Abstract
The iron, carbon and beryllium cross sections in JENDL-3 have been tested by the continuous energy Monte Carlo analysis of the neutron shielding benchmark experiments. The iron cross sections have been tested with analysis of the ORNL and the Winfrith experiments using the fission neutron sources, and also the LLNL iron experiment using the D-T neutron source. The carbon and beryllium cross sections have been tested with the JAERI-FNS TOF experiments using the D-T neutron source.
Revision of the subroutine TALLYD and an appropriate weight-window-parameter assignment have been accomplished in the MCNP code. In consequence, the FSD for each energy bin is reduced so small that the Monte Carlo results for neutron energy spectra could be recognized to be reliable.
The Monte Carlo calculations with JENDL-3 indicate a good agreement with the benchmark experiments in a wide energy range, as a whole. Particularly, for the Winfrith iron experiment, the results with JENDL-3 give better agreement, just below the iron 24 keV window, than that with ENDF/B-IV. For the JAERI-FNS TOF graphite experiment, the calculated angular fluxes with JENDL-3 give closer agreement than that with ENDF/B-IV at several peaks and dips caused by the inelastic scattering.
However, distinct underestimation is observed in the calculated energy spectrum with JENDL-3 between 0.8 and 3.0 MeV for the two iron experiments using fission neutron sources.
