Popis: |
Perhaps the two most abstract concepts in Monte Carlo sampling are variance reduction and next-event estimators. Variance reduction techniques enable significant enhancement of statistical convergence. They are then illustrated with safeguard examples of a lead slowing-down spectrometer and a Cf shuffler using weight windows, time splitting, and other techniques. Next-event estimators enable robust calculation of highly improbable events and challenging problems such as distributed sources in a UF6 enrichment cask model. Advanced methods are also required to model various safeguard detectors. These are illustrated with 3He pulse-height tallies, 3He coincidence tallies, and 10Be-lined detectors for 3He replacement. |