Contents lists available at ScienceDirect Nuclear Engineering and Design journal homepage: www.elsevier.com/locate/nucengdes Sensitivity analysis for the distribution of maximum responses by seismic isolation system parameters using the stochastic response database Seunghyun Eem a, ⁎ , Jung Han Kim b, ⁎ a Structural and Seismic Safety Research Team, Korea Atomic Energy Research Institute, Republic of Korea b Pusan National University, Department of Civil Engineering, Republic of Korea ARTICLE INFO Keywords: Seismic isolation Isolated nuclear power plants Stochastic response database Seismic response Sensitivity analysis ABSTRACT The seismic performance of nuclear power plants can be improved by installing the proposed seismic isolation system. Due to the characteristics of this system, while acceleration seismic response and base shear can be reduced, the displacement seismic response of the isolated structure increases. Therefore, the restriction of excessive displacement responses becomes critical to the design and safety assessment of isolated structures. Such seismic design and safety assessment requires a seismic response distribution, which itself necessitates a large number of nonlinear dynamic analyses. However, the stochastic response database can directly estimate the maximum seismic response distribution of an isolated structure without nonlinear dynamic analyses. In this study, a sensitivity analysis was performed using the stochastic response database for the isolated structures that affect the maximum seismic response distribution. First, the important seismic isolation parameters were se- lected, followed by an investigation into the effects these parameters have on the mean and standard deviation of the maximum seismic response distribution of the isolated structure. It is believed that this information will be useful in the seismic design and safety assessment of nuclear power plants with seismic isolation systems. 1. Introduction Research is underway to apply a seismic isolation system to nuclear power plants to improve seismic safety. This system isolates the ground motion from the plant structure to mitigate earthquake loading (Eem et al., 2011), and various research on the seismic mitigation effects of the isolation system have been conducted (Eem and Jung 2015; Dusi, 2012, Perotti et al., 2013; Sarebanha et al., 2018). Generally, research has focused on numerical models of seismic isolators (Mozaheb and Behnamfar, 2018; Kumar et al., 2014; Grant et al., 2004), floor response spectra (Jung et al., 2017; Chen et al., 2014), probabilistic seismic safety assessment (Yu et al., 2018; Kumar et al., 2017; Eem et al., 2013), and so on. For proper seismic design and safety assessment, it is necessary to analyze the seismic response distribution of nuclear power plants through numerous seismic response analyses. Damage to an isolated plant can be caused by earthquake-induced displacement, which is an important seismic response (Kumar et al., 2017; Eem and Jung, 2018). For this reason, maximum seismic response distributions are required; however, this means that a large number of nonlinear dynamic analyses need to be carried out because of the inherent nonlinearity of the base isolation system. In response to this problem, a stochastic response database (SRD) has been developed that can directly estimate the seismic response distribution of an isolated structure to reduce the ef- fort required by nonlinear dynamic analyses. The SRD is useful to ob- tain seismic response distributions as well as to conduct seismic risk analyses of isolated nuclear power plants, including specific isolated structures (Eem and Jung, 2015; Eem and Jung, 2018). In this study, the parameters of the seismic isolation system that affect the maximum seismic responses of isolated structures are in- vestigated through the SRD. This database can provide the mean and standard deviation of the maximum seismic response distribution of isolated structures. By identifying the parameters of the seismic isola- tion system affecting the maximum seismic response with the SRD, the behavior of isolated nuclear power plants can be predicted more effi- ciently. 2. Methods In order to perform a sensitivity analysis of the response distribution of the seismic isolation system, it is necessary to confirm which para- meters most strongly affect the seismic response of the isolated struc- tures. Such parameters affecting response behavior can be found from an equivalent model of the given isolated structure (Eem and Jung, https://doi.org/10.1016/j.nucengdes.2019.03.020 Received 19 November 2018; Received in revised form 12 March 2019; Accepted 20 March 2019 ⁎ Corresponding authors. E-mail addresses: eemsh@kaeri.re.kr (S. Eem), jhankim@pusan.ac.kr (J.H. Kim). Nuclear Engineering and Design 347 (2019) 53–58 Available online 28 March 2019 0029-5493/ © 2019 Elsevier B.V. All rights reserved. T