Modeling the Production of Beta-Delayed Gamma Rays for the Detection of Special Nuclear Materials

Abstract

The objective of this LDRD project was to develop one or more models for the production of β-delayed γ rays following neutron-induced fission of a special nuclear material (SNM) and to define a standardized formatting scheme which will allow them to be incorporated into some of the modern, general-purpose Monte Carlo transport codes currently being used to simulate inspection techniques proposed for detecting fissionable material hidden in seagoing cargo containers. In this report, we will describe a Monte Carlo model for β-delayed γ-ray emission following the fission of SNM that can accommodate arbitrary time-dependent fission rates and photon collection histories. The model involves direct sampling of the independent fission yield distributions of the system, the branching ratios for decay of individual fission products and spectral distributions representing photon emission from each fission product and for each decay mode. While computationally intensive, it will be shown that this model can provide reasonably detailed estimates of the spectra that would be recorded by an arbitrary spectrometer and may prove quite useful in assessing the quality of evaluated data libraries and identifying gaps in the libraries. The accuracy of the model will be illustrated by comparing calculated and experimental spectra from the decay of short-lived fission products following the reactions 235 U(n th , f) and 239 Pu(n th , f). For general-purpose transport calculations, where a detailed consideration of the large number of individual γ-ray transitions in a spectrum may not be necessary, it will be shown that a simple parameterization of the γ-ray source function can be defined which provides high-quality average spectral distributions that should suffice for calculations describing photons being transported through thick attenuating media. Finally, a proposal for ENDF-compatible formats that describe each of the models and allow for their straightforward use in Monte Carlo codes will be presented.

User Manual for MC2-3 Gamma Library Generation

“…Unfortunately, the yield data for delayed gamma are only given for 235 U and 239 Pu in the latest ENDF/B VII.1 library. Even though it reproduces the gamma ray spectra resulting from neutron-induced fission using 1,000 decay nuclides and 20,000 discrete-emission energies from England and Rider [9] and the NuDat nuclear database [10], the results are limited by the lack of experimental data and the uncertainties of fission product yields [11]. For these reasons, the PreGAMMA code calculates the delayed photon yield data using the latest decay and fission yield sublibraries.…”

Analysis of a Gross Counting Decision Metric for use in Threat Detection During Cargo Container Inspection

“…T (crit) at its origin analytically intractable from any practical point of view; however, as intrepid scientists we needn't be discouraged by this minor inconvenience, because this is a perfect example the sort of problem that God created Monte Carlo codes to address! Modern, general-purpose Monte Carlo radiation transport algorithms such as LLNL's COG code package [6] are now capable of simulating not only the physical layout of proposed cargo inspection systems such as the NCW, but also the detailed characteristics of the system's highintensity D(d,n) 3 He neutron source [7] and the β-delayed γ-ray spectra emitted during the decay of neutron-induced SNM fission products [8]. A series of Monte Carlo simulations of the NCW system could easily be done with different physical configurations of SNM placed in the most challenging locations in a container with "worst-probable-case" cargo configurations to obtain credible values for the limiting number of γ rays that one might expect to see at the detector interface in those scenarios.…”

Nuclear car wash status report, August 2005

“…N cycles of 30 sec with beam on followed by 100 sec with beam off (during which delayed gammas are recorded)). Coupled with our unique b-delayed g-ray database [10]. This should allow us to model actual or proposed Active Cargo Interrogation experiments at LLNL with a hitherto unprecedented level of accuracy.…”

“…Implementing these capabilities in a parallel processor was completed under the current DHS support. Details of these developments have been described elsewhere [10][11][12].…”