THE STRUCTURAL DESIGN OF TALL AND SPECIAL BUILDINGS

Struct. Design Tall Spec. Build. 22, 74–91 (2013)

Published online 3 December 2010 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/tal.658

Seismic response of R/C structures subjected to simulated ground

motions compatible with design spectrum

Dae-Han Jun*,†
Department of Civil and Architectural Engineering, Dongseo University, Busan, South Korea

SUMMARY
In seismic response analysis of building structures, the input ground motions have considerable effect on
the nonlinear seismic response characteristics of structures. The characteristics of soil and the locality of
the site where those ground motions were recorded affect the contents of ground motion time histories.
This study describes a generation of synthetic ground motion time histories compatible with seismic design
spectrum, and also evaluates the seismic response results of multi-story reinforced concrete structures by
the simulated ground motions. The simulated ground motion time histories have identical phase angles to
the recorded ground motions, and their overall response spectra are compatible with seismic design spec-
trum with 5% critical viscous damping. The input ground motions applied to this study have identical
elastic acceleration response spectra, but have different phase angles. The purpose of this study was to
investigate their validity as input ground motion for nonlinear seismic response analysis of building struc-
tures. As expected, the response quantities by simulated ground motions presented better stability than
those by real recorded ground motions. It was concluded that the simulated earthquake waves generated
in this paper are applicable as input ground motions for a seismic response analysis of building structures.
Copyright © 2010 John Wiley & Sons, Ltd.

1. INTRODUCTION

1.1. Background
An earthquake acceleration wave is only to represent the time histories of free field shaking of a
specific site caused by an earthquake event. In other words, any one input motion adopted in the
seismic response history analysis of building structures is nothing but a ground motion on a specific
free field caused by an earthquake event (Architectural Institute of Japan (AIJ), 1992). A single earth-
quake event can generate various ground motion time histories with different characteristics. There-
fore, any one ground motion does not necessarily represent typical time histories to guarantee the
seismic safety of building structures (AIJ, 2004). It is impossible to predict ground motion character-
istics that may occur in the future at a construction site because the property of the ground motion is
interrelated with many factors such as fault mechanism, seismic wave propagation from source to site
and the amplification characteristics of ground (Study Group on Theoretical Earthquake Waves, 1994).
The important factors of ground motions affecting structure’s response results are peak ground accel-
eration, frequency contents, duration of ground motion and shapes of waveform. Though required to
set input ground motions for general seismic design including these factors, it is not available at this
time (Stewart et al., 2001). Also, the input ground motions for seismic design need to correspond
appropriately to various structural materials and systems. The seismic design guidelines provide an
acceleration response spectrum for estimating the design seismic force of a structure. Accordingly,
the input ground motion applied to the dynamic response analysis of structures would be appropriate
for the ground motion history which is highly related with design seismic force. For this purpose,
new techniques have been studied by many seismic engineers to produce artificial ground motions

* Correspondence to: Dae-Han Jun, Department of Civil and Architectural Engineering, Dongseo University, San 69,
Jurye-2dong, Sasang-gu, Busan, South Korea
†
E-mail: jdh@dongseo.ac.kr

Copyright © 2010 John Wiley & Sons, Ltd.

 SEISMIC RESPONSE OF R/C STRUCTURES SUBJECTED TO SIMULATED GROUND MOTIONS 75

(Tsai, 1972; Preumont, 1984; Barenberg, 1989; Jun and Inoue, 1991a), and general software has been
used to simulate the artificial ground motions (Vanmarke and Gasparini, 1976). However, the results
of these studies seem to be inefficient in generating various artificial ground motions reflecting dif-
ferent site conditions and ground motion factors as they require much time and effort.
Meanwhile, the input ground motions can be expressed as design response spectrum. One of the
advantages of describing ground motions as a response spectrum is that it can simultaneously deter-
mine the seismic force and lateral displacement of structures. In general, the only response spectrum
corresponding to any one time history wave can be calculated. On the contrary, a specified design
response spectrum can not determine the time history wave corresponding to it. Design response
spectrum is prepared by statistical data using many time history waves, and it does not represent the
response spectrum of a specific time history wave.
The manhattanization of buildings and the irregularities of a plan and an elevation make the
dynamic behaviour of structures subject to ground motions complicated, and the evaluation of the
structures’ seismic performance requires dynamic time history response analysis. For time history
analysis, the selection of an appropriate input ground motion is important and the input motions should
be selected from real records including magnitude, distance of epicentre, site conditions and other
factors that control the ground motion properties. The duration of input ground motion, especially for
high-rise buildings, has also been recognized as an important factor to increase the input energy
required for long-period buildings (Moehle, 2006).
A difference of input ground motions causes considerable divergence in the analysis of time history
responses; therefore, appropriate scaling of the input ground motions is necessary in seismic response
history analysis. There are two scaling methods: one using peak value of ground accelerations as
baseline, the other using ground motion consistent with design spectrum. There is currently no con-
sensus on which approach, scaling or spectrum matching, is preferable for nonlinear seismic response
analysis (Kurama and Farrow, 2003; Aschheim et al., 2007; Han et al., 2007; Hancock et al., 2008).
Ground motion scaling procedures using the peak value of input ground accelerations maintains
original ground motion history characteristics including the response spectrum of each recorded
ground motion. It is, however, recommended to use not less than seven input ground motion records
to prevent the response values of structure from being biased by response spectrum characteristics of
any one ground motion. Especially when applied for nonlinear seismic response analysis, the scale
factor of 2 or below is required for amplification of the ground motion component because the non-
linear response appears a biased one-directional response characteristics dominated by the peak value
of the ground accelerations. In addition, it requires relatively large factor to compensate insufficient
energy for long period structure as it is difficult to obtain ground motion with sufficient input energy
in the long period range for high-rise buildings. In this case, a shortcoming of higher mode magnifica-
tion effect can occur as a result of unusual scaling up of relatively short-period spectrum (PEER
Ground Motion Selection and Modification Working Group, 2009).
The scaling method of input ground motion using design spectrum can be accomplished to perform
response history analysis with less ground motions, but it is a question how to effect on nonlinear
seismic response because the design spectrum is based on an elastic spectrum. In particular, a question
is indicated on the input energy magnification of the artificial motions corresponding to the response
spectrum (Naeim and Lew, 1995).

1.2. Purpose
This study proposes a new process to easily simulate the ground motions having response spectrum
characteristics close to standard design response spectrum shapes and also identical phase angles with
the ground motions recorded in the past earthquakes. In particular, this study also attempts to create
a large number of artificial time histories having design response spectra including various phase
angles characteristics of recorded ground motions. Nonlinear response analysis of single degree of
freedom system was performed by the simulated ground motions to examine the feasibility of it as
the input ground motion.
In seismic response analysis of multi-story buildings, the selection of input ground motion and
adjustment of input intensity level of the selected ground motion are most important for estimating

Copyright © 2010 John Wiley & Sons, Ltd. Struct. Design Tall Spec. Build. 22, 74–91 (2013)
DOI: 10.1002/tal
76 D.-H. JUN

the response results of the structure. The purpose of this study is to evaluate nonlinear response char-
acteristics of real buildings through nonlinear time history analysis on multi-story reinforced concrete
structures by inputting simulated seismic waves identical as response spectra, which was focused on
design response spectrum as scaling method of the input ground motions. This study also is to evalu-
ate its feasibility as input ground motions for the nonlinear time history analysis of actual buildings
by identifying relationships between design response spectra and nonlinear seismic response results
of the input ground motions.

2. GENERATION OF ARTIFICIAL GROUND MOTION

2.1. How to develop artificial ground motion

The simulation approach used in this study was to manipulate acceleration amplitudes to make accel-
eration response spectrum of the ground motion, which is closely related to the response of structures
match design response spectrum while maintaining same phase angles characteristics of the ground
motions recorded in the past earthquakes.
Recorded ground acceleration time histories (üg(t)) from the past earthquake can be expressed by
discrete and finite Fourier approximate formula as follows.
n −1
ug ( t ) = ∑ k =0 ak cos (ω k + φk )
 (1)

where, ak: kth Fourier amplitude

ωk: kth Fourier circular frequency
φk: kth Fourier phase angles
From the parameters obtained in Equation (1), artificial ground motion histories can be generated
by Fourier inverse transform, keeping phase angles(φk) and after adjusting Fourier amplitude (ak)
between periods T = 0.02~10.0 s to synchronize with set target for design acceleration response
spectrum. Fourier amplitude can be adjusted as shown in Equation (2) below.

SAT
a′k = ak (2)
SA

where, a′k: adjusted Fourier amplitude

ak: Fourier amplitude of the original recorded ground motion
SAT: target elastic acceleration response spectrum
SA: elastic acceleration response spectrum of real recorded ground motion or of adjusted
ground motions
In Equation (2), conducting Fourier inverse transform using the adjusted amplitude generates modi-
fied artificial ground motion acceleration. After seismic response analysis utilizing this adjusted
motion to recalculate acceleration response spectrum, the suitability of the design acceleration response
spectrum can be verified. This calculation process is continued until the response spectrum of final
motion is close enough to target response spectrum. Figure 1 represents a flow chart in simulating
artificial ground motion time histories.

2.2. Setting convergence criteria of suitability

This study was to generate the artificial ground motions close to design acceleration response spec-
trum. The criteria to determine the suitability for target response spectrum and response spectrum of
simulated ground motions can be based on the following three conditions (Jun and Inoue, 1991b).

2.2.1. Minimum response spectrum ratio (εi)

The ratio (εi) of the target acceleration response spectrum to the acceleration response spectrum of
single degree of freedom system having damping ratio of 5% for each period is calculated, and its
minimum value is chosen to satisfy the following Equation (3).

Copyright © 2010 John Wiley & Sons, Ltd. Struct. Design Tall Spec. Build. 22, 74–91 (2013)
DOI: 10.1002/tal
 SEISMIC RESPONSE OF R/C STRUCTURES SUBJECTED TO SIMULATED GROUND MOTIONS 77

Start

Setting range of periods and selecting

target design spectrum

Reading recorded motion data

Fourier transform
-Amplitude ak and phase angles φ k

The process of reference

Jun and Inoue, 1991b

The process of this study

Modify Fourier amplitude ak⇒ak’

Fourier inverse transform:

⇒ acceleration data

Calculate SDOF response spectra

No Estimate No
the goodness of fitting

Yes
End

Figure 1. Flow chart for generation of artificial ground motions.

s A (Ti , 0.05)
ε min = ≥ 0.85 (3)
S AT (TI , 0.05) min

where, SA(Ti, 0.05): acceleration response spectrum at period Ti with damping ratio 5%
SAT(Ti, 0.05): target acceleration response spectrum at period Ti with damping ratio 5%

2.2.2. Coefficient of variation for target acceleration response spectra

The ratio of coefficient of variations for acceleration response spectrum to the target acceleration
response spectrum of simulated ground motions should satisfy Equation (4) below.

∑
n
(ε i − 1.0 )2
v= i =1
≤ 0.05 (4)
N

where,

S A (Ti , 0.05)
εi = : the ratio of spectra (5)
S AT (Ti , 0.05)

N: total number of periodic components to determine errors

2.2.3. Deviation of average

Deviation of average of the ratio of spectra by periodic components should be set to satisfy the fol-
lowing Equation (6).

1 − ε min ≤ 0.02 (6)

Copyright © 2010 John Wiley & Sons, Ltd. Struct. Design Tall Spec. Build. 22, 74–91 (2013)
DOI: 10.1002/tal
78 D.-H. JUN

Table 1. Comparison of maximum acceleration values for recorded and simulated ground motions (unit:
cm/sec2, sec).

Recorded ground motion Simulated ground motion

Type of ground
motion wave Max Time SF1* SF2** Max Time SF1* SF2**
JMA Kobe 1995 NS 819.1 4.94 0.75 1.16 415.3 5.54 2.06 1.55
JMA Kobe 1995 EW 617.1 8.46 0.99 1.65 401.1 8.47 2.09 2.05
Taft 1952 NS 152.7 9.10 5.73 4.30 369.3 6.62 2.03 1.32
Taft 1952 EW 175.9 3.70 6.73 6.00 426.3 3.71 1.98 2.08
EL Centro 1940 NS 341.7 2.12 2.21 2.48 324.0 2.08 1.96 1.92
EL Centro 1940 EW 210.1 11.44 3.41 2.48 321.6 2.03 2.09 1.50
Mexico city 1985 NS 98.0 24.16 6.00 1.10 353.4 39.86 1.94 1.70
Mexico city 1985 EW 167.9 28.08 3.98 0.79 336.6 33.38 2.05 1.65
SF1, Scaling Factor for elastic response analysis; SF2, Scaling Factor for nonlinear response analysis.

where,

∑
N
ε
i =1 i
ε ave = (7)
N

Determination of the suitability was evaluated with N = 250 over the period range Ti = 0.02 s through
Ti = 10.0 s for calculation of target elastic acceleration response spectrum.

2.3. Selection of recorded ground motions

The ground motions used in this study were selected from the real recordings of ground motions
greater than magnitude scale of six. The characteristics of earthquake such as fault mechanism, wave
propagation path and site characteristics were not considered, and it has been commonly used as input
ground motions for seismic design or representative ground motions, which caused severe damage to
buildings were selected. Table 1 lists these records with the recorded peak ground accelerations and
the simulated peak values.

2.4. Selection of design acceleration spectrum

The target acceleration response spectrum to develop simulated ground motions was set in reference
to seismic design regulations (AIJ, 2004) as follows.
(1) Ti < 0.16 s: SAT (Ti, 0.05) = 320 + 3000 Ti
(2) 0.16 ≤ Ti < 0.64 s: SAT (Ti, 0.05) = 800
(3) 0.64 s ≤ Ti: SAT (Ti, 0.05) = 512/Ti
This is represented in Figure 2. The design acceleration response spectrum is defined by the vibra-
tion period (Ti) for damping ratio h = 5%. The variation of the size of the acceleration response spectra
may result in different design acceleration response spectra at other site where the site conditions and
seismic activities are different.

3. CHARACTERISTICS OF SIMULATED GROUND MOTIONS

3.1. Comparison of simulated ground motions with recorded ground motions

Figures 3 and 4 show comparison between simulated ground motion with original recorded ground
motion. Lower figures compare also target design spectra and response spectra of simulated ground
motions. In the figure, it is apparent that original recorded ground motions and simulated
ground motions have similar trends as phase angle characteristics are same. In addition, it is notable

Copyright © 2010 John Wiley & Sons, Ltd. Struct. Design Tall Spec. Build. 22, 74–91 (2013)
DOI: 10.1002/tal
 SEISMIC RESPONSE OF R/C STRUCTURES SUBJECTED TO SIMULATED GROUND MOTIONS 79

Figure 2. Response spectra and design spectra of input ground motions.

Figure 3. El Centro 1940 NS component.

that response spectra of recorded ground motions are adjusted in the proximity of the design response
spectrum.
Table 1 presents comparison of maximum acceleration value and its occurrence time of recorded
ground motions and simulated ground motions that were developed for target response spectrum. The

Copyright © 2010 John Wiley & Sons, Ltd. Struct. Design Tall Spec. Build. 22, 74–91 (2013)
DOI: 10.1002/tal
80 D.-H. JUN

Figure 4. Mexico City 1985 EW component.

Copyright © 2010 John Wiley & Sons, Ltd. Struct. Design Tall Spec. Build. 22, 74–91 (2013)
DOI: 10.1002/tal
 SEISMIC RESPONSE OF R/C STRUCTURES SUBJECTED TO SIMULATED GROUND MOTIONS 81

maximum acceleration value of the simulated ground motions adequately developed for design
response spectrum is in the range of 322–426 cm/s2. The occurrence time of the maximum accelera-
tion values of recorded ground motions and simulated ground motions was observed to occur at the
almost same time except for El Centro 1940 EW component.

3.2. Nonlinear response characteristics of simulated ground motions

Seismic response characteristics using simulated ground motion wave and recorded ground motion
wave was examined by the nonlinear response analysis of a single degree of freedom system. In
general, yield strength coefficient (Cy) has been widely used as an indicator to represent nonlinear
response characteristics of input ground motions. Yield strength coefficient is a normalized factor
obtained from the yield strength divided by total weight of the structural system after calculating the
yield strength of dynamic model in consistent with ductility of the same ground motion intensity. That
is, yield strength coefficient (Cy) can be expressed as follows when the yield strength of structural
model in the single degree of freedom system is Qy, total weight of the structural system, WT:

Qy Qy
Cy = = (8)
WT mg

where, g: gravity acceleration

m: mass
In the single degree of freedom system, yield strength coefficient (Cy) is expressed as the ratio
of yield strength of structural system to gravity acceleration when the mass (m = 1) is normalized.
It is called constant ductility response spectrum obtaining consistent ductility to represent
damage index of the structure for given ground motion. The yield strength of a structural system
for a specified ductility factor can be obtained using the corresponding yield strength coefficient Cy
value.
Figures 5 and 6 represent comparison of original recorded ground motions and simulated input
ground motions with respect to yield strength coefficient (Cy) using bilinear hysteresis model (ratio
of initial stiffness and post-yielding stiffness a = 5%). In the figures, as ductility of 1 means elastic
state, the response of recorded ground motions produced large differences depending on the charac-
teristics of the input ground motions. On the contrary, the response of simulated ground motions
indicated almost identical response characteristic, which was due to the fact that simulated ground
motions was developed close to design response spectrum.

4. NONLINEAR RESPONSE ANALYSIS OF MULTI-STORY FRAME STRUCTURES

4.1. Analytical model

The analytical model of multi-story frame structures used in this study was a standard plane frame
model for reinforced concrete moment resistant system; it can be generally applied in seismic design
of building structure as shown in Figure 7. Modelling of the column member for nonlinear analysis
utilized fibre models. The sectional dimensions and reinforcement of column and beam used in the
structural model are presented in Table 2 and Figure 8, respectively. The characteristics of material
are: For steel bar, elastic stiffness Es = 196 GPa, yield strength fy = 392 MPa; for concrete, elastic
stiffness Ec = 23 GPa and design strength fck = 24 MPa. The material stress–strain relationship of
concrete and reinforcement used in the analysis are presented in Figures 9 and 10.

4.2. Analysis method

The nonlinear response analysis of ground motions was performed using CANNY-2004 software (Li,
2004). Nonlinear time history analysis was used for nonlinear dynamic analysis and Newmark β
method (β = 0.25, γ = 0.5) was used for numerical integration method. A Rayleigh damping was

Copyright © 2010 John Wiley & Sons, Ltd. Struct. Design Tall Spec. Build. 22, 74–91 (2013)
DOI: 10.1002/tal
82 D.-H. JUN

Figure 5. Mexico City 1985 EW component ground motion.

Table 2. Member cross-section and reinforcement.

Story Size B (cm) D (cm) Reinforcement

Column (C1, C2)
1–3 60 60 12—D22
4–6 50 60 12—D22
7–10 50 50 8—D22
Beam (G1)
All 35 60 6—D19 (upper bar), 4—D19 (lower bar)

applied in the nonlinear time history analysis and horizontal input ground motion was used to perform
nonlinear analysis.

4.3. Elastic seismic response analysis

In elastic seismic response analysis, scaling of input ground motion is simple as the maximum
response value is proportional to a prescribed ground shaking intensity level. Elastic response quanti-
ties of multi-story frame structures are determined by the dynamic property of the structural model

Copyright © 2010 John Wiley & Sons, Ltd. Struct. Design Tall Spec. Build. 22, 74–91 (2013)
DOI: 10.1002/tal
 SEISMIC RESPONSE OF R/C STRUCTURES SUBJECTED TO SIMULATED GROUND MOTIONS 83

Figure 6. El Centro 1940 NS component ground motion.

and the elastic response spectrum characteristics of the input ground motions. The purpose of this
research was to examine the variations of response results on each floor of the multi-degree of freedom
system inputting different ground motions for the same structural model.
Here, the scaling of the input ground motion was adjusted to obtain elastic response displacement
of 36.5 cm on the top floor (deformation angle H/100 radian, where H: total building height). Scaling
factors (SF1) for elastic seismic response analysis of each input ground motion are shown in Table
1. The scale factors are modified through trial and error such that the roof displacement calculated
from the scaled input ground motions accords with the target response displacement. It indicates that
the recorded ground motions have the substantial change of SF1 reflecting elastic response spectrum
characteristics and dynamic property of the dynamic model. However, it was previously expected that
the SF1 value would be constant because the response spectrum characteristics was scaled regularly
for the simulated ground motions. However, a slight variation was observed as the response spectrum
characteristics of each wave cannot be exactly consistent with the theoretical values of the simulated
ground motions.
The seismic response results of the buildings for height-wise distribution of story displacement,
interstory drift and story seismic force are shown in Figures 11–14 for simulated ground motions and
recorded ground motions, respectively. In Figure 11, the displacement distribution of lower floor was
similar because the lateral displacement of the top floor was the same. In the case of simulated ground
motions, the response values of story displacement on all stories showed less than 3% differences,

Copyright © 2010 John Wiley & Sons, Ltd. Struct. Design Tall Spec. Build. 22, 74–91 (2013)
DOI: 10.1002/tal
84 D.-H. JUN

Figure 7. Structural model.

B
Column

B
Beam
Figure 8. Member cross-sections.

whereas the recorded ground motions in the middle stories had approximately 15% differences
depending on the type of input ground motion. Figure 12 presents interstory deformation angle dis-
tributions with input ground motions. It was verified that the response values by the simulated ground
motions were less variable compared with those by the recorded ground motions.

Copyright © 2010 John Wiley & Sons, Ltd. Struct. Design Tall Spec. Build. 22, 74–91 (2013)
DOI: 10.1002/tal
 SEISMIC RESPONSE OF R/C STRUCTURES SUBJECTED TO SIMULATED GROUND MOTIONS 85

Figure 9. Stress–strain relationship of concrete.

Figure 10. Stress–strain relationship of steel bar.

Figures 13 and 14 represent the distribution of story seismic force and shear force on each story.
Each figure shows that the response variance of the simulated ground motion is apparently smaller
than that of the recorded ground motion, regardless of the input ground motions. It is also shown that
the response variation of story force is greater on upper and lower stories compared with the middle
stories depending on the type of the input ground motion.

4.4. Nonlinear seismic response analysis

Nonlinear seismic response is different from elastic seismic response and the response results are very
complicated because of characteristics of input ground motion, dynamic property of the structures
and the influence of hysteresis model for structural component. Particularly, hysteresis model for
structural member is a critical factor to determine response characteristics of multi-story frame struc-
tures. Therefore, the purpose of this analysis is to evaluate the nonlinear responses of each story of
structures by changing only the input ground motion characteristics for multi-story frame structures
that have same dynamic model and hysteresis characteristics. The nonlinear seismic response analysis
was evaluated for story displacement, interstory deformation, story seismic force and story shear force
just as the elastic seismic response analysis.

Copyright © 2010 John Wiley & Sons, Ltd. Struct. Design Tall Spec. Build. 22, 74–91 (2013)
DOI: 10.1002/tal
86 D.-H. JUN

Figure 11. Distribution of story displacement by elastic analysis.

Figure 12. Distribution of interstory deformation angle by elastic analysis.

Here, the intensity of the input ground motion was scaled to obtain the elastic response displace-
ment of 36.5 cm on the top floor (deformation angle H/100 radian, where H: total building height)
followed by evaluation of nonlinear response results of multi-story frame structures. Scaling factors
(SF2) for nonlinear response analysis of each input ground motion are shown in Table 1. In Table 1,

Copyright © 2010 John Wiley & Sons, Ltd. Struct. Design Tall Spec. Build. 22, 74–91 (2013)
DOI: 10.1002/tal
 SEISMIC RESPONSE OF R/C STRUCTURES SUBJECTED TO SIMULATED GROUND MOTIONS 87

Figure 13. Distribution of story seismic force by elastic analysis.

Figure 14. Distribution of story shear force by elastic analysis.

the recorded ground motions show a wide difference of scaling factors between 0.79 and 6.00 depend-
ing on the type of the ground motion. For simulated ground motions, the SF2 values were relatively
less variable with a coefficient range of 1.32–2.08 because the response spectrum was primarily scaled
constantly for simulated ground motions.

Copyright © 2010 John Wiley & Sons, Ltd. Struct. Design Tall Spec. Build. 22, 74–91 (2013)
DOI: 10.1002/tal
88 D.-H. JUN

The seismic responses of the buildings for distribution of story displacement, interstory drift and
story seismic force are shown in Figures 15–18 for simulated ground motions and recorded ground
motions, respectively. In Figure 15, the story displacement distribution of lower floor was similar as
the roof displacement was adjusted same. In the case of the simulated ground motion waves, the
response quantities of story displacement on all stories showed less than 5% differences, whereas the

Figure 15. Distribution of story displacement by nonlinear analysis.

Figure 16. Distribution of interstory deformation angle by nonlinear analysis.

Copyright © 2010 John Wiley & Sons, Ltd. Struct. Design Tall Spec. Build. 22, 74–91 (2013)
DOI: 10.1002/tal
 SEISMIC RESPONSE OF R/C STRUCTURES SUBJECTED TO SIMULATED GROUND MOTIONS 89

Figure 17. Distribution of story seismic force by nonlinear analysis.

Figure 18. Distribution of story shear force by nonlinear analysis.

recorded ground motion waves in the middle stories had approximately 20% differences depending
on the type of input ground motion. Both the simulated ground motion and the recorded ground motion
resulted in greater difference for nonlinear response compared with elastic response. Figure 16
presents interstory drift angle distributions according to different input ground motions. It was con-

Copyright © 2010 John Wiley & Sons, Ltd. Struct. Design Tall Spec. Build. 22, 74–91 (2013)
DOI: 10.1002/tal
90 D.-H. JUN

firmed that the response results by the simulated ground motions were less variable compared with
those by recorded ground motions. Recorded ground motions, in particular, indicated that the response
variations were greater in the middle stories. This could be due to the effect of the higher mode of
the recorded ground motions.
Figures 17 and 18 represent the distribution of story seismic force and story shear force on each
floor. Each figure shows that the response variances of the simulated ground motion are smaller than
those of the recorded ground motion, regardless of types of input ground motions. Apparently, several
waves of the recorded ground motions showed large response values compared with other ground
motions. As such, these ground motions are classified as inappropriate waves to be chosen as input
ground motions for seismic design as indicated in other research (Zhai and Xie, 2007). Likewise, the
response values of the recorded ground motions in middle stories show greater variance than those
of simulated ground motions. These results may be possible because the recorded ground motions
had great effect on response results of the short-period range corresponding to higher mode. In case
of the simulated ground motion, the decrease of the short period component and amplification of the
long period component resulted in relatively small effect on response results of short-period
structure.
When examined in the basis-only results of the analytical model used in this paper, it was found
from the presented figures that the response variances of simulated ground motions were smaller than
those of recorded ground motions. However, the input intensity of the simulated ground motions,
which is the scaling up of the response spectrum, was not always proportional to the response value
of each story, and the nonlinear response value was dependent on the property of the ground motions.
Furthermore, it needs to be examined how the response distribution of recorded ground motions in
the middle stories had greater variance. The small variance of the response distribution of the simu-
lated ground motions on all stories was thought to be caused by the decreased influence of the specific
period component included in the ground motions. In the future, more research needs to be performed
with various structural models and recorded ground motions. Also, the proposed response analysis
method in this study is necessary to verify if the simulated ground motion histories generated by this
study can be used as input ground motion for seismic design.

5. CONCLUSION

The seismic safety of building structures can be evaluated by the nonlinear behaviour of structures
induced by ground motions. The characteristics of input ground motions and dynamic property of the
structure are important factors to affect seismic response of structures. Ground motions used in the
seismic response analysis of structure include various characteristics depending on fault mechanism
of earthquake, wave propagation and amplification of soil type. As the result of the analysis, it is a
difficult task to quantitatively examine all affecting factors. In seismic design, design response spec-
trum generally represents the response characteristics of structural model. In this study, a simulated
ground motion suitable for design spectrum was developed and its feasibility for the input ground
motion was evaluated through nonlinear seismic response analysis of the multi-story frame structures.
The results obtained from this study can be summarized as follows.
(1) The response results of simulated ground motions by each floor presented better stability than
those by recorded ground motions.
(2) The simulated ground motions scaled to design elastic spectrum was confirmed to show less
differences in nonlinear responses by each floor.
(3) The simulated ground motions generated in this paper can be applied as the input ground
motions for a nonlinear response analysis of high-rise building structures.

ACKNOWLEDGEMENT

This work was supported by the Korea Science and Engineering Foundation (KOSEF) grant funded
by the Korea government (MEST) (No. 2009-0071519).

Copyright © 2010 John Wiley & Sons, Ltd. Struct. Design Tall Spec. Build. 22, 74–91 (2013)
DOI: 10.1002/tal
 SEISMIC RESPONSE OF R/C STRUCTURES SUBJECTED TO SIMULATED GROUND MOTIONS 91

REFERENCES
AIJ (Architectural Institute of Japan). 1992. Seismic Loading-Strong Motion Prediction and Building Response. Maruzen Corp.:
Tokyo (in Japanese).
AIJ (Architectural Institute of Japan). 2004. Guidelines for Performance Evaluation of Earthquake Resistant Reinforced Con-
crete Buildings (Draft). Maruzen Corp.: Tokyo (in Japanese).
Aschheim M, Tjhin T, Cormartin C, Hamburger R, Inel M. 2007. The scaled nonlinear dynamic procedure. Engineering
Structures 29: 1422–1441.
Barenberg ME. 1989. Inelastic response of a spectrum-compatible artificial accelerogram. Earthquake Spectra 3: 477–493.
Han SW, Kim ES, Hwang SM. 2007. Variability of seismic demands according to different sets of earthquake ground motions.
Structural Design of Tall and Special Buildings 16(3): 321–332.
Hancock J, Bommer J, Stafford PJ. 2008. Numbers of scaled and matched accelerograms required for inelastic dynamic
analyses. Earthquake Engineering and Structural Dynamics 37: 1585–1607.
Jun D-H, Inoue Y. 1991a. Inelastic response of single-degree-of-freedom systems subjected to artificial ground motions. Journal
of Structural Engineering 37B: 111–118 (in Japanese).
Jun D-H, Inoue Y. 1991b. Inelastic response of SDOF systems subjected to spectrum compatible artificial accelerograms. In
The 3rd EASEC, Shanghai, China, 23–26 April; 711–716.
Kurama YC, Farrow KT. 2003. Ground motion scaling methods for different site conditions and structure characteristics.
Earthquake Engineering and Structural Dynamics 32(15): 2423–2450.
Li K-N. 2004. CANNY: 3-Dimensional Nonlinear Static/dynamic Structural Analysis Computer Program—-User Manual.
CANNY Structural Analysis: Vancouver, British Columbia, Canada.
Moehle JP. 2006. Seismic analysis, design, and review for tall buildings. Structural Design of Tall and Special Buildings 15:
495–513.
Naeim F, Lew M. 1995. On the use of design spectrum compatible time histories. Earthquake Spectra 11(1): 111–127.
PEER Ground Motion Selection and Modification Working Group. 2009. Evaluation Ground Motion Selection and Modifica-
tion Methods: Predicting Median Interstory Drift Response of Buildings, Haselton, CB ed. PEER-2009/01, PEER, Univer-
sity of California: Berkeley, CA.
Preumont A. 1984. The generation of spectrum compatible accelerograms for the design of nuclear power plants. Earthquake
Engineering and Structural Dynamics 12: 481–497.
Stewart JP, Chiou S, Bray JD, Graves RW, Somerville PG, Abrahamson NA. 2001. Ground Motion Evaluation Procedures for
Performance-based Design. PEER-2001/09, PEER, University of California: Berkeley, CA.
Study Group on Theoretical Earthquake Waves. 1994. Seismic Motion-its Synthesis and Processing. Kajima Publication: Tokyo
(in Japanese).
Tsai NC. 1972. Spectrum compatible motions for design purposes. Journal of Engineering Mechanics ASCE 98(EM2): 345–
356.
Vanmarke EH, Gasparini DA. 1976. A Program for Artificial Motion Generation, User’s Manual and Documentation. Depart-
ment of Civil Engineering, Massachusetts Institute of Technology: Cambridge, MA.
Zhai C-H, Xie L-L. 2007. A new approach of selecting real input ground motions for seismic design: the most unfavorable
real seismic design ground motions. Earthquake Engineering and Structural Dynamics February 36: 1009–1027.

Copyright © 2010 John Wiley & Sons, Ltd. Struct. Design Tall Spec. Build. 22, 74–91 (2013)
DOI: 10.1002/tal