Global Results from Deterministic and Stochastic Analysis of the MUSE-4 Experiments on the Neutronics of Accelerator-Driven Systems

Le résumé fourni par la source

The MUSE-4 program is a series of zero-power experiments carried out at the Commissariat à l’Energie Atomique Cadarache MASURCA nuclear facility from 2001 to 2004 to study the neutronics of accelerator-driven systems (ADSs). The program has investigated the coupling of a multiplying medium to neutron sources of 2.6 or 14 MeV provided by an accelerator (GENEPI) via D(d,n)3He or T(d,n)4He nuclear fusion reactions, respectively. The fuel was UO2-PuO2, the simulated coolant was sodium or lead, and the multiplication factor keff ranged from 1 to 0.95. The aim of the experiment was to develop new measurement techniques specific to ADSs and to test the performances of neutronic calculations codes for such systems.The interpretation of the MUSE-4 experiment has shown that the physical parameters of the system are globally well reproduced by calculations performed with the ERANOS code system, which proves good agreement with both the measurements and the reference Monte Carlo calculations; this concerns the critical mass, the delayed neutron fraction, the fission rate shapes, and the spectral indices. This is a particularly remarkable issue for ERANOS and its associated libraries, which had never been tested for such situations.Concerning the nuclear data, JEF-based cross sections provide a better agreement on critical mass than other libraries. A sensitivity of several measured parameters to the elastic and inelastic cross section of lead have been demonstrated, and possible biases on these cross sections have been indicated.We have shown that several methods based on deterministic or stochastic calculations allow us to relate the experimental neutron population decay after a source pulse with the reactivity of the system; these reactivity determination techniques are in good agreement with standard reactivity measurement techniques.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.