Systematic effects on cross section data derived from reaction rates in reactor spectra and a re-analysis of 241Am reactor activation measurements

Methodologies to derive cross section data from spectrum integrated reaction rates were studied. The Westcott convention and some of its approximations were considered. Mostly measurements without and with transmission filter are combined to determine the reaction cross section at thermal energy together with the resonance integral. The accuracy of the results strongly depends on the assumptions that are made about the neutron energy distribution, which is mostly parameterised as a sum of a thermal and an epi-thermal component. Resonance integrals derived from such data can be strongly biased and should only be used in case no other data are available. The cross section at thermal energy can be biased for reaction cross sections which are dominated by low energy resonances. The amplitude of the effect is related to the lower energy limit that is used for the epi-thermal component of the neutron energy distribution. It is less affected by the assumptions on the shape of the energy distribution. When the energy dependence of the cross section is known and information about the neutron energy distribution is available, a method to correct for a bias on the cross section at thermal energy is proposed. Reactor activation measurements to determine the thermal 241Am(n,γ) cross section reported in the literature were reviewed. In case enough information was available, the results were corrected to account for possible biases and included in a least squares fit. These data combined with results of time-of-flight measurements give a capture cross section 720 (14) b for 241Am(n,γ) at thermal energy.