A study of influence of the initial source distribution on convergence of multiplication factor (keff) in monte carlo calculation

Abstract

Convergence of the fission source spatial distribution in the fission source is one of the important criteria in the Monte Carlo calculation for the fission reactor criticality calculation to get unbiased value in the estimated multiplication factor (keff). The closer the initial source approximates the shape of the fundamental mode, the fewer cycles need to skip and this reduces the time required in a computer calculation. To study the influence of the initial source distribution on convergence of keff in Monte Carlo calculation, 10 different cases were conducted with five different initial point source positions with and without including inactive cycle in the MCNP5 code. Then, the Shannon entropy analysis and keff analysis with a reference keff value as a comparison were performed. From the concept, if the Monte Carlo calculation begins with more initial point sources, the numerical result will converge to a better final value rather than if insufficient or bad location initial point source was used and this concept was applied to get the reference keff value with 1 pcm standard deviation value. From this study, different locations of initial point source give different influence on the convergence of average keff value in Monte Carlo calculation. As a conclusion as shown in the results of case B and case C, the closer the initial point source to the center of the fission reactor core, the keff value converged to a better final average value. In the meantime, the Shannon entropy analysis could not show a clear indication on convergence rate of fission source spatial distribution for different locations of initial point sources and needs a better source convergence diagnostic tool for the convergence analysis particularly for convergence rate study.