Nuclear data uncertainty in iterative neutron spectrum unfolding

Abstract

ABSTRACT We developed an algorithm for propagating the uncertainty of dosimetry cross sections to that of the neutron spectrum obtained through multiple-foil activation and unfolding process. The algorithm derives the uncertainty of the solution spectrum as the standard deviation of a set of solution spectra obtained by unfolding with numerous sets of dosimetry cross sections, which are randomly generated from the evaluated cross sections and their covariances in a data library. We simulated a multiple-foil activation experiment in a C(d,n) neutron field by PHITS code for some dosimetry reactions to test the algorithm. The GRAVEL code analyzed production amounts to unfold neutron spectrum and then our new algorithm derived the uncertainties of the resultant neutron spectrum originated in the nuclear data uncertainty. Further, we found the uncertainty of the solution spectrum becomes smaller if we ignore the off-diagonal elements of the covariance matrices of the dosimetry cross sections.