Monte Carlo simulations to advance characterisation of landmines by pulsed fast/thermal neutron analysis

https://doi.org/10.1016/j.apradiso.2004.02.014Get rights and content

Abstract

The performance of a detection system based on the pulsed fast/thermal neutron analysis technique was assessed using Monte Carlo simulations. The aim was to develop and implement simulation methods, to support and advance the data analysis techniques of the characteristic γ-ray spectra, potentially leading to elemental characterisation of innocuous objects using the full spectrum analysis (FSA) approach. The simulations were carried out with a simplified tool, based on a 14 MeV DT pulse-neutron source and a bismuth–germanate detector. A MCNP-based method for de-coupling the radiation transport in mixed (n,γ) fields, to generate separate sets of standard detector γ-ray responses for individual elements, is outlined. When normalised and experimentally benchmarked in terms of the pulse-neutron source production rate, the standard spectra can be incorporated into algorithms for the FSA of in situ measurements and elemental fingerprinting of the inspected object.

Introduction

Urged by the latest tragic events and their aftermath, the world's attention has been re-captured to detection and characterisation of narcotics, explosives, contraband, landmines and unexploded ordnance (Hussein and Waller, 2000; BBC, 2003). For these applications, the development of nuclear radiation–detection techniques has strongly been stimulated to provide elemental characterisation and “fingerprint” signature of inspected objects.

However, the actual quantities measured through nuclear techniques do not always correspond to the geological/material parameters desired, a difference that is important to understand. Thus, additional information is required to relate the measurements made with nuclear techniques to representative parameters. As the extent to which laboratory calibrations can mimic complex in situ conditions is limited, Monte Carlo (MC) simulations may significantly contribute to the solution of this problem.

Although MC-based techniques are becoming more and more a prelude to experimental and design work, their applications to humanitarian de-mining (HD) are only very recent. The MC code MCNP (Briesmeister, 1999) was applied on a simple idealised geometry of a land-mine detector exposed to a 241Am–Be neutron source (Hussein, 1999). The results provide some basic conclusions in terms of detector response for thermal and fast neutrons as well as γ-rays produced in activation and by scattering. Maučec and de Meijer (2002) have assessed the MCNP as a feasibility tool for quantification of a thermal-neutron backscattering-based detector system for detection of non-metallic landmines. Several other authors (e.g. ElAgib and Csikai, 1999; Hlaváč, 1999; Pesente et al., 2001; Tickner (1999), Tickner (2000)) reported on the application of MC simulations for design of probing tools and optimisation of detector responses using various codes. Maučec and de Meijer (2001) proposed further development of the perturbation MC method, to advance the data interpretation and elemental characterisation in HD. Recently, a project was funded by the International Atomic Energy Agency (IAEA) (Rigollet et al., 2003) aiming at the development of MC simulation techniques to optimise the performance of the bismuth–germanate (BGO) detector system of pulse-neutron elemental analysis (PELAN) (Vourvopoulos and Womble, 2001).

In this paper, a development of MC simulation techniques at Nuclear Geophysics Division of Kernfysisch Versneller Instituut, Groningen, the Netherlands (NGD/KVI), to enhance the data analysis and interpretation of a pulse-neutron-based system is outlined. The work is based on the MCNP calculation of BGO detector responses in mixed (n,γ) fields and set-up of a database of standard responses required for the Full Spectrum Analysis (FSA) of γ-ray spectra.

The MC modelling of a simplified land-mine detector set-up, using a 14 MeV pulse-neutron source is presented in Section 2. In Section 3 the computational methods, aiming at the assessment of the neutron-induced background in the BGO detector and the generation of standard detector responses to the elements relevant for application of nuclear methods in HD are described and followed by the discussion of the results in Section 4. The last section provides conclusions and some guidelines for the future work.

Section snippets

MC modelling aspects

The general-purpose MC code MCNP4C was used for modelling the pulse-neutron DT (deuterium–tritium; PNDT) system schematically depicted in Fig. 1. The ∼14 MeV neutron source (108 n/s), radiating “almost” isotropically, is located a few centimetres above the conical neutron shield. The exact energy-spatial distribution of the DT source applied in our simulations was taken from (IAEA, 1987) and shown in Fig. 2.

A cylindrical, 7.6 cm×∅5 cm, BGO detector (ρBGO=7.13 g/cm3, Z=84), with an energy resolution

Computational methods

Initially, the MC simulations of the PNDT tool for the elemental characterisation of explosives and landmines have combined the assessment of the neutron-induced background in the BGO detector and the calculation of standard responses of the BGO detector to γ-rays (i.e. standard spectra, per unit elemental concentration of specific element).

Results

Originally, the spectra, calculated by MCNP are given normalised per starting particle. The standard spectra in Fig. 6, Fig. 7, Fig. 8, Fig. 9 are given in counts per unit (atom) concentration and smoothed with 5-point running average. By applying MC computational techniques, as described in Section 3, it was possible to decompose the combined scintillation detector response to γ-rays on two components, induced by neutron inelastic scattering and capture events. However, before being used for

Conclusions and discussion

In this paper, the MC modelling of the simplified 14 MeV pulse-neutron source based system for characterisation of landmines is presented. The simulations indicate that the pulsed fast/thermal neutron analysis (PFTNA) technique requires relatively high concentrations (∼20%) of soil moisture, to efficiently moderate fast neutrons (∼14 MeV) to lower energies (∼1 eV), where they can induce the characteristic thermal neutron capture γ-rays, thus providing the information on elemental composition. This

Acknowledgements

The research work was performed under IAEA Co-ordinated Research Project (CRP) “The application of nuclear methods to Anti-Personnel Landmine detection”. The authors would like to acknowledge James Tickner, for his assistance in providing MCNP neutron cross-section data for Germanium isotopes and for valuable discussions on the simulations of neutron-induced scintillator responses.

References (21)

There are more references available in the full text version of this article.

Cited by (17)

  • Monte Carlo simulation of explosive detection system based on a Deuterium-Deuterium (D-D) neutron generator

    2014, Applied Radiation and Isotopes
    Citation Excerpt :

    The general design and performances of an EDS system were previously determined using the MCNP Monte Carlo program (Nunes et al., 2002). Several types of neutron sources can be used for explosive detection systems such as 252Cf, Pu–Be, Am–Be (Nunes et al., 2002; Elsheikh et al., 2012), D–D and D–T neutron generators (Mauceca and Rigolleta, 2004; El Kanawati et al., 2011). Neutron generators face lower regulatory hurdles than radioisotope sources, and because they can be turned off, are inherently safer.For neutron generators, the D–D reaction has some advantages over the D–T reaction.

  • D-T neutron generators as a feasibility tool for landmine detection based on neutron backscattering method

    2014, Annals of Nuclear Energy
    Citation Excerpt :

    On the other words, by the nuclear techniques, the problem of landmine detection is reduced to the problem of identification of light elements. There are several landmine detection methods based on nuclear techniques such as neutron energy moderation, neutron-induced γ-ray emission, neutron and γ-ray attenuation, and fast neutron backscattering (Cinausero et al., 2004; Brooks et al., 2004b; Pazirandeh et al., 2006; Takahashi et al., 2011a; Maucec and Rigollet, 2004; Maucec and de Meijer, 2002; Bom et al., 2006; Brooks et al., 2004a; Datema et al., 2002b; Rezaei Ochbelagh et al., 2007; Pesente et al., 2001; Vourvopoulos and Womble, 2001). In all of these methods, two principle techniques are used based on elemental mapping.

  • Monte Carlo simulations of a D-T neutron generator shielding for landmine detection

    2011, Radiation Measurements
    Citation Excerpt :

    One possible approach to the detection of these potential threats is the use of nuclear techniques including nondestructive analysis with fast/thermal neutrons with detection based on the analysis of gamma rays emitted during the interaction of neutrons with the material of interest. In the field of landmine detection by nuclear techniques various requirements and results have been reported (Hussein and Waller, 2000; Datema et al., 2002; Brooks et al., 2004; Maucec and Rigollet, 2004; Hussein et al., 2005; Vourvopoulos and Sullivan, 2006; ElAgib et al., 2009). A detailed understanding of the source of background in the gamma ray energy region 9–12 MeV in the case of buried landmine detection is needed when the 10.8 MeV gamma ray transition emitted in the capture reaction of neutrons by nitrogen nuclei is used to tag the presence of explosives (Viesti et al., 1999).

  • Landmine detection method combined with backscattering neutrons and capture γ-rays from hydrogen

    2011, Applied Radiation and Isotopes
    Citation Excerpt :

    The capture γ-ray method can identify the nuclides by prompt γ-ray analysis (PGA) (Takahashi et al., 2011). In the past (Viesti et al., 1999; Vourvopoulos and Womble, 2001; Cinausero et al., 2004; Maucec and Rigollet, 2004; Takahashi et al., 2011), the detection of 10.8 MeV capture γ-rays from nitrogen contained in explosives was conducted for the landmine detection because 10.8 MeV high-energy capture γ-rays are rare in natural background. However, the detection of 10.8 MeV capture γ-rays was difficult due to low detection efficiency of the high-energy γ-rays.

  • Development of landmine detection system based on the measurement of radiation from landmines

    2010, Applied Radiation and Isotopes
    Citation Excerpt :

    For these reasons, the detection of 2.2 MeV capture γ-rays from hydrogen and 5.3 and 10.8 MeV capture γ-rays from nitrogen are expected to be effective for landmine detection. In the past (Viesti et al., 1999; Vourvopoulos and Womble, 2001; Cinausero et al., 2004; Maucec and Rigollet, 2004), some detection systems were developed on the basis of capture γ-ray detection. However, most of these were intended for the detection of anti-tank landmines, and they were less effective in humanitarian landmine detection.

  • Monte-Carlo simulations of elastically backscattered neutrons from hidden explosives using three different neutron sources

    2009, Applied Radiation and Isotopes
View all citing articles on Scopus
View full text
Copyright © 2004 Elsevier Ltd. All rights reserved.