Multivariate High Order Statistics of Measurements of the Temporal Evolution of Fission Chain-Reactions

Mattingly, J.K.

Oak Ridge National Laboratory

关键词: Fission;    Statistics;    73 Nuclear Physics And Radiation Physics;    Dimensions;    Prompt Neutrons;   

DOI  :  10.2172/814036
RP-ID  :  ORNL/TM-2001/45
RP-ID  :  AC05-00OR22725
RP-ID  :  814036

美国|英语

来源: UNT Digital Library

【 摘 要 】

The development of high order statistical analyses applied to measurements of the temporal evolution of fission chain-reactions is described. These statistics are derived via application of Bayes' rule to conditional probabilities describing a sequence of events in a fissile system beginning with the initiation of a chain-reaction by source neutrons and ending with counting events in a collection of neutron-sensitive detectors. Two types of initiating neutron sources are considered: (1) a directly observable source introduced by the experimenter (active initiation), and (2) a source that is intrinsic to the system and is not directly observable (passive initiation). The resulting statistics describe the temporal distribution of the population of prompt neutrons in terms of the time-delays between members of a collection (an n-tuplet) of correlated detector counts, that, in turn, may be collectively correlated with a detected active source neutron emission. These developments are a unification and extension of Rossi-a, pulsed neutron, and neutron noise methods, each of which measure the temporal distribution of pairs of correlated events, to produce a method that measures the temporal distribution of n-tuplets of correlated counts of arbitrary dimension n. In general the technique should expand present capabilities in the analysis of neutron counting measurements.