This is not the latest version of this item. The latest version can be found here.
Converting point-wise nuclear cross sections to pole representation using regularized vector fitting
Name
Forget_Converting point-wise.pdf
Description
Accepted version
Size
1.73 MB
Format
Adobe PDF
Checksum (MD5)
39820a0e71b43a0e2cc302f84b3c8dbf
Author(s)
Peng, Xingjie
Ducru, Pablo
Liu, Shichang
Forget, Benoit
Liang, Jingang
Smith, Kord
Date Issued
2018
Journal
Computer Physics Communications
Publisher
Elsevier BV
Version
Author's final manuscript
Abstract
© 2017 Elsevier B.V. Direct Doppler broadening of nuclear cross sections in Monte Carlo codes has been widely sought for coupled reactor simulations. One recent approach proposed analytical broadening using a pole representation of the commonly used resonance models and the introduction of a local windowing scheme to improve performance (Hwang, 1987; Forget et al., 2014; Josey et al., 2015, 2016). This pole representation has been achieved in the past by converting resonance parameters in the evaluation nuclear data library into poles and residues. However, cross sections of some isotopes are only provided as point-wise data in ENDF/B-VII.1 library. To convert these isotopes to pole representation, a recent approach has been proposed using the relaxed vector fitting (RVF) algorithm (Gustavsen and Semlyen, 1999; Gustavsen, 2006; Liu et al., 2018). This approach however needs to specify ahead of time the number of poles. This article addresses this issue by adding a poles and residues filtering step to the RVF procedure. This regularized VF (ReV-Fit) algorithm is shown to efficiently converge the poles close to the physical ones, eliminating most of the superfluous poles, and thus enabling the conversion of point-wise nuclear cross sections.
Terms of Use
Creative Commons Attribution-NonCommercial-NoDerivs License
Persistent DSpace Link
DOI of Published Version
10.1016/J.CPC.2017.12.004
