Initialising ...
Initialising ...
Initialising ...
Initialising ...
Initialising ...
Initialising ...
Initialising ...
Quinay, P. E. B.*; Ichimura, Tsuyoshi*; Hori, Muneo*; Nishida, Akemi; Yoshimura, Shinobu*
Bulletin of the Seismological Society of America, 103(3), p.2094 - 2110, 2013/06
Times Cited Count：11 Percentile：34.79(Geochemistry & Geophysics)We studied the seismic structural response of a fault-structure system using a large-scale, highfidelity model with multiscale analysis and parallel simulation. To reduce the computation cost of simulating seismic wave propagation (Ichimura and Hori, 2006), we extended the seismic wave propagation simulation tool to parallel computing which uses a distributed-memory computer. We developed a method for prepartitioning the three-dimensional model and performing the neighbor-independent submodel hybrid-grid meshing. We then verified and validated the extended simulation tool. Finally, we demonstrated its advantages for computing the dynamic responses of a fault-structure system model that includes a building structure of nuclear power plant, using a high fidelity model which is based on actual settings.
Quinay, P. E.*; Ichimura, Tsuyoshi*; Hori, Muneo*; Nishida, Akemi; Yoshimura, Shinobu*
Proceedings of 15th World Conference on Earthquake Engineering (WCEE-15) (USB Flash Drive), 8 Pages, 2012/09
In this paper we discuss the development of seismic response estimation method for nuclear power plant building structures based on fault-structure system. In analysing this system, we model the source process, seismic wave propagation in the crust and soil structure, and the dynamic response of the building structure. Due to the varied length-scales involved in simulating from the fault rupture to the building dynamic response, the computation cost of analysis is high. Thus, we perform the analysis by combining a multiscale approach called the Macro-Micro Analysis (MMA), and distributed-memory parallel computing. As an application example, we compute the response of a model of a nuclear power plant building with the 2007 Chuetsu-oki earthquake as the source. We target accuracy of the forward modelling up to 1.0 Hz. Advantages for the analysis of NPP building in the structure-level and in the finite element level are discussed.
Quinay, P. E. B.; Ichimura, Tsuyoshi*; Hori, Muneo*; Wijerathne, M. L. L.*; Nishida, Akemi
Progress in Nuclear Science and Technology (Internet), 2, p.516 - 523, 2011/10
In this paper, we conducted a numerical analysis based on a fault-structure system. We developed a full three dimensional model that includes from a fault to structures at the surface. However, huge computational cost is required to solve these components simultaneously. Ichimura and Hori (2009a) presented a macro-micro analysis method based on singular perturbation expansion to reduce computational cost. We implemented a highly tuned finite-element code for macro-micro analysis systems to handle large-scale wave propagation in a complicated crust structure as well as the seismic response of a structure at the surface. After verifying the accuracy by comparing the results with analytical Green's function solutions, we demonstrate a seismic response using a model of a nuclear power plant structure. Our methodology can enable detailed prediction of the seismic response of a nuclear facility, permitting the generation of more reliable estimates of the seismic safety of nuclear facilities.
Quinay, P. E. B.; Ichimura, Tsuyoshi*; Hori, Muneo*; Nishida, Akemi
Proceedings of 3rd ASEAN Civil Engineering Conference/3rd ASEAN Environmental Engineering Conference/1st Seminar on Asian Water Environments 2010, 5 Pages, 2010/11
In this study, a physical modeling-based approach, the fault-structure system (FSS) analysis is presented for estimating the response of nuclear power plant structure under a scenario earthquake. To efficiently solve the FSS, a method based on multi-scale approach, the Macro-Micro Analysis (MMA) method (Ichimura and Hori, 2007) is used. Results show that the complicated response of the structure can be modeled by using the FSS. In order to obtain high resolution crust structure information for FSS analysis, an inversion method that minimizes an objective function by step-wise increase in target frequency is presented. This study takes advantage of the present computational resources to obtain high resolution outputs for seismic analysis of important structures.
