Year
2008
Abstract
The paper presents the evaluation of various approaches to simulate non destructive assay of plutonium samples, including MOX, by neutron multiplicity counting, performed within the framework of the co-operation agreement between IRSN and the US DOE. Basically, two calculation schemes are described for simulating the neutron emissions from the source up to the count rates of interest (singles, doubles and triples) obtained in the final electronic acquisition device. These calculation schemes are based on the use of the multi-particle transport code MCNPX 2.6.c. The 1st calculation scheme results from the direct calculation of the values of interest within the MCNPX code. It is based on the number of captures in the counters for the singles rates, and on the distribution of the number of coincident captures in the counters for practical values of predelay and gate width, in order to assess the doubles and triples count rates. The 2nd calculation scheme, which needs to post-process the data, uses MCNPX to generate pulse output files. These files contain a list of histories reproducing the life of the neutrons emitted by a primary event, including the neutrons induced by secondary events, from their birth at time zero up to their capture in the counters, i.e. when they are detected. Those results are post-processed in two steps: the first step is the generation of a typical neutron pulse train; the second step is an analysis in a data processing software notably simulating a multiplicity shift register in order to derive the resulting singles, doubles and triples rates. The validation of the simulation procedure is firstly addressed by a comparison of the aforementioned calculation schemes, and secondly by direct comparison with experimental data.