Accelerating MCNP using the Adjoint Flux

Please enable javascript in your browser.

Shaddel, Poorshad | 2017

618 Viewed

Type of Document: M.Sc. Thesis
Language: Farsi
Document No: 49470 (46)
University: Sharif University of Technology
Department: Energy Engineering
Advisor(s): Vosoughi, Naser; Ayyoubzadeh, Mohsen
Abstract:
Nuclear engineering simulation requires computer codes which are capable of particle transportation and presenting valid answers. The Monte Carlo method is one of the particle transportation methods and the MCNP code is also a common and efficient code using Monte Carlo method. We usually come across statistical errors in Monte Carlo simulation for reducing which a number of methods have been presented which are called variance reduction methods. In simulating deep penetrating problems, analog Monte Carlo method is weak which leads to considerable error. The chosen variance reduction method for this study is the one which uses adjoint flux. In this study, a 3D deterministic transport code (IR-SN-3D) and its coupling with Monte Carlo method in order to obtain the adjoint flux and the variance reduction parameters are used to accelerate MCNP calculations. After coupling Monte Carlo with the deterministic transport code, a notable improvement was observed for the one energy group for the chosen geometries but as for the two energy group, the figure of merit was not as improved due to the usage of the weight window card
Keywords:
Variance Reduction ; Monte Carlo N-Particle (MCNP)Code ; Monte Carlo Method ; Deterministic Calculations ; Adjoint Flux ; Deterinistic Transport Code

No TOC