Year
2015
Abstract
The design and calibration of passive neutron multiplicity counters have become increasingly reliant on modeling using sophisticated Monte Carlo radiation transport codes. Modern codes (such as Monte Carlo) are able to faithfully reproduce the geometry, materials, and physical processes taking place in the item and detector system. However, the limiting factor in using these codes to perform absolute performance predictions is our imperfect knowledge of basic nuclear data used in these codes. We are interested in pushing current state-of-the-art modeling for nondestructive assay applications applied to safeguards. In this paper we consider one aspect: how the uncertainty in the spectral representation of 252Cf neutrons propagates into the uncertainty on predicted detection efficiency. Instead of using the Watt representation adopted by default in MCNP, we use a more exact, piecewise description from the evaluation made by Mannhart [1]. We use MCNP simulations along with experiments of the Oak Ridge National Laboratory Large-Volume Active Well Coincidence Counter (LVAWCC) to quantify the uncertainty in the prompt neutron fission spectrum of 252Cf. We also comment on the need for comprehensive data of similar quality for other fissioning systems.