This article appeared in a journal published by Elsevier.

The attached
copy is furnished to the author for internal non-commercial research
and education use, including for instruction at the authors institution
and sharing with colleagues.
Other uses, including reproduction and distribution, or selling or
licensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of the
article (e.g. in Word or Tex form) to their personal website or
institutional repository. Authors requiring further information
regarding Elsevier’s archiving and manuscript policies are
encouraged to visit:
http://www.elsevier.com/copyright
 Author's personal copy

Soil Dynamics and Earthquake Engineering 31 (2011) 77–90

Contents lists available at ScienceDirect

Soil Dynamics and Earthquake Engineering

journal homepage: www.elsevier.com/locate/soildyn

Evaluation of ASCE-41, ATC-40 and N2 static pushover methods based on

optimally designed buildings
Nikos D. Lagaros a,n, Michalis Fragiadakis b,c
a
Institute of Structural Analysis & Seismic Research, National Technical University of Athens, Zografou Campus, Athens 15780, Greece
b
Department of Civil and Environmental Engineering, University of Cyprus, P.O. BOX 20537, 1678 Nicosia, Cyprus
c
Department of Civil Engineering, University of Thessaly, Pedion Areos, Volos 38334, Greece

a r t i c l e in fo abstract

Article history: Alternative static pushover methods for the seismic design of new structures are assessed with the aid
Received 10 February 2010 of advanced computational tools. The current state-of-practice static pushover methods as suggested in
Received in revised form the provisions of European and American regulations are implemented in this comparative study.
6 August 2010
In particular the static pushover methods are: the displacement coefficient method of ASCE-41, the
Accepted 20 August 2010
ATC-40 capacity spectrum method and the N2 method of Eurocode 8. Such analysis methods are
typically recommended for the performance assessment of existing structures, and therefore most of
the existing comparative studies are focused on the performance of one or more structures. Therefore,
contrary to previous research studies, we use static pushover methods to perform design and we then
compare the capacity of the outcome designs with reference to the results of nonlinear response history
analysis. This alternative approach pinpoints the pros and cons of each method since the discrepancies
between static and dynamic analysis are propagated to the properties of the final structure. All methods
are implemented in an optimum performance-based design framework to obtain the lower-bound
designs for two regular and two irregular reinforced concrete building configurations. The outcome
designs are compared with respect to the maximum interstorey drift and maximum roof drift demand
obtained with the Incremental Dynamic Analysis method. To allow the comparison, also the life-cycle
cost of each design is calculated; i.e. a parameter that is used to measure the damage cost due to future
earthquakes that will occur during the design life of the structure. The problem of finding the lower
bound designs is handled with an Evolutionary type optimization algorithm.
& 2010 Elsevier Ltd. All rights reserved.

1. Introduction (SPO) methods offer simplicity and reduced computational effort.

Performance-based earthquake engineering has put forward
the need for high level analysis procedures. Recent design codes
and guidelines introduce comprehensive frameworks for dynamic
and nonlinear analysis procedures [1–3]. The choice of the
analysis procedure to be adopted depends on several parameters,
such as the importance of the structure, the performance level,
the structural characteristics (e.g. regularity, complexity, fre-
quency properties), the amount of data available for developing a
structural model, etc.
Static pushover analysis (SPO) requires a mathematical model
that directly incorporates the nonlinear load-deformation char-
acteristics of the individual components and the elements of a
building. The structure is subjected to monotonically increasing
lateral forces that represent the seismic inertia forces. Compared
to nonlinear response history analysis (NRHA), Static Pushover
n
Corresponding author.
E-mail address: nlagaros@central.ntua.gr (N.D. Lagaros).
Despite their ease-of-use, such methods can provide important
information regarding the capacity of a structural system, while
their limitation is mainly related to the level of violation of their
underlying assumptions in real-life applications. The assumption
that the response of a multi-degree-of-freedom system is directly
related to the response of an equivalent single-degree-of-freedom
(SDOF) system, in several cases, is not accurate enough, since
apart from the fundamental mode, higher modes may contribute
to the response. Furthermore, the lateral load pattern is usually
applied without taking into consideration the member yielding
and its influence on the modification of the building properties as
the lateral forces are incremented.
A number of comparative studies assessing the performance of
the methods have been published in the past. After the capacity
spectrum method was adopted by ATC-40 [1], Fajfar [4] and
Chopra and Goel [5] pointed out that the ATC-40 procedure
significantly underestimated the deformation demands of sys-
tems for a wide range of periods when used for the Type A
idealized hysteretic damping model. Furthermore, Lin et al. [6]
compared the FEMA 273 [7] displacement coefficient method

0267-7261/$ - see front matter & 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.soildyn.2010.08.007
 Author's personal copy
78 N.D. Lagaros, M. Fragiadakis / Soil Dynamics and Earthquake Engineering 31 (2011) 77–90
with the ATC-40 procedures against experimental data on RC
columns. They conclude that compared to the test data both
methods introduce bias that is on average +28% and  20%,
respectively. Bosco et al. [8] discuss the differences and compare
the N2 and the ASCE-41 methods for a wide range of steel frames.
Their comparison is based on nonlinear response history analysis,
corresponding to hazard levels of increasing intensity. Cardone [9]
examined the ATC-40, the N2 and the displacement coefficient
method, against shaking table results, reaching to useful conclu-
sions about the performance of each method. Finally, Pinho et al.
[10] compared various nonlinear static procedures, including
multimodal or adaptive methods, on continuous span bridges.
In this study we compare the current state-of-practice static
pushover methods according to the provisions of contemporary
design codes and guidelines. The SPO methods chosen are the
ATC-40 [1] capacity spectrum method, the displacement coeffi-
cient method of ASCE-41 [3] and the acceleration-displacement
version of the N2 method [4] that has been recently adopted in
Eurocode 8 (2004). Contrary to other publications where
investigations on static pushover methods are carried out
assuming one or more predefined structural designs, we imple-
ment all SPO methods on an automated structural design
environment with the aid of an optimization algorithm. Thus,
we use the three methods in order to perform design and we
finally compare the capacity of the outcome buildings by means
of Incremental Dynamic Analysis and life-cycle cost assessment.
This alternative procedure pinpoints the pros and cons of each
method propagating the discrepancies between the static and the
dynamic analysis to the properties of the final structure.
2. Structural models and numerical simulation
The performance of the current state-of-practice SPO methods
is examined with respect to the two multi-storey reinforced
concrete, three-dimensional (3D) building structures of Fig. 1. The
Fig. 1. Plan views: (a) regular building configuration and (b) irregular building configuration.
first configuration is regular in plan and the second is irregular,
while both can be practically considered as regular in elevation
regardless of the fact that the first storey is 1.0 m higher than the
rest (Fig. 2). We consider two versions of each plan configuration
one with three stories and another with six stories, as shown in
Fig. 2. Thus we have in total four buildings that for the remainder
of the paper will be denoted as: RG3 (three-storey, regular frame),
RG6 (six-storey, regular frame), IRG3 (three-storey, irregular
frame) and IRG6 (six-storey, irregular frame). These are building
configurations typical of south Mediterranean countries (e.g.
Greece), where the vast majority of dwellings are RC buildings,
with similar characteristics.
The slab thickness for all four buildings is equal to 15 cm. The
slab contributes to the moment of inertia of the beams with an
effective flange width that varies between 0.65 and 1.4 m. In
addition to the self-weight of the beams and the slab, a
distributed dead load of 2 kN/m2 due to floor finishing and
partition walls are considered. The live load is assumed equal to
1.5 kN/m2. For the non-seismic load combination, the dead and
live loads are multiplied by 1.35 and 1.5, respectively; while
according to EC8 for the seismic load combination and ordinary-
use buildings the corresponding coefficients are equal to 1.0 and
0.30 [11], respectively. For the RC buildings considered, concrete
C20/25 (nominal cylindrical strength 20 MPa, modulus of elasti-
city 30 GPa) and S500 steel (nominal yield stress 500 MPa,
modulus of elasticity 210 GPa) are assumed.
All analyses were performed using the OpenSEES platform
[12]. Each member is modeled with a single force-based, fibre
beam-column element. This element provides a good balance
between accuracy and computational cost. The modified Kent–
Park model, where the monotonic envelope of concrete in
compression follows the model of Kent and Park [13] as extended
by Scott et al. [14], is employed for the simulation of the concrete
fibres. This model was chosen because it allows for an accurate
prediction of the demand for flexure-dominated RC members
despite its relatively simple formulation. The transient behavior of
 Author's personal copy
N.D. Lagaros, M. Fragiadakis / Soil Dynamics and Earthquake Engineering 31 (2011) 77–90
Fig. 2. Typical side views for the frames of Fig. 1: (a) three-storey building configuration and (b) six-storey building configuration.
the reinforcing bars was simulated with the Menegotto-Pinto
model [15], while the effects of shear and bond-slip are neglected.
The effect of gravity loads and second-order effects are considered
using the complete geometric stiffness matrix.
3. Performance-based design
Performance-based earthquake engineering (PBEE) implies the
design, evaluation, construction and maintenance of engineering
facilities in order to meet the objectives set by the society and the
owners/users of a facility [16]. In the case of earthquakes, the aim
is to make structures having a predictable and reliable perfor-
mance, or in other words, they should be able to resist
earthquakes with quantifiable confidence. Therefore, the modern
conceptual approach of seismic structural design is that the
structures should meet performance-based objectives for a
number of different hazard levels ranging from earthquakes with
a small intensity and with a small return period, to more
destructive events with large return periods. The current state
of practice in performance-based earthquake engineering is
defined by US guidelines [1,3]. These guidelines do not differ
conceptually and introduce procedures that can be considered as
the first significant diversification from prescriptive building
design codes. Many of the current codes for the design of new
buildings are only partially performance-based, since they
attempt to tie all design criteria to one performance level, usually
to that of Life Safety or Collapse Prevention.
For nonlinear structural analysis it is essential to formulate
structural models that incorporate all the essential characteristics
of the problem to be examined and can estimate the demand
within acceptable accuracy. In order to evaluate the demand,
appropriate engineering demand parameters (EDPs) are neces-
sary. As an EDP any response variable can be used, such as stress
resultants, displacements, chord rotations, among others. Accord-
ing to ASCE-41 the actions can be either force or deformation-
controlled depending on the capacity of the members to deform
inelastically. The capacity of force-controlled members should be
assessed using formulas based on stress resultants (e.g. EC2 [17]),
while for deformation-controlled actions an appropriate EDP
must be chosen. EDPs may be interstorey drifts, inelastic
79
deformations, section curvatures, floor accelerations and velo-
cities, etc. [18,19].
4. Static pushover methods for seismic design
The purpose of the nonlinear static procedure is to assess
structural performance in terms of strength and deformation
capacity globally as well as at the element level. The structural
model is ‘‘pushed’’ according to a predefined lateral load pattern.
In order to determine the target displacement in multiple hazard
levels required by the performance-based design framework,
typically one of the following methods is adopted: the Capacity
Spectrum method of ATC-40 [1], the Coefficient method of ASCE-
41 [3] and the N2 method of EC8 [2]. According to ASCE-41, apart
from a first-mode based lateral load pattern the use of a uniform
pattern is also suggested. In the numerical results that follow only
the first-mode pattern was taken into consideration. For 3D
structures the properties of the lateral load pattern have to be
extracted from the mode that refers to the direction under
consideration.
4.1. The displacement coefficient method (ASCE-41)
When pushover analysis, or adopting the ASCE-41 terminol-
ogy, the nonlinear static procedure (NSP) is implemented, the
target displacement, which is the displacement during a given
seismic event of a characteristic node on the top of a structure,
typically in the roof, is defined with the aid of the formula:
Te2
dt ¼ C0 C1 C2 C3 Sa g ð1Þ
4p2
where C0, C1, C2 and C3 are modification factors, discussed in the
FEMA-440 [20] guidelines and Te is the effective fundamental
period of the building.
4.2. The capacity spectrum method (ATC-40)
The capacity spectrum method (CSM) was initially proposed
by Freeman [21]. The method compares the capacity of a structure
to resist lateral forces to the demand given by a response
 Author's personal copy
80 N.D. Lagaros, M. Fragiadakis / Soil Dynamics and Earthquake Engineering 31 (2011) 77–90
spectrum. The response spectrum represents the demand while
the pushover curve (or the ‘capacity curve’) represents the
available capacity. Among the three variations of the method
discussed in ATC-40, we examine Procedure A. The steps of the
method are briefly summarized as follows: (i) Perform pushover
analysis and determine the capacity curve in base shear (Vb)
versus roof displacement of the building (D). This diagram is then
converted to acceleration–displacement terms (AD) using an
equivalent single degree of system (ESDOF). The conversion is
performed using the first mode participation factor C0 (D* ¼D/C0)
and the modal mass (A¼Vb/M). (ii) Plot the capacity diagram on
the same graph with the 5%-damped elastic response spectrum
that is also in AD format. (iii) Select a trial peak deformation
demand dt and determine the corresponding pseudo-acceleration
A from the capacity diagram, initially assuming z ¼5%. (iv)
Compute ductility m ¼D*/uy and calculate the hysteretic damping
zh as zh ¼2(m 1)/pm. The equivalent damping ratio is evaluated
from a relationship of the form:zeq ¼ zel þ kzh , where k is a
damping modification factor that depends on the hysteretic
behavior of the system. Update the estimate of dt using the elastic
demand diagram for zeq. (v) Check for convergence the displace-
ment dt . When convergence has been achieved the target
displacement of the MDOF system is equal to dt ¼ C0 dt .
4.3. The N2 method (EC8)
The N2 method was initially proposed by Fajfar [22,23] and
was later expressed in a displacement-acceleration format [4].
Recently, the method has been included in the Eurocode 8 (2004).
Conceptually the method is a variation of Capacity Spectrum
Method that instead of highly damped spectra uses an R–m–T
relationship. The method, as implemented in EC8, consists of the
following steps: (i) Perform pushover analysis and obtain the
capacity curve in Vb–D terms, (ii) convert the pushover curve of
the MDOF system to the capacity diagram of an ESDOF system
and approximate the capacity curve with an idealized elasto-
perfectly plastic relationship to get the period Te of the ESDOF, (iii)
the target displacement is then calculated as
 2
Te
det ¼ Sa ðTe Þ ð2Þ
2p
where Sa(Te) is the elastic acceleration response spectrum at the
period Te. To determine the target displacement dt , different
expressions are suggested for the short and the medium to long-
period ranges, thus:
 T* oTC (short period range): If Fy =m ZSa ðTe Þ, the response is
elastic and thus dt ¼ det and dt ¼ C0 dt . Otherwise the response is
nonlinear and the ESDOF maximum displacement is calculated as
 
d TC
dt ¼ et 1 þðqu 1Þ Zdet ð3Þ
qu Te
where qu is the ratio of the acceleration of the structure with
unlimited elastic capacity times the modal mass m* over its yield
force, or simply: qu ¼ Sa ðTe Þm =Fy .
 T* ZTC (medium and long period range): The target
displacement of the inelastic system is equal to that of an elastic
structure, thus dt ¼ det . The displacement of the MDOF system is
always calculated as dt ¼ C0 dt .
5. Preliminary evaluation of static pushover methods
Initially we examine the influence of different state-of-practice
SPO methods on the variation of two commonly adopted
engineering demand parameters (EDPs), the maximum interstor-
ey drift (ymax) and the maximum roof drift (yroof). We perform a
preliminary analysis to compare the three methods on a given RC
building configuration, thus following the most common and
conceptually straightforward approach to compare two analysis
methods. The results of the SPO methods are compared to those of
the Incremental Dynamic Analysis (IDA) method [24] for the
regular in plan, three-storey structure (RG3).
For this comparison we adopt an arbitrary design of the RG3
building configuration. Thus, the columns are considered to have
dimensions 40  40 cm2 and 8Ø22 reinforcement, while the
beams are of 25  50 cm2 with 6Ø20 bars. In total the weight of
the reinforcement steel is equal to 1910 and 1705 kg while the
concrete volume is equal to 12.8 and 14.4 m3 for the columns and
beams, respectively. The base shear was obtained from the
response spectrum for soil type B (characteristic periods TB ¼0.15
s, TC ¼0.50 s and TD ¼2.00 s), while the PGA considered is equal to
0.32 g. Moreover, the importance factor gI was taken equal to 1.0,
while the damping correction factor is equal to 1.0, since a
damping ratio of 5% has been considered.
IDA is here performed in the form of multi-stripe analysis,
according to the notation of Jalayer [25]. To this cause a suite of
twenty strong ground motion records is adopted. The twenty
records and their properties are listed in Table 1, where PGAlong
and PGAtran stands for the peek ground acceleration along the
longitudinal and transverse directions, respectively. The records
are scaled to four levels of seismic intensity, represented by a
common intensity measure (IM). The chosen IM is the 5%-
damped, first mode spectral acceleration, denoted as Sa(T1,5%).
Therefore, all records are scaled to the same Sa(T1,5%) values,
obtained from the EC8 elastic spectrum for peak ground
acceleration (PGA) values equal to 0.16, 0.24, 0.30 and 0.50 g.
Following the terminology of ASCE-41, the PGA values refer to the
following performance levels: operational (O), immediate occu-
pancy (IO), life safety (LS) and collapse prevention (CP),
respectively. In order to preserve the relative scale of the two
components of the records, the scaling factor of every record is
obtained from the component with the larger PGA value and is the
same for both components of the record. When IDA is performed,
we obtain 16%, 50% and 84% fractiles of the chosen EDP. The 50%
fractile is the median demand, conditional on the intensity
measure IM, while the 16%, 84% fractiles are used to provide a
measure of dispersion around the median.
The EDPs examined are the maximum interstorey drift, ymax,
X
and the maximum roof drift in the two building directions, yroof
Y
and yroof . According to the recommendation of ASCE-41, the
multidirectional effect is accounted as the sum of the 100% of the
response due to loading in the longitudinal direction and the 30%
of the response in the transverse direction, and vice versa.
Although this recommendation refers to forces, it was here
adopted to combine the drift ymax values. Thus, the ymax values
shown are the maximum values of the SRSS combination of the
X Y
corresponding ymax and ymax values, along the height of the
building:
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ymax ¼ maxð ðyXmax Þ2 þ ð0:3yYmax Þ2 , Y
ðymax Þ2 þ ð0:3ymax Þ2 Þ
X
ð4Þ
Fig. 3 shows the demand in terms of the maximum interstorey
drift, ymax, for the four stripes of the chosen intensity measure,
Sa(T1,5%) and Fig. 4 presents the demand in terms of the
X
maximum roof drift in the two building directions, yroof and
Y
yroof . The building’s response when subjected to the twenty
records of Table 1 and the corresponding 16%, 50% and 84%
fractile values are plotted against the demand values of the SPO
procedures.
Both figures show the variability of the response estimates due
to the different SPO methods and also the record-to-record
 Author's personal copy

N.D. Lagaros, M. Fragiadakis / Soil Dynamics and Earthquake Engineering 31 (2011) 77–90 81

Table 1
Characteristics of the 20 records.

a b c d
Record and Station R (km) EpiD (km) Duration (s) PGAlong (g) PGAtran (g) Soile Fault ruptf

Superstition Hills 1987 (B) (M ¼ 6.7)

1. El Centro Imp. Co Cent 18.5 35.83 40.00 0.36 0.26 A SS
2. Wildlife Liquefaction Array 24.1 29.41 44.00 0.18 0.21 A SS
Imperial Valley 1979, (M ¼6.5)
3. Chihuahua 8.4 18.88 40.00 0.27 0.25 A SS
4. Compuertas 15.3 24.43 36.00 0.19 0.15 A SS
5. Plaster City 31.1 54.26 18.75 0.04 0.06 A SS
6. El Centro Array #12 18.85 31.99 39.00 0.14 0.12 A SS
7. El Centro Array #13 22.83 35.95 39.50 0.12 0.14 A SS
San Fernando 1971 (M¼ 6.6)
8. LA, Hollywood Stor. Lot 25.9 39.49 28.00 0.21 0.17 A RN
Northridge 1994 (M¼ 6.7)
9. Leona Valley #2 37.2 51.88 32.00 0.09 0.06 A RN
10. LA, Baldwin Hills 29.9 28.20 40.00 0.24 0.17 C RN
11. Lake Hughes #1 89.67 93.22 32.00 0.09 0.08 A RN
12. LA, Hollywood Stor FF 114.62 118.26 40.00 0.23 0.36 A RN
13. LA, Centinela St. 31.53 32.72 30.00 0.46 0.32 A RN
Loma Prieta 1989 (M¼ 6.9)
14. Hollister Diff Array 24.8 45.10 39.64 0.27 0.28 A RO
15. WAHO 17.5 12.56 24.96 0.37 0.64 C RO
16. Halls Valley 30.5 36.31 39.95 0.13 0.10 B RO
17. Agnews State Hospital 24.6 40.12 40.00 0.17 0.16 A RO
18. Anderson Dam (Downstream) 4.4 16.67 39.61 0.24 0.24 B RO
19. Coyote Lake Dam (Downstream) 20.8 30.89 39.95 0.16 0.18 B RO
20. Hollister – South & Pine 27.93 48.24 60.00 0.37 0.18 A RO

a
Campbell’s R distance.
b
Distance from the recording site to epicenter.
c
Long: longitidunal direction.
d
Trans: transverse direction.
e
Campbell’s site classification: A (form soil), B (very firm soil), C (soft rock).
f
Fault rupture mechanism: SS (Strike Slip), RN (Reverse-Normal), RO (Reverse-Oblique).

As it can be seen from Fig. 4, N2 and ASCE-41 methods also

coincide for all intensity levels when yroof is the chosen EDP. On
X
1.2 the other hand, for every intensity level yroof obtained by the SPO
procedures is between the 50% and 16% fractiles while
Y
the corresponding yroof is below the 84% fractile. In the case of
1 the capacity spectrum method along the X-direction, the values
Sa (T1,5%) (g)

in the lower intensity levels coincide with the values of the 16%
0.8 medians while for the higher intensities it exceeds all maximum
values of the twenty records. Along the Y-direction, though,
similarly to N2 and ASCE-41 the values coincide with those of the
0.6 84% fractile. Therefore, since both ASCE-41 and N2 have as their
single records basis the equal displacement rule, for buildings that do not
IDA 16,50,84% fractiles deviate considerably from the conditions that the equal displace-
0.4 ATC40
ment rule applies, i.e. regular, first-mode dominated buildings of
EC8-N2
ASCE-41
medium to long period or for cases where the response is nearly
0.2 elastic (small R values) the discrepancies are expected to be small.
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
maximum interstorey drift, θmax (%) 6. Lower bound performance-based design
Fig. 3. Comparison of the SPO methods against the 16, 50 and 84% values of IDA in
terms of the maximum interstorey drift (ymax) (RG3 structure). The ultimate objective of our study is to compare lower-bound
designs, or in other words comparing the designs that satisfy the
code requirements in the most cost-effective way, i.e. those with
minimum cross-section and reinforcement dimensions. For this
reason, a structural optimization problem is formulated and the
variability of IDA. In particular, for all four intensity levels the N2 designs obtained are then assessed.
and the ASCE-41 results practically coincide with the median ymax
of the IDA method. On the other hand the capacity spectrum 6.1. Heuristic design of reinforce concrete structures
method of ATC-40 produces slightly higher predictions for the
first three Sa(T1,5%) intensities, while as the IM level increases, the The formulation of a structural optimization problem, which
demand approaches the 16% fractile. constitutes the basis for the performance evaluation of different
 Author's personal copy
82 N.D. Lagaros, M. Fragiadakis / Soil Dynamics and Earthquake Engineering 31 (2011) 77–90
1.2
1 1
Sa (T1 ,5%) (g)
0.8
0.6
0.4
0.2
0 0.2 0.4 0.6 0.8 1 1.2
roof drift, θXroof (%)
Fig. 4. Comparison of the SPO methods against the 16%, 50% and 84% values of IDA in terms of the maximum roof drift (yroof) (RG3 structure): (a) X-direction and
(b) Y-direction.
reinforced concrete building designs, is defined as follows:
mins A F CTOT ðt,sÞ
subject to gjSERV ðsÞ r 0 j ¼ 1,:::,m
gjPBD ðsÞ r 0 j ¼ 1,:::,k
where s represents the design vector with the cross-section
dimensions of all columns and beams, F is the feasible region
where all the serviceability and performance-based constraint
functions (gSERV and gPB) are satisfied. The objective function
considered is the total cost CTOT of the design and its calculation is
discussed in the following section. For the solution of the
optimization problem at hand an evolutionary type of algorithm
(EA) is employed. A detailed description of the EA method
implemented in this study can be found in Lagaros et al. [26].
The columns and the beams have a rectangular cross-section
and are separated into groups. For the regular building the
structural elements (beams and columns) for both building
configurations are separated into 9 groups (8 for the columns
and 1 for the beams). Therefore, we have 27 design variables
which correspond to the dimensions of the rectangular cross-
section and the longitudinal reinforcement. Similarly, the IRG3
and the IRG6 buildings are separated into 12 groups, 10 groups for
the columns and 2 for the beams, thus resulting to 36 design
variables in total.
6.2. Life-cycle cost analysis
Life-cycle cost analysis (LCCA) is a significant tool in economic
analysis. LCCA constitutes the core of the investment decision-
framework for structure and infrastructure assistant management
where it is used to evaluate alternative investment options [27].
The total expected cost CTOT of a structure, may refer either to the
design life period of a new structure or to the remaining life
period of a retrofitted structure. This cost can be expressed as a
function of the time and the design vector as follows [28]:
CTOT ðt,sÞ ¼ CIN ðsÞ þCLS ðt,sÞ
where CIN is the initial cost of a new or a retrofitted structure, CLS
is the present value of the expected limit-state cost; s is the
1.2
Sa (T1 ,5%) (g)
0.8
0.6
single records
IDA 16,50,84% fractiles
0.4 ATC40
EC8-N2
ASCE-41
0.2
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
maximum interstorey drift, θmax (%)
design vector corresponding to the design loads, resistance and
material properties and t is the time period. The term initial cost
refers to the cost just after the construction of a new structure.
ð5Þ The initial cost is related to the material and the labor cost for the
construction of the building, including concrete, steel reinforce-
ment, labor cost for placement as well as the cost of the non-
structural components. The term limit-state cost refers to the
potential damage cost from earthquakes that may occur during
the life of the structure. It accounts for the cost of the repairs after
an earthquake, the cost of loss of contents, the cost of injury
recovery or human fatality and other direct or indirect economic
losses. The quantification of the losses in economical terms
depends on several socio-economic parameters. The expected
limit-state cost (CLS), for the i-th limit-state, can be formulated as
follows:
i i i i i i i
CLS ¼ Cdam þCcon þ Cren þCinc þCinj þCfat ð7Þ
i i
where Cdam is the damage repair cost, Ccon is the loss of contents
i i
cost, Cren is the loss of rental cost and Cinc is the income loss cost,
i i
Cinj is the cost of injuries and Cfat is the cost of human fatality.
Details about the calculation formula for each limit-state cost
along with the values of the basic cost for each category used in
our study can be found in the works of Wen and Kang [29] and
Lagaros [30]. The values of the mean damage index, loss of
function, down time, expected minor injury rate, expected serious
injury rate and expected death rate we adopted, are based on
Ellingwood and Wen [31], Kang and Wen [32], ATC-13 [33] and
FEMA 227 [34]. In Eq. (7), the most difficult cost to quantify is the
i
cost corresponding to the loss of a human life (Cfat ). There are
several approaches for its calculation, ranging from purely
economic reasoning to more sensitive that consider the loss of a
person irreplaceable. Therefore, the estimation of the cost of
exceedance of the collapse prevention damage state will vary
considerably according to which approach is adopted.
It is generally accepted that maximum interstorey drift, ymax,
can be used to quantify the structural damage and also the non-
ð6Þ
structural but drift-sensitive damage. Based on analytical and
experimental data Ghobarah [35] examined the correlation
between drift and damage of various structural elements and
 Author's personal copy

N.D. Lagaros, M. Fragiadakis / Soil Dynamics and Earthquake Engineering 31 (2011) 77–90 83

systems. He determined the relationship between interstorey form (Table 3):

drift and various damage levels of different reinforced concrete
i i
elements and structural systems. Pi ðymax ymax Þ ¼ a1 ðymax Þa2 ð10Þ
Based on the Poisson model of earthquake occurrence and
the assumption that after a major damage-inducing seismic event,
the building is immediately retrofitted to its original intact
conditions, Wen and Kang [28,29] proposed the following formula
for the expected life-cycle cost considering N damage states:

n X
N
CLS ðt,sÞ ¼ ð1elt Þ i
CLS Pi ð8Þ
l i¼1

i i
Pðymax 4 ymax Þ ¼ ð1=tÞln½1Pðymax 4 ymax Þ
i iþ1
Pi ¼ Pðymax 4 ymax ÞPðymax 4 ymax Þ ð9Þ

i
According to Eq. (8) the limit-state costs CLS of Eq. (8) are used
to calculateCLS ðt,sÞ, where Pi is the probability of the ith damage
state being violated given the occurrence of an earthquake. Eq. (9)
assumes the maximum interstorey drift ymax as the characteristic
i
EDP, while P(ymax 4 ymax ) is the exceedance probability given
i
occurrence. ymax and ymax are the drift ratios that define the lower
i
and upper bounds of the ith damage state and Pi (ymax  ymax ) is
the annual exceedance probability of the maximum interstorey
i
drift value ymax . Finally, n is the annual occurrence rate of
significant earthquakes modeled as a Poisson variable. The
exponential component of Eq. (8) is used to express CLS in present
value, where l is the annual momentary discount rate. The annual
momentary discount rate is typically taken to be constant and
equal to 5% [36], while its value lies in the range of 3–6%. Each
damage state is defined by appropriate drift ratio limits depend-
ing on the type of structure, e.g. steel or RC structure (Table 2).
When one of the limit-state drifts is reached, the corresponding
limit-state is assumed to be exceeded. Parameters a1 and a2 are
i
obtained by best fit of known pairs Pi  ymax and the annual
exceedance probability Pi is obtained from a relationship of the
Fig. 5. Flowchart of the proposed performance-based design procedure [37].

Table 2
Limit-state drift ratio limits for bare RC moment resisting frames according to Table 4
Ghobarah [35]. Initial and total cost for the four building configurations with each of the SPO
methods.
Limit-state Interstorey drift limit (%)
ATC-40 EC8-N2 ASCE-41
(I) – None ymax r0.1
(II) – Slight 0.1 o ymax r0.2 CIN (CTOT) in 1000h CIN (CTOT) in 1000h CIN (CTOT) in 1000h
(III) – Light 0.2 o ymax r0.4
(IV) – Moderate 0.4 o ymax r1.0 RG3 269.86 (296.58) 270.11 (293.49) 280.11 (300.95)
(V) – Heavy 1.0 o ymax r1.8 RG6 551.51 (620.73) 536.05 (613.24) 531.77 (586.14)
(VI) – Major 1.8 o ymax r 3.0 IRG3 510.81 (628.31) 513.41 (584.06) 514.64 (584.29)
(VII) – Collapsed 3.0 o ymax IRG6 1027.41 (1271.96) 1011.23 (1187.09) 1012.66 (1165.49)

Table 3
Limit-state parameters for cost evaluation.

FEMA-227 [34] ATC-13 [33]

Performance
Damage state Expected
level Mean damage Expected minor Expected death Loss of function
serious injury Down time (%)
index (%) injury rate rate (%)
rate

I None 0 0 0 0 0 0
II Slight 0.5 3.0  10  5 4.0  10  6 1.0  10  6 0.9 0.9
III Light 5 3.0  10  4 4.0  10  5 1.0  10  5 3.33 3.33
IV Moderate 20 3.0  10  3 4.0  10  4 1.0  10  4 12.4 12.4
V Heavy 45 3.0  10  2 4.0  10  3 1.0  10  3 34.8 34.8
VI Major 80 3.0  10  1 4.0  10  2 1.0  10  2 65.4 65.4
VII Collapsed 100 4.0  10  1 4.0  10  1 2.0  10  1 100 100
 Author's personal copy

84 N.D. Lagaros, M. Fragiadakis / Soil Dynamics and Earthquake Engineering 31 (2011) 77–90

6.3. Step-by-step analysis
For all building configurations our parametric studies are
performed in two stages: (i) the building is designed with the
aid of a structural optimization algorithm and (ii) life-cycle
cost analysis is carried out based on IDA. In the first stage
the optimum design is computed employing the EA(m + l)
optimization scheme [26] with ten parent and offspring
(m ¼ l ¼10) design vectors. Following the performance-based
design framework described above, three optimization problems

1.6 1.6

1.3 1.3
Sa (T1,5%) (g)

Sa (T1,5%) (g)
1 1

0.7 0.7 ATC-40

EC8-N2
ASCE-41
0.4 0.4
0 0.4 0.8 1.2 1.6 2 0 0.4 0.8 1.2 1.6 2
maximum interstorey drift, θmax (%) roof drift, θroof (%)

Fig. 6. 3-storey regular building (RG3): median Sa(T1,5%) capacities with respect to (a) maximum interstorey drift (ymax) and (b) roof drift (yroof).

1 1

0.8 0.8
Sa (T1,5%) (g)

Sa (T1,5%) (g)

0.6 0.6

0.4 0.4 ATC-40

EC8-N2
ASCE-41
0.2 0.2
0 0.4 0.8 1.2 1.6 2 0 0.4 0.8 1.2 1.6 2
maximum interstorey drift, θmax (%) roof drift, θroof (%)

Fig. 7. 6-storey regular building (RG6): median Sa(T1,5%) capacities with respect to (a) maximum interstorey drift (ymax) and (b) roof drift (yroof).

1.4 1

1.1 0.8
Sa (T1,5%) (g)

Sa (T1,5%) (g)

0.8 0.6

0.5 0.4 ATC-40

EC8-N2
ASCE-41
0.2 0.2
0 0.4 0.8 1.2 1.6 2 0 0.5 1 1.5 2 2.5
maximum interstorey drift, θmax (%) roof drift, θroof (%)

Fig. 8. 3-storey irregular building (IRG3): median Sa(T1,5%) capacities with respect to (a) maximum interstorey drift (ymax) and (b) roof drift (yroof).
 Author's personal copy

N.D. Lagaros, M. Fragiadakis / Soil Dynamics and Earthquake Engineering 31 (2011) 77–90 85

are defined. The detailed description of the exact steps followed Moreover, for all four building configurations we calculate the
for the seismic design of the buildings can be found in Lagaros total life-cycle cost which is the objective function to be
and Papadrakakis [37], while a flowchart of this algorithm minimized. In the second stage, life-cycle cost analysis is
is presented in Fig. 5. According to Fig. 5 the static pushover performed on the optimum designs using IDA as the method to
methods are nested in the iterative optimum design algorithm. assess the capacity.

1 1

0.8 0.8
Sa (T1,5%) (g)

Sa (T1,5%) (g)
0.6 0.6

0.4 0.4 ATC-40

EC8-N2
ASCE-41
0.2 0.2
0 0.5 1 1.5 2 2.5 0 0.5 1 1.5 2 2.5
maximum interstorey drift, θmax (%) roof drift, θroof (%)

Fig. 9. 6-storey irregular building (IRG6): median Sa(T1,5%) capacities with respect to (a) maximum interstorey drift (ymax) and (b) roof drift (yroof).

305 300.95

300
296.58
293.49 300.76
295 296.29
Cost (1000 €)

290 293.28

285

280 280.10

CIN
275
TOT (I)
270 269.86 270.11
TOT (II)
265
ATC40 EC8-N2 ASCE 41
Designs

Fig. 10. 3-storey regular building (RG3): (a) initial (CIN) and total expected (CTOT) life-cycle costs for the three SPO-based designs and (b) contribution of the initial cost and
limit-state cost components to the total expected life-cycle cost.

86 N.D. Lagaros, M. Fragiadakis / Soil Dynamics and Earthquake Engineering 31 (2011) 77–90
7. Numerical results
The optimum designs achieved when the three SPO methods
are implemented in a seismic design algorithm are discussed in
this section. The design spectrum used for the design corresponds
to soil type B (characteristic periods TB ¼0.15 s, TC ¼ 0.50 s and
TD ¼2.00 s). Moreover, the importance factor gI was taken equal to
1.0, while the damping correction factor Z is equal to 1.0, since a
damping ratio of 5% has been considered. Depending on the
performance objective examined, the spectrum is scaled to
the following PGA values: 0.16, 0.24, 0.30 and 0.50 g, while the
corresponding limit-state values are equal to 0.5%, 1.0%, 1.5%
and 2.5%.
The properties of the building designs obtained are shown
in Tables A1–A4 of the Appendix. In the bottom of the Tables
the initial cost (CIN) and the total cost (CTOT, Eq. (6)) are shown,
which are also summarized in Table 4. The initial cost of
the structure CIN is related to material cost, which includes
concrete, steel reinforcement, labor costs for the construction
of the building and the cost of the non-structural elements.
Additionally, the cost of the contents is included. The third,
fourth and fifth column of Tables A1–A4 show the cross-section
dimensions and the reinforcement which is assumed symmetric
for every column cross-section. In the sections that follow, we
assess the designs obtained with each of the three state-of-the-art
SPO methods with respect to their (i) initial cost, (ii) seismic
demand and (iii) total and life-cycle cost.
640
620.73
620
619.84
Cost (1000 €)
600
580
560
CIN
540 TOT (I)
TOT (II)
520
ATC40
Fig. 11. 6-storey regular building (RG6): (a) initial (CIN) and total expected (CTOT) life-cycle costs for the three SPO-based designs and (b) contribution of the initial cost and
limit-state cost components to the total expected life-cycle cost.
Author's personal copy
7.1. Comparison with reference to the initial cost (CIN)
Table 4 compares the costs of each of the four building
configurations obtained with the SPO methods that we compare.
For the RG3 building the ATC-40 and the EC8 method resulted to
designs that practically have the same initial cost, which is
slightly lower (3.5%) than that of the ASCE-41 method. For the
RG6 building the situation is reversed, but again the differences
with respect to the CIN are quite insignificant, with the ATC-40 and
the EC8 designs compared to the ASCE-41 case being slightly
more expensive by 3.5% and 1%, respectively. For the irregular
buildings there are small differences only for the IRG6 building
when the ATC-40 procedure is adopted. In this case, the cost is
approximately 1.5% higher than that of the other methods, while
for the RG3 building the CIN values are very close. Therefore, from
Table 4 it is evident that when any of the three SPO methods is
adopted for design, the outcome building will practically have
the same initial cost with minor differences that depend on the
problem itself. These differences will depend on the inaccuracy
on the estimation of the maximum interstorey drift imposed by
each method.
7.2. Comparison with respect to the seismic demand
Figs. 6–9 show the median demand for given Sa(T1,5%)
capacities obtained by IDA, where the ground motion records
are scaled to the same intensity level, expressed in Sa(T1,5%)
613.24
612.49
586.14
585.60
551.51
536.05
531.77
EC8-N2 ASCE 41
Designs
 Author's personal copy
N.D. Lagaros, M. Fragiadakis / Soil Dynamics and Earthquake Engineering 31 (2011) 77–90
terms. The EDPs considered are the maximum interstorey drift
(ymax) and the roof drift (yroof), both calculated according to
Eq. (4). The results shown have been obtained with the twenty
records of Table 1, while for every intensity level we show only
the median values obtained for each building design. For the RG3
structure the ATC-40 procedure produces larger median demands
compared to the other two methods, regardless of the EDP (Fig. 6).
The EC8 and the ASCE-41 approaches practically coincide when
ymax is the EDP but some small differences are observed if the
comparison is performed with respect to yroof. Similarly for the
RG6 building, the ATC-40 procedure causes larger demands
(Fig. 7), while the difference on the median demands is significant
for both ymax and yroof comparisons. Especially for the yroof case
(Fig. 7b) the EC8 produces slightly larger demands compared to
the ASCE-41 approach.
For the six-storey buildings (RG6, IRG6) the ASCE-41 approach
produces the smallest drift demands as was also the case with the
RG3 structure, while the larger demands are those of the ATC-40
method. The EC8 approach fans between the demands of the other
two SPO methods and is closer to the ATC-40 prediction for the
regular case and to ASCE-41 for the irregular building. In general,
from Figs. 6–9, it is seen that the ATC-40 method produces buildings
more flexible and therefore with larger demands, compared to the
other methods, while and the opposite applies for the ASCE-41
method. This behavior, to some extent, explains the findings of the
comparison with respect to the initial cost. The design with largest
CIN is expected to have the smallest drift demand.
640 628.31
620
623.52
600
Cost (1000 €)
580
560
540
520
510.81
500
480
ATC40
Fig. 12. 3-storey irregular building (IRG3): (a) initial (CIN) and total expected (CTOT) life-cycle costs for the three SPO-based designs and (b) contribution of the initial cost
and limit-state cost components to the total expected life-cycle cost.
87
7.3. Comparison with respect to the total and life-cycle cost
Figs. 10–13 show schematically the comparison of the SPO
methods with respect to the initial cost and the limit-state and total
life-cycle costs. For the life-cycle cost calculations we have adopted
the ymax threshold values of Ghobarah [35], while the corresponding
demands are obtained with Incremental Dynamic Analysis.
For the RG3 building the ASCE-41 results to a slightly increase,
which is 1.5–2.5% of the total cost CTOT compared to the other SPO
methods (Fig. 10a). The contribution of the initial and the limit-
state cost components to the total life-cycle cost, according to
Eq. (7), are shown in Fig. 10b. The initial cost is the dominant
contributor for all optimum designs; while the second and third
most important cost parameters are the damage and income
costs, respectively. This explains why although the ASCE-41 has
the smallest drift demand it has also the largest initial and life-
cycle cost. This is of value, because typically the design with the
smallest drift demand has also the largest life-cycle cost and the
smallest initial cost [38].
For the RG6 building the ATC-40 design has the largest CTOT,
while Fig. 11b implies that the most important cost parameter is
the initial cost, which is approximately 90% of the total. Moreover,
according to Fig. 11a the design of ASCE-41 is the best, since the
corresponding costs are increased by 6% and 4.5% for the ATC-40
and the EC8 approaches, respectively. With respect to Fig. 11b, the
damage and the income costs have also a major contribution, as
was the case in the RG3 building.
CIN
TOT (I)
TOT (II)
584.06 584.29
583.17 583.07
513.41 514.64
EC8-N2 ASCE 41
Designs
 Author's personal copy

88 N.D. Lagaros, M. Fragiadakis / Soil Dynamics and Earthquake Engineering 31 (2011) 77–90

, ,
, ,
,
, ,

, ,
,
,

, ,
,
,

Fig. 13. 6-storey irregular building (IRG6): (a) initial (CIN) and total expected (CTOT) life-cycle costs for the three SPO-based designs and (b) contribution of the initial cost
and limit-state cost components to the total expected life-cycle cost.

Figs. 12 and 13 show the performance of the three design
methods for the irregular buildings (IRG3, IRG6). For both building
configurations and regardless of the SPO method, initial cost is
dominant in the calculation of the total life-cycle cost (Figs. 12
and 13b). Furthermore, although the initial cost is practically the
same regardless of the SPO method the ATC-40 design has larger
total cost for 7.5% for the three-storey building and 9% for the six-
storey one, while the differences between the other two methods
are very small.
8. Conclusions
In this study we performed an investigation on the influence of
alternative static pushover methods when adopted for the seismic
design of new structures. The numerical investigation was
performed on four multi-storey reinforced concrete buildings,
with symmetrical and non-symmetrical plans, that have been
optimally designed. The current trend in evaluating investment
practices in structures is to take into consideration the total life-
cycle cost rather than the initial construction cost. In the present
study life-cycle cost analysis is employed for the assessment of
the seismic performance of buildings designed using either the
capacity spectrum method of ATC-40, the N2 method of EC8 or
the displacement coefficient method of ASCE-41.
Based on the limited number of buildings examined, we were
able to compare the three SPO methods with respect to the
properties of the outcome designs for RC buildings. Although the
focus of the study is not to determine which of the three methods
is more accurate, our results indicate that the ATC-40 method
tends to overestimate the demand, while the differences between
the EC8 and the ASCE-41 methods are rather small for low and
medium levels of seismic intensity. More specifically the main
findings of this study are summarized as follows:
1. The initial cost of reinforced concrete structures designed
based on ATC-40 varies compared to the designs obtained
according to N2 and ASCE-41 up to 3.7%, while similar
variability is encountered between the designs of N2 and
ASCE-41. On the other hand comparing the total life-cycle cost,
the designs of ATC-40 varies in the range of 1.0–9.0% compared
to the design of N2 and ASCE-41. The corresponding variability
between the designs of N2 and ASCE-41 is up to 4.5%.
2. Comparing, the contributing parts of the total life-cycle cost
(Figs. 10–13), it is concluded that the initial cost, in all test
examples, is the first dominant contributor for all designs. The
second and the third most important cost parameters are the
damage and income costs, respectively. It is also worth mention-
ing that, although the four designs differ significantly, injury and
fatality costs represent only a small percentage of the total cost.
3. Depending on the design method employed increasing
construction cost does not always mean that seismic safety
is also increases.
Appendix A
See Tables A1–A4.
 Author's personal copy

N.D. Lagaros, M. Fragiadakis / Soil Dynamics and Earthquake Engineering 31 (2011) 77–90 89

Table A1
3-storey regular building (RG3): optimum designs.

Designs

ATC-40 EC8-N2 ASCE-41

Columns
h1  b1 (reinforcement) 0.25  0.70 (10Ø28) 0.40  0.60 (10Ø20) 0.25  0.45 (6Ø18)
h2  b2 (reinforcement) 0.60  0.50 (15Ø28) 0.55  0.70 (8Ø28) 0.30  0.60 (8Ø22)
h3  b3 (reinforcement) 0.30  0.50 (6Ø18) 0.25  0.65 (8Ø20) 0.70  0.70 (24Ø28)
h4  b4 (reinforcement) 0.30  0.40 (6Ø20) 0.25  0.30 (4Ø20) 0.65  0.35 (10Ø18)
h5  b5 (reinforcement) 0.25  0.25 (4Ø18) 0.70  0.65 (14Ø22) 0.30  0.45 (8Ø24)
h6  b6 (reinforcement) 0.60  0.70 (10Ø22) 0.35  0.50 (10Ø28) 0.70  0.40 (8Ø22)
h7  b7 (reinforcement) 0.60  0.45 (16Ø28) 0.25  0.25 (4Ø18) 0.25  0.55 (8Ø22)
h8  b8 (reinforcement) 0.65  0.30 (8Ø20) 0.70  0.40 (4Ø18) 0.65  0.30 (8Ø22)

Beams
h  b (reinforcement) 0.65  0.25 (8Ø24) 0.65  0.35 (10Ø24) 0.65  0.35 (8Ø24)

CIN (1000 h) 269.86 270.11 280.11

CTOT (1000 h) 296.58 293.49 300.95

Table A2
6-storey regular building (RG6): optimum designs.

Designs

ATC-40 EC8-N2 ASCE-41

Columns
h1  b1 (reinforcement) 0.25  0.35 (6Ø24) 0.60  0.25 (6Ø18) 0.30  0.30 (6Ø24)
h2  b2 (reinforcement) 0.25  0.40 (4Ø18) 0.60  0.60 (8Ø22) 0.45  0.40 (6Ø28)
h3  b3 (reinforcement) 0.25  0.25 (4Ø18) 0.25  0.25 (4Ø28) 0.40  0.70 (16Ø30)
h4  b4 (reinforcement) 0.45  0.70 (16Ø30) 0.25  0.45 (6Ø24) 0.55  0.70 (8Ø24)
h5  b5 (reinforcement) 0.70  0.25 (8Ø28) 0.35  0.55 (4Ø24) 0.70  0.60 (10Ø22)
h6  b6 (reinforcement) 0.30  0.65 (14Ø24) 0.70  0.70 (18Ø30) 0.25  0.70 (10Ø24)
h7  b7 (reinforcement) 0.70  0.45 (14Ø30) 0.70  0.50 (8Ø24) 0.70  0.25 (8Ø24)
h8  b8 (reinforcement) 0.70  0.25 (10Ø28) 0.50  0.25 (4Ø20) 0.70  0.55 (10Ø28)

Beams
h  b (reinforcement) 0.60  0.35 (8Ø28) 0.65  0.35 (8Ø22) 0.65  0.30 (6Ø20)

CIN (1000 h) 551.51 536.05 531.77

CTOT (1000 h) 620.73 613.24 586.14

Table A3
3-storey irregular building (IRG3): optimum designs.

Designs

ATC-40 EC8-N2 ASCE-41

Columns
h1  b1 (reinforcement) 0.30  0.45 (8Ø22) 0.50  0.60 (10Ø28) 0.25  0.25 (4Ø18)
h2  b2 (reinforcement) 0.55  0.70 (6Ø28) 0.70  0.30 (8Ø24) 0.60  0.25 (6Ø24)
h3  b3 (reinforcement) 0.40  0.70 (6Ø24) 0.40  0.30 (4Ø24) 0.40  0.60 (10Ø28)
h4  b4 (reinforcement) 0.55  0.65 (8Ø24) 0.25  0.70 (6Ø24) 0.65  0.45 (6Ø24)
h5  b5 (reinforcement) 0.55  0.55 (8Ø24) 0.35  0.55 (8Ø24) 0.50  0.25 (4Ø20)
h6  b6 (reinforcement) 0.55  0.45 (6Ø24) 0.25  0.25 (4Ø18) 0.50  0.25 (6Ø22)
h7  b7 (reinforcement) 0.25  0.65 (6Ø24) 0.45  0.35 (4Ø24) 0.50  0.50 (6Ø24)
h8  b8 (reinforcement) 0.40  0.25 (4Ø18) 0.35  0.60 (6Ø28) 0.45  0.65 (8Ø28)
h9  b9 (reinforcement) 0.70  0.50 (10Ø28) 0.30  0.55 (10Ø24) 0.70  0.60 (10Ø24)
h10  b10 (reinforcement) 0.65  0.65 (8Ø28) 0.60  0.55 (8Ø24) 0.70  0.30 (6Ø24)

Beams
h1  b1 (reinforcement) 0.55  0.30 (6Ø20) 0.65  0.35 (6Ø22) 0.40  0.25 (6Ø20)
h2  b2 (reinforcement) 0.40  0.25 (4Ø18) 0.40  0.25 (6Ø20) 0.55  0.25 (6Ø20)

CIN (1000 h) 510.81 513.41 514.64

CTOT (1000 h) 628.31 584.06 584.29
 Author's personal copy

90 N.D. Lagaros, M. Fragiadakis / Soil Dynamics and Earthquake Engineering 31 (2011) 77–90

Table A4
6-storey irregular building (IRG6): optimum designs.

Designs

ATC-40 EC8-N2 ASCE-41

Columns
h1  b1 (reinforcement) 0.40  0.50 (4Ø24) 0.45  0.45 (6Ø28) 0.45  0.50 (6Ø28)
h2  b2 (reinforcement) 0.55  0.65 (8Ø28) 0.60  0.60 (10Ø28) 0.70  0.55 (10Ø28)
h3  b3 (reinforcement) 0.40  0.25 (4Ø20) 0.50  0.25 (4Ø24) 0.45  0.35 (6Ø24)
h4  b4 (reinforcement) 0.50  0.70 (10Ø28) 0.60  0.60 (8Ø24) 0.55  0.70 (10Ø28)
h5  b5 (reinforcement) 0.65  0.25 (6Ø20) 0.65  0.35 (10Ø24) 0.60  0.35 (8Ø24)
h6  b6 (reinforcement) 0.70  0.35 (6Ø28) 0.55  0.60 (10Ø24) 0.70  0.60 (12Ø24)
h7  b7 (reinforcement) 0.30  0.40 (8Ø24) 0.55  0.25 (4Ø24) 0.55  0.30 (6Ø24)
h8  b8 (reinforcement) 0.55  0.25 (8Ø22) 0.60  0.30 (4Ø24) 0.60  0.25 (4Ø24)
h9  b9 (reinforcement) 0.30  0.35 (6Ø22) 0.70  0.40 (6Ø24) 0.55  0.45 (10Ø22)
h10  b10 (reinforcement) 0.60  0.70 (8Ø28) 0.60  0.70 (8Ø28) 0.60  0.70 (8Ø28)

Beams
h1  b1 (reinforcement) 0.65  0.35 (6Ø24) 0.65  0.35 (6Ø24) 0.65  0.35 (6Ø24)
h2  b2 (reinforcement) 0.40  0.25 (4Ø22) 0.40  0.25 (4Ø22) 0.40  0.25 (4Ø22)

CIN (1000 h) 1027.41 1011.23 1012.66

CTOT (1000 h) 1271.96 1187.09 1165.49

References
[1] ATC-40. Seismic evaluation and retrofit of concrete buildings. Redwood City:
Applied Technology Council; 1996.
[2] EC8. Eurocode 8: Design of structures for earthquake resistance. European
Committee for Standardisation: Brussels, Belgium, The European Standard EN
1998-1, 2004.
[3] ASCE/SEI Standard 41-06. Seismic Rehabilitation of Existing Buildings.
prepublication edition. Structural Engineering Institute, American Society of
Civil Engineers; 2006.
[4] Fajfar P. Capacity spectrum method based on inelastic demand spectra.
Earthquake Engineering and Structural Dynamics 1999;28(9):979–99.
[5] Chopra AK, Goel RK. Evaluation of NSP to estimate seismic deformation:
SDF systems. Journal of Structural Engineering New York 2000;126(4):
482–490.
[6] Lin Y-Y, Chang K-C, Wang Y-L. Comparison of displacement coeffi-
cient method and capacity spectrum method with experimental results
of RC columns. Earthquake Engineering and Structural Dynamics 2004;33:
35–48.
[7] FEMA 273. NEHRP Guidelines for the Seismic Rehabilitation of Buildings.
Washington, DC: Building Seismic Safety Council for the Federal Emergency
Management Agency; 1997.
[8] Bosco M, Ghersi A, Marino EM. On the evaluation of seismic response of
structures by nonlinear static methods. Earthquake Engineering and
Structural Dynamics 2009;38(13):1465–82.
[9] Cardone D. Nonlinear static methods vs. experimental shaking table test
results. Journal of Earthquake Engineering 2007;11(6):847–75.
[10] Pinho R, Monteiro R, Casarotti C, Delgado R. Assessment of continuous span
bridges through nonlinear static procedures. Earthquake Spectra 2009;25(1):
143–59.
[11] EC0. Eurocode 0. Basis of structural design: Annex A1. European Committee
for Standardisation: Brussels, Belgium. The European Standard EN 2005. 1990.
[12] McKenna F, Fenves GL. The OpenSees Command Language Manual. (1.2. edn).
PEER; 2001.
[13] Kent DC, Park R. Flexural members with confined concrete. Journal of
Structural Division 1971;97(7):1969–90.
[14] Scott BD, Park R, Priestley MJN. Stress–strain behavior of concrete confined
by overlapping hoops at low and high strain rates. ACI Journal 1982;79:
13–27.
[15] Menegotto M, Pinto PE. Method of analysis for cyclically loaded reinforced
concrete plane frames including changes in geometry and non-elastic
behavior of elements under combined normal force and bending. In:
Proceedings, IABSE symposium on resistance and ultimate deformability of
structures acted on by well defined repeated loads, p. 15–22.
[16] Krawinkler H, Miranda E. Performance-base earthquake engineering. In:
Bozorgnia Y, Bertero Vitelmo V, editors. Earthquake Engineering: From Engineer-
ing Seismology to Performance-based Earthquake Engineering. CRC Press; 2004.
[17] EC2. Eurocode 2. Design of concrete structures—Part 1: General rules and
rules for buildings. The European Standard EN 1992-1-1. European
Committee for Standardisation: Brussels, Belgium; 2004.
[18] Fragiadakis M, Papadrakakis M. Modeling, analysis and reliability of
seismically excited structures: computational issues. International Journal
of Computational Methods 2008;5(4):483–511.
[19] Mitropoulou ChCh, Lagaros ND, Papadrakakis M. Economic building design
based on energy dissipation: a critical assessment. Bulletin of Earthquake
Engineering, doi:10.1007/s10518-010-9182-x.
[20] FEMA-440. Improvement of nonlinear static seismic analysis procedures.
Washington, DC: Federal Emergency Management Agency; 2005.
[21] Freeman SA. Development and use of capacity spectrum method. In:
Proceedings of the 6 US national conference on earthquake engineering,
1998.
[22] Fajfar P, Fischinger M. N2-a method for non-linear seismic analysis of regular
buildings. In: Proceedings of the 9th world conference on earthquake
engineering, Maruzen, Tokyo, 1988, Kyoto, 1989. p. 111–6.
[23] Fajfar P, Gaspersic P. The N2 method for the seismic damage analysis for
RC buildings. Earthquake Engineering and Structural Dynamics 1996;25:
23–67.
[24] Vamvatsikos D, Cornell CA. Incremental dynamic analysis. Earthquake
Engineering and Structural Dynamics 2002;31(3):491–514.
[25] Jalayer F. Direct probabilistic seismic analysis: implementing nonlinear
dynamic assessments. PhD thesis. Department of Civil and Environmental
Engineering, Stanford University, CA, 2003.
[26] Lagaros ND, Fragiadakis M, Papadrakakis M. Optimum design of shell
structures with stiffening beams. AIAA Journal 2004;42(1):175–84.
[27] Sahely HR, Kennedy CA, Adams BJ. Developing sustainability criteria for
urban infrastructure systems. Canadian Journal of Civil Engineering
2005;32(1):72–85.
[28] Wen YK, Kang YJ. Minimum building life-cycle cost design criteria. I:
Methodology. Journal of Structural Engineering 2001;127(3):330–7.
[29] Wen YK, Kang YJ. Minimum building life-cycle cost design criteria. II:
Applications. Journal of Structural Engineering 2001;127(3):338–46.
[30] Lagaros ND. Life-cycle cost analysis of construction practices. Bulletin of
Earthquake Engineering 2007;5:425–42.
[31] Ellingwood BR, Wen Y-K. Risk-benefit-based design decisions for low-
probability/high consequence earthquake events in mid-America. Progress
in Structural Engineering and Materials 2005;7(2):56–70.
[32] Kang Y-J, Wen YK. Minimum life-cycle cost structural design against natural
hazards, Structural Research Series No. 629. Department of Civil and
Environmental Engineering, University of Illinois at Urbana-Champaign,
Urbana, IL, 2000.
[33] ATC-13. Earthquake damage evaluation data for California. Redwood City,
CA: Applied Technology Council; 1985.
[34] FEMA 227. A benefit–cost model for the seismic rehabilitation of buildings.
Washington, DC: Federal Emergency Management Agency, Building Seismic
Safety Council; 1992.
[35] Ghobarah A. On drift limits associated with different damage levels. In:
International workshop on performance-based seismic design, June 28–July
1, 2004.
[36] Rackwitz R. The effect of discounting, different mortality reduction schemes
and predictive cohort life tables on risk acceptability criteria. Reliability
Engineering and System Safety 2006;91(4):469–84.
[37] Lagaros ND, Papadrakakis M. Seismic design of RC structures: a critical
assessment in the framework of multi-objective optimization. Earthquake
Engineering and Structural Dynamics 2007;36(12):1623–39.
[38] Fragiadakis M, Lagaros ND, Papadrakakis M. Performance-based multi-
objective optimum design of steel structures considering life-cycle cost.
Structural and Multidisciplinary Optimization 2006;32:1.