eDirect
Nu c l e a r E n g i n e e r i n g a n d T e c h n o l o g y 4 8 ( 2 0 1 6 ) 1 2 3 7e1 2 5 1
Available online at Scienc
Nuclear Engineering and Technology
journal homepage: www.elsevier .com/locate/net
Original Article
Potentiality of Using Vertical and Three-Dimensional Isolation Systems in NuclearStructures
Zhiguang Zhou a,*, Jenna Wong b, and Stephen Mahin c
a Research Institute of Structural Engineering and Disaster Reduction, Tongji University, 1239 Siping Road, Shanghai
200092, PR Chinab Lawrence Berkeley National Laboratories, 1 Cyclotron Road, MS74R316C, Berkeley, CA 94720, USAc Department of Civil and Environmental Engineering, 777 Davis Hall, University of California, Berkeley, CA 94720,
USA
a r t i c l e i n f o
Article history:
Received 26 September 2015
Received in revised form
30 January 2016
Accepted 19 March 2016
Available online 5 April 2016
Keywords:
3D Isolation Systems
Isolator
Nuclear Power Plant
Response Spectrum
Rocking Effect
Vertical Isolation
* Corresponding author.E-mail address: [email protected] (Z.
http://dx.doi.org/10.1016/j.net.2016.03.0051738-5733/Copyright © 2016, Published by Elthe CC BY-NC-ND license (http://creativecom
a b s t r a c t
Although the horizontal component of an earthquake response can be significantly
reduced through the use of conventional seismic isolators, the vertical component of
excitation is still transmitted directly into the structure. Records from instrumented
structures, and some recent tests and analyses have actually seen increases in vertical
responses in base isolated structures under the combined effects of horizontal and
vertical ground motions. This issue becomes a great concern to facilities such as a
Nuclear Power Plants (NPP), with specialized equipment and machinery that is not only
expensive, but critical to safe operation. As such, there is considerable interest
worldwide in vertical and three-dimensional (3D) isolation systems. This paper ex-
amines several vertical and 3D isolation systems that have been proposed and their
potential application to modern nuclear facilities. In particular, a series of case study
analyses of a modern NPP model are performed to examine the benefits and challenges
associated with 3D isolation compared with horizontal isolation. It was found that
compared with the general horizontal isolators, isolators that have vertical frequencies
of no more than 3 Hz can effectively reduce the vertical in-structure responses for the
studied NPP model. Among the studied cases, the case that has a vertical isolation
frequency of 3 Hz is the one that can keep the horizontal period of the isolators as the
first period while having the most flexible vertical isolator properties. When the ver-
tical frequency of isolators reduces to 1 Hz, the rocking effect is obvious and rocking
restraining devices are necessary.
Copyright © 2016, Published by Elsevier Korea LLC on behalf of Korean Nuclear Society. This
is an open access article under the CC BY-NC-ND license (http://creativecommons.org/
licenses/by-nc-nd/4.0/).
Zhou).
sevier Korea LLC on behamons.org/licenses/by-nc
lf of Korean Nuclear Society. This is an open access article under-nd/4.0/).
Fig. 1 e Sketch of thick-rubber-layer bearing after ref [14].
Nu c l e a r E n g i n e e r i n g a n d T e c h n o l o g y 4 8 ( 2 0 1 6 ) 1 2 3 7e1 2 5 11238
1. Introduction
Conventional isolation systems are generally intended to
reduce seismic demands due to the horizontal components of
ground shaking. However, they do not prevent vertical seismic
forces from being transmitted directly into the structure.
Under certain circumstances, isolation systems can amplify
or add to the vertical vibrations similar to that experienced in
a fixed-base structure. For instance, elastomeric bearings can
have flexibilities by themselves or in combination with the
flexibility of the structural system shift the effective vertical
frequency of the isolated system into an amplified range of the
vertical pseudo-acceleration spectrum. As such, the vertical
response could beworse than that experienced by a fixed-base
structure. Sliding bearings are typically stiff in the vertical
direction, so the vertical response may be similar to a fixed-
base structure. However, if vertical excitations become large,
uplift can occur unless tension-capable bearings are used.
While moderate uplift may be acceptable in some applica-
tions, the uplift and reseating behavior of the bearing system
may result in impact loads that produce additional vertical
vibrations in the structure.
Various attempts have been made to provide enhanced
protection against the vertical component of response by: (1)
using a complete three-dimensional (3D) seismic isolation
solution; and (2) adding localized vertical isolation systems
to individual parts of a horizontally isolated structure
[1e11].
Early efforts to develop 3D isolation systems focused on
modifying the design parameters of laminated rubber bear-
ings. In 1986, Kajima Corporation (3-1, Motoakasaka 1-chome,
Minato-ku, Tokyo 107-8388, Japan) utilized this approach to
construct a two-story reinforced concrete (RC) acoustic labo-
ratory building in Japan [12]. This approach was also investi-
gated by the USA nuclear industry using laminated rubber
bearings [13]. More recently, a type of laminated thick rubber
bearing was adopted for the seismic isolation design of the
Japan sodium-cooled fast reactor (JSFR) [14].
Moving beyond design parameter modifications, other 3D
systems have been introduced. The GERB System consists of
helical springs that are flexible horizontally and vertically. This
system was used in the residential and industrial sector for
various applications. In Japan, a number of important devel-
opment studies have been completed related to 3D isolation of
nuclear facilities. A project was started in 2000 for the devel-
opment of 3D seismic isolation technologies for use in the Jap-
anese fast breeder reactors (FBR), under the sponsorship of the
Japanese Ministry of Economy, Trade and Industry. This was
motivated to achieve more economical designs for the FBR de-
signs thancouldbe achievedusing only thehorizontal isolation
systems. Threepromising ideas for 3D isolationwere examined
bytheFBRproject, i.e., “RollingSealTypeAirSpring,” “Hydraulic
3D Isolation System,” and “Cable Reinforced Air Spring” [2].
Kozo Keikaku Engineering Inc. (4-38-13 Honcho, Nakano-ku,
Tokyo164-0012, Japan) in conjunction with Shimizu Corpora-
tion (2-16-1 Kyobashi, Chuo-ku, Tokyo 104-8370, Japan) has
extended the basic ideas from these 3D isolation projects and
applied it to an actual three-story reinforced concrete apart-
ment building in Tokyo, Japan [3,4]. The 3D isolation system
installed in the building performed as expected in the 2011 East
Japan Earthquake.
Vertical isolation systems provide flexible supports in the
vertical direction by a combination of metallic or air springs
and supplemental damping devices. For example, the Euro-
pean FBR project considers isolating the reactor vault from the
horizontally isolated base mat using vertical springs [15]. For
the Japanese FBR, a vertical isolation system was explored
using a series of coned disk springs surrounding a central
vertical guide [2].
These vertical and 3D isolation systems and their potential
application to modern nuclear facilities are examined in this
paper. Moreover, a series of case study analyses of a modern
nuclear power plant (NPP) model are performed to examine
the benefits and challenges associated with 3D isolation
compared with horizontal isolation.
2. Choices of vertical and 3D isolationsystems for nuclear structures
2.1. Thick-rubber-layer bearing
3D isolation systems can be achieved by using thick rubber
layers for rubber bearings [16]. A sketch of thick rubber-layer
bearing is illustrated in Fig. 1. In 1986, Kajima Corporation
built a two-story RC acoustic laboratory building supported on
18 steel laminated natural rubber bearings [12]. The bearings
were designed to bemore flexible in the vertical direction than
other bearings used in Japan. Fourteen round steel bars were
used to provide damping. In addition, oil dampers were added
to reduce vertical and rocking motions during earthquakes.
The vertical isolation frequency is 5 Hz. The effectiveness of
the isolation system was demonstrated during both earth-
quakes and traffic vibrations.
Three dimensional isolation system using laminated rub-
ber bearings was also investigated for the USA nuclear in-
dustry [13,17]. The target horizontal and vertical frequencies
of the proposed system were 0.5 Hz and 3 Hz, respectively.
The investigation showed that rubber bearings could be
designed to provide isolation in the horizontal and vertical
directions.
Mitsubishi FBR Systems Inc. (2-34-17 Jingumae, Shibuya-
ku, Tokyo 150-0001, Japan) has proposed a thick rubber-layer
Nu c l e a r E n g i n e e r i n g a n d T e c h n o l o g y 4 8 ( 2 0 1 6 ) 1 2 3 7e1 2 5 1 1239
bearing design for JSFR recently [14]. The target horizontal and
vertical frequencies are 0.29 Hz and 8 Hz, respectively. The
diameter of the bearing is set to 1,600mm, and the bearing has
10 layers of laminated rubber with each layer being 30-mm
thick. Okamura et al. [14] concluded that this bearing design
can reduce the response of the reactor system in comparison
with the previous design for JSFR. Tests of a 1/8-reduced-scale
bearing model (diameter, 200 mm) were carried out confirm-
ing the applicability of the bearings.
2.2. The GERB system
The GERB system consists of large helical steel springs ar-
ranged in an assembly that is flexible both horizontally and
vertically, as illustrated in Fig. 2. The spring assembly is
essentially undamped and typically would be used in
conjunction with supplemental dampers. The vertical fre-
quency is approximately three to five times the horizontal
frequency [18,19]. This is to avoid excessive movement in the
vertical direction due to variations in live load, wind, or other
lateral loads. The GERB system had been implemented in two
residential buildings in California, USA before 1994. The
buildings were shaken strongly in the 1994 Northridge Earth-
quake [20]. There is a limitation to this systemdue to the strong
coupling between vertical and horizontalmotion. If the bearing
Fig. 2 e Sketch of GERB system after ref [18,19].
Rolling seal rubber
Upper basement
Air compartment
Contact part
Stoper(damper)
Lower basement
To atmosphereElastomeric-based bearing
Fig. 3 e Sketch of rolling seal ty
offsets laterally, vertical vibrations will affect the lateral
displacement of the bearing due to geometric nonlinearities.
2.3. Rolling seal type air spring
Suhara [7] developed a 3D seismic isolation device that uses
laminated rubber bearing as a horizontal isolation device and
rolling seal type air spring as a vertical isolation device, as
illustrated in Fig. 3. The air compartment was about 1.4 m in
diameter and 3 m tall. The design air pressure is 1.6 MPa for
normal operating conditions and 2.0 MPa for the earthquake
conditions. This results in a vertical frequency of 0.5 Hz. The
issues of concern for this device were the ultimate strength of
the rolling seal (rubber) against internal pressure, deteriora-
tion of the seal due to aging, and the smoothness of motion
during combined axial and lateral motion. More information
on this system can be found in Suhara's [7] study.
Because the center of mass of the FBR is higher than the
isolation plane, a plant isolatedwith 3D bearings at its basewill
tend to pitch and roll. To suppress this motion, rocking sup-
pression systems were also investigated [7]. Rocking can in-
crease the horizontal displacements and accelerations in the
upper parts of the plant, and it can also increase the vertical
displacement demands for the isolator bearings located closest
to the outer edge of the plant. Thus, a variety of systems using
mechanical pantograph linkage apparatus [8], pulley systems,
and hydraulic systems that permit vertical and lateral motion
of the upper mat without pitch or roll have been explored. The
system cited by Inoue et al. [2] as the most favorable candidate
uses a separate system of interconnected hydraulic cylinders
placed around the perimeter of the plant to control rocking.
These are interconnected in a crossover arrangement whereby
out-of-phase motion on opposite sides of the plant is sup-
pressed while in-phase motion is not restricted. Small-scale
vibration tests have demonstrated that this system can sup-
press rocking motion. More information on the rocking sup-
pression system can be found in Shimada et al. [8] 2005.
Among the 3D isolation systems examined by the FBR
project, Inoue et al. [2] indicate that the rolling seal air spring
Upper cylinder
Air supply
Lower cylinder
Normalcondition
Deformation condition
pe air spring after ref [7,8].
Fig. 5 e Sketch of hydraulic three-dimensional isolation
system after ref [22]. (A) Isolator. (B) Accumulator unit.
Nu c l e a r E n g i n e e r i n g a n d T e c h n o l o g y 4 8 ( 2 0 1 6 ) 1 2 3 7e1 2 5 11240
in combination with the rocking suppression system was the
most promising system.
2.4. Cable reinforced air spring
Cable reinforced air spring provides both horizontal and ver-
tical isolation in a single air pressure activated device [2]. The
device is composed of two cylinders of different diameters, as
illustrated in Fig. 4. The inner cylinder is attached to the lower
mat, while the upper cylinder is attached to the upper mat. A
rubber sheet interconnects the two cylinders. Polyester fabric
and a set of load carrying wire cables reinforce the sheet. The
difference between the inner diameter of the outermost cyl-
inder and the outside diameter of the inner cylinder is roughly
twice the bearing horizontal displacement capacity.
The vertical restoring stiffness is provided by the bulk
modulus of the pressurized air in the device. The horizontal
restoring force comes from the difference in air pressure
acting on the U-shaped rubber gasket between the inner and
outer cylinders. As the upper cylinder moves towards the
lower cylinder, the rubber gasket hangs down lower on the
side with the narrower gap; on the opposite side, the gasket is
stretched and thus does not hang down so far. Thus, there is
net pressure on the gasket tending to make the inner cylinder
recenter inside the outer cylinder. For a proposed design, the
inner diameter is 6 m and the outer diameter is 8 m, resulting
in a lateral displacement capacity less than 1 m. The height of
the inner cylinder is 3 m. The internal air pressure is 1.4 MPa
for operational conditions. The vertical and horizontal effec-
tive isolated frequencies are 0.35 Hz and 0.27 Hz, respectively.
Issues of concern for this bearing are the ultimate strength
of the wire reinforced bladder, verification of the vertical and
horizontal restoring force characteristics, and smooth and
predictable behavior under 3D motion. Some experiments on
the bearing have been conducted. There is more information
on this type of bearing in the Kageyama et al. [21] study.
2.5. Hydraulic 3D isolation system
A hydraulic 3D isolation system consists of a natural rubber
bearing upon which a vertically oriented hydraulic cylinder is
placed, as illustrated in Fig. 5. The hydraulic cylinder is con-
nected to a small bank of accumulators that are partially filled
with pressurized nitrogen gas. The gas is pressure regulated to
provide sufficient vertical load strength for the isolator unit
under service operating loads. During the earthquake, the gas
compresses. By providing an appropriate volume of gas, the
accumulator acts like an air spring. The piping system
Fig. 4 e Sketch of cable reinforced air spring after ref [2,21].
between the hydraulic cylinder and the accumulator bank,
and a flow-restricting orifice in the piping system, generates
the desired viscous damping. For safety reasons, a noncom-
bustible hydraulic fluid is proposed.
This system is designed for the same load as the previous
rolling seal type air spring system, but the operating pressure
under service operations is raised to 15 MPa and under
earthquake conditions is raised to 20 MPa. The issues of
concern cited for this system were the leak-tightness of the
seal system between the piston and cylinder, friction char-
acteristics of the piston cylinder system with lateral loads
applied, and the confidence that the desired damping prop-
erties could be achieved. Some reduced scale tests were car-
ried out on this system. There is more information on this
system in the Kashiwazaki et al. [22] study.
2.6. Coned disk springs
Providing vertical isolation for individual components within
a horizontally isolated NPP against the vertical components of
Nu c l e a r E n g i n e e r i n g a n d T e c h n o l o g y 4 8 ( 2 0 1 6 ) 1 2 3 7e1 2 5 1 1241
motion is an intermediate and perhaps more economical
approach to achieve 3D protection. In the Ministry of Econ-
omy, Trade and Industry program for 3D isolation, the concept
of a common deck isolation system was proposed. In this
system, a flat slab platform is constructed within the reactor
containment structure. It is assumed that the structure is
isolated to resist horizontal components of ground motion.
The common deck is supported on a series of vertical isolation
devices (lateral displacements of the deck relative to the plant
are restrained). The reactor vessel can be suspended through a
hole in the deck, and other components sensitive to vertical
motion can be supported on the deck. It is expected that the
isolation plane would be located at the elevation of the center
of mass of the items supported on the deck so that rocking
would not be an issue, and that the isolators would be posi-
tioned to avoid mass eccentricities in the vertical direction.
For the vertical isolation systems, a different approach
than what was considered for the overall plant was assumed.
In this case, stacks of large bore and coned disk springs (Fig. 6)
were considered [5,6]. For the cases cited, the springs were
configured to provide a vertical isolated frequency of 1 Hz.
This was selected to provide a more limited relative vertical
displacement movement for piping systems in the plant. The
spring units have an outer diameter of 1 m and an inner
diameter of 0.5 m. They are 27-mm thick and bent into a cone
shape. Seventy cone disks are used to make up one isolator;
five parallel stacks of 14 disks in series. These stacked cones
result in a total height of 2.2 mwith a stroke limit of ±100mm.
Damping is provided by separate steel yielding dampers. Dy-
namic analysis suggests that the vertical response is similar to
that observed for horizontal isolation; the spectral response
for frequencies higher than that of the isolation system are
lower than for the fixed base system, and the spectral
response around the isolation frequency is higher than for the
fixed base system. Inoue et al. [2] and Morishita et al. [5]
concluded that this system is effective.
Fujita et al. [23] carried out a set of shaking table tests with
small diameter natural rubber bearings providing isolation in
the horizontal direction. On top of these bearings, a set of
coned disk springs and yielding dampers were used to provide
full 3D isolation. The specimen had a horizontal isolated fre-
quency of 0.5 Hz and a vertical isolated frequency of 3 Hz. The
later fixed base frequency of the super structurewas 10Hz and
Damper
Common deck
Coned disk spring
Inside middle washer
Outside middle washerCenter guide
Fig. 6 e Sketch of coned disk spring after ref [5,6,23].
above 50 Hz in the horizontal and vertical directions, respec-
tively. Several levels of excitation were imposed. The results
indicated that the system worked as expected, with little
rocking. For a large amplitude El Centro motion, it is noted
that the floor spectral ordinates in the vertical direction are
significantly reduced for periods less than 0.15 seconds (6 Hz)
and much lower than the fixed base one at the natural fre-
quency of the fixed base building. Larger reductions were
obtained in the horizontal direction. Although there was a
consistent reduction in floor spectra in the vertical direction in
the high frequency of response range when the excitations
were moderate and large, it was noted that there was minor
amplification of response in the vertical direction for low
amplitudes of excitation. A summary of various vertical and
3D isolation systems is illustrated in Table 1.
3. Analysis results using different verticaland 3D isolation systems
3.1. Model summary
A representative but simplified stick lumped mass model is
introduced to represent a Generation III Pressurized Water
Reactor NPP in the case study analyses. Linear modal time
history analyses, conducted using SAP2000 (Computers and
Structures, Inc. 1646 N. California Blvd., Suite 600 Walnut
Creek, CA 94596, USA), are used to examine the effect of
different 3D isolation systems compared with a general hori-
zontal isolation system in reducing the floor response spectra
in the vertical direction. The 3D finite element model is illus-
trated in Fig. 7. The Nuclear Island consists of the reactor
containment building (RCB) and the auxiliary building, which
share a common basemat. The shape of the auxiliary building
is a rectangular type with maximum dimensions of
103.6m� 102.4m. It wraps around the RCBwith a gap equal to
5 cm and the radius of circular cavity for the RCB is 24 m. The
RCB consists of the containment shell (CS), the reactor
containment internal structures, and the reactor coolant
systems. The first horizontal frequency and the first vertical
frequency of the CS are 3.85 Hz and 11.02 Hz, respectively. The
first horizontal frequency and the first vertical frequency of
the primary shield wall (representative of the internal struc-
tures) are 8.13 Hz and 29.65 Hz, respectively.
There are 454 isolator elements in the model, as illustrated
in Fig. 7B. The input motion is illustrated in Fig. 8. It is
generated synthetically to be consistent with European Utility
Requirements for light water reactor NPPs, and is scaled to
have 0.5 g horizontal and 0.33 g vertical peak ground
accelerations.
The isolator properties used in the analyses are illustrated
in Table 2, where Case 0 is the fixed-base case; Case 1 uses
general horizontal isolation systems; and Cases 2e6 use
some form of 3D isolation. The vertical frequency of the
isolation system (fv) varies from 63.2 Hz to 0.432 Hz, while
the horizontal frequency remains constant at 0.432 Hz. The
damping for the response history analysis is defined as a
constant value of 0.04 over all the modes except for the
modes of isolators which are overwritten with the values
listed in Table 2.
Table
1e
Sum
mary
ofvariousverticalandth
ree-d
imensional(3D)isolationsy
stem
s.
Isolationsy
stem
Relatedrese
arch&
developmentefforts
Verticalfrequency
Horizo
ntal
freq
uency
Obse
rvations
Thickru
bber-layerbearing
[14]
2-story
RCaco
ustic
laboratory
buildingbyKajimain
1986;
Japanso
dium-cooledfast
reactor
3e8Hz
0.3e0.5
Hz
Effectiveness
demonstratedin
experiments
TheGERBSystem
[18,19]
2residentialbuildingsbuiltin
Californ
ia3e5timesth
e
horizo
ntalfrequency
eShakenstro
ngly
inth
e1994NorthridgeEarthquake,
stro
ngco
uplingbetw
eenvertical&
horizo
ntalmotion
Rollingse
altypeair
spring
[7,8]
Developmentofse
ismic
isolationsy
stem
for
nextgenerationNPPin
Japan
0.5
Hz
0.36Hz
Effectiveifco
mbinedwithro
ckingsu
ppressionsy
stem
Cable
reinforcedair
spring
[2,21]
0.35Hz
0.27Hz
Issu
es:
theultim
ate
strength
ofth
ewirereinforcedbladder,
etc.
Hydraulic3D
isolation
system
[22]
0.5e0.67Hz
0.54Hz
Issu
es:
leak-tightn
ess
ofth
ese
alsy
stem,etc.
Coneddisksp
rings[5,6,23]
METIpro
gram
for3D
isolationin
Japan
1e3Hz
0.5
Hz
Effective(m
inoramplifica
tionofresp
onse
inth
e
verticaldirectionforlow
amplitu
desofexcitation).
METI,MinisterofEco
nomy,TradeandIndustry;NPP,nuclearpowerplant;RC,reactorco
ntainment.
Nu c l e a r E n g i n e e r i n g a n d T e c h n o l o g y 4 8 ( 2 0 1 6 ) 1 2 3 7e1 2 5 11242
Modal frequencies for the different cases are illustrated
in Table 3. For the fixed-base case, the first two modes
correspond to the horizontal translation modes of the RCB.
The third mode corresponds to the horizontal translation
mode of the auxiliary building, and the 10th mode corre-
sponds to the vertical translation mode of the RCB. From
Case 1 to Case 4 (i.e., fv varies from 63.2 Hz to 3 Hz), the first
mode corresponds to the horizontal translation mode of
the isolators. However, the first mode shows a complex
characteristic of both horizontal translation and rocking
for Case 5 and Case 6. It means that rocking plays an
important role in the first mode when fv reduces to 1 Hz.
The first vertical translation mode for the different cases is
marked as shaded cell in Table 3. The modal direction is
also shown in Table 3 when the modal participating mass
ratio is greater than 1% in that direction. It can be found
that the first vertical translation mode is controlled by the
vertical stiffness of the isolators for Cases 2e6. Among
these cases, Case 4 maintains the horizontal period of the
isolators as the fundamental period while having the most
flexible vertical isolator properties.
3.2. Response of the isolators
As there are 454 isolator elements in the model, it is infea-
sible to illustrate the response results of all the isolators.
Therefore, the response of the center isolator is discussed in
this section. Response horizontal and vertical deformation
comparison of the center isolator among the different cases
are illustrated in Figs. 9Ae9D, respectively. It can be seen that
response horizontal orbits of the isolator are almost the
same for the caseswhen fv is between 63.2 Hz and 3 Hz, while
are different for the cases when fv is between 3 Hz and
0.432 Hz. As to the vertical deformation of the isolator, the
amplitude of the deformation increases with the decreasing
of fv.
Response isolator axial force comparison of the center
isolator among the different cases is illustrated in Figs. 10A
and 10B, while the shear force X and the shear force Y are
illustrated in Figs. 10C and 10D and Figs. 10E and 10F,
respectively. It can be seen that the isolator axial force shows
limited difference for the cases when fv is between 63.2 Hz
and 3 Hz, while it shows an obvious difference for the cases
when fv is between 3 Hz and 0.432 Hz. The isolator shear
forces are almost the same for the cases that fv is between
63.2 Hz and 3 Hz, while decrease with the value of fv for the
cases that fv is between 3 Hz and 0.432 Hz.
3.3. Response spectral accelerations of thesuperstructure
Vertical pseudo spectral accelerations of the fixed-base case
are illustrated in Fig. 11. The unit of pseudo spectral accel-
erations is set as g, the acceleration of gravity. It can be found
that the in-structure response for the CS near 11 Hz is highly
amplified, and for the Primary Shield Wall near 30 Hz is
amplified by the input motion. Referring to the modal fre-
quencies for the different cases, it is clear that the floor
spectra are amplified in the vertical direction at frequencies
corresponding to the vertical direction modes of the
Layout of the isolators
(A) (B)
3D View of the FE model
Top of the reactor containment building(EL. 101 m) Top of the primary shield
wall (EL. 58.2 m)
Bottom of the reactor containment building (EL. 23.8 m)
Fig. 7 e Three-dimensional (3D) finite element (FE) model. (A) 3D view of the FE model. (B) Layout of the isolators.
EL., elevation level.
0.60.40.2
00.20.40.6
0 5 10 15 20 25
Acc
el. (
g)
Time (s)
0.60.40.2
00.20.40.6
0 5 10 15 20 25
Acc
el. (
g)
Time (s)
0.40.30.20.1
00.10.20.30.4
0 5 10 15 20 25
Acc
el. (
g)
Time (s)
––– –
–– –
–––
(A) (B) (C)
Fig. 8 e Input motion. (A) X component. (B) Y component. (C) Z component. Accel., acceleration.
Table 2 e Isolator properties used in the analyses.
Caseno.
Supposed isolation system Vertical properties Horizontal properties
Frequency fv(Hz)
Stiffness kv(kN/m)
Dampingxv
Frequency fh(Hz)
Stiffness kh(kN/m)
Dampingxh
Case 0 Fixed base
Case 1 Lead rubber bearing 63.2 1.68E þ 08 2% 0.432 7.62E þ 03 20%
Case 2 Thick rubber-layer bearing 8 2.70E þ 06 4%
Case 3 Thick rubber-layer bearing 5 1.05E þ 06 10%
Case 4 Coned disk spring/GERB system þ LRB 3 3.79E þ 05 20%
Case 5 Air spring/hydraulic system þ LRB 1 4.21E þ 04 20%
Case 6 0.432 7.62E þ 03 20%
LRB, Lead-plug Rubber Bearing.
Nu c l e a r E n g i n e e r i n g a n d T e c h n o l o g y 4 8 ( 2 0 1 6 ) 1 2 3 7e1 2 5 1 1243
Containment Shell and the Primary ShieldWall. As illustrated
in Fig. 12, vertical floor spectral accelerations of the Contain-
ment Shell are amplified for the cases when fv is between
63.2 Hz and 5 Hz, reduced for the case when fv is 3 Hz at the
frequency of 5~20Hz, and reduced for the case when fv is no
more than 1 Hz over a broad frequency range. Comparison of
vertical floor response spectra at three specific positions for
the different cases is illustrated in Fig. 13. It can be seen that
the vertical floor response spectra have little difference for the
fixed-base case and the case when fv is 63.2 Hz. At the top of
Table 3 e Modal frequencies for the different cases.
Mode no. 1 2 3 4 5 6 7 8 9 10 11
Case 0 Freq. (Hz) 3.850 3.852 5.926 6.282 7.373 8.382 9.444 9.915 10.706 11.020 11.617
direction UYa UX UY UX RZ RZ UX RZ UY UZ
Case 1 Freq. (Hz) 0.432 0.432 0.445 3.677 3.691 7.747 7.869 7.989 9.577 10.023 10.690
direction UY UX RZ RX RY RX RY RY UZ
Case 2 Freq. (Hz) 0.432 0.432 0.445 3.305 3.410 6.503 6.888 7.878 7.971 8.563 9.143
direction UY UX RZ RX RY RX RY UZ RY RX
Case 3 Freq. (Hz) 0.431 0.431 0.445 3.083 3.219 4.407 4.793 4.855 7.876 8.146 8.362
direction UY UX RZ RX RY RY RX UZ RX
Case 4 Freq. (Hz) 0.430 0.430 0.445 2.432 2.613 3.005 3.520 3.647 7.876 8.039 8.194
direction UY UX RZ RX RY UZ RX RY
Case 5 Freq. (Hz) 0.407 0.412 0.445 0.908 0.995 1.018 3.302 3.451 7.875 7.961 8.118
direction UY UX RZ RY RX UZ
Case 6 Freq. (Hz) 0.288 0.307 0.440 0.444 0.555 0.583 3.288 3.441 7.875 7.950 8.111
direction RX UX UZ RZ UX RX
a RX, RY, and RZmean rotation in X, Y, and Z direction, respectively. UX, UY, and UZmean translation in X, Y, and Z direction, respectively. The
direction is shown only when the modal participating mass ratio is > 1%.
Freq., frequency.
Shaded cell: the first vertical translation frequency.
0.1
0.05
0
0.05
0.1
0.1 0.05 0 0.05 0.1
Isol
ator
def
orm
atio
n Y
(m)
Isolator deformation X (m)
fv = 63.2 Hz fv = 8 Hzfv = 5 Hz fv = 3 Hz
0.1
0.05
0
0.05
0.1
0.1 0.05 0 0.05 0.1
Isol
ator
def
orm
atio
n Y
(m)
Isolator deformation X (m)
fv = 3 Hz fv = 1 Hz fv = 0.432 Hz
0.0120.0090.0060.003
00.0030.0060.0090.012
0 5 10 15 20
Isol
ator
def
orm
atio
n Z
(m)
Time (s)
fv = 63.2 Hz fv = 8 Hzfv = 5 Hz fv = 3 Hz
0.10.080.060.040.02
00.020.040.060.080.1
0 5 10 15 20
Isol
ator
def
orm
atio
n Z
(m)
Time (s)
fv = 3 Hz fv = 1 Hz fv = 0.432 Hz
– –
–– –
–– –
––
––––
–––
(A) (B)
(C) (D)
Fig. 9 e Response deformation comparison of the center isolator among the different cases. (A) Horizontal deformation for
Case 1e4. (B) Horizontal deformation for Case 4e6. (C) Vertical deformation for Case 1e4. (D) Vertical deformation for Case
4e6. fv, vertical frequency of the isolation system.
Nu c l e a r E n g i n e e r i n g a n d T e c h n o l o g y 4 8 ( 2 0 1 6 ) 1 2 3 7e1 2 5 11244
the CS, the case when fv is 63.2 Hz has the maximum vertical
floor response spectra near 11 Hz for the input motion. At the
top of the primary shield wall, the case when fv is 63.2 Hz has
the maximum vertical floor response spectra near 20 Hz,
while other cases take the local maximum spectra at other
frequencies. At the bottom of the CS, the case when fv is 3 Hz
has the maximum vertical floor response spectra near 3 Hz.
Considering that the natural vertical frequency of reactor
6.0E + 6
4.0E + 6
2.0E + 6
0.0E + 0
2.0E + 6
4.0E + 6
6.0E + 6
0 5 10 15 20
Isol
ator
axi
al fo
rce
(N)
Time (s)
fv = 63.2 Hz fv = 8 Hz fv = 5 Hz fv = 3 Hz
6.0E + 6
4.0E + 6
2.0E + 6
0.0E + 0
2.0E + 6
4.0E + 6
6.0E + 6
0 5 10 15 20
Isol
ator
axi
al fo
rce
(N)
Time (s)
fv = 3 Hz fv = 1 Hz fv = 0.432 Hz
8.0E + 56.0E + 54.0E + 52.0E + 50.0E + 02.0E + 54.0E + 56.0E + 58.0E + 5
0 5 10 15 20
Isol
ator
shea
r for
ce X
(N)
Time (s)
fv = 63.2 Hz fv = 8 Hz fv = 5 Hz fv = 3 Hz
8.0E + 56.0E + 54.0E + 52.0E + 50.0E + 02.0E + 54.0E + 56.0E + 58.0E + 5
0 5 10 15 20
Isol
ator
shea
r for
ce X
(N)
Time (s)
fv = 3 Hz fv = 1 Hz fv = 0.432 Hz
8.0E + 56.0E + 5
4.0E + 52.0E + 50.0E + 02.0E + 54.0E + 56.0E + 58.0E + 5
0 5 10 15 20
Isol
ator
shea
r for
ce Y
(N)
Time (s)
fv = 63.2 Hz fv = 8 Hz fv = 5 Hz fv = 3 Hz
8.0E + 56.0E + 54.0E + 52.0E + 50.0E + 02.0E + 54.0E + 56.0E + 58.0E + 5
0 5 10 15 20
Isol
ator
shea
r for
ce Y
(N)
Time (s)
fv = 3 Hz fv = 1 Hz fv = 0.432 Hz
– –
––
– –
–
–
–
–
–––
–––
–––
–––
(A) (B)
(C) (D)
(E) (F)
Fig. 10 e Response force comparison of the center isolator among the different cases. (A) Isolator axial force for Case 1e4. (B)
Isolator axial force for Case 4e6. (C) Isolator shear force X for Case 1e4. (D) Isolator shear force X for Case 4e6. (E) Isolator
shear force Y for Case 1e4. (F) Isolator shear force Y for Case 4e6. fv, vertical frequency of the isolation system.
Nu c l e a r E n g i n e e r i n g a n d T e c h n o l o g y 4 8 ( 2 0 1 6 ) 1 2 3 7e1 2 5 1 1245
systems is about 10 Hz, it can be concluded that vertical floor
response spectra can be effectively reduced for the cases
when fv is no more than 3 Hz.
Although the input motion has X component and Y
component in a horizontal direction, the horizontal floor
response spectra in X direction can present the characteris-
tics of the response in a horizontal direction and was chosen
as the output variable for the horizontal responses. Hori-
zontal floor response spectral accelerations in X direction of
the fixed-base case are illustrated in Fig. 14. Similar to the
vertical direction, it can be found that the floor spectra are
amplified in the horizontal direction at frequencies
corresponding to the horizontal direction modes of the CS
and the primary shield wall. For the cases with isolators,
there is a substantial reduction of the horizontal floor spectra
over a broad frequency range, as illustrated in Fig. 15. A
comparison of horizontal floor response spectra at three
specific positions for the different cases is illustrated in
Fig. 16. Horizontal floor response spectral accelerations in X
direction of the CS increase with elevation near 11 Hz for the
cases when fv is 63.2 Hz and 8 Hz, while this trend tends to
disappear for other cases. This is considered to be the
coupling effect between the vertical and the horizontal re-
sponses for the cases when fv is 63.2 Hz and 8 Hz, while the
0
2
4
6
8
10
12
0.1 1 10 100
PSA
(g)
Frequency (Hz)
Input motion EL.101 m EL.66.7 mEL.47.5 m EL.34.7 m EL.23.8 m
00.20.40.60.8
11.21.41.61.8
2
0.1 1 10 100
PSA
(g)
Frequency (Hz)
Input motion EL.58.2 m EL.47.5 mEL.34.7 m EL.23.8 m
(A) (B)
Fig. 11 e Vertical floor response spectral accelerations of the fixed-base model. (A) Containment shell. (B) Primary shield
wall. EL., elevation level; PSA, pseudo spectral accelerations.
Nu c l e a r E n g i n e e r i n g a n d T e c h n o l o g y 4 8 ( 2 0 1 6 ) 1 2 3 7e1 2 5 11246
vertical response is mitigated for the cases when fv is < 5 Hz.
One interesting point is that for the case when fv is 1 Hz, the
horizontal floor spectra near 0.432e0.8 Hz (frequency of the
isolators) tends to increase with the elevation and the trend
changes at other frequencies. For the case when fv is
0.432 Hz, the horizontal floor spectra near 0.288 Hz (the
0
2
4
6
8
10
12
0.1 1 10 100
PSA
(g)
Frequency (Hz)
Input motion EL.101 mEL.66.7 m EL.47.5 mEL.34.7 m EL.23.8 m
00.5
11.5
22.5
33.5
44.5
5
0.1 1
PSA
(g)
Frequ
Input motionEL.66.7 mEL.34.7 m
0
0.5
1
1.5
2
2.5
0.1 1 10 100
PSA
(g)
Frequency (Hz)
Input motion EL.101 mEL.66.7 m EL.47.5 mEL.34.7 m EL.23.8 m
00.10.20.30.40.50.60.70.80.9
1
0.1 1
PSA
(g)
Freque
Input motionEL.66.7 mEL.34.7 m
(A) (B)
(D) (E)
Fig. 12 e Vertical floor response spectral accelerations of the co
fv¼ 8 Hz. (C) fv¼ 5 Hz. (D) fv¼ 3 Hz. (E) fv¼ 1 Hz. (F) fv¼ 0.432 H
system; PSA, pseudo spectral accelerations.
natural frequency of that case) tends to increase with the
elevation, while the horizontal floor spectra decreases first
and then increases as the elevation increases at the fre-
quencies greater than 0.5 Hz. This is due to the rocking effect
of the superstructures. Thus, rocking restraining devices are
necessary for such cases.
10 100ency (Hz)
EL.101 mEL.47.5 mEL.23.8 m
0
0.5
1
1.5
2
2.5
3
0.1 1 10 100
PSA
(g)
Frequency (Hz)
Input motion EL.101 mEL.66.7 m EL.47.5 mEL.34.7 m EL.23.8 m
10 100ncy (Hz)
EL.101 mEL.47.5 mEL.23.8 m
00.10.20.30.40.50.60.70.80.9
1
0.1 1 10 100
PSA
(g)
Frequency (Hz)
Input motion EL.101 mEL.66.7 m EL.47.5 mEL.34.7 m EL.23.8 m
(C)
(F)
ntainment shell for the isolated model. (A) fv¼ 63.2 Hz. (B)
z. EL., elevation level; fv, vertical frequency of the isolation
0
2
4
6
8
10
12
0.1 1 10 100
PSA
(g)
Frequency (Hz)
Input motion fv = 63.2 Hzfv = 8 Hz fv = 5 Hzfv = 3 Hz fv = 1 Hzfv = 0.432 Hz Fixed base
0
0.5
1
1.5
2
2.5
0.1 1 10 100PS
A (g
)Frequency (Hz)
Input motion fv = 63.2 Hzfv = 8 Hz fv = 5 Hzfv = 3 Hz fv = 1 Hzfv = 0.432 Hz Fixed base
0
0.5
1
1.5
2
2.5
0.1 1 10 100
PSA
(g)
Frequency (Hz)
Input motion fv = 63.2 Hzfv = 8 Hz fv = 5 Hzfv = 3 Hz fv = 1 Hzfv = 0.432 Hz Fixed base
(A) (B) (C)
Fig. 13 e Comparison of vertical floor response spectral accelerations for the different cases. (A) At the top of the
containment wall. (B) At the bottom of the containment wall. (C) At the top of the primary shield wall. fv, vertical frequency of
the isolation system; PSA, pseudo spectral accelerations.
0
2
4
6
8
10
12
0.1 1 10 100
PSA
(g)
Frequency (Hz)
Input motion EL.101 m EL.66.7 mEL.47.5 m EL.34.7 m EL.23.8 m
02468
101214161820
0.1 1 10 100
PSA
(g)
Frequency (Hz)
Input motion EL.58.2 m EL.47.5 mEL.34.7 m EL.23.8 m
(A) (B)
Fig. 14 e Horizontal floor response spectral accelerations in X direction of the fixed-base model. (A) Containment shell. (B)
Primary shield wall. EL., elevation level; PSA, pseudo spectral accelerations.
Nu c l e a r E n g i n e e r i n g a n d T e c h n o l o g y 4 8 ( 2 0 1 6 ) 1 2 3 7e1 2 5 1 1247
3.4. Response displacements and spectral displacementsof the superstructure
Comparisons of response displacements at the top and at the
bottom of the CS are illustrated in Figs. 17 and 18, respectively,
for the different cases. Response displacement difference be-
tween the top and the bottom of the CS is illustrated in Fig. 19.
For the caseswhen fv is between 63.2 Hz and 3 Hz, the response
displacement of the superstructures is in translationmode. For
the caseswhen fv is 1Hzor 0.432Hz, the responsedisplacement
of the superstructures shows a rocking mode, and the vertical
response displacement has a larger amplitude while the
displacement difference between the top and the bottom has a
smaller amplitude compared with the other cases. It is in
agreement with the isolation characteristics in the vertical di-
rection for the different cases. Comparing the response dis-
placements for thevertical andhorizontal directionsat thebase
of the structure, it suggests that the vertical displacements of
the bearings are as large or even larger than the horizontal
displacements in the case where fv is 0.432 Hz.
Comparisons of response spectral displacements at the top
and at the bottom of the CS are illustrated in Fig. 20 and in
Fig. 21, respectively, for the different cases. In the horizontal
direction, the response spectral displacement is reduced at
the frequencies over 0.7 Hz, is amplified at frequencies less
than 0.7 Hz, and shows different trends at the top and at the
bottom for the different cases because of the rocking effect. In
the vertical direction, the response spectral displacement is
amplified near the frequency fv, but is negligible at the inter-
ested frequencies over 10 Hz.
4. Discussion
Verticaland3Disolationsystemsandtheirpotentialapplication
to modern nuclear facilities are investigated in this paper. Two
approaches to isolate vertical groundmotion canbe selected for
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
0.1 1 10 100
PSA
(g)
Frequency (Hz)
Input motion EL.101 mEL.66.7 m EL.47.5 mEL.34.7 m EL.23.8 m
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
0.1 1 10 100PS
A (g
)Frequency (Hz)
Input motion EL.101 mEL.66.7 m EL.47.5 mEL.34.7 m EL.23.8 m
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
0.1 1 10 100
PSA
(g)
Frequency (Hz)
Input motion EL.101 mEL.66.7 m EL.47.5 mEL.34.7 m EL.23.8 m
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
0.1 1 10 100
PSA
(g)
Frequency (Hz)
Input motion EL.101 mEL.66.7 m EL.47.5 mEL.34.7 m EL.23.8 m
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
0.1 1 10 100
PSA
(g)
Frequency (Hz)
Input motion EL.101 mEL.66.7 m EL.47.5 mEL.34.7 m EL.23.8 m
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
0.1 1 10 100
PSA
(g)
Frequency (Hz)
Input motion EL.101 mEL.66.7 m EL.47.5 mEL.34.7 m EL.23.8 m
(A) (B) (C)
(D) (E) (F)
Fig. 15 e Horizontal floor response spectral accelerations in X direction of the containment shell for the isolated model. (A)
Vertical frequency of the isolation system (fv )¼ 63.2 Hz. (B) fv¼ 8 Hz. (C) fv¼ 5 Hz. (D) fv¼ 3 Hz. (E) fv¼ 1 Hz. (F) fv¼ 0.432 Hz.
EL., elevation level; PSA, pseudo spectral accelerations.
0
0.1
0.2
0.3
0.4
0.5
0.6
0.1 1 10 100
PSA
(g)
Frequency (Hz)
Input motion fv = 63.2 Hzfv = 8 Hz fv = 5 Hzfv = 3 Hz fv = 1 Hzfv = 0.432 Hz Fixed base
0
0.1
0.2
0.3
0.4
0.5
0.6
0.1 1 10 100
PSA
(g)
Frequency (Hz)
Input motion fv = 63.2 Hzfv = 8 Hz fv = 5 Hzfv = 3 Hz fv = 1 Hzfv = 0.432 Hz Fixed base
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
0.1 1 10 100
PSA
(g)
Frequency (Hz)
Input motion fv = 63.2 Hzfv = 8 Hz fv = 5 Hzfv = 3 Hz fv = 1 Hzfv = 0.432 Hz Fixed base
(A) (B) (C)
Fig. 16 e Comparison of horizontal floor response spectral accelerations in X direction for the different cases. (A) At the top of
the containment shell. (B) At the bottom of the containment shell. (C) At the top of the primary shield wall. fv, vertical
frequency of the isolation system; PSA, pseudo spectral accelerations.
Nu c l e a r E n g i n e e r i n g a n d T e c h n o l o g y 4 8 ( 2 0 1 6 ) 1 2 3 7e1 2 5 11248
0.20.15
0.10.05
00.05
0.10.15
0.2
0 5 10 15 20
Hor
izon
tal d
ispl
acem
ent (
m)
Time (s)
fv = 63.2 Hz fv = 8 Hz fv = 5 Hzfv = 3 Hz fv = 1 Hz fv = 0.432 HzFixed base
0.120.1
0.080.060.040.02
00.020.040.060.08
0.1
0 5 10 15 20
Ver
tical
dis
plac
emen
t (m
)
Time (s)
fv = 63.2 Hz fv = 8 Hz fv = 5 Hzfv = 3 Hz fv = 1 Hz fv = 0.432 HzFixed base
– ––
––––
––
–
(A) (B)
Fig. 17 e Comparison of response displacement at the top of the containment shell. (A) Horizontal (X direction). (B) Vertical.
fv, vertical frequency of the isolation system.
0.120.1
0.080.060.040.02
00.020.040.060.08
0.1
0 5 10 15 20
Hor
izon
tal d
ispl
acem
ent (
m)
Time (s)
fv = 63.2 Hz fv = 8 Hz fv = 5 Hzfv = 3 Hz fv = 1 Hz fv = 0.432 HzFixed base
0.120.1
0.080.060.040.02
00.020.040.060.080.1
0 5 10 15 20
Ver
tical
dis
plac
emen
t (m
)
Time (s)
fv = 63.2 Hz fv = 8 Hz fv = 5 Hzfv = 3 Hz fv = 1 Hz fv = 0.432 HzFixed base
– ––
––––
–––––
(B)(A)
Fig. 18 e Comparison of response displacement at the bottom of the containment shell. (A) Horizontal (X direction).
(B) Vertical. fv, vertical frequency of the isolation system.
0.15
0.1
0.05
0
0.05
0.1
0.15
0 5 10 15 20
Hor
izon
tal d
ispl
acem
ent (
m)
Time (s)
fv = 63.2 Hz fv = 8 Hz fv = 5 Hzfv = 3 Hz fv = 1 Hz fv = 0.432 Hz
0.0050.0040.0030.0020.001
00.0010.0020.0030.004
0 5 10 15 20
Ver
tical
dis
plac
emen
t (m
)
Time (s)
fv = 63.2 Hz fv = 8 Hz fv = 5 Hzfv = 3 Hz fv = 1 Hz fv = 0.432 Hz
– –––––
–
–
(A) (B)
Fig. 19 e Response displacement difference between the top and the bottom of the containment shell. (A) Horizontal
(X direction). (B) Vertical. fv, vertical frequency of the isolation system.
Nu c l e a r E n g i n e e r i n g a n d T e c h n o l o g y 4 8 ( 2 0 1 6 ) 1 2 3 7e1 2 5 1 1249
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.1 1 10 100Hor
izon
tal s
pect
ral d
ispl
acem
ent (
m)
Frequency (Hz)
Input motion fv = 63.2 Hz fv = 8 Hzfv = 5 Hz fv = 3 Hz fv = 1 Hzfv = 0.432 Hz Fixed base
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.1 1 10 100
Ver
tical
spec
tral d
ispl
acem
ent (
m)
Frequency (Hz)
Input motion fv = 63.2 Hz fv = 8 Hzfv = 5 Hz fv = 3 Hz fv = 1 Hzfv = 0.432 Hz Fixed base
(A) (B)
Fig. 20 e Comparison of response spectral displacements at the top of the containment shell. (A) Horizontal (X direction). (B)
Vertical. fv, vertical frequency of the isolation system.
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.1 1 10 100
Hor
izon
tal s
pect
ral d
ispl
acem
ent (
m)
Frequency (Hz)
Input motion fv = 63.2 Hz fv = 8 Hzfv = 5 Hz fv = 3 Hz fv = 1 Hzfv = 0.432 Hz Fixed base
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.1 1 10 100
Ver
tical
spec
tral d
ispl
acem
ent (
m)
Frequency (Hz)
Input motion fv = 63.2 Hz fv = 8 Hzfv =5 Hz fv = 3 Hz fv = 1 Hzfv = 0.432 Hz Fixed base
(A) (B)
Fig. 21 e Comparison of response spectral displacements at the bottom of containment shell. (A) Horizontal. (B) Vertical. fv,
vertical frequency of the isolation system.
Nu c l e a r E n g i n e e r i n g a n d T e c h n o l o g y 4 8 ( 2 0 1 6 ) 1 2 3 7e1 2 5 11250
consideration for the next generation nuclear power plant. One
is where the entire building utilizes 3D base isolation technol-
ogy, and the other iswhere vertical isolation is provided for key
components within a horizontally isolated building. While no
NPPhasbeenconstructedusingverticalor3Disolationsystems,
there have beenmany efforts to develop them.
Moreover, a series of case study analyses of a Generation
III pressurized water reactor NPP model are performed to
examine the benefits and challenges associated with 3D
isolation compared with horizontal isolation. It is found that:
(1) the vertical response spectrum of the superstructure is
amplified near the frequency of the vertical isolation system.
But compared with the general horizontal isolators, isolators
that have vertical frequency fv of no more than 3 Hz can
effectively reduce the vertical in-structure responses for the
studied NPP model.; (2) among the studied cases, the case
when fv is 3 Hz is the one that can keep the horizontal period
of the isolators as the first period while having the most
flexible vertical isolator properties. For the cases when fv is
between 63.2 Hz and 3 Hz, the response displacement of the
superstructures is in translation mode. Thus, a 3D isolation
system which has a vertical frequency of about 3 Hz is
feasible for the studied NPP model; (3) Comparing the
response displacement for the vertical and horizontal di-
rections at the base of the structure suggests the vertical
bearing displacements are as large as, or even larger than,
the horizontal displacements in the case fv being 0.432 Hz;
and (4) when the vertical frequency of isolators reduces to
1 Hz, the rocking effect is obvious and maximum horizontal
in-structure responses can decrease first and then increase
with increasing elevation. The fundamental frequency of the
system is no longer controlled by the horizontal stiffness of
isolators. Rocking restraining devices will be required for
such cases.
For structures containing equipment and other items
sensitive to accelerations, such as in NPPs, considerable
attention should be placed on bearing stability and the hori-
zontalevertical coupling behavior that can occur in different
isolation systems to a varying degree. Linear analysis was
used in this study; however, results show that future research
Nu c l e a r E n g i n e e r i n g a n d T e c h n o l o g y 4 8 ( 2 0 1 6 ) 1 2 3 7e1 2 5 1 1251
on this subject will benefit from nonlinear analysis and an
advanced bearing model that can account for the coupling
behavior. Moreover, European Utility Requirements motions
are used for this study, but analysis considering long-period
motions or other ground motion characteristics such as
those described in Nuclear Regulatory Commission RG1.60 is
expected for future work.
Conflicts of interest
All authors have no conflicts of interest to declare.
Acknowledgments
The authors wish to gratefully acknowledge the support for
thiswork by theNational Natural Science Foundation of China
under Grant Numbers 51478357 and 51308418, National Sci-
ence and Technology Pillar ProgramNumber 2015BAK17B04, a
cooperative research program funded by KEPCO Engineering
and Construction Company Inc., and a research program
funded by the Electric Power Research Institute.
r e f e r e n c e s
[1] J. Wong, Z. Zhou, S. Mahin, Seismic Isolation of NuclearPower Plants, Report No. 3002000561, Electric Power ResearchInstitute Inc., Palo Alto, CA, 2013.
[2] K. Inoue, M. Fushimi, S. Moro, M. Morishita, S. Kitamura,T. Fujita, Development of three-dimensional seismicisolation system for next generation nuclear power plant, in:Proceedings of the 13th World Conference on EarthquakeEngineering, Vancouver, BC, Canada, 2004.
[3] J. Suhara, R. Matsumoto, H. Torita, Y. Tsuyuki, T. Kamei,O. Takahashi, Y. Kunimatsu, H. Aida, T. Fujita, Constructionof civil building using three dimensional seismic isolationsystem: part 2, tests for three dimensional seismic isolationsystem, in: Proceedings, 14th World Conference onEarthquake Engineering, Beijing, China, 2008.
[4] O. Takahashi, H. Aida, J. Suhara, R. Matsumoto, Y. Tsuyuki,T. Fujita, Construction of civil building using threedimensional seismic isolation system: part 1, design ofbuilding using three dimensional seismic isolation system,in: Proceedings, 14th World Conference on EarthquakeEngineering, Beijing, China, 2008.
[5] M. Morishita, S. Kitamura, Y. Kamishima, Structure of 3-dimensional seismic isolated FBR plant with verticalcomponent isolation system, in: Proceedings, 17th SMiRT,Paper #K09e2, Prague, Czech Republic, 2003.
[6] M.Morishita, S. Kitamura, S. Moro, Y. Kamishima, S. Takahiro,Study on 3-dimensional seismic isolation system for nextgeneration nuclear power plantdvertical component isolationsystemwith coned disk spring, Technical Report No. 620, in:Proceedings, 13thWorld Conference on EarthquakeEngineering, Vancouver, BC, Canada, 2004.
[7] J. Suhara, Research on 3D base isolation system applied tonew power reactor 3D seismic isolation device with rolling
seal type air spring: part 1, in: Proceedings, SMiRT 17, Paper#K09e4, Prague, Czech Republic, 2003.
[8] T. Shimada, J. Suhara, K. Inoue, Three dimensional seismicisolation system for next-generation nuclear power plantwith rolling seal type air spring and hydraulic rockingsuppression system, in: ASME, Proceedings, Pressure Vesselsand Piping Conference, Denver, CO, 2005.
[9] S. Ogiso, K. Nakamura, M. Suzuki, S. Moro, Development of 3Dseismic isolator using metallic bellows, in: Proceedings of17th International Conference on Structural Mechanics inReactor Technology (SMiRT 17), Prague, Czech Republic, 2003.
[10] T. Somaki, T. Nakatogawa, A. Miyamoto, K. Sugiyama,Y. Oyobe, K. Tamachi, Development of 3-dimensional baseisolation system for nuclear power plants, in: Proceedings of16th International Conference on Structural Mechanics inReactor Technology (SMiRT 16), Washington DC, 2001.
[11] S. Okamura, S. Kitamura, K. Takahashi, T. Somaki,Experimental study on vertical component isolation system,in: Proceedings of 18th International Conference onStructural Mechanics in Reactor Technology (SMiRT18),Beijing, China, 2005.
[12] J.M. Kelly, Base Isolation in Japan, Report No. UCB/EERC-88/20, University of California, Berkeley, CA, 1988.
[13] I.D. Aiken, J.M. Kelly, F.F. Tajirian, Mechanics of Low ShapeFactor Elastomeric Seismic Isolation Bearings, Report No.UCB/EERC-89/13, University of California, Berkeley, CA, 1989.
[14] S. Okamura, Y. Kamishima, K. Negishi, Y. Sakamoto,S. Kitamura, S. Kotake, Seismic isolation design for JSFR,J. Nucl. Sci. Technol. 48 (2011) 688e692.
[15] International Atomic Energy Agency (IAEA), Verification ofAnalysis Methods for Predicting the Behavior of SeismicallyIsolated Nuclear Structures, Final Report of a CoordinatedResearch Project 1996e1999, IAEA-TECDOC-1288, IAEA,Vienna, Austria, 2002.
[16] G.P. Warn, A.S. Whittaker, M.C. Constantinou, Verticalstiffness of elastomeric and lead-rubber seismic isolationbearings, J. Struct. Eng. 133 (2007) 1227e1236.
[17] F.F. Tajirian, J.M. Kelly, I.D. Aiken, W. Veljovich, Elastomericbearings for three-dimensional seismic isolation”, in:Proceedings of the 1990 ASME PVP Conference, Nashville,TN, 1990.
[18] F. Naeim, J.M. Kelly, Design of Seismic Isolated Structures:From Theory to Practice, John Wiley and Sons Inc.,Chichester, UK, 1999.
[19] J.M. Kelly, E. Quiroz, Testing and Evaluation of CEGB IsolationSystem, Report No. 90e004, Earthquake EngineeringResearch Center, University of California, Berkeley, CA, 1990.
[20] N. Makris, H.S. Deoskar, Prediction of observed response ofbase-isolated structure, J. Struct. Eng. 122 (1996) 485e493.
[21] M. Kageyama, T. Iba, K. Umeki, T. Somaki, S. Moro,Development of three-dimensional base isolation systemwith cable reinforcing air spring, in: Proceedings, 17thSMiRT, Paper #K09e5, Prague, Czech Republic, 2003.
[22] A. Kashiwazaki, T. Shimada, T. Fujiwaka, S. Moro, Study on3-dimensional base isolation system applying to new typepower plant reactor (hydraulic 3-dimensional base isolationsystem: no.1), in: Proceedings, 17th SMiRT, Paper #K09e2,Prague, Czech Republic, 2003.
[23] T. Fujita, S. Fujita, Y. Watanabe, A. Kato, S. Suzuki,K. Fukumori, Fundamental study of three-dimensionalseismic isolation system for nuclear power plants, in:Proceedings, 11th World Conference on EarthquakeEngineering, Acapulco, Mexico, 1996.