Results for the C5G7 Benchmark Problem using a Subelement Option of the Variational Nodal Method

Abstract

Recently, an OECD benchmark problem was proposed to test the abilities of modern reactor physics codes to perform three-dimensional transport calculations without spatial homogenization of the fuel-coolant interfaces. To solve this benchmark problem we used the Argonne National Laboratory nodal transport code, VARIANT. VARIANT solves the even-parity transport equation on a nodal grid coupled with odd-parity Lagrange multipliers at the node interfaces. Each spatial node thus constitutes a primal hybrid finite element, with separate spatial approximations within the node and along the interface. The angular variables in the transport equation are treated with an expansion in either spherical harmonics or simplified spherical harmonics. To treat the spatial heterogeneities defined in the benchmark problem, we replaced the basis set of spatial polynomial trial functions used previously in VARIANT with a finite element spatial approximation. This approach subdivides the spatial node into finite elements with continuous, piecewise linear or quadratic trial functions; we refer to them as subelements. By allowing step changes in cross sections at the subelement interfaces, the spatial heterogeneities within the node can be treated explicitly. Using this new form of the VARIANT code we solved the two-dimensional and three-dimensional problems specified by the OECD benchmark and compared the results to reference multigroup Monte Carlo solutions.

Department(s)

Nuclear Engineering and Radiation Science

Comments

U.S. Department of Energy, Grant DE-FG07-98ID13632

International Standard Book Number (ISBN)

978-089448672-2

Document Type

Article - Conference proceedings

Document Version

Citation

File Type

text

Language(s)

English

Rights

© 2024 American Nuclear Society, All rights reserved.

Publication Date

01 Jan 2002

This document is currently not available here.

Share

 
COinS