Uncertainty Analysis Method for Activation Neutron Spectra Based on Random Sampling

To improve the accuracy of neutron spectrum measurement via activation method and address the critical issue of uncertainty analysis, this study proposed a random sampling-based uncertainty analysis method integrating experimentally measured activation rates, cross-section covariance, and preset spectra for the SD spectrum unfolding program developed using an iterative method. By developing an iterative method-driven SD spectrum unfolding program, the impacts of activation rates, cross-section covariance, and preset spectra on spectrum unfolding uncertainties were systematically quantified by the system. A multi-factor coupling analysis model was constructed based on the principles of random sampling and validated using neutron spectrum experimental data from the China Experimental Fast Reactor (CEFR). The results demonstrate that: Neutron spectrum uncertainties exhibit non-uniform distribution across the full energy range, with the maximum uncertainty of spectrum unfolding in the thermal region reaching 24%, while uncertainties in the fast region remain below 8%. The preset spectrum dominates the unfolding uncertainties (contributing over 90%), whereas cross-section covariance and activation rates have negligible effects (contributing less than 5%). Compared to traditional methods, the random sampling-based analysis method provides a more comprehensive analysis of uncertainty sources, offering not only a reliable reference for enhancing neutron spectrum measurement accuracy but also serving as an effective tool for evaluating the performance of spectrum unfolding programs.