Analysis of prompt emissions following photon interrogation of fissionable material
Active interrogation systems are at the forefront of technologies for the detection of concealed nuclear material. Because such systems are large and complex, Monte Carlo is the preferred method for their simulation and design. The MCNP-PoliMi code simulates time-analysis quantities resulting in very accurate detector response predictions. However, this code has not been applied to problems involving photonuclear physics. The goal of this work is to develop, implement, and assess this capability in the MCNP-PoliMi code system. This thesis has three main contributions: (i) implementing photofission neutron energy distribution models into MCNP-PoliMi; (ii) developing a MCNP-PoliMi postprocessor subroutine to model the time-width of the accelerator pulse; and (iii) developing a novel method to assess the sensitivity of photoneutron production to perturbations in cross-section data. The enhanced version of MCNP-PoliMi was then used to simulate the results of various photon interrogation experiments. Only minor discrepancies are observed at low photon energies, however the discrepancies increased dramatically (up to 100%) as photon energy increased (up to 20 MeV). Many possible explanations for these discrepancies were analyzed (e.g., the MCNP-PoliMi detection model, the simulation geometry, the bremsstrahlung source specification, and detector deadtime effects), none of which resolved the discrepancies. Observed uncertainties in cross-section data suggested another possible explanation for these discrepancies that needed to be explored. Sensitivity and perturbation analysis of photonuclear cross sections have not been reported in the literature in the past. Furthermore, the MCNPX perturbation routines are not available for photonuclear applications. Therefore, a new methodology was developed to address this issue. The results of this analysis showed that the uncertainty in the results of the simulated experiments such as those considered here would be less than 10% even if the uncertainty in the cross-section data reaches 20%. This leaves the observed discrepancies unexplained and requires future work in this area. The energy distribution of the photoneutrons suggests another possible explanation for the discrepancies. Future experiments and analysis in this direction would further develop MCNP-PoliMi as a valuable simulation tool for the design and analysis of active interrogation systems.
Pozzi, Purdue University.
Off-Campus Purdue Users:
To access this dissertation, please log in to our