VERIFICATION OF ITER SHIELDING CAPABILITY AND FENDL DATA BENCHMARKINGTHROUGH ANALYSIS OF BULK SHIELDING EXPERIMENT ON LARGE SS316 WATER ASSEMBLY BOMBARDED WITH 14 MEV NEUTRONS/

The recently developed FENDL-1 database, both in multigroup form (FEND L/MC-1.0) and continuous energy form (FENDL/MC-1.0) has been tested th rough analyzing a fusion integral experiment performed at the FNS faci lity, Japan, on a large bulk shielding assembly made of multilayers of SS316 and water. The assembly is a replica that simulates ITER shield ing blanket and is bombarded by a 14 MeV source placed at 30 cm from t he cylindrical assembly and housed inside a SS316 cylindrical can. Thi s activity is undertaken as part of co-operation with JAERI on executi ng the required neutronics R&D tasks for ITER shield design. The objec tives are (a) benchmarking FENDL-1 data and identifying any flaws that may exist in this newly developed database, and (b) examining the ran ge of discrepancy between the calculated nuclear parameters inside the assembly and the measured ones in terms of the ratio of the calculate d-to-experimental (CIE) data. Both differential and integral experimen tal data were analyzed along the central axis of the similar to 120 cm D x 140 cm L assembly. The analyses with the multigroup data, MG also included library derived from ENDF/B-VI data base for comparison purp oses. The MCNP Monte Carlo (MC) code was used with the FENDL/MC-1 data . The largest range of discrepancy between calculated and measured res ponses (reaction rates, neutron spectra, gamma ray heating, etc.) was found to be similar to 20-30% even though in most cases this discrepan cy falls within the experimental errors. (C) 1998 Elsevier Science S.A . All rights reserved.