7th National Congress on Civil Engineering, 7-8 May 2013

University of Sistan and Baluchestan, Zahedan, Iran

Progressive collapse in asymmetric

RC moment frame buildings

Negar Afsharhasani1, Abdolreza S.Moghadam2, Salar Manie3

1.M.SC, International Institute of Earthquake Engineering and Seismology, Tehran, Iran
2.Assistant Professor, International Institute of Earthquake Engineering and Seismology,
Tehran, Iran
3.Ph.D candidate of earthquake Engineering faculty member Islamic Azad University, Sanandaj
Branch, sanandaj, Iran

n.afshar@iiees.ac.ir

Abstract
The present paper deals with nonlinear assessment of progressive collapse in low-rise threedimensional (3D) reinforced concrete (RC) frame buildings with mass irregularities in plan. To
this end, 3 story 3D frame buildings, with 0%(symmetrical), 10% and 20% mass eccentricity in
plan and designed based on the provisions of the Iranian seismic code of practice (Standard
No.2800), were modeled using strength and stiffness degrading material models adopted in
FEMA P-695. Then incremental dynamic analyses (IDA) were performed to quantify the frames
performances from linear elastic through the overall collapse level of response along with
assessing the IDA curve of each record. The results show that response of asymmetrical frames
are quite different from the symmetrical ones. Also, it is observed that irregularities in plan may
cause a very irregular pattern of progressive yielding in critical sections of all asymmetric
structures, especially in frames located on the flexible side of the plan. Such a poor response is
shown to be intensified with increasing eccentricity.
Keywords: nonlinear analysis, Progressive collapse, mass irregularity, reinforced concrete,
performance assessment

1.

INTRODUCTION

The definition of progressive collapse has evolved over time and in different codes but essentially it is
a phenomenon in which an initial local failure spread from element to element and eventually results
in the collapse of the whole structure or to an extent disproportionate to the original failure. Some
researchers also distinguish between the terms of progressive collapse and disproportionate collapse
[1].
The cause of progressive collapse phenomena can be due to human-made hazards (blast or explosion,
vehicle impact, fire, etc.) or natural hazards such as earthquakes. Earthquake loading can generate
strong lateral forces and stress reversals. These load effects can overload structural members which
result in the loss of one or more load-carrying members, which may then lead to failure of additional
structural members in other parts of the system and the unzipping effect of progressive collapse of the

7th National Congress on Civil Engineering, 7-8 May 2013

University of Sistan and Baluchestan, Zahedan, Iran

entire system. Observations of earthquake damage in past earthquakes show that seismic loads can
cause structural damage that results in loss of support in the structures [2,3]. The initial failure of
individual structural elements or components itself can propagate to other adjacent load resisting
member in variety of ways [3].
On the other hand, one of the main factors that affect building performance during earthquake
excitation is building asymmetry. Asymmetric buildings often sustain more extensive damage in
compare with symmetric ones. In this study asymmetry in models is produced by mass irregularities in
plan. For modeling structures with mass irregularities in plan it would be better to define nodal
masses. Asymmetric models include two types of distribution: a) 10% mass eccentricity b) 20% mass
eccentricity. This study deal with effect of seismic loads on progressive collapse of asymmetric
buildings.

2.

SIMULATION OF STRUCTURAL COLLAPSE

The results of nonlinear analysis are dependent on defining inelastic behavior. Various methods for
considering non-linear behavior of structures has led to different approaches to modeling. One method
of modeling the non-linear behavior of structures is using lumped plastic hinges, In addition to
simplicity, this modeling approach has no limitation of other methods of modeling, such as restriction
in buckling of reinforcement as in fiber element modeling. The hinges are defined at the end of beams
and columns of flexural reinforced concrete frame structures.

Figure1. Concentrated plasticity mode

lumped plastic hinges are hinges with zero length, parameters needed to define the specifications of
backbone of these hinged are: Effective stiffness, Flexural strength, Plastic rotation capacity, Postcapping rotation capacity, Post-yeild hardening stiffness, Cyclic Energy Dissipation. This backbone
curve represents member behavior from elastic mode up to fracture. (Figure1)
The purpose of this study is the impact of increased lateral load on developing of critical areas causing
the instability of the structure. The response of each structure is simulated using the Opensees
Software (PEER 2006) with the following modeling features (Figure 2-a) for beam column elements
with concentrated plastic rotational hinges at each end. Lumped plasticity phenomenological models
are used in order to capture softening post-peak response, which is difficult to simulate using fiber-

7th National Congress on Civil Engineering, 7-8 May 2013

University of Sistan and Baluchestan, Zahedan, Iran

element type models. The seismic resisting lateral system is represented by a three-dimensional, threebay frame model. The plastic hinge models for the beam-columns are described by the monotonic
trilinear backbone curve developed by Ibarra et al (2005)[4].

Figure2. Schematic view of analytical model for frame structures, showing: (a) generalized threedimensional model configuration and (b) nonlinear material feature of beam-column hinges.

This backbone and associated hysteretic rules provide for versatile modeling of cyclic behavior
(Figure 2-b). An important aspect of this model is the negative stiffness associated with the post-peak
response, which enables modeling of strain softening behavior associated with phenomena such as
concrete crushing and rebar buckling and fracture. The model also capture four basic modes of cyclic
deterioration: strength deterioration of the inelastic strain hardening branch, strength deterioration of
the post-peak strain softening branch, accelerated reloading stiffness deterioration, and unloading
stiffness deterioration. The element hinge model requires specification of seven parameter to control
both the monotonic and cyclic behavior of the element:
and two deterioration
parameters ( and C).
The calibration of these parameters for reinforced concrete column is part of a separate study
(Haselton et. al 2006) of more than 250 columns from the PEER Structural Performance
Database(PEER 2005)[5]. Simulation of global sideway collapse is based on the incremental Dynamic
Analysis(IDA) technique(Vamvatsikos and Cornell, 2003). The basic process of this technique is as
follows: (a) a strong ground motion is selected and scaled to a specified ground motion intensity, such
as spectral acceleration at the first mode period, (b) the peak response quantity, such as peak interstory
drift ratio(IDR), is recorded and associated with the specified earthquake intensity, (C) steps (a)-(b)
are repeated by scaling up the intensity of the input ground motion until the building becomes
dynamically unstable and a collapse occurs, and finally (d) steps (a)-(c) are repeated for additional
ground motion. In this study, the IDA analyses are based on a set of 8 pairs of recorded ground
motions, each with two orthogonal components[5].

7th National Congress on Civil Engineering, 7-8 May 2013

University of Sistan and Baluchestan, Zahedan, Iran

3.

OVERVIEW OF COLLAPSE ASSESSMENT PROCEDURE

A progressive collapse analysis is needed to determine the capability of a structure to resist abnormal
loadings. When a structure undergoes progressive collapse, the response of the structure is affected by
dynamic effects. This requires the dynamic behavior of a structure to be taken into account in the
progressive collapse analysis. It is also expected that nonlinear structural behavior can significantly
affect the progressive collapse behavior of a structure since before reaching the collapse condition a
structure and its member components must have exceed its elastic limits. Considering these two
observations, it can be concluded that the nonlinear static analysis and nonlinear dynamic analysis are
the two most appropriate methods for evaluation of progressive collapse behavior of structures among
the available analysis methodologies [3]. In this study assessing the initial failure is based on
structural performance level (collapse prevention) defined by ASCE41/06 (Figure 3) and the global
sideway collapse capacity is based on the incremental dynamic analysis (IDA) technique.

Figure 3. Qualitative performance levels of ASCE41-06[7]

Structural Performance Level S-5, Collapse Prevention, shall be defined as the post-earthquake
damage state that includes damage to structural components such that the structure continues to
support gravity loads but retains no margin against collapse [7], recommended drift for this structural
performance is 0.04.
In IDA, the structural model, which capture both material and geometric nonlinearities, is analyzed for
specific ground motion record. This time-history analysis is for specific ground motion record. This
time-history analysis is repeated, each time increasing the scale factor on the input ground motion,
until the record causes structural collapse, as identified by runaway interstory drift displacement [6].
This study consider the 0.04 drift for initial structural damage and obtain ultimate collapse capacity
from IDA curve then assess the effect of mass irregularities on progressive collapse.

4.

INCREMENTAL DYNAMIC ANALYSIS

The incremental Dynamic Analysis concept has been developed by the analogy of passing from single
static analysis to the incremental static pushover, where the seismic loading is scaled. The concept has
been mentioned as early as 1977 by Bertero, and has been cast in several forms in the work of many
researchers. It has been adopted by the U.S. federal Emergency Management Agency(FEMA)
guideline as the Incremental Dynamic Analysis (IDA) and established as the state-of-the art method to
determine the global collapse capacity[8]. Table-1 shows Summary of Parameters of recorded Ground
Motions for the Far-Field record set.

7th National Congress on Civil Engineering, 7-8 May 2013

University of Sistan and Baluchestan, Zahedan, Iran

Table 1- Summary of PEER-NGA Database information and Parameters of recorded Ground Motions for
the Far-Field record set

1
2
3
4
5
6
7
8

Record NO

Year

Recording Station

Epicentere

PGAmax
(g)

PGVmax
(cm/s)

829
752
68
125
169
721
725
900

1992
1989
1971
1976
1979
1987
1987
1992

7
6.9
6.6
6.5
6.5
6.5
6.5
7.3

Rio Dell
Capitola
Overpass
LA Tolmezzo
Hollywood
Delta
Stor
El Centro
Imp.Road
Co
Poe
Yermo
Fire
(temp)
Station

22.7
9.8
39.5
20.2
33.7
35.8
11.2
86

0.55
0.53
0.21
0.35
0.35
0.36
0.45
0.24

44
35
19
31
33
46
36
52

5. INCREMENTAL DYNAMIC ANALYSIS RESULTS

Incremental Dynamic Analysis is repeated, each time increasing the scale factor on the input
ground motion, until the record causes structural collapse, as identified by runaway interstory
drift displacement. Figure 4 contains the results of incremental dynamic analysis response
plot of spectral acceleration versus maximum interstory drift at mass center for
0%(symmetrical), 10% and 20% mass eccentricity in plan.

7th National Congress on Civil Engineering, 7-8 May 2013

University of Sistan and Baluchestan, Zahedan, Iran
Figure 4-Incremental dynamic analysis response plot of spectral acceleration versus maximum
interstory drift at mass center a)0% mass eccentricity b)10%mass eccentricity c)20%mass
eccentricity

For assessing the effect of mass irregularities in plan Figure 5 contains the results of incremental
dynamic analysis response plot of spectral acceleration versus maximum interstory drift at stiff edge
for 0%(symmetrical), 10% and 20% mass eccentricity in plan .

Figure 5-Incremental dynamic analysis response plot of spectral acceleration versus maximum interstory
drift at stiff edge: a) 0%mass eccentricity b)10%mass eccentricity c) 20%mass eccentricity

Therewith Figure 6 contains the results of incremental dynamic analysis response plot of spectral
acceleration versus maximum interstory drift at Flexible edge for 0%(symmetrical), 10% and 20%
mass eccentricity in plan. The result for 20% eccentricity is not complete because of absurd output of
incremental dynamic analysis.
This study consider the 0.04 drift for initial structural damage and obtain ultimate interstory drift of
collapse capacity from IDA curve for each record then calculated the ratio between maximum inter
story drift and 0.04 then calculated average (Figure 7) of these values between selected records at
mass center, stiff edge and flexible edge. Damage index in this study is ratio of maximum interstory
drift to whatsoever this ratio increases the safety margin against progressive collapse decrease.

7th National Congress on Civil Engineering, 7-8 May 2013

University of Sistan and Baluchestan, Zahedan, Iran

Figure 6- Incremental dynamic analysis response plot of spectral acceleration versus maximum interstory
drift at Flexible edge : a)0% mass eccentricity b)10%mass eccentricity c)20%mass eccentricity

Figure 7- Average value of (max interstory drift/0.04)

7th National Congress on Civil Engineering, 7-8 May 2013

University of Sistan and Baluchestan, Zahedan, Iran

6.

Concluding Remarks

The results obtained for low-rise three-dimensional (3D) reinforced concrete (RC) building frames
with mass irregularities in plan (0% -symmetrical ,10% and 20% mass eccentricity )show that:
Dynamic response (maximum interstory drift) of 3D asymmetrical structures associated with the
supposed location of response(e.g. mass center, stiff edge, flexible edge,)
Dynamic response (maximum interstory drift)of 0% mass eccentricity is constant at mass center, stiff
edge and flexible edge as expected.
Safety margin against progressive collapse is sensitive to mass eccentricity.
Damage index of 10% mass eccentricity in comparison with 0% mass eccentricity increases about 37
percent in mass center , about 32 percent in flexible edge and about 8 percent in stiff edge.
Damage index of 20% mass eccentricity in comparison with 0%mass eccentricity increases about 19
percent in mass center and increases about 37 percent in stiff edge.

7.

References

1- Nair, R. S., Preventing Disproportionate Collapse. Journal of Performance of Constructed Facilities

ASCE, 20(4), (2006), pp. 309-314.
2- Moehle, J. P., Elwood, K. j., and Sezen, H.,Gravity Load Collapse of Building Frames during
Earthquakes. Proceedings of S. M. Uzumeri Symposium: Behavior and Design of Concrete Structures
for Seismic Performance, ACI Special Publication (SP-197),(2002), pp. 215-238
3-Wibowo, H. and Lau, D.T., Seismic Progressive Collapse: Qualitative Point of View, Civil
Engineering Dimension, Vol. 11,No 1,March (2009), 8-14
4-OpenSees. Open System For Earthquake Engineering Simulation , Pacific Earthquake
Engineering Research Center ,http://peer.berkeley.edu//(2007)
5-AbbieB.Liel, Curt B. Haselton, Gregory G.Deierlein, The Effectiveness of seismic building code
provisions on reducing collapse risk of reinforced concrete moment frame buildings, 4th International
Conference on Earthquake Engineering Taipei, Taiwan , October 12-13, (2006)
6-AbbieB.Liel, Curt B. Haselton, Gregory G.Deierlein, Jack W.Baker, Incorporating Modeling
uncertainties in the assessment of seismic collapse risk of buildings, Structural Safety 31(2009)
Science Direct.
7-ASCE 41 Seismic Rehabilotation of Existing Buildins ASCE/SEI41/06. ASCE Standard No.
ASCE/SEI41-06(2007)
8-Alexandra Tolis, influence of modeling on the seismic Performance Assessment an Existing Three
Story RC building, National University of Athens,(2009)