Magnitude of bias in Monte Carlo eigenvalue calculations
Conference
·
OSTI ID:5701979
Most Monte Carlo eigenvalue calculations are based on power iteration methods, like those used in analytical algorithms. But if N/sub H/, the number of histories in each generation is fixed, then such Monte Carlo calculations will be biased. Various arguments lead to the conclusion that eigenvalue and shape biases are both proportional to 1/N/sub H/, but little more is known about their magnitudes. Numerical experiments on simple matrices suggest that the biases are small, but information more relevant to real reactor calculations is very sparse. In fact to determine the bias in real reactor calculations is quite expensive. It seems worthwhile, therefore, to try to understand the Monte Carlo biases in systems more realistic than arbitrary matrices, but simpler than real reactors. For this reason biases in simple one-group model problems have been computed.
- Research Organization:
- Argonne National Lab., IL (USA)
- DOE Contract Number:
- W-31109-ENG-38
- OSTI ID:
- 5701979
- Report Number(s):
- CONF-831047-59; ON: DE83015372
- Country of Publication:
- United States
- Language:
- English
Similar Records
Estimation of fission source bias in Monte Carlo eigenvalue calculations
Error estimations and their biases in Monte Carlo eigenvalue calculations
Variational Variance Reduction for Monte Carlo Eigenvalue and Eigenfunction Problems
Conference
·
Thu Dec 31 23:00:00 EST 1992
· Transactions of the American Nuclear Society; (United States)
·
OSTI ID:6844680
Error estimations and their biases in Monte Carlo eigenvalue calculations
Journal Article
·
Tue Dec 31 23:00:00 EST 1996
· Nuclear Science and Engineering
·
OSTI ID:445398
Variational Variance Reduction for Monte Carlo Eigenvalue and Eigenfunction Problems
Journal Article
·
Sat Feb 14 23:00:00 EST 2004
· Nuclear Science and Engineering
·
OSTI ID:20804909