Monte Carlo small-sample perturbation calculations
Conference
·
OSTI ID:6363564
Two different Monte Carlo methods have been developed for benchmark computations of small-sample-worths in simplified geometries. The first is basically a standard Monte Carlo perturbation method in which neutrons are steered towards the sample by roulette and splitting. One finds, however, that two variance reduction methods are required to make this sort of perturbation calculation feasible. First, neutrons that have passed through the sample must be exempted from roulette. Second, neutrons must be forced to undergo scattering collisions in the sample. Even when such methods are invoked, however, it is still necessary to exaggerate the volume fraction of the sample by drastically reducing the size of the core. The benchmark calculations are then used to test more approximate methods, and not directly to analyze experiments. In the second method the flux at the surface of the sample is assumed to be known. Neutrons entering the sample are drawn from this known flux and tracking by Monte Carlo. The effect of the sample or the fission rate is then inferred from the histories of these neutrons. The characteristics of both of these methods are explored empirically.
- Research Organization:
- Argonne National Lab., IL (USA)
- DOE Contract Number:
- W-31109-ENG-38
- OSTI ID:
- 6363564
- Report Number(s):
- CONF-830304-17; ON: DE83009588
- Country of Publication:
- United States
- Language:
- English
Similar Records
MONTE CARLO METHODS
A MONTE CARLO MULTI-BEND DUCT PROCEDURE
Calculations of reactivity perturbations using correlated sampling Monte Carlo
Technical Report
·
Mon Sep 01 00:00:00 EDT 1958
·
OSTI ID:4302729
A MONTE CARLO MULTI-BEND DUCT PROCEDURE
Technical Report
·
Sat Sep 15 00:00:00 EDT 1962
·
OSTI ID:4719323
Calculations of reactivity perturbations using correlated sampling Monte Carlo
Journal Article
·
Fri Dec 30 23:00:00 EST 1994
· Transactions of the American Nuclear Society
·
OSTI ID:89101