordance with the NEHRP recommended seismic provisions [13]. The design base shear
W14X34
(508X559) (508X559)
(508X559) (508X559)
(508X508) (457X508)
W14X68
W14X68
W12X45
4@3658 W18X40
W21X50
W24X62
W24X68
W14X500 W14X455 W14X370 W14X283 W14X257
W14X500 W14X455 W14X370 W14X283 W14X257
W14X370 W14X370 W14X283 W14X257 W14X233

W14X74
W14X74
W12X53
W24X62
W27X84
4267
W30X99
4@7315
W36X135
(a) 5-story Frame
8@3960
W36X135
W36X135
W36X135
W21X50
W36X160
(508X559) (508X559) (508X559)
(508X559) (508X559) (508X559)
(508X508) (508X508) (457X508)

W21X50
W14X82 W14X82 W14X68
W14X82 W14X82 W14X68
W12X58 W12X58 W12X45
W21X50 W36X160
5500
6@3658
W24X68
W24X68
5@9150
W24X76
(c) 9-story Frame
W24X76
4267
4@7315
(b) 7-story Frame
Figure 3. Elevation and member sizes (dimensions in millimetres).
coecients for strength calculations were 0.088 and 0:075 g, respectively. The frames were
designed for steel beams and steel-encased reinforced concrete (SRC) columns [12]. A reduced
beam section with 50% reduction in the beam ange width was used to reduce the demand
to the connection. The specied 28-day concrete strength was 34 MPa (5 ksi) for all SRC
columns. The specied yield strength of reinforcement was 414 MPa (60 ksi), and A572 Gr.
50 steel was used for both steel beams and columns.
The 9-storey oce building [14] with steel moment-resisting frames in both directions was
designed by an SAC-commissioned consulting rm, which followed the 1994 UBC [15]. The
design nominal yield strength of the beams and columns were 248 MPa (36 ksi) and 345 MPa
(50ksi), respectively. The expected yield strengths of 339MPa (49:2ksi) and 397MPa (57:6ksi)
were used to compute their member capacity in dynamic analyses. The elevation and member
sizes of the three frames are shown in Figure 3. The computer program DRAIN-2DX [16] was
used for analysis, and a bi-linear moment-rotation relationship with 0 and 2% strain hardening
was used to model beams and columns, respectively, in the MDOF model. Rayleigh damping
with 5% damping for the rst two modes was assumed. The calculated frame weight (half of
building weight), W , and natural periods, Ti , of the rst two modes are listed in Table II.
EQUIVALENT SDOF SYSTEMS

To estimate the energy demand of an MDOF system from the energy response spectra (or Va
spectra), the ductility factor of the equivalent SDOF system for each of the rst two modes
Table II. Frame properties.

Frame W (KN) T1 (sec) T2 (sec) 1 2
5-Storey 7 476 1.47 0.53 799.1 295.7
7-Storey 10 608 1.85 0.66 952.2 337.6
9-Storey 44 214 2.14 0.80 1928.9 696.3
needs to be established rst. The procedure includes the following: (i) convert the MDOF
system into an SDOF system for each of the rst two modes with the corresponding elastic
mode shape, (ii) perform a static pushover analysis to determine the yield strength coecient,
Cy , for each mode, and (iii) determine the modal ductility factor from the Cy spectrum. The
absorbed energy of the MDOF system then can be evaluated by summing the energy of each
mode, which is obtained from the Va spectrum using the modal ductility factor. This procedure
is presented below.
The equation of motion for an MDOF frame is
M] + C]˙ + f = −M1]g (6)
where M is the diagonal mass matrix, C is the damping matrix, f is the restoring force
vector at each oor, and 1 is a unit vector. ] is a vector of relative displacement and can be
expressed as

N
]= M n xn (7)
n=1
where Mn is the elastic modal shape, xn is the generalized displacement of the nth mode, and
N is the number of stories. Pre-multiplying both sides of Equation (7) by MnT M, where MnT is
the transpose of Mn , the generalized displacement xn is
xn = MnT M] (8)
where Mn can be normalized such that mn = MnT MMn = 1. Substituting Equation (7) into Equa-
tion (6) and premultiplying each term by MnT give
cn fn
x n + ẋn + = −n g (9)
mn mn
where n ( = MnT M1=mn ) is the participation factor, and cn (= MnT CMn ) is the generalized damping
coecient. The generalized restoring force is
fn = MnT f (10)
To obtain the generalized yield force, fyn , for the nth mode, a static pushover analysis is
performed with the following lateral load:
L n = M Mn s = q n s (11)
where q n is the lateral load pattern, and s is the scalar factor. At each load step, the generalized
force fn is calculated by premultiplying Equation (11) by MnT , and the generalized displacement
xn is obtained from Equation (8) by substituting ] for the resulting lateral displacement
14
1
5-story Frame
Fr e
7-story Frame
ized Force, fn (kN)

12
1
9-story Frame
10
1
Second Mode
8
Generalized
6
4
First Mode
2
0
0 200 400 600 800

Generalized Displacement, xn (m)
(a) Original Case
Generalized
Force, f n MDOF Pushover Analysis (Actual)
Bi-linear Approximation (Idealized)
Second Mode
f y2
First Mode
f y1
k2
k1
x y2 x y1 Generalized
Displacement, xn
(b) Idealized Case
Figure 4. Generalized force-generalized displacement relationships.
Table III. Generalized modal yield forces and Cyn values.

Frame First mode Second mode
fy1 (kN) Cy1 fy2 (kN) Cy2
5-Storey 2.15 0.27 5.17 1.78
7-Storey 1.90 0.21 5.72 1.73
9-Storey 4.81 0.25 10.18 1.49
vector. Relationships between the generalized force and generalized displacement for the three
example frames are shown in Figure 4(a). Each curve is then approximated by a bi-linear
relationship like that shown in Figure 4(b) to determine the generalized yield force, fyn ,
listed in Table III. Note that the initial slope, kn , in Figure 4(b) is related to the frequency
Table IV. Modal-ductility factors for three building frames.

Record name 5-Storey frame 7-Storey frame 9-Storey frame
First Second First Second First Second
mode mode mode mode mode mode
1974 Tabas: LA30 2.8 1.2 3.0 0.9 2.5 1.3
1994 Northridge: Sylmar County Hospital 2.5 0.9 2.6 0.8 1.5 0.7
1989 Loma Prieta: Saratoga 1.6 0.3 1.6 0.4 1.3 0.3
of the nth mode:
fn MnT Ln MnT q n s MnT (K]) MnT KMn xn

kn = = = = = = MnT KMn = !n2 mn (12)
xn xn xn xn xn
where K the stiness matrix of the MDOF system, and !n is the Rayleigh natural frequency
[17]. The non-dimensional yield strength coecient, Cyn , is dened as
fyn
Cyn = (13)
n mn g
n in Equation (13), listed in Table II, is required in the denominator because the intensity of
the ground motion in Equation (9) is n times that in Equation (1). Based on the calculated
Cyn values listed in Table III and the natural periods listed in Table II, the Cy spectra in Figure
2(a) then can be used to determine the ductility factors of the equivalent SDOF systems for
the rst two modes (see Table IV).
ENERGY DEMAND EVALUATION
Based on the ductility factor and natural period in each mode, the Va value for each equivalent
SDOF system can be determined from the Va response spectra in Figure 2(c). When the
ductility factor is less than one (i.e. elastic response), the pseudo-velocity, V , is used instead.
From Equation (5), the absorbed energy for each of the rst two modes is
Ea1 = 12 m1 (Va1 1 )2 (14)
Ea2 = 12 m2 (Va2 2 )2 (15)
The total absorbed energy, Eam , computed from an MDOF time-history analysis is compared
with Ea1 and Ea2 for each frame in Figure 5. For the frames studied, it is observed that
the total absorbed energy in an MDOF system can be estimated by summing the absorbed
energies of the rst two modes. That is, Eam can be approximated as
Eam = Ea1 + Ea2 (16)
Figure 5. Comparison of absorbed energy.
(a) 5-story Frame (b) 7-story Frame (c) 9-story Frame
Figure 6. 1; base and absorbed energy relationships.
ENERGY DISTRIBUTION EVALUATION
The energy distributions along the height of the frames will be developed in this
section based on static pushover analyses with two modal load patterns q1 and q 2 . The
corresponding energy shapes are then used to distribute energies Ea1 and Ea2 along the frame
height.
First-mode pushover analysis
Performing a static pushover analysis based on the rst-mode load pattern q1 , Figure 6 shows
the relationship between the cumulative rotation, 1; base , at the column base and the absorbed
energy of each frame. 1; base is the average value of the cumulative rotations at the base
of exterior and interior columns. Each column-base cumulative rotation is calculated as the
absorbed energy normalized by the exural strength, Mpc , of the column. Three zones are
characterized from three distinct slopes at load steps A and B, which are shown by solid
marks. The rst zone (before step A) represents the system within the elastic range. The
second zone (between A and B) is typied by the formation of beam plastic hinges at lower
0.0 0.2 0.4 0.6 0.8 1.0

Relative Height
0.0 0.5 1.0 1.5 2.0 0.0 0.5 1.0 1.5 2.0 0.0 0.5 1.0 1.5 2.0
η1 (x0.01 rad) η1 (x0.01 rad) η1 (x0.01 rad)
Elastic Case; Formation of Beam Plastic Hinges at Lower Floor Levels;

Formation of First Plastic Hinge at Column Base
Figure 7. 1 proles from rst-mode pushover analysis.
oor levels, and the third zone (beyond B) indicates that plastic hinges have formed at the
base of the rst-storey columns.
Figure 7 shows the distribution of the cumulative rotation, 1 , along the frame height at
dierent load steps, marked by 1, 2, and 3 in Figure 6. Both the cumulative rotation at the
base of the columns and the cumulative rotation of the beams, 1; beam , are used to construct
the proles. Assuming same beam size on each oor, 1; beam are calculated as
2m
Ea1; i
1; beam = i=1 (17)
2mMpb
where m is the number of bays, Ea1; i is the absorbed energy in each beam end, and Mpb is
the beam exural strength.
For each frame in Figure 7, the rst distribution shape (solid line) represents the elastic
case at load step 1 in Figure 6, while the second shape shows the distribution at load step 2
in Figure 6. The plastic hinge usually does not form at the base of the columns (step 3 in
Figure 6) until the lateral load is increased to a certain level; the corresponding distribution
is shown as the third shape. The 1 proles in Figure 7 clearly show that, for the rst-mode
pushover analysis, energy demands at higher deformation levels are concentrated in the lower
stories.
Two limits for the rst storey drift ratio that corresponds to the initiation of beam yielding
(step A in Figure 6) and column base yielding (step B in Figure 6) can be established from
the rst-mode pushover analysis. The results are summarized in Table V, in which SDR 1; beam
is the rst storey drift ratio at step A, and SDR 1; base is the rst storey drift ratio at step
B. For a well-designed frame, SDR 1; base is generally larger than SDR 1; beam due to the strong
column-weak beam criterion adopted in modern seismic provisions.
First-mode energy shapes
Three energy shapes ( 1 ; 2 and 3 ) in Figure 8 are obtained by normalizing the cumulative
rotation on each oor by the second oor’s rotation in Figure 7. It is observed that these
Table V. First storey drift ratios at two loading steps.

Limit (%) 5-Storey frame 7-Storey frame 9-Storey frame
SDR 1; beam 0.91 0.95 0.95
SDR 1; base 1.37 1.30 1.52
0.0 0.2 0.4 0.6 0.8 1.0
Relative Height
1 1 1

2 2 2
3 3 3
0.0 0.5 1.0 1.5 0.0 0.5 1.0 1.5 0.0 0.5 1.0 1.5
Ratio Ratio Ratio
Figure 8. Energy shapes for rst-mode response.
Table VI. SDR 1 for frames subjected to three ground motions.

Record 5-Storey frame (%) 7-Storey frame (%) 9-Storey frame (%)
1974 Tabas: LA30 2.71 3.23 2.50
1994 Northridge: Sylmar
County Hospital 2.42 2.58 2.36
1989 Loma Prieta: Saratoga 1.46 1.70 1.15
shapes have dierent distribution patterns at dierent load stages. The percentage of energy
absorbed at the base of the columns and upper oor levels decreases with the load amplitude.
Once these energy shapes and two limits for the rst storey drift ratio (Table V) are
established, the next step is to determine the appropriate shape for energy distribution. The
rst storey drift ratio, SDR 1 , is used for this purpose, because it can reect the extent of
inelastic action at the base of the column. For a given earthquake ground motion, SDR 1
produced by the rst-mode response can be estimated as follows:
Ds 1 M11
SDR 1 = (18)
H1
where H1 is the rst storey height, Ds , the maximum drift, is obtained from Figure 2(b) based
on the rst-mode period and ductility factor, and M11 is the component of the mode shape M1
at the second oor level.
Once SDR 1 is calculated (see Table VI), the energy shape for distributing the rst-mode
energy is determined from the following rule:
Shape 1( 1 ) if SDR 1 ¡SDR 1; beam (19)
q2
Floor “A”
Figure 9. Frame deformation with second-mode loading.
Shape 2( 2 ) if SDR 1; beam 6SDR 1 ¡SDR 1; base (20)

Shape 3( 3 ) if SDR 1 ¿SDR 1; base (21)
The rst-mode energy Ea1 is then distributed to each oor level and the base of the frame with
the chosen energy shape, k . For small levels of strain hardening, the vector of cumulative
rotation along the frame height can be determined as
Ea1
W1 = k (22)
k M̂p
T
m+1 T

m
m
m
M̂p = Mpc ; 2 Mpb2 ; : : : ; 2 Mpbj ; : : : ; 2 MpbN (23)
i=1 i=1 i=1 i=1
m+1
where M̂p is a plastic moment vector;
m i=1 Mpc is the summation of plastic moment capacities
of the columns at the base, and 2 i=1 Mpbj is the summation of beam plastic moment capacity
at the jth oor.
Second-mode pushover analysis
Performing a pushover analysis with q 2 ( = MM2 ) as the lateral load pattern, Figure 9 shows
the typical deformation pattern when plastic hinges are formed in the frame. Floor ‘A’ in the
gure represents the transition level below which inelastic action does not occur; this oor
is generally near the location where the lateral load changes its sign. Note that the formation
of column plastic hinges above oor ‘A’ represents a very severe situation. It occurs when
a signicant amount of energy is imparted to the structure in the second mode. Figure 10
shows the relationship between the cumulative rotation, 2; base , at the base of the column and
the absorbed energy of each frame. 2; base is the average value of the cumulative rotations at
the base of exterior and interior columns. Each column base cumulative rotation is calculated
as the absorbed energy normalized by the exural strength, Mpc , of the column. Three zones
are characterized from three slopes at load steps A and B, which are shown by solid marks.
The rst zone (before step A) represents the system within the elastic range. The second zone
(between A and B) is typied by the formation of beam plastic hinges at upper oor levels,
and the third zone (beyond B) indicates that column plastic hinges have formed above oor
‘A’ in Figure 9.
0.0 0.1 0.2 0.3 0.4 0.5
0.0 0.1 0.2 0.3 0.4 0.5
0.0 0.1 0.2 0.3 0.4 0.5

η2,base (×0.01 rad)
B 3
B 3 2
B
2 A 3
A 2
A
1
1 1
0.0 0.5 1.0 1.5 2.0 0.0 0.5 1.0 1.5 2.0 0.0 0.5 1.0 1.5 2.0
Absorbed Energy (MN-m) Absorbed Energy (MN-m) Absorbed Energy (MN-m)
Figure 10. 2; base and absorbed energy relationships.

0.0 0.2 0.4 0.6 0.8 1.0
Relative Height
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
η2 (x0.01 rad) η2 (x0.01 rad) η2 (x0.01 rad)
Elastic Case; Form ation of Beam Plastic Hinges at Upper Floor Levels;
Form ation of First Colum n Plastic Hinge above Floor “A”
Figure 11. 2 proles from second-mode pushover analysis.
The proles of cumulative rotation W2 , containing 2; base and 2; beam , from the pushover
analysis are shown in Figure 11, in which the cumulative rotation of the beams, 2; beam , is
computed as
2m
Ea2; i
2; beam = i=1 (24)
2mMpb
where Ea2; i is the absorbed energy in each beam end. The rst distribution shape (solid line)
represents the elastic case at load step 1 in Figure 10, while the second shape shows the
distribution at load step 2. The third shape is the distribution at load step 3 when the rst
column plastic hinge is formed right above oor ‘A’.
A comparison of Figure 7 and Figure 11 shows that the rst-mode load pattern produces
high energy demand at lower oor levels, while energy produced by the second-mode load
pattern is concentrated at upper oor levels. The second-mode pushover analysis usually does
not result in yielding at the base of the columns, implying that the base yielding observed in
the MDOF non-linear time-history analysis is mainly caused by the rst-mode response.
0.0 0.2 0.4 0.6 0.8 1.0

Relative Height
1 1 1
2 2 2
3 3 3
0.0 0.5 1.0 1.5 0.0 0.5 1.0 1.5 0.0 0.5 1.0 1.5
Ratio Ratio Ratio
Figure 12. Energy shapes for second-mode response.
Second-mode energy shapes
Normalizing each prole in Figure 11 by its maximum value, three normalized energy shapes
( 1 ; 2 , and 3 ) for each frame are obtained in Figure 12. 1 is the shape for distributing
seismic absorbed energy Ea2 when the frame responds elastically. 2 is used as the only shape
for distributing Ea2 for the inelastic case, because shapes 2 and 3 are similar. Energy Ea2
is then distributed to each oor level based on the following expression:
Ea2
W2 = (25)
T M̂p k
k
where index k can be 1 or 2 determined based on the ductility factor of the second mode.
Combining the eects from both modes, the total cumulative rotations in the beams and at
the base of the columns are
W a = W1 + W 2 (26)
Comparisons
The a distributions in Figure 13 show good correlation between the results of MDOF non-
linear time-history analysis and the proposed procedure. The energy absorbed at lower oor
levels can be predicted by considering only the rst-mode response. However, the second-
mode eect is more signicant at upper oor levels in the 7- and 9- storey frames, and the
resulting energy concentration in these oors can be predicted only when the second-mode
eect is considered. This observation is supported by a more thorough study in which a total
of four frames and six ground motions with varying shaking intensities were used [12].
SUMMARY AND CONCLUSIONS
A procedure, similar to the response spectrum method for elastic dynamic analysis, to compute
the total energy demand and to distribute it along the height of a multistorey frame from
inelastic energy spectra is presented. The energy demand can be estimated without performing
0.0 0.2 0.4 0.6 0.8 1.0

Relative Height
0 5 10 15 0 5 10 15 0 5 10 15
ηa (x0.01 rad) ηa (x0.01 rad) ηa (x0.01 rad)
(a) 5-story Frame

0.0 0.2 0.4 0.6 0.8 1.0
Relative Height
0 5 10 15 0 5 10 15 0 5 10 15
(b) 7-story Frame

0.0 0.2 0.4 0.6 0.8 1.0
Relative Height
0 5 10 15 0 5 10 15 0 5 10 15
(c) 9-story Frame
1st Mode; 1st Mode+2nd Mode; MDOF Dynamic Analysis
Figure 13. Comparison of a distributions.
an MDOF inelastic time-history analysis. The procedure, which takes into account the higher
mode eect for energy distribution in a low- to medium-rise frame, is summarized below:
(1) The MDOF frame is rst converted into an equivalent SDOF system for each of the rst
two modes by elastic mode shapes.
(2) For each equivalent SDOF system, the generalized modal yield force is determined from
a static pushover analysis (Figure 4) with a lateral load (Equation (11)). Based on the
modal yield strength coecient (Equation (13)) and period, the modal ductility factor
can be determined from the constant-ductility yield strength response spectrum (Figure
2(a)).
(3) Based on the modal ductility factor, participation factor and natural period, the absorbed
energy of each mode is evaluated from the constant-ductility energy spectra (Figure
2(c)). The energy spectra can be constructed based on an attenuation approach [4].
(4) The absorbed energy of each mode is then distributed along the frame height based on
the specic energy shape (Equations (22) and (25)), which is established from a static
pushover analysis. The resulting energy distribution, expressed in the form of cumulative
rotation, is the summation of contribution from each mode (Equation (26)).
This study shows that the proposed procedure can predict the damage distribution of low-
to medium-rise frames when response of the rst two modes are considered (Figure 13).
Although the majority of energy is contributed by the rst mode (Figure 5), for some frames
and earthquake ground motions used in this study the eect of second mode (Figures 9 and
11) needs to be included to predict the damage in the upper stories.
REFERENCES
1. McCabe SL, Hall WJ. Damage and reserve capacity evaluation of structures subjected to strong earthquake
ground motion. Proceedings of the 10th World Conference on Earthquake Engineering, Madrid, Spain, 1992,
3653–3658.
2. Akiyama H. Earthquake-resistant Limit-state Design for Buildings. University of Tokyo Press: Japan, 1985.
3. Leelataviwat S, Goel SC, Stojadinovic B. Toward performance-based seismic design of structures. Earthquake
Spectra 1999; 15(3):435 – 461.
4. Chou C-C, Uang C-M. Establishing absorbed energy spectra—an attenuation approach. Earthquake Engineering
and Structural Dynamics 2000; 29(10):1441 – 1455.
5. Lawson RS, Vance V, Krawinkler H. Nonlinear static push-over analysis—why, when, and how? Proceedings
of the 5th U. S. National Conference on Earthquake Engineering, vol. I. 1994; 283 –292.
6. Fajfar P, Gaspersic P. The N2 method for the seismic damage analysis of RC buildings. Earthquake Engineering
and Structural Dynamics 1996; 25:31 – 46.
7. Seneviratna GDPK, Krawinkler H. Evaluation of inelastic MDOF eects for seismic design. Report No. 120,
Department of Civil Engineering, Stanford University, Stanford, CA, 1997.
8. Shen J, Akbas B. Seismic energy demand in steel moment frames. Journal of Earthquake Engineering 1999;
3(4):519 – 559.
9. Chopra AK, Goel RK. A modal pushover analysis procedure to estimate seismic demands for buildings: theory
and preliminary evaluation. Report No. 2001/03, Pacic Earthquake Engineering Research Center, University
of California, Berkeley, CA, 2001.
10. Somerville P. Development of ground motion time histories for phase 2 of the FEMA/SAC steel project. Report
No. SAC/BD-97/04, SAC Joint Venture, Sacramento, CA, 1997.
11. Uang C-M, Bertero VV. Evaluation of seismic energy in structures. Earthquake Engineering and Structural
Dynamics 1990; 19(1):77– 90.
12. Chou C-C. An energy-based seismic evaluation procedure for moment-resisting frames. Doctoral Dissertation,
Department of Structural Engineering, University of California, San Diego, La Jolla, CA, 2001.
13. BSSC, NEHRP Recommended Provisions for the Development of Seismic Regulations for New Buildings,
FEMA-222A, Washington, DC, 1994.
14. Krawinkler H, Gupta A. Storey drift demands for steel moment frame structures in dierent seismic regions.
Proceedings of the 6th U. S. National Conference on Earthquake Engineering, Oakland, CA, 1998.
15. ICBO, Uniform Building Code (UBC). International Conference of Building Ocials, Whittier, CA, 1994.
16. Prakash V, Powell GH, Filippou FC. DRAIN-2DX. User Guide, University of California, Berkeley, CA, 1993.
17. Chopra AK. Dynamics of Structures: Theory and Applications to Earthquake Engineering. Prentice-Hall: New
Jersey, 1995.

Procedure For Evaluating Seismic Energy Demand of Framed Structures

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Procedure For Evaluating Seismic Energy Demand of Framed Structures

Uploaded by

Copyright:

Available Formats

EARTHQUAKE ENGINEERING AND STRUCTURAL DYNAMICS

Earthquake Engng Struct. Dyn. 2003; 32:229–244 (DOI: 10.1002/eqe.221)

A procedure for evaluating seismic energy demand

Chung-Che Chou and Chia-Ming Uang∗; †

Received 30 April 2001

Table I. Earthquake data.

Record name Earthquake magnitude PGA (cm/sec2 )

CONSTANT-DUCTILITY RESPONSE SPECTRA FOR SDOF SYSTEMS

(b) Northridge: Sylmar

Figure 1. Ground motion acceleration and velocity time histories.

Tabas: LA30 Northridge: Sylmar Loma Prieta: Saratoga

Tabas: LA30 Northridge: Sylmar Loma Prieta: Saratoga

Tabas: LA30 Northridge: Sylmar Loma Prieta: Saratoga

Figure 2. Constant-ductility response spectra.

DESIGN OF MOMENT-RESISTING FRAMES

W14X500 W14X455 W14X370 W14X283 W14X257

W14X500 W14X455 W14X370 W14X283 W14X257

W14X370 W14X370 W14X283 W14X257 W14X233

(508X559) (508X559) (508X559)

(508X508) (508X508) (457X508)

W14X82 W14X82 W14X68

W12X58 W12X58 W12X45

(b) 7-story Frame

Figure 3. Elevation and member sizes (dimensions in millimetres).

EQUIVALENT SDOF SYSTEMS

Table II. Frame properties.

ized Force, fn (kN)

0 200 400 600 800

(a) Original Case

(b) Idealized Case

Figure 4. Generalized force-generalized displacement relationships.

Table III. Generalized modal yield forces and Cyn values.

Table IV. Modal-ductility factors for three building frames.

of the nth mode:

fn MnT Ln MnT q n s MnT (K]) MnT KMn xn

ENERGY DEMAND EVALUATION

Ea1 = 12 m1 (Va1 1 )2 (14)

Ea2 = 12 m2 (Va2 2 )2 (15)

Eam = Ea1 + Ea2 (16)

Figure 5. Comparison of absorbed energy.

(a) 5-story Frame (b) 7-story Frame (c) 9-story Frame

Figure 6. 1; base and absorbed energy relationships.

ENERGY DISTRIBUTION EVALUATION

First-mode pushover analysis

0.0 0.2 0.4 0.6 0.8 1.0

Elastic Case; Formation of Beam Plastic Hinges at Lower Floor Levels;

Figure 7. 1 proles from rst-mode pushover analysis.

First-mode energy shapes

Table V. First storey drift ratios at two loading steps.

Figure 8. Energy shapes for rst-mode response.

Table VI. SDR 1 for frames subjected to three ground motions.

Figure 9. Frame deformation with second-mode loading.

Shape 2( 2 ) if SDR 1; beam 6SDR 1 ¡SDR 1; base (20)

0.0 0.1 0.2 0.3 0.4 0.5

0.0 0.1 0.2 0.3 0.4 0.5

0.0 0.1 0.2 0.3 0.4 0.5

Figure 10. 2; base and absorbed energy relationships.

Figure 11. 2 proles from second-mode pushover analysis.

0.0 0.2 0.4 0.6 0.8 1.0

Figure 12. Energy shapes for second-mode response.

Second-mode energy shapes

SUMMARY AND CONCLUSIONS

Tabas: LA30 Northridge: Sylmar Loma Prieta: Saratoga

Table III. Generalized modal yield forces and Cyn values.

Ea1 = 12 m1 (Va1 1 )2 (14)

Ea2 = 12 m2 (Va2 2 )2 (15)