On-the-fly doppler broadening of unresolved resonance region cross sections via probability band interpolation

In this work we present a scheme for computing temperature-dependent unresolved resonance region cross sections in Monte Carlo neutron transport simulations. This approach relies on the generation of equiprobable cross section magnitude bands on an energy-temperature mesh. The bands are then interpo...

Full description

Bibliographic Details
Main Authors: Brown, Forrest B., Walsh, Jonathan Alan, Forget, Benoit Robert Yves, Smith, Kord S.
Other Authors: Massachusetts Institute of Technology. Department of Nuclear Science and Engineering
Format: Article
Language:en_US
Published: American Nuclear Society (ANS) 2017
Online Access:http://hdl.handle.net/1721.1/109720
https://orcid.org/0000-0002-2542-1149
https://orcid.org/0000-0003-1459-7672
https://orcid.org/0000-0003-2497-4312
Description
Summary:In this work we present a scheme for computing temperature-dependent unresolved resonance region cross sections in Monte Carlo neutron transport simulations. This approach relies on the generation of equiprobable cross section magnitude bands on an energy-temperature mesh. The bands are then interpolated in energy and temperature to obtain a cross section value. This is in contrast to the typical procedure of pre-generating probability tables at all temperatures present in the simulation. As part of this work, a flexible probability table generation capability is integrated into the continuous-energy neutron transport code OpenMC [1]. Both single-level and multi-level Breit-Wigner formalisms are supported, as is modeling of the resonance structure of competitive reactions. A user-specified cross section band tolerance is enabled with batch statistics. Probability tables are generated for all 268 ENDF/B-VII.1 [2] isotopes that have an unresolved resonance region evaluation. Integral benchmark simulations of the Big Ten critical assembly show that, for a system that is sensitive to the unresolved resonance region, a temperature interval of ∼200 K around 293.6 K is sufficient to reproduce the keff value that is obtained with probability tables generated exactly at room temperature. A finer mesh of < 50 K is required to reproduce some cross section values at the common target relative difference of 0.1%