R matrix theory and light element standards evaluation

R matrix theory was used for the evaluation of the standard cross-sections for the neutron induced reactions 6Li(n,t), 10B(n,α1γ) and 10B(n,α). Use of this model is justified because the R matrix theory gives an accurate parameterization of the multichannel-multilevel nuclear reaction cross-sections in the resonance region, reproducing the shape of the cross-sections even in the presence of strong interference effects. The influence of the three-body break up channel on the standard reactions can be accommodated by the inclusion of a pseudo-channel. All observable quantities for which measured data exist can be calculated by R matrix theory, and therefore may be included in the fitting procedure. The list of such observables in the evaluation of standard reaction cross-sections includes all non-negligible partial reaction cross sections, the total cross-section, angular distributions of secondary particles and polarizations. Data for all reactions that form the 7Li and 11B systems were analysed, including inverse and charged particle induced reactions. Comparisons of the results of evaluations obtained with the EDA and RAC R matrix fits for the 6Li(n,t), 10B(n,α0), 10B(n,α1) and 10B(n,α) reactions are shown as ratios to the 1987 GMA combined fit used to obtain the ENDF/B-VI standards. The different procedures in the codes led to evaluated data. Fortunately, these discrepancies generally occur in an energy range higher than that for which the cross-sections are used as standards. Uncertainties of the evaluated data obtained with EDA are lower in most cases than those estimated with RAC. The figures demonstrate clearly the same trends in the new EDA and RAC fits when compared with the 1987 fits obtained using EDA, and reveal the larger differences in the R matrix evaluations for these standard reactions. Such differences cannot be removed by further analysis, or by improving the consistency of the procedures used in EDA and RAC. The difference between the two fits can be related to the uncertainty of the evaluation method; consequently an additional component of the uncertainty equal to half the difference between the two fits was added to the components of the total uncertainty of the R matrix evaluation