Impact of missing resonances in determining group cross sections and uncertainties

In the resolved- and unresolved-resonance regions, reconstruction of the pointwise cross sections is frequently carried out by using a cross-section formalism derived from the R-matrix theory that uses evaluated resonance parameters. The uncertainties in the cross section are due to the uncertainties in the resonance parameters. For reactor applications, group cross sections are produced by weighting the pointwise cross sections with a neutron flux spectrum. Consequently, the uncertainty in the group cross sections is also dependent on the uncertainty in the resonance parameters. Resonance parameters can be obtained by fitting the experimental data using generalized least-squares technique in conjunction with R-matrix theory. Such an approach is used in the computer code SAMMY.[1] Lack of high-quality resolution in the experimental data is a key problem that prevents the evaluators from determining resonance parameters for all individual resonances. This paper addresses the impact of missing resonances in determining average group cross sections and their corresponding uncertainties. To this end, a numerical study has been performed in the energy range 1500 to 1900 eV with a mock-up set of Reich-Moore resonance parameters for ²³⁵U. A similar study was previously done to determine the effect of missing resonances in energy self-shielding calculations of the group cross sections.[2].