Initialising ...
Initialising ...
Initialising ...
Initialising ...
Initialising ...
Initialising ...
Initialising ...
Hashimoto, Shintaro
no journal, ,
no abstracts in English
Hashimoto, Shintaro; Sato, Tatsuhiko
no journal, ,
PHITS can simulate a behavior of various particles such as neutrons and photons, and has been used for shielding calculations in accelerator facilities. Reliability of PHITS based on the Monte Carlo method can be usually evaluated by statistical uncertainties related with the number of trials. However, uncertainty of input parameters such as densities and compositions of shielding materials might influence on results of the shielding calculation. Estimation of the influence as systematic uncertainties gives quantitative discussion of the calculation. We proposed a new method to estimate systematic uncertainties with ANOVA, and then applied the method to analysis of neutron shielding calculations. In the analysis, the effect of the uncertainty of water contents in concrete materials on neutron fluences could be estimated quantitatively.
Hashimoto, Shintaro; Sato, Tatsuhiko
no journal, ,
Particle transport simulation code PHITS has been successfully applied to shielding calculations in accelerator facilities. We developed a function to estimate reliability of PHITS results as systematic uncertainty based on ANOVA (analysis of variance), where it is assumed that the results show a normal distribution. In this study, we applied the bootstrap method to analysis of non-normal distribution, which is rarely caused when uncertainty of nuclear reaction models has the biased distribution. The confidence interval estimated by the bootstrap method was more appropriate than that obtained by the assumption of the normal distribution for use as an indicator of the reliability of a PHITS result in a neutron shielding calculation. However, the interval by the bootstrap method was narrow compared to the distribution of the original result.