Full-Scale 3D Vibration Simulator of an Entire Nuclear Power Plant on Simple Orchestration Application Framework

Guehee Kim (), Kohei Nakajima (), Takayuki Tatekawa (), Naoya Teshima (), Yoshio Suzuki () and Hiroshi Takemiya ()
Additional contact information
Guehee Kim: Japan Atomic Energy Agency, Center for Computational Science and e-Systems
Kohei Nakajima: Japan Atomic Energy Agency, Center for Computational Science and e-Systems
Takayuki Tatekawa: Japan Atomic Energy Agency, Center for Computational Science and e-Systems
Naoya Teshima: Japan Atomic Energy Agency, Center for Computational Science and e-Systems
Yoshio Suzuki: Japan Atomic Energy Agency, Center for Computational Science and e-Systems
Hiroshi Takemiya: Japan Atomic Energy Agency, Center for Computational Science and e-Systems

A chapter in High Performance Computing on Vector Systems 2010, 2010, pp 93-106 from Springer

Abstract: Abstract So far, we have developed grid-enabled application for “Full-Scale 3D Vibration Simulator for an Entire Nuclear Power Plant” which is simulation platform to analyze seismic response of a whole digitalized nuclear power plant. In the 3D Vibration Simulator, components of a nuclear power plant are treated in hierarchical manner in which large components are grouped at primary level and small components such as pipes are grouped at secondary level and boundary condition data from the large components simulation are used as input data of small components simulation. In this work, to make the whole simulation more efficient than the previous sequential scenario in which after large components simulation is completed, small components simulation starts, we introduce pipelined data-transfer scenario in which boundary condition data are transferred each time step while all components simulation are run in parallel. In realization of the 3D Vibration Simulator in the introduced scenario, we confronted two challenges: first, clearance of job’s idle time to be wasted for only waiting data which takes from a few ten minutes to a few hours per each time step and second, immediate resubmission of abnormal ended jobs for a long time simulation under the introduced scenario. To address these challenges, we proposed two solutions: as first solution, we set policy by which jobs of small components are submitted after all necessary data per each time step arrive and executed only that time step, which process is repeated whenever next time step input data arrive and as second solution, we make an abnormal ended job automatically resubmitted. Since there were no pre-existing grid technologies to provide sufficient functionalities to enable these solutions to be possible from the previous grid-enabled application, we developed Simple Orchestration Application Framework (SOAF) and upgraded the previous grid-enabled application by implementing the SOAF. Using the upgraded grid-enable application, we performed seismic analysis of High Temperature Engineering Test Reactor at O-arai R&D center of Japan Atomic Energy Agency and confirmed that the simulation were performed in pipelined data-transfer scenario effectively using computing resources without idle time for about a week simulation period resubmitting abnormally ended jobs. In this paper, the details of all of this work will be described.

Keywords: Nuclear Power Plant; High Performance Computing; Secondary Level; Primary Level; Reactor Pressure Vessel (search for similar items in EconPapers)
Date: 2010
References: Add references at CitEc
Citations:

There are no downloads for this item, see the EconPapers FAQ for hints about obtaining it.

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:spr:sprchp:978-3-642-11851-7_7

Ordering information: This item can be ordered from
http://www.springer.com/9783642118517

DOI: 10.1007/978-3-642-11851-7_7

Access Statistics for this chapter

More chapters in Springer Books from Springer
Bibliographic data for series maintained by Sonal Shukla () and Springer Nature Abstracting and Indexing ().