Methods for U. S. shielding calculations: applications to FFTF and CRBR designs

The primary components of the U.S. reactor shielding methodology consist of: (1) computer code systems based on discrete ordinates or Monte Carlo radiation transport calculational methods; (2) a data base of neutron and gamma-ray interaction and gamma-ray-production cross sections used as input in the codes; (3) a capability for processing the cross sections into multigroup or point energy formats as required by the codes; (4) large-scale integral shielding experiments designed to test cross-section data or techniques utilized in the calculations; and (5) a ''sensitivity'' analysis capability that can identify the most important interactions in a transport calculation and assign uncertainties to the calculated result that are based on uncertainties in all of the input data. The required accuracy for the methodology is to within 5 to 10% for responses at locations near the core to within a factor of 2 for responses at distant locations. Under these criteria, the methodology has proved to be adequate for in-vessel LMFBR calculations of neutron transport through deep sodium and thick iron and stainless steel shields, of neutron streaming through lower axial coolant channels and primary pipe chaseways, and of the effects of fuel stored within the reactor vessel. For ex-vessel LMFBR problems, themore » methodology requires considerable improvement, the areas of concern including neutron streaming through heating and ventilation ducts, through the cavity surrounding the reactor vessel, and through gaps around rotating plugs in the reactor heat, as well as gamma-ray streaming through plant shield penetrations.« less