Nuclear Engineering and Technology 52 (2020) 192e205

Contents lists available at ScienceDirect

Nuclear Engineering and Technology

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

Original Article

Identifying significant earthquake intensity measures for evaluating

seismic damage and fragility of nuclear power plant structures
Duy-Duan Nguyen a, b, Bidhek Thusa a, Tong-Seok Han c, Tae-Hyung Lee a, *
a
Department of Civil and Environmental Engineering, Konkuk University, Seoul 05029, South Korea
b
Department of Civil Engineering, Vinh University, Vinh 461010, Vietnam
c
Department of Civil and Environmental Engineering, Yonsei University, Seoul 03722, South Korea

a r t i c l e i n f o a b s t r a c t

Article history: Seismic design practices and seismic response analyses of civil structures and nuclear power plants
Received 15 March 2019 (NPPs) have conventionally used the peak ground acceleration (PGA) or spectral acceleration (Sa) as an
Received in revised form intensity measure (IM) of an earthquake. However, there are many other earthquake IMs that were
14 May 2019
proposed by various researchers. The aim of this study is to investigate the correlation between seismic
Accepted 12 June 2019
Available online 13 June 2019
responses of NPP components and 23 earthquake IMs and identify the best IMs for correlating with
damage of NPP structures. Particularly, low- and high-frequency ground motion records are separately
accounted in correlation analyses. An advanced power reactor NPP in Korea, APR1400, is selected for
Keywords:
Nuclear power plant
numerical analyses where containment and auxiliary buildings are modeled using SAP2000. Floor dis-
Earthquake intensity measure placements and accelerations are monitored for the non- and base-isolated NPP structures while shear
High-frequency ground motion deformations of the base isolator are additionally monitored for the base-isolated NPP. A series of
Pearson's correlation coefficient Pearson's correlation coefficients are calculated to recognize the correlation between each of the 23
Seismic fragility curve earthquake IMs and responses of NPP structures. The numerical results demonstrate that there is a
significant difference in the correlation between earthquake IMs and seismic responses of non-isolated
NPP structures considering low- and high-frequency ground motion groups. Meanwhile, a trivial
discrepancy of the correlation is observed in the case of the base-isolated NPP subjected to the two
groups of ground motions. Moreover, a selection of PGA or Sa for seismic response analyses of NPP
structures in the high-frequency seismic regions may not be the best option. Additionally, a set of fragility
curves are thereafter developed for the base-isolated NPP based on the shear deformation of lead rubber
bearing (LRB) with respect to the strongly correlated IMs. The results reveal that the probability of
damage to the structure is higher for low-frequency earthquakes compared with that of high-frequency
ground motions.
© 2019 Korean Nuclear Society, Published by Elsevier Korea LLC. This is an open access article under the
CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

1. Introduction automatically. The 2016 Gyeongju and 2017 Pohang earthquakes

The 2011 Great East Japan Earthquake caused significant dam-
age to infrastructures in the region and immediately made a shut-
down state of the eleven reactors at four nuclear power plants
(NPPs) in Fukushima Daiichi. In the 2011 Mineral Virginia US
earthquake, the ground motion exceeded the level of the operating
basis and design basis earthquakes at the North Anna NPP. It was
the first US nuclear plant to be shut down due to an earthquake. The
1999 Chichi earthquake caused three nuclear reactors in Northern
Taiwan (Kuosheng and Chinshan areas) were shut down
* Corresponding author.
E-mail address: thlee@konkuk.ac.kr (T.-H. Lee).
occurred in South Korea, nearby the NPPs, showed that the peak
ground acceleration can be larger than 0.3 g, the safe shutdown
earthquake design level of NPPs. Therefore, studies on seismic
response analyses of NPP structures are always needed.
The peak ground acceleration (PGA), velocity (PGV) or spectral
acceleration (Sa) is commonly selected as an intensity measure (IM)
of an earthquake for the seismic performance analysis of civil en-
gineering structures. However, several studies showed that such
parameters might not always be the best selections for seismic
responses and damage analysis of civil structures [1e9]. Addition-
ally, many other IMs were proposed by researchers and these IMs
might be strongly correlated with seismic damage of different kinds
of structures. In specific, the correlation between earthquake IMs

https://doi.org/10.1016/j.net.2019.06.013
1738-5733/© 2019 Korean Nuclear Society, Published by Elsevier Korea LLC. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/
licenses/by-nc-nd/4.0/).
 D.-D. Nguyen et al. / Nuclear Engineering and Technology 52 (2020) 192e205
and seismic damage of building structures was the most sufficiently
studied [6e15]. Cao and Ronagh [8,9] they pointed out that velocity
spectrum intensity (VSI), Housner intensity, and Sa are the best
indicators correlated to seismic damage of low-rise buildings.
Additionally, various researchers demonstrated that other IMs also
have a strong correlation with the seismic response of buildings
such as frequency-dependent parameters [10], Arias intensity [11],
Sa(T1) and velocity-based parameters [12], and PGV and spectral
response parameters [13,14]. An improved earthquake IM was
proposed by Lu et al. [15] for super high-rise buildings, which
showed to be more reliable and efficient than PGA and Sa(T1) in
estimating collapse simulation.
The relation between different earthquake IMs and responses of
other civil structures were also investigated numerously. Optimal
IMs for seismic responses of different types of bridges are con-
ducted by Padgett et al. [16], Jahangiri et al. [17], Zelaschi et al. [18],
Avşar et al. [19], and Nguyen et al. [20]. Correlation analyses be-
tween ground motion parameters and damage of mountain tunnels
were implemented by Chen and Wei [1]. Nguyen et al. [21]
demonstrated that PGV/Vs30 is the optimum seismic IM for devel-
oping the fragility curves of underground structures. Phan and
Paolacci [22] identified the efficient IMs for seismic response
analysis of liquid steel storage tanks.
Seismic probabilistic risk assessment (SPRA) and seismic margin
assessment (SMA) are two kinds of effective approaches for safety
examination of NPPs under seismic events. The seismic fragility
evaluation is a crucial step in the seismic risk assessment of NPPs.
During the evaluation process, to represent the seismic capacity of
structures and equipment, high confident of low probability of failure
(HCLPF) values and fragility curves are developed. The HCLPF value is
defined as an earthquake level corresponding approximately to a 95%
confidence that the probability of failure is less than 5%. There are
many methods to derive fragility curves as well as calculate the
HCLPF such as probabilistic fragility analysis (PFA), conservative
deterministic failure margin method (CDFM), and test results.
For the PFA methodology [23,24], the fragility curves and HCLPF
are derived from a function of seismic median capacity and com-
posite standard deviation. In this method, numerous seismic fac-
tors are considered as uncertainty parameters, which are ground
motion, damping, modeling, mode shape combination, time history
simulation, strength, inelastic model, and soil-structure interaction.
Each factor is estimated in a statistic way to establish a median and
a composite standard deviation. It is noted that the PFA method-
ology is reflected in the calculation process of seismic margin fac-
tors. While the PFA method calculates the seismic margin factors
using a lognormal distribution in probabilistic methodology, CDFM
method [25] determines the seismic margin factor based on finite
element analysis. The three common deterministic seismic factors
are the strength, inelastic energy absorption, and damping. The test
data method [26] is widely used to evaluate electrical equipment
with sufficient experimental data. Based on the response curves
obtained from various failure modes, the lower bound of the curves
can represent the deterministic fragility level. Zero period accel-
eration and average spectral acceleration are two indicators for
fragility level in both deterministic and probabilistic approaches.
Recently, many studies performed the seismic fragility assess-
ment of NPP structures and equipment adopting the conventional
aforementioned methodologies [27e33] or modified conventional
method [34,35]. Additionally, many studies adopted general ap-
proaches such as the maximum likelihood estimation (MLE), the
incremental dynamic analysis (IDA), and the regression, which have
been used in seismic fragility assessment of general civil engi-
neering structures (e.g. buildings, bridges, and underground
structures), for NPP structures and components [27,31,33,37e40].
Some studies used the ground motions compatible to uniform
193
hazard response spectrum for developing fragility curves
[27,29,36,37], while other studies used arbitrary ground
[30,31,33,39]. The advantage of using arbitrary ground motions is to
consider a wide range of magnitude, period, rupture distance, or
amplitude as uncertainty characteristics in fragility analysis. In
particular, Mandal et al. [30,31] performed comparative studies on
fragility assessments using the conventional method for NPPs and
the general approaches for civil structures. They used a set of
arbitrary ground motions for analyses and concluded that the IDA-
and MLE-based fragility analyses give highly reasonable estimates
compared to the conventional method, which produces an over-
estimation of the fragility by a great margin.
Until now, correlations between seismic IMs and responses of
buildings and other civil engineering structures were systemati-
cally studied, but those of NPP structures are still not studied yet.
Therefore, the purpose of this study is to investigate the relation-
ship between 23 earthquake IMs and seismic responses of NPP
structures to identify the significant IMs to estimate damage of NPP.
In particular, significant IMs for high-frequency and low-frequency
ground motions are distinguished where a ground motion is
referred to as a high-frequency motion when its' dominant fre-
quency is over 10 Hz. An NPP containing an advanced pressurized
water nuclear reactor, APR1400, designed in South Korea is selected
as a case study example. Both NPP structures with and without base
isolators are considered. Structural models of NPP structures are
developed in SAP2000 program where lumped-mass stick model is
used to model the containment and auxiliary buildings. A series of
time history analysis is performed to monitor structural responses
such as floor accelerations, floor displacements, and shear defor-
mation of the base isolator. Thereafter, Pearson's correlation co-
efficients are calculated for each IM and strongly correlated IMs
with seismic responses of NPP structures are identified. Finally, a
set of fragility curves are developed based on the shear deformation
of LRB with respect to the strongly correlated IMs.
2. Earthquake intensity measures and ground motions
records
2.1. Earthquake intensity measures
Earthquake IMs are crucial values for representing the charac-
teristics of a seismic ground motion. In general, IMs characterize
the amplitude, frequency content, and duration of motions [41].
The optimum IMs can reduce the deviation of seismic structural
performances and predict the responses of structures accurately
[15]. In this study, we select 23 available IMs proposed by many
researchers for correlation analyses. These IMs are calculated for
every ground motion time histories using SeismoSignal program
[42]. The IMs adopted in this study and their definitions are pre-
sented in Table 1. It should be noted that the “intensity measure”
(IM) terminology used herein represents all investigated ground
motion parameters in this paper.
2.2. Selected ground motion records
For this study, 80 ground motion records are selected from
worldwide earthquakes that provided by the PEER center [53] and
KMA [54]. Two groups of ground motions including low- and high-
frequency content are classified separately, in which 40 records are
in the low-frequency motion group and 40 records are in the high-
frequency group. The low-frequency ground motions are collected
from worldwide historic earthquakes while the high-frequency
ground motion group consists of some earthquake events in
North America and Korea, which can be categorized as mid-to low-
seismicity zones. The frequency content of a ground motion is
194 D.-D. Nguyen et al. / Nuclear Engineering and Technology 52 (2020) 192e205

Table 1
Selected earthquake intensity measures.

No. Earthquake parameter Definition Unit Reference

1 Peak ground acceleration PGA ¼ max ja(t)j g e

2 Peak ground velocity PGV ¼ max jv(t)j m/s e
3 Peak ground displacement PGD ¼ max jd(t)j m e
4 Ratio of PGV/PGA PGV/PGA s Kramer 1996 [41]
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
Z ttot
5 Root-mean-square of 1 g Dobry et al., 1978 [43]
acceleration Arms ¼ aðtÞ2 dt
ttot 0
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
Z ttot ffi
6 Root-mean-square of velocity 1 m/s Kramer 1996 [41]
Vrms ¼ vðtÞ2 dt
ttot 0
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
Z ttot
7 Root-mean-square of 1 m Kramer 1996 [41]
displacement Drms ¼ dðtÞ2 dt
ttot 0
Z ttot
8 Arias intensity p m/s Arias 1970 [44]
Ia ¼ aðtÞ2 dt
2g 0
9 Characteristic intensity 2=3 pffiffiffiffiffiffiffi m1.5/ Park et al., 1985 [45]
Ic ¼ ðArms Þ ttot
s2.5
R ttot
10 Specific energy density SED¼ 0 vðtÞ dt 2 m2/s e
11 Cumulative absolute velocity R t  
 m/s Benjamin and Associates 1988 [46]
CAV ¼ 0tot aðtÞdt
R
12 Acceleration spectrum intensity ASI ¼ 0:5 Sa ðx ¼ 0:05; TÞdT g*s Housner 1952 [47] & Thun et al., 1988
0:1
[48]
13 Velocity spectrum intensity R 2:5 m Housner 1952 [47] & Thun et al., 1988
VSI ¼ 0:1 Sv ðx ¼ 0:05; TÞdT
[48]
14 Housner spectrum intensity R 2:5 m Housner 1952 [47]
HI ¼ 0:1 PSv ðx ¼ 0:05; TÞdT
15 Sustained maximum SMA ¼ the 3rd of PGA g Nuttli 1979 [49]
acceleration
16 Sustained maximum velocity SMV ¼ the 3rd of PGV m/s Nuttli 1979 [49]
17 Effective peak acceleration meanðS0:10:5
a ðx ¼ 0:05ÞÞ g Benjamin and Associates 1988 [46]
EPA ¼
2:5
18 Spectral acceleration at T1 Sa ðT1 Þ g Shome et al., 1988 [50]
19 Spectral velocity at T1 Sv ðT1 Þ m/s e
20 Spectral displacement at T1 Sd ðT1 Þ m e
21 A95 parameter A95 ¼ 0.764 I 0:438
a g Sarma & Yang 1987 [51]
22 Predominant period Tp s Kramer 1996 [41]
23 Mean period SC 2i =fi Þ s Rathje et al., 1998 [52]
Tm ¼ ; Ci is the Fourier amplitude, fi is the discrete frequency corresponding
C 2i
to Ci

normally identified based on the response spectrum. In this study,
10 Hz is defined as the threshold to distinguish between low- and
high-frequency earthquakes. In other words, the ground motions
with large values of spectral accelerations fallen in the frequency
range approximately larger than 10 Hz is considered as a high-
frequency ground motion, otherwise discerned as low-frequency
motions [55,56]. Fig. 1 shows the response spectra of two groups
of ground motion records used in this study.
3. Structural modeling
3.1. Numerical modeling of NPP structures
The Advanced Power Reactor 1400 (ARP1400), an advanced
pressurized water nuclear reactor designed in Korea is employed
for a numerical example in this study. Fig. 2 shows a general view
and cutting view of structural configuration of the APR1400 plant.
We focus our modeling on the reactor containment building and
the auxiliary building. The lumped-mass stick models of the
structure are implemented in SAP2000 [57], a commercial struc-
tural analysis program. The containment and auxiliary buildings
are modeled in terms of a simplified model using elastic beam el-
ements. The equivalent section properties are calculated based on
the designed cross sections of the structures [58,59]. The lumped
masses are also assigned into associated element nodes. Elastic
shell elements are applied for the base-mat foundation. The
structural properties of the APR1400 NPP in the lumped-mass
beam stick model are shown in Table 2.

Fig. 1. Response spectra of selected ground motion records.

 D.-D. Nguyen et al. / Nuclear Engineering and Technology 52 (2020) 192e205 195

Fig. 2. Advanced power reactor 1400.

Table 2
Properties of the APR1400 NPP in the lumped-mass beam stick model.

Node Height frombase-mat (m) Nodal mass(ton) Area(m2) Moment of inertia(m4) Shear area(m2) Torsional constant (m4)

Reactor containment building

1 16.76 87.07 202.90 56299.85, 56299.85 101.45, 101.45 112634.22

2 20.27 166.52 202.90 56299.85, 56299.85 101.45, 101.45 112634.22
3 23.46 185.42 202.90 56299.85, 56299.85 101.45, 101.45 112634.22
4 27.73 189.29 202.90 56299.85, 56299.85 101.45, 101.45 112634.22
5 31.09 170.39 202.90 56299.85, 56299.85 101.45, 101.45 112634.22
6 34.59 234.68 202.90 56299.85, 56299.85 101.45, 101.45 112634.22
7 40.53 314.15 202.90 56299.85, 56299.85 101.45, 101.45 112634.22
8 47.24 333.05 202.90 56299.85, 56299.85 101.45, 101.45 112634.22
9 53.94 318.02 202.90 56299.85, 56299.85 101.45, 101.45 112634.22
10 60.65 310.43 202.90 56299.85, 56299.85 101.45, 101.45 112634.22
11 66.44 376.80 202.90 56299.85, 56299.85 101.45, 101.45 112634.22
12 70.56 279.92 179.76 47591.20, 47591.20 89.89, 89.89 95199.65
13 78.63 355.52 179.76 35861.70, 35861.70 89.89, 89.89 71732.03
14 86.72 352.09 166.11 12825.63, 12825.63 83.03, 83.03 25651.25
15 94.64 147.80

Internal structure

1201 16.76 184.17 833.15 51055.67, 79896.93 662.77, 662.77 164989.72

1202 18.28 341.40 883.97 51262.81, 81942.48 704.29,704.29 168389.13
1203 20.26 796.48 857.92 51149.25, 80710.1 684.03, 684.03 165957.82
1204 23.46 523.67 313.78 9908.34, 21253.36 221.77, 221.77 37753.43
1204 25.75 273.10 254.60 9816.63, 19442.95 171.14, 171.14 35811.75
1205 27.73 296.29 221.94 9515.25, 19384.02 144.33, 144.33 35233.94
1206 31.09 296.86 261.38 9848.14, 20166.32 175.93,175.93 36571.27
1207 32.61 355.47 202.76 9630.81, 18524.13 130.81, 130.81 34566.87
1208 34.59 80.51 202.76 9630.81, 18524.13 130.81, 130.81 34566.87
1209 36.57 264.64 202.76 9630.81, 18524.13 130.75, 130.75 34566.87
1210 40.53 255.68 103.23 1932.60, 4666.70 94.90, 94.90 7888.57
1211 46.32 271.76 97.93 1918.14, 4642.60 92.25, 92.25 7840.37
1212 51.20 81.38

Auxiliary structure

1001 16.76 4608.78 1660.45 530766.3, 405355.4 770.35, 611.30 239729.53

1002 23.46 5265.66 1503.17 466685.8, 332679.3 658.91, 582.41 198363.83
1003 36.57 4680.11 1529.65 464101.2, 340852.6 659.42, 589.37 205791.05
1004 41.91 4150.72 1363.35 368330.8, 292325.7 565.41, 558.81 187226.65
1005 40.53 3218.83 842.16 241060.5, 158771.4 358.88, 329.06 173747.81
1006 52.42 2000.28 579.06 142392.3, 121479.2 212.0, 261.6 76553.8
1007 57.92 1659.49 371.98 75850.4, 86362.9 160.3, 192.0 52753.99
1008 63.39 957.38

For the base-isolated model, 486 LRBs are installed beneath the
base mat for improving the seismic performance of the NPP
structures. Fig. 3 shows the finite element models of the non- and
base-isolated NPP structures in SAP2000. Fig. 4 shows the layout of
LRBs and the bilinear response model of LRBs due to shear forces.
The mechanical properties of LRBs are described in Table 3.
The results of the eigenvalue analysis are presented in Table 4.
For the non-isolated NPP model, the 1st and 2nd modes are the
horizontal translation modes of the containment building, while,
the 3rd and 4th modes represent the translational modes of the
auxiliary building, as shown in Fig. 5. For the base-isolated NPP
model, the 1st and 2nd modes are translational modes of the su-
perstructure, while the 3rd mode represents the rotational mode
and the 4th mode is the translational model of the containment
building, as shown in Fig. 6. These results are also in good agree-
ment with the findings elsewhere [60e62].
3.2. Verification of the lumped-mass stick model (LMSM)
The LMSM is verified with a 3D finite element model (3D FEM)
using ANSYS software. A set of time-history analyses for the non-
isolated reactor containment building is performed to compare
196 D.-D. Nguyen et al. / Nuclear Engineering and Technology 52 (2020) 192e205

Fig. 3. Finite element lumped-mass stick model of APR1400: Non-isolated (left) and base-isolated model (right).

structures are expected to vibrate within an elastic range during a

design earthquake. All ground motions were applied to the NPP
models in the horizontal direction. The seismic responses of non-
isolated NPP structures are obtained in terms of the maximum
drift ratio and floor acceleration at each component, while for the
base-isolated model, the additional shear deformation behavior of
LRBs is observed. Figs. 10 and 11 show representative examples of
the maximum floor displacements and examples of the displace-
ment time-history responses of the NPP structure, respectively, for
low-and high-frequency earthquakes. The acceleration floor
Fig. 4. Layout of LRBs and their bilinear response model.
response spectra at the top of the NPP components due to earth-
quakes are also plotted in Fig. 12. Fig. 13 shows shear force versus
shear deformation behaviors of the LRB under different earth-
Table 3 quakes. The responses of the NPP components are obtained for
Mechanical properties of LRBs. every ground motion records.
Property Value Unit
A relationship between seismic responses of NPP structures and
earthquake IMs is needed to identify the strong and the weak
Elastic stiffness, Ku 537,703 kN/m
correlation indicators. In this study, the Pearson's coefficient is used
Hardening stiffness, Kd 4204.84 kN/m
Yield strength, Fy 1009.65 kN to reflect the correlation between seismic responses of NPP struc-
Characteristic strength, Qd 1001.03 kN tures and earthquake IMs. The correlation coefficient given by Ang
Vertical stiffness, Kv 12.896 kN/m and Tang [63] is defined as
Effective stiffness, Keff 8970 kN/m
Equivalent damping ratio, x 0.335 P
1 ðxi  xÞðyi  yÞ
r¼ (1)
n1 sx sy
Table 4
where x, y, are the sample means of variables xi and yi. xi represents
Eigenvalue analysis results.
the seismic responses of NPP components, while yi represents the
Mode shape Natural frequency (Hz) values of intensity measures. sx , sy are the sample standard de-
Fixed base Isolated base viations of x and y, determined by
Translational Y of superstructure e 0.476 e Mode 1 rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
Translational X of superstructure e 0.477 e Mode 2 1 h i 1 h i
Rotational Z of superstructure e 0.709 e Mode 3 sx ¼ ðSxi Þ2  nðxÞ2 ; sy ¼ ðSyi Þ2  nðyÞ2
Translational Y of RCB 3.858 e Mode 1 3.786 e Mode 4
n1 n1
Translational X of RCB 3.859 e Mode 2 3.812 e Mode 5 (2)
Translational Y of auxiliary bldg. 5.077 e Mode 3 e
Translational X of auxiliary bldg. 5.351 e Mode 4 e Fig. 14 shows the representative results of the calculated cor-
relation coefficients for the non-isolated NPP structures associated
with low- and high-frequency ground motions. It is found that for
the low-frequency motions, the A95 and PGA are the strongest
the seismic behaviors of two models. Fig. 7 shows the 3D FEM of the correlated parameters, followed by sustained maximum accelera-
containment building modeled in ANSYS. Time-history responses tion (SMA) and effective peak acceleration (EPA). The weakest
and floor response spectra of LMSM and 3D FEM are shown in correlated IMs are specific energy density (SED), root mean square
Figs. 8 and 9, respectively. It can be found that the results of the two of displacement (Drms), PGD, and Sv(T1). For the high-frequency
models are in good agreement, demonstrating that the LMSM is motion group, the good correlation indicators are SED, character-
sufficient for time-history analyses of NPP structures. istic intensity (Ic), and Arias intensity (Ia). The poor correlation IMs
are PGV/PGA, predominant period (Tp), and mean period (Tm). The
4. Seismic responses of NPP structures and correlation similar tendency is observed for all NPP components modeled.
analyses Fig. 15 shows the representative results of the correlation co-
efficients for the base-isolated NPP model. It can be demonstrated
A series of linear time-history analyses are performed since NPP that for low-frequency ground motions, SED, Ia, and Ic are strongly
 D.-D. Nguyen et al. / Nuclear Engineering and Technology 52 (2020) 192e205 197

Fig. 5. Mode shapes of the non-isolated NPP model.

Fig. 6. Mode shapes of the base-isolated NPP model.

correlated parameters. Meanwhile, the lowest correlation IMs are and Tm, are less than 1.7 s and 0.7 s for the LF and HF groups, respec-
PGV/PGA, Tp, and Tm. Also, this trend is obtained for the base-isolated tively. It is smaller than the fundamental period of the base-isolated
NPP structures associated with the high-frequency motions. structures, consequently, causing a less damage to the structure.
Tables 5 and 6 summarize the strongly and weakly correlated
IMs with seismic responses of NPP structures under low- and high-
5. Fragility analyses
frequency earthquakes, respectively. It is highlighted that SED, Ia,
and Ic are the best IMs in correlating with seismic responses of the
5.1. Limit states
base-isolated NPP structures for both low- and high-frequency
earthquakes and the non-isolated structure in HF excitations.
The seismic fragility is the conditional probability that a struc-
A95, PGA, and SMA are the also good indicators for correlation with
tural system exceeds a limit state when subjected to a specified
damage of the non-isolated NPP structures in LF seismic zones. In
level of earthquake intensity. For the base-isolated NPP structure,
contrast, Tp, Tm, and PGV/PGA are the weakest correlated IMs with
the superstructure should behave within the elastic range during
seismic responses of the NPP structures under HF earthquakes and
the seismic excitation. However, the base isolator will experience a
the base-isolated structures in LF earthquakes. Additionally, Drms,
large deformation that may exceed the yield deformation and reach
PGD, and Sv(T1) are also shown to be low correlation with seismic
the ultimate deformation capacity. For developing fragility curves, a
damage of the non-isolated NPP structure in LF earthquakes.
set of limit states should be pre-defined according to damage levels
For the non-isolated NPP model with regard to the low-
of components. In this study, the damage levels of an LRB with
frequency motions, the strongly correlated intensity measures
respect to the shear strain is considered as the limit states. The
(PGA, A95, SMA, EPA) seem to be resulted from their definitions, in
shear strain of LRB, g is expressed by
which these measures are related to the force (or the acceleration).
Thus, it governs the floor responses of the fixed base structures for D
high amplitude of ground motions. Whereas, the poor correlation g¼ (3)
H
parameters (Drms and PGD), which are directly related to the
deformation (or the displacement), seems to be less sensitive to the where D and H are the maximum lateral deformation and the
relatively rigid structures. height of LRB, respectively. According to recent experimental
For the base-isolated structures, most of input seismic energy is studies [64e69], the LRB can reach to an ultimate capacity of
dissipated through the hysteretic behavior of LRBs. Therefore, it is beyond 400% shear strain. Therefore, we adopt these results to
obvious to find that the energy-based IMs (SED, Ia) are the strong define three limit states, if the shear strain exceeds 100% (i.e.
correlated parameters. Meanwhile, the period-based parameters, Tp D  40 cm), the slight limit state (LS-1) is specified. Similarly, if the
198 D.-D. Nguyen et al. / Nuclear Engineering and Technology 52 (2020) 192e205

Shinozuka et al. [73] is used in this study. In this approach, the

fragility function is assumed as a log-normal cumulative distribu-
tion function (CDF) expressed by
 
lnða=ck Þ
Fk ðaÞ ¼ F (4)
zk

where a is the earthquake intensity measure, ck and zk are the

median and the log-standard deviation of the log-normal CDF, and
Fð Þ is the standard normal CDF. In Equation (4), the subscript k
indicates the k-th limit state when more than one limit state is
considered. In the MLE, ck and zk are determined by maximizing the
likelihood function. This function is defined by
YN
L¼ i¼1
½Fk ðai Þxi ½1  Fk ðai Þð1xi Þ (5)

where Fk ðaÞ increases when damage occurs and 1  Fk ðaÞ, the

probability of not experiencing a damage, increases when damage
does not occur for the earthquake intensity of ai . In Equation (5), N
is the number of ground motions considered and xi is a Bernoulli
random variable that indicates whether the structure is damaged or
not where 0 indicates no damage and 1 indicates damage. ck and zk
are determined so that Equation (5) is maximized with respect to ck
and zk as follows
Fig. 7. 3D FEM of the reactor containment building.
vL vL
¼ ¼ 0; k ¼ 1; /; NLS (6)
vck vzk
shear strain goes beyond 300% (i.e. D  120 cm) and 400% (i.e.
D  160 cm), the moderate (LS-2) and extensive (LS-3) limit states where NLS is the number of limit states.
are established. This approach was also used in previous studies
[39,40,70e72]. Three limit states are summarized in Table 7. 5.3. Fragility curves

5.2. Maximum likelihood estimation Since LRB is the crucial element in the base-isolated NPP, a set of
fragility curves is developed for different limit states, which are
Among several methods to develop seismic fragility curves, the defined based on the shear strain of LRB. The two spectrally
maximum likelihood estimation (MLE) approach proposed by equivalent ground motion suites (LF and HF motions) are used to

Fig. 8. Comparison of time-history responses between two models under the 1940 El Centro earthquake.
 D.-D. Nguyen et al. / Nuclear Engineering and Technology 52 (2020) 192e205 199

Fig. 9. Floor response spectra of the structure under the 1940 El Centro earthquake.

Fig. 10. Maximum displacements at the top of the NPP structures under the LF and HF earthquakes.

Fig. 11. Displacement time-history responses at the top of the containment building under the LF and HF earthquakes.
200 D.-D. Nguyen et al. / Nuclear Engineering and Technology 52 (2020) 192e205

Fig. 12. Mean floor response spectra at the top of the NPP structures with different earthquakes.

Fig. 13. Responses of LRB under different earthquakes.

 D.-D. Nguyen et al. / Nuclear Engineering and Technology 52 (2020) 192e205 201

Fig. 14. Correlation between responses of non-isolated NPP and earthquake IMs.

perform the incremental dynamic analysis (IDA) [74]. In IDA, the respectively, under LF motions. Table 8 describes the fragility
selected ground motions are scaled to multiple levels of intensity, function parameters (i.e. median and standard deviation (SD)) for
herein, we select two strongly correlated IMs, which are SED and Ia, LF and HF earthquakes with different limit states.
to develop fragility curves for the base-isolated NPP structure. For
each analysis, the maximum deformation of LRB is monitored as
6. Conclusions
the engineering demand parameter. In total, 800 dynamic analyses
are performed (i.e. 2 suites x 40 records x 10 intensity levels).
Significant intensity measures of earthquakes that are strongly
Fig. 16 shows the IDA curves in terms of deformation of LRB
correlated to seismic responses of NPP structures are identified. We
versus IMs for LF and HF motions. The solid curves represent the
consider the effect of low-frequency and high-frequency ground
mean values. Due to higher input energy of the LF motions, the base
motions, separately. A series of time history analyses are performed
isolator is deformed in a larger range during a stronger seismic
for both non- and base-isolated numerical models of NPP struc-
intensity.
tures to monitor the floor displacement, floor acceleration, and
Fragility curves of base-isolated NPP model are developed for
shear deformation of LRB. Pearson's correlation coefficients are
three limit states considering two groups of ground motions. Fig. 17
calculated to identify the best earthquake IMs for indicating the
shows the fragility curves for three limit states with different
seismic damage of NPP structures. A set of fragility curves is also
earthquake groups. It can be observed that the base-isolated NPP
developed with respect to the well-correlated IMs. Based on the
structure might behave without damage if the level of SED and Ia
numerical results, the following conclusions are drawn.
are less than 0.2 m2/s and 1.5 m/s, respectively.
Fig. 18 shows a comparison of fragility curves for three LSs and
(1) For the non-isolated NPP structures subjected to low-
two groups of earthquakes. It can be observed that the structural
frequency ground motions, the best IMs for correlating
model under LF motions would be more vulnerable than that due to
with seismic responses of such structures are PGA, A95, and
HF earthquakes. This can be attributed to the reason that the
SMA. For high-frequency earthquakes, the best IMs are SED,
deformation of LRBs produced by LF motions is higher than that
Ic, and Ia.
under HF motions. It can be clearly observed that the median values
(2) For the base-isolated NPP structures subjected to low-
for LS-1 are 22% and 24% lower when LF motions are applied than
frequency ground motions, SED, Ia, and Ic are the strongly
applying HF ones for SED and Ia, respectively. Similarly, the median
correlated parameters. This trend is also obtained for the
values for LS-2 and LS-3 are reduced by 15% and 9% with respect to
base-isolated NPP structures associated with the high-
SED, respectively, and also decreased 12% and 8% with respect to Ia,
frequency ground motions.
202 D.-D. Nguyen et al. / Nuclear Engineering and Technology 52 (2020) 192e205

Fig. 15. Correlation between the responses of base-isolated NPP and earthquake IMs.

Table 5
Strongly correlated IMs with seismic responses of NPP structures.

Low-frequency earthquakes High-frequency earthquakes

Non-isolated model Base-isolated model Non-isolated model Base-isolated model

A95 SED SED SED

PGA Ia Ic Ia
SMA Ic Ia Ic

Table 6
Weakly correlated IMs with seismic responses of NPP structures.

Low-frequency earthquakes High-frequency earthquakes

Non-isolated model Base-isolated model Non-isolated model Base-isolated model

Drms Tp Tp Tp
PGD Tm Tm Tm
Sv(T1) PGV/PGA PGV/PGA PGV/PGA

Table 7
Definition of limit states.

Limit state Shear strain of LRB Maximum lateral deformation of LRB (mm)

Slight damage (LS-1) g  100% 400

Moderate damage (LS-2) g  300% 1200
Extensive damage (LS-3) g  400% 1600
 D.-D. Nguyen et al. / Nuclear Engineering and Technology 52 (2020) 192e205 203

Fig. 16. Incremental displacement of LRB with respect to SED and Ia.

Fig. 17. Fragility curves for limit states with various ground motion groups.

(3) A selection of PGA or Sa for correlation analyses with seismic
responses of the NPP structures in the high-frequency
seismic regions may not be the best option.
(4) The median values of fragility functions for three limit states
in the case of base-isolated NPP are reduced by 9%e22% and
8%e24% for SED and Ia, respectively, when low-frequency
ground motions applied than when high-frequency mo-
tions are applied. It implies that the considered NPP structure
is more vulnerable to low-frequency ground motions than
high-frequency ground motions.
Data availability
All the data supporting the key findings of this paper are pre-
sented in the figures and tables of the article. Request for other data
will be considered by the corresponding author.
Conflicts of interest
The authors declare that they have no conflicts of interest.
204 D.-D. Nguyen et al. / Nuclear Engineering and Technology 52 (2020) 192e205

Fig. 18. Comparison of fragility curves for low- and high-frequency excitations.

Table 8 698e704.
Fragility function parameters of the based-isolated NPP structure. [8] V. Cao, H.R. Ronagh, Correlation between seismic parameters of far-fault
motions and damage indices of low-rise reinforced concrete frames, Soil
Limit state LF earthquake HF earthquake Dynam. Earthq. Eng. 66 (2014) 102e112.
[9] V. Cao, H.R. Ronagh, Correlation between parameters of pulse-type motions
SED (m2/s) Ia (m/s) SED (m2/s) Ia (m/s)
and damage of low-rise RC frames, Earthq. Struct. 7 (3) (2014) 365e384.
Median SD Median SD Median SD Median SD [10] A. Massumi, F. Gholami, The influence of seismic intensity parameters on
structural damage of RC buildings using principal components analysis, Appl.
LS-1 0.489 0.442 6.206 0.570 0.625 0.549 8.123 0.643 Math. Model. 40 (3) (2016) 2161e2176.
LS-2 1.054 0.422 16.860 0.587 1.247 0.443 19.165 0.491 [11] N. Nanos, A. Elenas, P. Ponterosso, Correlation of different strong motion
LS-3 1.633 0.409 33.202 0.590 1.793 0.247 36.179 0.279 duration parameters and damage indicators of reinforced concrete structures,
in: The 14th World Conference on Earthquake Engineering, 2008. Beijing,
China, 2008.
[12] K. Kostinakis, A. Athanatopoulou, K. Morfidis, Correlation between ground
Acknowledgement motion intensity measures and seismic damage of 3D R/C buildings, Eng.
Struct. 82 (2015) 151e167.
[13] J. Pejovic, S. Jankovic, Selection of ground motion intensity measure for
This study is supported by project titled “Development of High reinforced concrete structure, Procedia Eng. 117 (2015) 588e595.
Reliability Seismic Monitoring System for Precise Analysis of High [14] J. Pejovic, N. Serdar, R. Pejovic, Optimal intensity measures for probabilistic
Frequency Earthquake Motion in Nuclear Power Plants”, assignment seismic demand models of RC high-rise buildings, Earthq. Struct. 13 (3) (2017)
221e230.
number 20171510101860. [15] X. Lu, L. Ye, X. Lu, M. Li, X. Ma, An improved ground motion intensity measure
for super high-rise buildings, Sci. China Technol. Sci. 56 (6) (2013)
1525e1533.
References [16] J.E. Padgett, B.G. Nielson, R. DesRoches, Selection of optimal intensity mea-
sures in probabilistic seismic demand models of highway bridge portfolios,
Earthq. Eng. Struct. Dyn. 37 (5) (2008) 711e725.
[1] Z. Chen, J. Wei, Correlation between ground motion parameters and lining
[17] V. Jahangiri, M. Yazdani, M.S. Marefat, Intensity measures for the seismic
damage indices for mountain tunnels, Nat. Hazards 65 (3) (2013) 1683e1702.
response assessment of plain concrete arch bridges, Bull. Earthq. Eng. (2018)
[2] M. Corigliano, C.G. Lai, G. Barla, January). Seismic vulnerability of rock tunnels
1e24.
using fragility curves, in: 11th ISRM Congress, International Society for Rock
[18] C. Zelaschi, R. Monteiro, R. Pinho, Critical assessment of intensity measures for
Mechanics and Rock Engineering, 2007.
seismic response of Italian RC bridge portfolios, J. Earthq. Eng. (2017) 1e21.
[3] S. Yaghmaei-Sabegh, Application of wavelet transforms on characterization of € Avşar, G. Ozdemir,
€
[19] O. Response of seismic-isolated bridges in relation to in-
inelastic displacement ratio spectra for pulse-like ground motions, J. Earthq.
tensity measures of ordinary and pulselike ground motions, J. Bridge Eng. 18
Eng. 16 (4) (2012) 561e578.
(3) (2011) 250e260.
[4] M. Ghayoomi, S. Dashti, Effect of ground motion characteristics on seismic
[20] D. Nguyen, H. Park, B. Thusa, T.-H. Lee, Identification of Earthquake Intensity
soil-foundation-structure interaction, Earthq. Spectra 31 (3) (2015)
Measures as the Seismic Damage Indicator of Bridges Considering Low- and
1789e1812.
High-Frequency Contents of Ground Motions, Submitted to Earthquakes And
[5] Y.Y. Zhang, Y. Ding, Y.T. Pang, Selection of optimal intensity measures in
Structures, 2019.
seismic damage analysis of cable-stayed bridges subjected to far-fault ground
[21] D.D. Nguyen, D. Park, S. Shamsher, V.Q. Nguyen, T.H. Lee, Seismic vulnerability
motions, J. Earthq. Tsunami 9 (01) (2015), 1550003.
assessment of rectangular cut-and-cover subway tunnels, Tunn. Undergr.
[6] A. Elenas, Correlation between seismic acceleration parameters and overall
Space Technol. 86 (2019) 247e261.
structural damage indices of buildings, Soil Dynam. Earthq. Eng. 20 (1e4)
[22] H.N. Phan, F. Paolacci, Efficient intensity measures for probabilistic seismic
(2000) 93e100.
response analysis of anchored above-ground liquid steel storage tanks, in:
[7] A. Elenas, K. Meskouris, Correlation study between seismic acceleration pa-
ASME 2016 Pressure Vessels and Piping Conference (pp. V005T09A010-
rameters and damage indices of structures, Eng. Struct. 23 (6) (2001)
 D.-D. Nguyen et al. / Nuclear Engineering and Technology 52 (2020) 192e205
V005T09A010), American Society of Mechanical Engineers, 2016.
[23] R.P. Kennedy, M.K. Ravindra, Seismic fragilities for nuclear power plant risk
studies, Nucl. Eng. Des. 79 (1) (1984) 47e68.
[24] EPRI, Methodology for Developing Seismic Fragilities, Report TR-103959, Palo
Alto, CA, USA, 1994.
[25] EPRI, A Methodology for Assessment of Nuclear Power Plant Seismic Margin,
Report EPRI-NP-6041-M-REV.1, 1991 (Palo Alto, CA, USA).
[26] U.S. NUREG, Nuclear Regulatory Commission, Seismic Fragility of Nuclear
Power Plant Components (Phase II), NUREG/CR-4659, vols. 2e4, Department
of Nuclear Energy, Brookhaven Laboratory, Long Island, NY, USA, 1987. BNL-
NUREG-52007.
[27] A. Ali, N.A. Hayah, D. Kim, S.G. Cho, Probabilistic seismic assessment of base-
isolated NPPs subjected to strong ground motions of Tohoku earthquake,
Nucl. Eng. Technol. 46 (5) (2014) 699e706.
[28] P.C. Basu, M.K. Ravindra, Y. Mihara, Component fragility for use in PSA of
nuclear power plant, Nucl. Eng. Des. 323 (2017) 209e227.
[29] S.G. Cho, Y.H. Joe, Seismic fragility analyses of nuclear power plant structures
based on the recorded earthquake data in Korea, Nucl. Eng. Des. 235 (17e19)
(2005) 1867e1874.
[30] T.K. Mandal, N.N. Pujari, S. Ghosh, A comparative study of seismic fragility
estimates using different numerical methods, in: 4th ECCOMAS Thematic
Conference on Computational Methods in Structural Dynamics and Earth-
quake Engineering, 2013.
[31] T.K. Mandal, S. Ghosh, N.N. Pujari, Seismic fragility analysis of a typical Indian
PHWR containment: comparison of fragility models, Struct. Saf. 58 (2016) 11e19.
[32] F. Perotti, M. Domaneschi, S. De Grandis, The numerical computation of
seismic fragility of base-isolated Nuclear Power Plants buildings, Nucl. Eng.
Des. 262 (2013) 189e200.
[33] T.T. Tran, A.T. Cao, T.H.X. Nguyen, D. Kim, Fragility assessment for electric
cabinet in nuclear power plant using response surface methodology, Nucl.
Eng. Technol. 51 (3) (2019) 894e903.
[34] Y.N. Huang, A.S. Whittaker, N. Luco, A probabilistic seismic risk assessment
procedure for nuclear power plants:(I) Methodology, Nucl. Eng. Des. 241 (9)
(2011) 3996e4003.
[35] Y.N. Huang, A.S. Whittaker, N. Luco, A probabilistic seismic risk assessment
procedure for nuclear power plants:(II) Application, Nucl. Eng. Des. 241 (9)
(2011) 3985e3995.
[36] I.K. Choi, Y.S. Choun, S.M. Ahn, J.M. Seo, Seismic fragility analysis of a CANDU
type NPP containment building for near-fault ground motions, KSCE J. Civ.
Eng. 10 (2) (2006) 105e112.
[37] I.K. Choi, Y.S. Choun, S.M. Ahn, J.M. Seo, Probabilistic seismic risk analysis of
CANDU containment structure for near-fault earthquakes, Nucl. Eng. Des. 238
(6) (2008) 1382e1391.
[38] S. De Grandis, M. Domaneschi, F. Perotti, A numerical procedure for
computing the fragility of NPP components under random seismic excitation,
Nucl. Eng. Des. 239 (11) (2009) 2491e2499.
[39] D.D. Nguyen, B. Thusa, T.H. Lee, Seismic fragility of base-isolated nuclear
power plant considering effects of near-fault ground motions, J. Korean Soc.
Hazard Mitig. 18 (7) (2018) 315e321.
[40] D.D. Nguyen, B. Thusa, T.H. Lee, Effects of significant duration of ground
motions on seismic responses of base-isolated nuclear power plant, J. Earthq.
Eng. Soc. Korea 23 (3) (2019) 149e157.
[41] S.L. Kramer, Geotechnical Earthquake Engineering, Prentice Hall, New Jersey,
USA, 1996.
[42] SeismoSignal - a computer program for signal processing of strong-motion
data, available from: http://www.seismosoft.com, 2017.
[43] R. Dobry, I.M. Idriss, E. Ng, Duration characteristics of horizontal components
of strong-motion earthquake records, Bull. Seismol. Soc. Am. 68 (5) (1978)
1487e1520.
[44] A. Arias, A Measure of Earthquake Intensity, Massachusetts Inst. of Tech.,
Cambridge, 1970 (Univ. of Chile, Santiago de Chile).
[45] Y.J. Park, A.H.S. Ang, Y.K. Wen, Seismic damage analysis of reinforced concrete
buildings, J. Struct. Eng. 111 (4) (1985) 740e757.
[46] J.R. Benjamin, A Criterion for Determining Exceedances of the Operating Basis
Earthquake, Report No. EPRI NP-5930, Electric Power Research Institute, Palo
Alto., California, USA, 1988.
[47] G.W. Housner, Spectrum intensities of strong-motion earthquakes, in: Sym-
posium on Earthquake and Blast Effects on Structures, 1952, pp. 20e36 (Los
Angeles, California, USA).
[48] J.L. Von Thun, Earthquake Ground Motions for Design and Analysis of Dams.
Earthquake Engineering and Soil Dynamics II-Recent Advances in Ground-
Motion Evaluation, 1988.
[49] O.W. Nuttli, The Relation of Sustained Maximum Ground Acceleration and Velocity
to Earthquake Intensity and Magnitude, Report 16, US Army Engineer Waterways
Experiment Station, Vicksburg, Mississippi, USA, 1979. Misc. Paper S-73-1.
205
[50] N. Shome, C.A. Cornell, P. Bazzurro, J.E. Carballo, Earthquakes, records, and
nonlinear responses, Earthq. Spectra 14 (3) (1998) 469e500.
[51] S.K. Sarma, K.S. Yang, An evaluation of strong motion records and a new
parameter A95, Earthq. Eng. Struct. Dyn. 15 (1) (1987) 119e132.
[52] E.M. Rathje, N.A. Abrahamson, J.D. Bray, Simplified frequency content esti-
mates of earthquake ground motions, J. Geotech. Geoenviron. Eng. 124 (2)
(1998) 150e159.
[53] PEER Ground Motion Database, 2018. http://peer.berkeley.edu/peer_ground_
motion_database.
[54] KMA, Korean Meteorological Administration, 2016, https://web.kma.go.kr/
eng/index.jsp
[55] EPRI, Program on Technology Innovation: the Effects of High Frequency
Ground Motion on Structures, Components, and Equipment in Nuclear Power
Plants, Report 1015108, Electrical Power Research Institute, Palo Alto, Cali-
fornia, USA, 2007.
[56] EPRI, Advanced Nuclear Technology: High-Frequency Seismic Loading Eval-
uation for Standard Nuclear Power Plants, Report 3002009429, Electrical
Power Research Institute, Palo Alto, California, USA, 2017.
[57] SAP2000 Ver 15, Computers and structures Inc, Berkeley, CA, USA, 2013.
[58] J.M. Kim, E.H. Lee, Development and Verification of Simplified Beam-Stick
Model of Seismically Isolated ARP1400 Nuclear Power Plant Structure,”
Research Report, Central Research Institute of KHNP, KETEP Project No.
2014151010170B, Korea, 2015.
[59] E.H. Lee, J.M. Kim, K.H. Joo, H. Kim, Evaluation of the soil-structure interaction
effect on seismically isolated nuclear power plant structures, J. Earthq. Eng.
Soc. Korea 20 (6) (2016) 379e389.
[60] G.J. Kim, K.K. Yang, B.S. Kim, H.J. Kim, S.J. Yun, J.K. Song, Seismic response
evaluation of seismically isolated nuclear power plant structure subjected to
Gyeong-Ju earthquake, J. Earthq. Eng. Soc. Korea 20 (7) (2016) 453e460.
[61] J.W. Jung, H.W. Jang, J.H. Kim, J.W. Hong, Effect of second hardening on floor
response spectrum of a base-isolated nuclear power plant, Nucl. Eng. Des. 322
(2017) 138e147.
[62] S.G. Cho, S.M. Yun, D. Kim, K.J. Hoo, Analyses of vertical seismic responses of
seismically isolated nuclear power plant structures supported by lead rubber
bearings, J. Earthq. Eng. Soc. Korea 19 (3) (2015) 133e143.
[63] A.H.S. Ang, W.H. Tang, Probability Concepts in Engineering: Emphasis on
Applications in Civil & Environmental Engineering, vol. 1, Wiley and Sons,
2007.
[64] J.H. Kim, M.K. Kim, I.K. Choi, Experimental study on the ultimate limit state of
a lead-rubber bearing, in: ASME 2016 Pressure Vessels and Piping Conference
(pp. V008T08A039-V008T08A039), American Society of Mechanical Engi-
neers, Vancouver, BC, Canada, 2016, 2016.
[65] H.P. Lee, M.S. Cho, J.Y. Park, K.S. Jang, Assessment of LRB seismic isolation
device for nuclear power plant, in: Transactions of the Korean Nuclear Society
Autumn Meating, 2013. Gyeongju, Korea, 2013.
[66] I. Choi, J. Kim, M. Kim, Performance based design of lrb systems for nuclear
power plants, in: Transaction of the 24th Structural Mechanics in Reactor
Technology (SMiRT-24) Conference, Busan, Korea, 2017, 2017.
[67] S. Eem, D. Hahm, Large strain nonlinear model of lead rubber bearings for
beyond design basis earthquakes, Nucl. Eng. Technol. 51 (2) (2019) 600e606.
[68] S. Yabana, K. Kanazawa, S. Nagata, S. Kitamura, T. Sano, Shaking table tests
with large test specimens of seismically isolated FBR plants: Part 3 - ultimate
behavior of upper structure and rubber bearings, in: ASME 2009 Pressure
Vessels and Piping Conference, American Society of Mechanical Engineers,
2009, pp. 179e187.
[69] T. Hiraki, S. Nagata, K. Kanazawa, T. Imaoka, T. Nakayama, Y.,. Umeki,
H. Shimizu, Development of an evaluation method for seismic isolation sys-
tems of nuclear power facilities: Part 9 - ultimate properties of full-scale lead
rubber bearings based on breaking test, in: ASME 2014 Pressure Vessels and
Piping Conference (pp. V008T08A012-V008T08A012), American Society of
Mechanical Engineers, 2014.
[70] J. Zhang, Y. Huo, Optimum isolation design for highway bridges using fragility
function method, in: The 14th World Conference on Earthquake Engineering,
WCEE), Beijing, China, 2008.
[71] J.H. Lee, J.K. Song, Seismic fragility analysis of seismically isolated nuclear
power plant structures using equivalent linear-and bilinear-lead rubber
bearing model, J. Earthq. Eng. Soc. Korea 19 (5) (2015) 207e217.
[72] S.H. Eem, H.J. Jung, M.K. Kim, I.K. Choi, Seismic fragility evaluation of isolated
NPP containment structure considering soil-structure interaction effect,
J. Earthq. Eng. Soc. Korea 17 (2) (2013) 53e59.
[73] M. Shinozuka, M.Q. Feng, J. Lee, T. Naganuma, Statistical analysis of fragility
curves, J. Eng. Mech. 126 (12) (2000) 1224e1231.
[74] D. Vamvatsikos, C.A. Cornell, Incremental dynamic analysis, Earthq. Eng.
Struct. Dyn. 31 (3) (2002) 491e514.