[2E08-10] Nuclear Data Evaluation, Uncertainty Evaluation

[2E09] Development of an estimation method of uncertainties in reactor-physics quantities by a random sampling of nuclear data covariance, and application to Godiva

*Maho Kawaguchi¹, Naoki Yamano¹, Tsunenori Inakura¹, Satoshi Chiba¹ (1. Tokyo Tech)

Keywords:uranium235, uncertainty, MCNP, Monte Carlo calculation

We have developed a computational scheme to estimate quantitatively covariance of integral quantities in the reactor physics based on the covariance information of nuclear data libraries. We can select quantities to be perturbed by MF- and MT-wise manner. We applied this method to Godiva, by adopting ^{234, 235, 238}U data from JEND-4, and Sandy code generated randomly-sampled nuclear data. Then, we processed each random file by NJOY, and MCNP calculation was performed for each perturbed file. We took integral quantities such as k_eff, fission rate and capture rate, and their covariance was derived.