The use of an Artificial Neural Network for on-line prediction of Pin-Cell Discontinuity Factors in PARCS

Tomasz Kozlowski, Deokjung Lee, Thomas J. Downar

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

During the last several years, the U.S. NRC has sponsored the development of multi-group, SP3 method in the Purdue Advanced Reactor Core Simulator (PARCS) primarily for MOX fuel analysis. These methods were implemented within the framework of pin-by-pin discretization using pin homogenized cross sections. Pin-Cell Discontinuity Factors (PDFs) were proposed in order to recover the error introduced by pin cell homogenization. The method showed the potential to improve the accuracy of the pin power prediction; however its application for practical core problems was limited because of the considerable amount of data required for whole core calculations and because of uncertainties in the application of PDFs to core conditions for which they were not generated. The work reported here is the development of innovative methods to implement PDFs for practical applications using an Artificial Neural Network (ANN). The work is demonstrated using the KAIST MOX benchmark.

Original languageEnglish (US)
Title of host publicationProceedings of the PHYSOR 2004
Subtitle of host publicationThe Physics of Fuel Cycles and Advanced Nuclear Systems - Global Developments
Pages613-623
Number of pages11
StatePublished - 2004
Externally publishedYes
EventPHYSOR 2004: The Physics of Fuel Cycles and Advanced Nuclear Systems - Global Developments - Chicago, IL, United States
Duration: Apr 25 2004Apr 29 2004

Publication series

NameProceedings of the PHYSOR 2004: The Physics of Fuel Cycles and Advanced Nuclear Systems - Global Developments

Other

OtherPHYSOR 2004: The Physics of Fuel Cycles and Advanced Nuclear Systems - Global Developments
Country/TerritoryUnited States
CityChicago, IL
Period4/25/044/29/04

Keywords

  • Discontinuity Factors
  • Neural Network
  • Pin Cell

ASJC Scopus subject areas

  • General Engineering

Fingerprint

Dive into the research topics of 'The use of an Artificial Neural Network for on-line prediction of Pin-Cell Discontinuity Factors in PARCS'. Together they form a unique fingerprint.

Cite this