Unexploded Ordnance identification—A gamma-ray spectral analysis method for Carbon, Nitrogen and Oxygen signals following tagged neutron interrogation

A novel gamma-ray spectral analysis method has been demonstrated to optimally extract the signals of the signature elements of explosives, carbon (C), nitrogen (N) and oxygen (O) from 57–155 mm projectiles following tagged neutron interrogation with 14 MeV neutrons. The method was implemented on Monte Carlo simulated, synthetic spectra of Unexploded Ordnance (UXO) that contained high explosive fillers (Composition B, TNT or Explosive D) within steel casings of appropriate thicknesses. The analysis technique defined three broad regions-of-interest (ROI) between 4–7.5 MeV of a spectrum and from a system of three equations for the three unknowns namely C, N and O, the maximum counts from each of these elements were extracted. Unlike conventional spectral analysis techniques, the present method included the Compton continuum under a spectrum. For a neutron output of ∼2×107 ns−1 and using four 12.7 cm diameter×12.7 cm NaI(Tl) detectors, the C/N and C/O gamma-ray counts ratios of the explosive fillers were vastly different from that of an inert substance like sand. Conversion of the counts ratios to elemental ratios could further discriminate the different types of explosive fillers. The interrogation time was kept at ten minutes for each projectile.