Optimization of Monte Carlo methods for calculational predictions of dosimetry measurements in "JOYO" MK-III core

Wootan, D. W.

One of the primary missions of the "JOYO" experimental fast reactor is to perform irradiation tests of fuel and structural materials to support the development of fast reactors. From 1983 to 2000 "JOYO" was operated with the MK-II irradiation test core. In 2003 the "JOYO" reactor upgrade to the MK-III core was completed to increase the irradiation testing capability. The MK-III core incorporates significant changes to the core size and arrangement, fuel enrichment, and reactor power level compared to the MK-II core. Accurate core calculational methods are required for predicting neutron fluence, related spectral information, and other key performance parameters for new fuels and materials irradiation tests in the MK-III core. An extensive set of neutron dosimeters ranging from the core center to beyond the reactor vessel were irradiated during the first two operating cycles of the MK-III core. This dosimetry data, along with calculational predictions, will allow a detailed characterization of the neutron spectra and flux distributions for this new core configuration. Previous methods applied at "JOYO" for predicting neutron fields included the MAGI three dimensional diffusion-theory based core management code system for the fuel region and the DORT two dimensional deterministic transport code for ex-core regions. Recently Monte Carlo transport calculations using the MCNP code have been introduced to account for heterogeneous effects. This paper describes the development of generalized whole-core Monte Carlo analysis methods, including heterogeneity effects, to provide accurate calculational predictions for the detailed irradiation environments of various "JOYO" irradiation test locations down to the level of individual specimens or dosimeters. Efficient calculational strategies using the continuous energy MCNP code were investigated and variance reduction techniques were optimized for each unique type of test environment. The accuracy of the methods was ...