Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
Abstract: In earthquake engineering, the performance-based design method is used to determine the expected performance level of struc-
tures under an earthquake effect. The level of performance is related to the damage situation that could occur in the structure after the
earthquake. In performance-based structural design, it is predicted that more than one damage level can emerge under one earthquake effect.
In this study, the seismic behavior of a reinforced-concrete shear-wall building that collapsed during the 2003 Bingöl earthquake was in-
vestigated by nonlinear static analysis and nonlinear dynamic analysis. The selected reinforced-concrete shear-wall structure is located in
Bingöl, Turkey. Local code was considered for assessing a seismic performance evaluation of the selected reinforced-concrete shear-wall
building. The performance goals of the reinforced-concrete shear-wall structure were evaluated by applying the pushover (incremental equiv-
alent earthquake load) method and procedures of the code and nonlinear dynamic analysis. According to the code, the reinforced-concrete
shear-wall building was not expected to satisfy life safety performance levels under a design earthquake. In this study, one collapsed building
was selected in order to test the reliability and usability of performance analysis methods under different earthquakes. DOI: 10.1061/(ASCE)
CF.1943-5509.0001117. © 2017 American Society of Civil Engineers.
Author keywords: Reinforced-concrete shear-wall structure; Pushover analysis; Nonlinear time-history analysis; Performance analysis.
Introduction most reliable analysis method among all the nonlinear analysis
methodologies. However, static pushover analysis has become im-
In Turkey, there are a large number of reinforced-concrete buildings portant due to its easy application compared to time-history analy-
that just meet or fall short of the standards of earthquake safety. sis. Many papers have been published on the topic of performance
In addition, many existing reinforced-concrete shear-wall buildings evaluation of existing reinforced-concrete buildings. Scawthorn and
in first-degree seismic zones need seismic evaluation due to non- Johnson (2000) investigated the predominant building type, which
compliance with old code requirements, updating of codes, and consists of midrise nonductile reinforced-concrete frames with hol-
building design practice. Their maintenance and reinforcement is low clay tile infill, thousands of which have collapsed in a pancake
not possible due to economic and technical reasons. A more real- mode. Krawinkler and Seneviratna’s (1998) research indicated that
istic form of earthquake safety evaluation for existing buildings has static pushover analysis may be less accurate for structures in which
come into question. In the Turkish Earthquake Code (TEC) in 2007 the story shear force versus story drift relationships are sensitive to
(TEC 2007), performance-based evaluations were emphasized by the applied load. Sucuoğlu (2006) investigated Chapter 7 of TEC
using advanced knowledge of earthquake engineering. Therefore, (2007), entitled “Assessment and Strengthening of Existing Build-
performance-based design procedures have been recently investi- ings,” which sets standards for the performance assessment and
gated for the structures. Several procedures for performance assess- rehabilitation of existing buildings. Kalkan and Kunnath (2007)
ment have been discussed in the literature. The most common evaluated the seismic deformation demands of multistory steel and
assessment procedures are explained in four main guidelines/codes: concrete moment frames using nonlinear procedures based on the
Applied Technology Council (ATC) 40 (ATC 1996), Federal Emer- spread hinge assumption. Inel et al. (2008) evaluated the seismic
gency Management Agency (FEMA) 356 (FEMA 2000), FEMA performance of the most common reinforced-concrete building
440 (FEMA 2005), and TEC (2007). The TEC (2007) code came stock in Turkey considering nonlinear behavior of the components.
into use in 2007. The nonlinear seismic performance of structures Duan and Hueste (2012) investigated the seismic performance of
under earthquake effects is determined by static pushover and time- a multistory reinforced-concrete frame building that was designed
history analyses. Pushover analysis allows for direct evaluation of according to the provisions of the Chinese seismic code.
the performance of the structure at each limit state (Tehranizadeh Bayraktar (2012) studied the performance of existing reinforced-
and Moshref 2011). Nonlinear dynamic analysis (NDA) is the concrete shear-wall buildings determined by the nonlinear analysis
methods defined in the Turkish Seismic Code.
1
Associate Professor, Dept. of Civil Engineering, Gümüşhane Univ., Inel and Meral (2016) evaluated the seismic performance of
Gümüşhane 29100, Turkey (corresponding author). E-mail: ozlem_cavdar@ existing low-rise and midrise reinforced-concrete buildings by com-
hotmail.com paring their displacement capacities and displacement demands
2
Associate Professor, Dept. of Civil Engineering, Gümüşhane Univ., under selected ground motions experienced in Turkey. Çavdar and
Gümüşhane 29100, Turkey. E-mail: ahmcavdar@hotmail.com Bayraktar (2014) investigated the nonlinear seismic behavior of a
3
Dept. of Civil Engineering, Gümüşhane Univ., Gümüşhane 29100,
collapsed reinforced-concrete residential building in the city of Van
Turkey. E-mail: badblues81@hotmail.com
Note. This manuscript was submitted on February 9, 2017; approved on in Turkey using static pushover and nonlinear time-history analyses.
July 11, 2017; published online on November 24, 2017. Discussion period In this study, nonlinear static pushover and nonlinear dynamic
open until April 24, 2018; separate discussions must be submitted for analyses were used to estimate the expected seismic performance of
individual papers. This paper is part of the Journal of Performance of a reinforced-concrete shear-wall building that collapsed during the
Constructed Facilities, © ASCE, ISSN 0887-3828. 2003 Bingöl earthquake in Turkey.
(2007). These are the minimum damage limit (MN), safety limit for each member. Beams and columns were modeled as frame
(GV), and collapsing limit (GC). The term MN defines the begin- elements connected to each other at the joints.
ning of the behavior beyond elasticity, GV defines the limit of the A typical floor plan is shown in Figs. 3 and 4. Because the
behavior beyond elasticity at which the section is capable of safely majority of buildings in Bingöl, Turkey, were constructed accord-
ensuring strength, and GC defines the limit of the behavior before ing to TEC (1975), the selected building was designed according
collapsing. This classification does not apply to elements dam- to this code, too. Because all the static projects are available, the
aged in a brittle condition. Elements in which the damage to critical reinforced-concrete properties of structural members are assumed
sections does not reach MN are within the minimum damage re- to be known completely.
gion, those in between MN and GV are within the marked damage The pushover and nonlinear dynamic analyses were performed
region, those in between GV and GC are in the advanced damage
by using the finite-element method in Structural Analysis Program
region, and those going beyond GC are within the collapsing region
2000 (SAP2000) (Wilson and Habibullah 1997). Beam and column
[Fig. 1(b)].
elements were modeled as nonlinear frame elements with lumped
Force
C
CP
LS 20 cm
B IO
(a)
3 cm
Ø8/15
D E
A Deformation
3Ø16
(a)
GV GÇ
Internal
50 cm
Force MN
5Ø14
Minimum Marked Advanced Collapsing
Damage Damage Damage Region
Region Region Region
Deformation 30 cm
(b) (b)
Fig. 1. Force-deformation relationship of a typical plastic hinge: Fig. 2. Geometry and reinforcement details of typical column and
(a) ATC 40, FEMA 273; (b) TEC (2007) beam: (a) 20 × 40 cm column; (b) 30 × 50 cm beam
© ASCE
04017122-3
Fig. 3. Typical basement plan
© ASCE
04017122-4
Fig. 4. Typical floor plan
Fig. 7. Plastic hinges occurring through the X-direction and Y-direction of the building for design earthquake after pushover analysis: (a) front view;
(b) side view
in Turkey. The Bingöl, Adana-Ceyhan, and Van earthquakes had It is seen from Figs. 12–14 that plastic hinges occurred through
destructive magnitudes 6.1, 6.2, and 7.1 Mw, respectively. The 2011 the X-direction and Y-direction as a result of NDA. It can be seen
Van earthquake was the largest natural disaster in Turkey after the from Fig. 12 that these hinges were concentrated on the upper
Kocaeli and Duzce earthquakes in 1999. For the Van earthquake, the floors.
official death toll was more than 600, with thousands of people
injured and thousands left homeless. The acceleration record from
Comparison of Performance Analysis Methods Used
Muradiye station, the nearest station, was used in the nonlinear
analysis. The performances of the upper-story elements under the design
The horizontal component [North–South (N–S)] of the acceler- earthquake are compared for static pushover analysis and NDA
ation time history and spectral acceleration records used in the in Figs. 15 and 16. The section damage regions determined with
analyses is shown in Figs. 9 and 10. Spectra for the acceleration the nonlinear static analysis methods defined in TEC (2007) show
record and TEC (2007) are shown in Fig. 11. similarities to each other. Generally, the result variations of the sec-
According to TEC (2007), in linear and nonlinear analysis, in tions differ by as much as three (minimum, marked, advanced)
the case of using three ground motions the maximum results (base damage regions. However, the results obtained from the nonlinear
shear forces) shall be used for the design of a reinforced-concrete dynamic method differ from those of the pushover analysis. It can
shear-wall building.
The aim of the nonlinear calculation within the scope of the time
definition method was to integrate the movement equation of the
(a)
(a)
(b)
(b)
(c)
(c)
Fig. 9. Acceleration time histories of (a) Bingöl earthquake, 2003;
(b) Adana-Ceyhan earthquake, 1998; (c) Van earthquake, 2011 (data Fig. 10. Spectral accelerations of (a) Bingöl earthquake, 2003;
from AFAD 2017) (b) Adana-Ceyhan earthquake, 1998; (c) Van earthquake, 2011
Fig. 11. Spectra for the acceleration record of the earthquakes and TEC (2007)
Table 3. Determination of Earthquake Analysis for Reinforced-Concrete be seen that NDA (GC level) yields more dangerous results than
Shear-Wall Building pushover analysis (MN level) by about three levels.
Weight of Base Ratio The results from pushover analysis show lower damage ratios
building shear (base shear/ for the upper-story beams than those of NDA. The targeted perfor-
Direction Earthquake (kN) (kN) weight) mance level for a reinforced-concrete shear-wall building deter-
X Adana 38,583 13,580 0.352
mined in TEC (2007) is the collapsing region (GC) level. In the
Bingöl 38,583 11,080 0.287 X-direction, 91.7% of the beams reached this level for NDA only.
Van 38,583 5,835 0.151 Only 46.9% of the beams reached the GC level for linear analysis
in the Y-direction. Thus, according to these results, the building
Y Adana 38,583 10,920 0.283 performance level was achieved in pushover analysis, whereas the
Bingöl 38,583 10,830 0.281
building collapsed according to NDA. This is a great controversy for
Van 38,583 5,580 0.145
the two methods. On the one hand, NDA takes into consideration
Fig. 12. Plastic hinges occurring through the X-direction and Y-direction of the building for Adana earthquake after dynamic analysis: (a) front view;
(b) side view
Fig. 13. Plastic hinges occurring through the X-direction and Y-direction of the building for Bingöl earthquake after dynamic analysis: (a) front view;
(b) side view
Fig. 14. Plastic hinges occurring through the X-direction and Y-direction of the building for Van earthquake after dynamic analysis: (a) side view;
(b) side view
repetitive and directional effects; on the other hand, pushover Torsion is greater in the upper stories than in the lower, as expected
analysis takes into consideration the effect at one direction and mon- in irregular structures because normal force levels absorbing tor-
otonic factors. The most important factor affecting it is torsional sional effects at the upper floors are less than those downstairs.
irregularity. One other detail increasing torsional action is the fact that the base-
Especially in the Y-direction, the torsional irregularity coeffi- ment floor shear walls were not distributed equally in two direc-
cient (ηbi ) is greater than 1.2 except at the basement, first floor, tions. If they were equally distributed, the torsional effect would
and second floor. This value is not a problem for pushover analysis. be decreased. Because the basement floor shear walls increase
However, there is torsional action according to TEC (2007). the large moment of inertia in the direction of calculation and
Fig. 18. Comparison of top displacements and base shear force for
Fig. 15. Comparison chart of methods for the X-direction for upper pushover and NDA for the Y-direction of the shear-wall reinforced-
story beams obtained by the design earthquake concrete building
Conclusions