Abstract
Seventeen integral fusion experiments have been performed so far within the USDOE/JAERI Collaborative Program on Fusion Neutronics. The main objective of these experiments is to verify the state-of-the-art neutron transport codes and nuclear data in predicting tritium production rates, in-system neutron spectra, activation reaction rates, nuclear heating, and gamma decay heating in a Li2O test assembly. In performing these experiments, the incident neutron source condition and the experimental geometrical arrangements for the test assembly were altered to study the impact of system changes on the prediction capability for the key neutronics parameters, particularly tritium production rate both locally and globally within predesignated zones in the breeding material. The test assembly itself was changed from a simple, one-material zone to a more prototypical blanket that included a stainless-steel first wall, neutron multiplier (beryllium) and coolant channels. The experiments proceeded through phase I and IIIA. In the latter phase, a line source was simulated by cyclic movement of the annular test assembly relative to the stationary point source that is located axially at the center of the inner cavity. In the latter phase, a better simulation has been achieved to the secondary energy, and angular distributions of the incident neutron source found in Tokamak plasmas. In this paper, the results obtained by the USA, quantified in terms of the calculated-to-experimental values (C/E's) for the key neutronics parameters, are discussed for all the experiments performed so far. The change in the trends of these C/E values as one moves from one phase to another is considered by statistically treating these C/E values to arrive at a mean value for the prediction uncertainty in each experiment and an average mean value to all the experiments. This was carried out for tritium production rate, in-system spectra, and other reaction rates.