Nuclear Engineering and Design: Sugandha Singh, Abhinav Gupta
Nuclear Engineering and Design: Sugandha Singh, Abhinav Gupta
Nuclear Engineering and Design: Sugandha Singh, Abhinav Gupta
A R T I C L E I N F O A B S T R A C T
Keywords: Recent ground motion studies conducted at nuclear facilities indicate that in Central and Eastern United States,
High frequency ground motions the ground motion response spectra exceed the safe–shutdown earthquake spectra at high frequencies. Even
Seismic response of electrical equipment though high–frequency ground motions do not cause structural damage, high–frequency accelerations may
Cabinets and control panels
propagate through the structure and interfere with the output of the safety–related equipment such as relays
Nonlinear mounting of cabinets
Effect of localized nonlinearities
required to ensure safe shutdown of the plant during a seismic event. Hence, it is essential to seismically qualify
electrical equipment subjected to high–frequency accelerations. The seismic qualification is conducted by
comparing in–cabinet response spectra evaluated at the equipment locations with the capacities of equipment
obtained from shake table tests. The in–cabinet response spectrum is obtained from analysis of nuclear power
plant building and electrical cabinet in which the equipment are mounted. This study is based on the hypothesis
that the high–frequency motions do not reach the equipment as the small displacements induced by such motions
are filtered out by the geometric nonlinearities. The effect of two different types of nonlinearities are studied: (1)
a gap in the connection between electrical cabinet and floor; (2) sliding friction between electrical cabinet’s base
and the floor. The results from an analyses of various different cases show that the high–frequency motions do
not reach the relays if the maximum displacement of building floor is less than the gap. Even if the displacement
is larger than the gap, the in–cabinet spectral accelerations are not excessively high. On the contrary, results from
a conventional linear analysis give excessively high unrealistic spectral accelerations.
1. Introduction ground motions have high-frequency content due to the presence of hard
rock soil profile unlike the motions recorded in Western United States.
1.1. Purpose and motivation The frequency content of ground motion response spectra in CEUS
ranges from 15 to 45 Hz as compared to the frequency content of
A nuclear plant should be able to safely shutdown and maintain USNRC’s Reg. Guide 1.60 design safe shutdown earthquake (SSE)
necessary operations to avoid core melt during or after an earthquake. spectra which ranges from 1 to 10 Hz (USNRC, 2014). An example of the
The safety-related electrical equipment are critical for safe shutdown of difference in response spectra at a site in CEUS is shown in Fig. 1. The
plant and thus, are required to maintain their functionality. Following seismic hazard studies in accordance with NTTF recommendation
the damages caused by the earthquake and subsequent tsunami at required the nuclear plants to evaluate the vulnerability of their SSCs to
Fukushima-Daiichi nuclear power station in Japan, USNRC appointed a high-frequency ground motions.
Near-Term Task Force (NTTF) to review the insights from Fukushima EPRI (2007) shows that the high-frequency ground motions do not
Daiichi accident and provide recommendations for enhancing the safety cause structural damage to the power plant or heavy industrial struc
of nuclear power plants in the United States (USNRC, 2014). NTTF’s tures and equipment because the frequency ratio (the ratio of frequency
Recommendation 2.1 requires the licensees to reevaluate the seismic of excitation to the natural frequency of the system) is much greater than
and flooding hazards at their sites and update the structures, systems, unity resulting in a low value of dynamic amplification factor and hence,
and components (SSCs) to safeguard against the updated hazards if lower response. On the other hand, structures with higher natural fre
necessary. Recent studies (SSHAC, 1997; EPRI, 2013; PEER, 2015) quencies do not undergo large relative displacements or stresses due to
indicate that in Central & Eastern United States (CEUS), earthquake small displacement amplitude of high-frequency ground motions.
* Corresponding author.
E-mail address: agupta1@ncsu.edu (A. Gupta).
https://doi.org/10.1016/j.nucengdes.2020.111046
Received 31 August 2020; Received in revised form 17 December 2020; Accepted 28 December 2020
Available online 29 January 2021
0029-5493/© 2021 Elsevier B.V. All rights reserved.
S. Singh and A. Gupta Nuclear Engineering and Design 374 (2021) 111046
2013). Historically, the relays have not been tested for excitations with
frequencies higher than 16 Hz. The acceleration sensitivity of relays to earthquake (1952). TAFT earthquake has frequency content between 1
high frequency accelerations is not well understood (EPRI, 2015). There and 6 Hz and peak spectral acceleration occurs at 2.8 Hz. The high-
has been limited observations that certain electrical devices that were frequency ground motion used in the analysis is obtained from the
seismically qualified for low frequency ground motions did not perform design spectrum of a nuclear power station on the Eastern US. It has a
their intended function during high frequency shaking (EPRI, 2007). frequency content in the range of 12–55 Hz and peak spectral acceler
Some of the more recent guidelines particularly in the non-nuclear ap ation occurs at 35 Hz. The ground motion response spectra for both
plications require testing of nonstructural systems up to 33 Hz but these motions are shown in Fig. 2 in which both the grounds motions are
guidelines are applicable to only newer facilities and are not necessarily normalized to have peak ground acceleration of 1 g. The peak ground
applicable to nuclear equipment. acceleration and displacement are given in table 1.
Engineers at a nuclear power plant qualify the electrical devices by
evaluating the seismic demands inside the electrical cabinets and
comparing it with the capacities. The method to evaluate seismic de 1.2. Summary of existing studies
mands amplifies the floor response spectrum or in–structure response
spectrum (ISRS). Amplification of ISRS due to the electrical cabinet is Many studies (Stafford, 1975; Djordjevic and O’Sullivan, 1990;
based on the type of the cabinet. The amplified spectrum is called Djordjevic, 1992; Rustogi and Gupta, 1998; Gupta et al., 1999; Yang
in–cabinet response spectrum (ICRS) which gives the seismic demands et al., 2002; Rustogi and Gupta, 2004; Sankaranarayanan, 2007; Cho
on the devices like relays. The method suggested in EPRI (2015) uses et al., 2011; Vlaski, 2013; Herve, 2014; Gupta et al., 2019) have focused
empirical data to calculate the amplification factors. The approach ig on understanding the seismic behavior of electrical cabinets to deter
nores many factors such as cabinet mounting arrangement, location of mine the factors that influence the in-cabinet response spectra. A sum
equipment on the cabinet, geometric nonlinearities, etc. In general, the mary of the main observations made in these studies is discussed below
use of empirical amplification factors have been found to be inaccurate in order to understand the dynamic behavior of cabinets.
which can often render some devices seismically unacceptable for use in Gupta et al. (1999) proposed a method for evaluating in-cabinet
a plant. The extensive seismic probabilistic risk assessment studies response spectra in which the response of a cabinet is represented by
(SPRA) conducted by the plants in United States as an outcome of the only a few significant modes. A significant mode which governs ICRS
NTTF’s Recommendation 2.1, experienced that several relays in the can either be a global cantilever mode of the cabinet or a local mode of
plants could not be seismically qualified because the ICRS generated the door, internal frame or back wall depending on where the relay is
using simple amplification of ISRS with high frequency content had very mounted. These mode shapes are calculated by Rayleigh-Ritz method to
large spectral accelerations. represent the cabinet behavior. However, for some cabinets, the signif
This paper presents a preliminary assessment of the effect of local icant global mode can also occur due to rigid body rocking of cabinet.
ized nonlinearities that exist in the cabinet mounting arrangements at The rigid body rocking of cabinet is explored in more detail by Yang
the base. It is intended to evaluate whether localized nonlinearities at et al. (2002) and Makris and Zhang (1999). These studies propose for
the base due to gaps that exist around anchor bolts would filter out the mulations to evaluate rocking stiffness for different mounting arrange
small displacements associated with high frequency ground motions. ments in cabinets which facilitates incorporation of cabinet rocking
The study is based on observations from the analysis of simple repre mode in Ritz vector approach. Rustogi and Gupta (2004) use modal data
sentative systems and compares the ICRS obtained from a linear analysis obtained from shake table and in-situ tests of cabinets to validate finite
to that from a nonlinear analysis. Two different ground motions are used element analysis and develop a basis for the Ritz vector approach. Gupta
in the analysis. Since low-frequency ground motions are also likely to and Yang (2002) modify the originally proposed Ritz vector approach by
occur in various regions, the systems are subjected to a low-frequency incorporating cabinet rocking mode and a partial rotational constraints
ground motion in addition to a high-frequency ground motion. The provided by structural members in a cabinets.
low-frequency ground motion used in the analysis is the TAFT The studies mentioned above do not consider two important aspects
associated with rocking behavior. First, when a cabinet is anchored and
2
S. Singh and A. Gupta Nuclear Engineering and Design 374 (2021) 111046
Fig. 3. Local mode shapes of cabinet door. Fig. 4. Local mode shape of cabinet internal frame.
3
S. Singh and A. Gupta Nuclear Engineering and Design 374 (2021) 111046
3. Proposed Premise/Hypothesis 4.1. Model-I: Gap between cabinet base and anchor bolt
As discussed above in detail, the seismic behavior of a cabinet de This model is used to represent the gap as shown in Fig. 6 in between
pends on the dynamic properties of the cabinet itself as well as the the bolt head and the cabinet base plate. Herve (2014) considered a
structural configuration of its mounting arrangement. Cabinets mounted spring-mass-damper system with a gap of 1 mm which is based on the
with anchor bolts undergoes rigid body rocking which usually affects the recommendations by EPRI (2007) and International Atomic Energy
fundamental mode of the cabinet. The rigid body rocking mode may be Association (IAEA, 2012) as the cut-off displacement for the case of
the significant mode by itself or in combination with a local mode. These high-frequency motions due to an airplane impact. The force–displace
factors affect the resulting in-cabinet response spectra and hence the ment relationship used in their study to represent this model considers a
frequency content of cabinet acceleration at the equipment level. This gap of 1 mm in both the tension and the compression.
paper focuses solely on the effects of rigid body rocking and sliding of Contrary to the recommendation by Herve (2014), the gap associated
cabinets. Hence, the cabinets models studied in this paper are idealized with sliding or rocking behavior in a cabinet would exhibit zero force
cases to understand the equipment response due to such factors. When a only in tension. The behavior in compression between the cabinet base
cabinet base is anchored to the floor and the cabinet undergoes rocking, plate and the anchor bolt is linear elastic with no gap. Furthermore, we
there can be a gap between the cabinet base plate and the bolt head. As consider two different magnitudes of gaps, 1 mm and 5 mm. Fig. 7 shows
the force–displacement relationship of this model. Fig. 6 shows that for
4
S. Singh and A. Gupta Nuclear Engineering and Design 374 (2021) 111046
displacement between 0 mm and a predetermined gap, there is zero friction force, t is the time instant, Δt is the time step used in the analysis
resisting force in case of tension whereas for displacements more than and f(t) is the spring force at time ‘t’.
the gap, the resisting force follows linear elastic relationship. Hence, it is
a nonlinear elastic model that exhibits geometric nonlinearity. The 5. Application case studies
model follows a simple set of equations of motion which are described
by Eqs. (1) and (2). In the equations below, m denotes the mass, c is the In order to achieve the objective of this study, analysis of various
damping constant, k is the stiffness, ugap is the gap length (1 mm or 5 cases are conducted by creating representative building-cabinet systems
mm), u is the displacement, u̇ is the velocity, ü is the acceleration of the and considering two different earthquake ground motions. For
cabinet, respectively. Also, üf is the floor acceleration which is an input simplicity and for facilitating appropriate interpretation of the phe
at the base of cabinet. nomenon, both the building and the cabinet are represented by a
respective SDOF system. The two ground motions considered in this
mü + cu̇ = − müf , 0 ≤ u ≤ ugap (1)
study are same as those given in Fig. 2. Even though Fig. 2 shows
spectrum curves normalized to a PGA of 1 g, the actual motions used in
mü + cu̇ + ku = − müf , u < 0 (2)
the numerical study correspond to the actual PGA values for these mo
tions. The low frequency ground motion represented by the Taft earth
mü + cu̇ + k(u − ugap ) = − müf , u > ugap (3)
quake record has a PGA of 0.159 g and that for the high frequency
ground motion, represented by the actual design spectrum of a nuclear
power plant on Eastern US, is equal to 0.415 g. At first, it may appear
4.2. Model-II: Friction effect
inappropriate to consider different PGA values for the two motions.
However, the problem being studied is a nonlinear problem in which the
Following USNRC’s Near Term Task Force (NTTF) recommendations
behavior is highly dependent upon the amplitudes of displacements.
2.1 and 2.2, many plants in CEUS have completed their Seismic Proba
Therefore, the actual motions are used without any modifications. As
bilistic Risk Assessment (SPRA) and found that the relays in many cab
noted in Fig. 2, the peak of the spectrum curve for the low frequency
inets do not qualify based on the acceptable risk goals. One of the
earthquake record occurs around 3 Hz and that for the high frequency
solutions suggested and pursued by the industry is based on utilizing the
motion close to 35 Hz. Consequently, two different variations of the
ductility (elastic-perfectly plastic) considerations. The model-II repre
building and the cabinet are considered. Typical buildings in nuclear
sents the effect of friction on seismic behavior of cabinets (Kon
power plants have low frequency modes for which the frequencies lie in
stantinidis and Nikfar, 2015). As discussed earlier, the diameter of bolt
the range of 1 to 5 Hz. Therefore, a 3 Hz SDOF system is selected to
hole in the base plate is often a little larger than the bolt diameter.
represent a low–frequency building (or a low–frequency cabinet).
Consequently, the friction forces would act at the interface of the base
Further, a 35 Hz SDOF system is considered to represent a high
plate and the floor before sliding begins. The cabinet would act as a
–frequency building (or a high–frequency cabinet). One may argue that
linear elastic material when the seismic force is less than the friction
the typical building frequencies are much less than 35 Hz. However, it
force. When the seismic force becomes greater than the friction force,
must be noted that every building will have higher order modes in the
the cabinet would exhibit a slip in the direction of motion. As the di
frequency range of interest in this study. These modes will exhibit
rection of motion changes, the cabinet would again exhibit a linear
appreciable amplifications when subjected to a high frequency ground
elastic behavior until the time seismic force exceeds the friction force.
motion. The high frequency building modes contribute primarily to the
Thus, the behavior is primarily hysteretic in nature and identical to that
floor motions at lower elevations in a building. Similarly, there are
of an elastic-perfectly plastic material. Fig. 8 shows the force–displace
hundreds of cabinets in a nuclear plant that are located at various
ment relationship for this model. The equations of motion for this model
different elevations throughout the building. Typically, a global rocking
can be written as follows:
cabinet mode is a low frequency mode. A local mode of the cabinet panel
mü + cu̇ + ku = − müf , |f (t)| ≤ fs max (4) or frame can be either a low frequency or a high frequency mode.
Therefore, the high frequency systems are meant to be representative of
mü + cu̇ + (ku − fs max ) = − müf , |f (t)| ≥ fs max (5) higher order modes for both the buildings and the cabinets. Typical
damping ratios for concrete buildings such as auxiliary building and
When the direction of velocity changes,
service building in nuclear plants are on the order of 5%. Therefore, the
f (t + Δt) = f (t) + (k(u(t + Δt) − u(t) ) ) (6) damping ratio of building is taken as 5%. Similarly, the damping ratio
for cabinets is typically taken as 2% in almost all plants and their SPRA
mü(t + Δt) + cu̇(t + Δt) + f (t + Δt) = − müf (t + Δt) (7) studies. Therefore, the damping ratio for cabinets is taken as 2%.
While the building model is linear-elastic in all cases, different
where k is the stiffness before cabinet slips, fs_max is the maximum static
5
S. Singh and A. Gupta Nuclear Engineering and Design 374 (2021) 111046
Fig. 9. Steps for analyzing in–cabinet response spectrum for each case.
6. Discussion of results
• Case-1 (L–L–L): Low–frequency earthquake, Low–frequency (3 Hz)
building, and Low–frequency (3 Hz) cabinet.
The results from the analyses for various application cases enumer
• Case-2 (H–L–L): Same as Case-1 except that the ground motion is
ated above are discussed in this section. The observations from these
high–frequency earthquake.
results are used to evaluate the validity of the hypothesis presented
• Case-3 (L–L–H): Low–frequency earthquake, Low–frequency (3 Hz)
earlier. To begin with, the two types of building models, low–frequency
building, and high–frequency (35 Hz) cabinet.
building with natural frequency of 3 Hz and high–frequency building
• Case-4 (H–L–H): Same as Case-3 except that the ground motion is
with natural frequency of 35 Hz are analyzed for both the low-frequency
high-frequency earthquake.
and the high-frequency ground motions. The peak floor acceleration for
each building induced by both ground motions are shown in Table 2.
6
S. Singh and A. Gupta Nuclear Engineering and Design 374 (2021) 111046
Table 3
Maximum Total and Relative Floor Displacement in ‘mm’ units.
Relative Displacement Total Displacement
Ground Motion Low-Frequency Building High-Frequency Building Ground Motion Low-Frequency Building High-Frequency Building
Fig. 10. ISRS of low–frequency building subjected to both ground motions. Fig. 12. Comparison of fixed-base cabinet ICRS with ISRS for case 1 (L–L–L)
and case 2 (H–L–L).
7
S. Singh and A. Gupta Nuclear Engineering and Design 374 (2021) 111046
Table 4
Maximum Total Cabinet Acceleration in ‘g’ units.
Case Number Linear 1 mm Gap 5 mm Gap Elastic Perfectly Plastic
Table 5
Maximum Total Cabinet Displacement in ‘mm’
Case Number Linear 1 mm Gap 5 mm Gap Elastic Perfectly Plastic
Fig. 14. Comparison of fixed-base cabinet ICRS with ISRS for case 5 (L–H–L)
and case 6 (H–H–L). 1 (L-L-L) 14.85 18.085 10.891 9.456
2 (H-L-L) 11.107 5.134 9.178 6.502
3 (L-L-H) 6.135 7.028 8.864 5.947
4 (H-L-H) 5.407 5.261 5.933 5.407
5 (L-H-L) 6.163 7.095 11.244 6.388
6 (H-H-L) 11.107 11.396 12.315 11.107
7 (L-H-H) 5.99 6.875 7.027 5.991
8 (H-H-H) 11.21 10.562 9.492 10.203
8
S. Singh and A. Gupta Nuclear Engineering and Design 374 (2021) 111046
Fig. 18. ICRS for 5 mm gap cabinet- nonlinear base case 1 (L–L–L). Fig. 19. ICRS for nonlinear base case 2 (H–L–L).
acceleration for 1 mm gap model is even higher than the fixed-base case gap. Second, an overall amplification can be seen in the peak spectral
because the maximum displacement of the floor is more than 1 mm gap. accelerations compared to GMRS. This observation is attributed to the
The maximum floor displacement is 6.14 mm. When the cabinet base, in maximum floor displacement being slightly greater than the gap.
this case, hits the anchor bolt, an impulse is generated due to the high Nonlinear Base Case − 2 (H–L–L): In this case, the ICRS are evaluated
velocity of the cabinet which acts as an additional load. Repeated im for same systems as in Figs. 17 and 18 but for a high frequency motion.
pulses thus resonate with natural frequency of cabinet and further adds The new ICRS curves are compared in Fig. 19. As discussed earlier, the
to the already large response from linear analysis at the 3 Hz value. In high frequency ground motion also contains some low frequency con
contrast, the peak spectral acceleration is much less for the case of 5 mm tent which resonates with the low–frequency building as well as low
gap as compared to the fixed-base case. Essentially, a 5 mm gap filters –frequency cabinet. Consequently, an amplification can be observed in
out most of the displacement of the cabinet. Even though the maximum the ICRS around 3 Hz for the case of fixed-base analysis. The response is
floor displacement (6.14 mm) is slightly larger than 5 mm, the floor comparatively less for the elastic-perfectly plastic case due to energy
motion for almost all the duration of this motion remains below 5 mm.
Therefore, no amplification is observed. Furthermore, this large floor
displacement occurs only for a very short duration during the entire
floor motion and the response due to the corresponding impulse is
almost nonexistent. In addition to the gap boundary condition, Fig. 17
also gives ICRS for the case of elastic-perfectly plastic boundary condi
tion (friction). As seen in the figure, the hysteresis loop dissipates some
energy and therefore the peak of ICRS is less than that of fixed base case.
The relatively low values of spectral accelerations in the ICRS for 5 mm
gap case make it difficult to observe the true variation in Fig. 17.
Therefore, Fig. 18 gives the ICRS for only the 5 mm gap case. This figure
can be used to make two observations. First, the peak spectral acceler
ation occurs at 1.7 Hz instead of 3 Hz. The frequency shifts because of
reduction in stiffness of the cabinet, i.e., essentially zero stiffness when
relative displacement of floor and cabinet is between zero and 5 mm
Fig. 20. ICRS for 5 mm gap cabinet-nonlinear base case 2 (H–L–L).
9
S. Singh and A. Gupta Nuclear Engineering and Design 374 (2021) 111046
10
S. Singh and A. Gupta Nuclear Engineering and Design 374 (2021) 111046
11
S. Singh and A. Gupta Nuclear Engineering and Design 374 (2021) 111046
7. Conclusions
Fig. 28. ICRS for nonlinear base case 8 (H–H–H; without fixed-base cabinet).
12
S. Singh and A. Gupta Nuclear Engineering and Design 374 (2021) 111046
the work reported in this paper. IEEE, 2013. IEEE Standard for Seismic Qualification of Equipment for Nuclear Power
Generating Stations. IEEE Std 344TM-2013.
Katona, T., Kennerknecht, H., Henkel, F.O., 1995. Earthquake design of switchgear
Acknowledgement cabinets of the VVER-440/213 at Paks. In: Transactions of the Thirteenth
International Conference on Structural Mechanics in Reactor Technology, 1995; III,
This research was supported by Center for Nuclear Energy Facilities Porto Alegre, Brazil, August 13–18.
Konstantinidis, D., Nikfar, F., 2015. Seismic response of sliding equipment and contents
and Structures at North Carolina State University. Resources for the in base-isolated buildings subjected to broadband ground motions. J. Int. Assoc.
Center come from the dues paid by member organizations and from the Earthquake Eng. 44 (6), 865–887. https://doi.org/10.1002/eqe.2490.
Civil, Construction, and Environmental Engineering Department and Lee, B.J., Abou-Jaoude, C., 1992. Effect of base uplift on dynamic response of electrical
and mechanical equipment. Seismic Eng. PVP 237-2, 145–150 (ASME).
College of Engineering in the University. Lee, B.J., Berak, E.G., Passalugo, P.N., 1990. Effect of base shim plates on seismic
qualification of electrical control panels. In: Proceedings of the 1990 Pressure Vessels
References and Piping Conference, PVP, vol. 197, ASME, pp. 301–306.
Llambias, J.M., Sevant, C.J., Shepherd, D.J., 1989. Non-linear response of electrical
cubicles for fragility estimation. Transactions of 10th International Conference on
Cho, S.G., Kim, D., Chaudhary, S., 2011. A simplified model for nonlinear seismic
Structural Mechanics in Reactor Technology, Anaheim, USA, August 22–27.
response analysis of equipment cabinets in nuclear power plants. Nuclear Eng. Des.
Makris, N., Zhang, J., 1999. Rocking Response and Overturning of Anchored Equipment
241(8), 2750–2757. https://doi.org/10.1016.
under Seismic Excitations. PEER 1999/06.
Djordjevic, W., 1992. Amplified response spectra for devices in electrical cabinets. Proc.
McKenna, F., Fenves, G., 2001. The OpenSEES Command Language Manual: version 1.2.
4th Symp. on Current Issues Related to Nuclear Power Plant Structures, Equipment
Pacific Earthquake Engineering Center, University of California, Berkeley.
and Piping, Orlando, Florida, USA.
PEER, 2015. NGA-East: Median Ground-Motion Models for the Central and Eastern North
Djordjevic, W., O’Sullivan, J.J., 1990. Guidelines for development of incabinet amplified
America Region. PEER Report No. 2015/04, April 2015.
response spectra for electrical bench boards and panels. 1990. In: Rep. No., EPRI-NP-
Rustogi, S.K., Gupta, A., 1998. Incabinet response spectra. Tech. Rep. No. C-NPP-SEP 21/
5223, Stevenson and Associates, Inc, Woburn, Mass.
98, 1998, Center for Nuclear Power Plant Structures, Equipment and Piping, N.C.
EPRI, 2007. Program on technology innovation: The effects of high-frequency ground
State Univ., Raleigh, N.C.
motion on structures, components, and equipment in nuclear power plants. Palo
Rustogi, S.K., Gupta, A., 2004. Modeling the dynamic behavior of electrical cabinets and
Alto, CA: 2007.1015108.
control panels: Experimental and analytical results. J. Struct. Eng 130 (3), 511–519.
EPRI, 2013. Seismic Evaluation Guidance: Screening, Prioritization and Implementation
SSHAC, 1997. Recommendations for Probabilistic Seismic Hazard Analysis: Guidance on
Details (SPID) for the Resolution of Fukushima Near-Term Task Force
Uncertainty and Use of Experts. NUREG/CR-6372, UCRL-ID- 122160, Vol. 1.
Recommendation 2.1: Seismic. Palo Alto, CA: 2013.1025287.
Sankaranarayanan, R., 2007. Seismic Response of Acceleration-Sensitive Nonstructural
EPRI, 2014. High frequency program: High frequency testing summary. Palo Alto, CA,
Components Mounted on Moment-Resisting Frame Structures. Ph.D. Dissertation.
2014.3002002997.
University of Maryland, College Park.
EPRI, 2015. High frequency program: Application guidance for functional confirmation
Stafford, J. R., 1975. Finite element predictions of the dynamic response of power plant
and fragility evaluation. Palo Alto, CA, 2015.3002004396.
control cabinets. In: Proc. 2nd ASCE Specialty Conf. on Structural Design of Nuclear
Gupta, A., Cho, S-.G., Hong, K-.J., Han, M., 2019. Current State of in-cabinet response
Plant Facilities, New York.
spectra for seismic qualification of equipment in nuclear power plants. Nuclear Eng.
USNRC, 2014. Regulatory Guide 1.60: Design Response Spectra for Seismic Design of
Des. 343, 269–275.
Nuclear Power Plant. Rev. 2, July 2014.
Gupta, A., Rustogi, S.K., Gupta, A.K., 1999. Ritz vector approach for evaluating incabinet
USNRC, Feb 2014. Recommendation for enhancing reactor safety in the 21st century.
response spectra. Nuclear Eng. Des. 190 (3), 255–272. https://doi.org/10.1016/
USNRC, Tech. Rep. 46.
S0029-5493(99)00076-X.
Vlaski, V., 2013. Reduction of External Hazard (Fast Impact) Induced Vibrations. F/AB-
Gupta, A., Yang, J., 2002. Modified Ritz vector approach for dynamic properties of
58. 1st Conference on Technical Innovation in Nuclear Civil Engineering TINCE
electrical cabinets and control panels. Nuclear Eng. Des. 217(1), 49–62. https://doi.
2013, Paris, October 28–31.
org/10.1016/S0029-5493(02)00133-4.
Yang, J., Rustogi, S.K., Gupta, A., 2002. Rocking stiffness of mounting arrangements in
Herve, G., 2014. Improvement of the evaluation of high frequency content in the
electrical cabinets and control panels. Nuclear Eng. Des. 219 (2), 127–141. https://
calculation of impact floor response spectra. 2nd Conference on Technical
doi.org/10.1016/S0029-5493(02)00279-0.
Innovation in Nuclear Civil Engineering TINCE 2014, Paris.
IAEA, 2012. Safety Assessment of NPP Structures against Human Induced External
Events. IAEA SAFETY REPORT SERIES, DD1086 Draft, Rev.: R1- 1, 6 November.
13