Complexity Measures And Occurrence Of The "Breakpoint" In The Neutron And Gamma-Rays Time Series Measured With Organic Scintillators

This paper deals with the first analysis of the neutron and gamma time series measured with organic scintillators from plutonium samples by using information measures. Fast neutron detection with organic scintillators has been widely used for various nuclear safeguards applications and homeland security. One of the significant attributes of special nuclear materials (SNM) is the high multiplicity events in a short period of time. The time distributions of neutron and gamma-rays events for the plutonium metal plates designed as fuel plates for the Idaho National Laboratory (INL) Zero Power were measured with the Fast Neutron Multiplicity Counter (FNMC) consisting of 8 EJ-309 liquid scintillators and 8 stilbene detectors. Since the neutron correlated counts within the coincidence window of 40 ns are related to 240Pu effective mass of plutonium metal plates, it is of interest to investigate the randomness of the measured neutron and gamma-rays events. To access such information, we resort to complexity measures in the hope of being able to connect complexity values with the reliability of detection. That was done through (i) application of Kolmogorov complexity (KC) and its derivatives [Kolmogorov spectrum and its highest value (KCM) and running complexity (RKC)] and (ii) establishing the "breaking point" after which there exists a sharp drop in the running Kolmogorov complexity of the neutron and gammarays time series. It was found that the complexity of all the time series detected from the sample with 5, 9, 11, 13, and 15 plutonium plates had the high almost identical values of KC while the sample with 3 plates had by onethird smaller KC values than all the others. These calculations were supplemented by the Lypaunov exponents for a time series and the National Institute of Standards and Technology (NIST)tests. The low KC values can be addressed to the different sources of uncertainties in measuring procedure with the sample consisting of three plates. The complexity measures applied in this study are capable of revealing aspects of information that would otherwise remain hidden to the one-off complexity estimate.