Importance Sampling Implemented in the Code PRIZMA for Deep Penetration and Detection Problems in Reactor Physics
Russian Federal Nuclear Center Zababakhin All-Russian Research Institute of Applied Physics (RFNC-VNIITF) Snezhinsk, Russia 456770
* Corresponding Author, E-mail: firstname.lastname@example.org
At RFNC-VNIITF a code PRIZMA which has been developed for more than 30 years is used to model radiation transport by the Monte Carlo method. The code implements individual and coupled tracking of neutrons, photons, electrons, positrons and ions in 1D, 2D or 3D geometry. Attendance estimators are used for tallying, i.e., the estimators whose scores are only nonzero from particles which cross a region or surface of interest. Importance sampling is used to make deep penetration and detection calculations more effective. However, its application to reactor analysis appeared peculiar and required further development.
The paper reviews methods used for deep penetration and detection calculations by PRIZMA. It describes in what these calculations differ when applied to reactor analysis and how we compute approximated importance functions and parameters for biased distributions. Methods to control the statistical weight of particles are also discussed.
A number of test and applied calculations which were done for the purpose of verification are provided. They are shown to agree either with asymptotic solutions if exist, or with results of analog calculations or predictions by other codes. The applied calculations include the estimation of ex-core detector response from neutron sources arranged in the core, and the estimation of in-core detector response.
Key words: particle transport / Monte Carlo method / importance sampling / deep penetration / detection / PRIZMA code
© Owned by the authors, published by EDP Sciences, 2014