Application of IRDF2002 dosimetry data to shielding benchmarks using the Monte Carlo code MCBEND

The IRDF2002 dosimetry cross section data are processed into very fine energy groups and added to the detector response library for the Monte Carlo code MCBEND. The IRDF2002 covariances are processed into broad groups to form a MCBEND detector covariance library. A number of benchmark experiments are analysed using MCBEND with both IRDF2002 and IRDF90 data and the results compared. Uncertainties due to material and detector cross sections are calculated automatically. The results demonstrate that the IRDF2002 115 In(n,n � ) 115m In dosimetry cross sections give improved agreement between calculation and measurement in an iron benchmark and that use of IRDF2002 data allows straightforward, explicit and accurate calculation of single resonance detector reaction-rates, e.g., 197 Au(n,γ) 198 Au. Nuclear data form an integral part of dosimetry analyses: neutron transport is determined by material nuclear data (cross sections and secondary angle/energy distributions) and indi- cated by detector reaction-rates which are also dependent on detector cross sections. Thus improvements in detector cross sections may have a significant effect on such analyses. The uncertainties on detector reaction-rates due to uncertainties in the material data and detector cross sections are also fundamental to dosimetry. Generally one of the purposes of dosimetry analyses is to benchmark computer codes and nuclear data by comparing calculated and measured detector reaction-rates. These can be on reactor plant or in material benchmark experiments such as those contained in the SINBAD database (1). The results may be used to assign correction factors and also uncertainty when predicting quantities such as displacements per atom (dpa) using the same computer code and data. Improvements in detector cross sections will help to improve the agreement between calculated and measured reaction-rates thus reducing correction factors and assigned uncertainties. Uncertainties on material cross sections will affect both detector reaction-rates and quantities such as dpa but those on detector cross sections will only affect the detector results. This should be taken into account when assigning correc- tion factors or uncertainties. Traditionally computation of uncertainty on detector reaction-rates has been a somewhat laborious process and thus the domain of the expert rather than the regular analyst. Hence automatic calculation of the contributions to this uncertainty is an important step forward in making uncertainty analysis more accessible. One particular difficulty for the dosimetry analyst is the accurate calculation of reaction-rates for detectors that contain large resonances, such as 55 Mn(n,γ) 56 Mn and 197 Au (n,γ) 198 Au. The resonance causes the flux to be suppressed significantly and this suppression needs to be taken into a Presenting author, e-mail: christopher.dean@sercoassurance.com account in the calculation. This can be done either by applying a dilute cross section and modelling the foil in its environment or by applying a suppressed cross section and not including the foil. Adequate resolution of the resonance is important.