Neutron tomography using projection data obtained by Monte Carlo simulation for nondestructive evaluation

Abstract

This work present the application of a computer package for generating of projection data for neutron computerized tomography, and in second part, discusses an application of neutron tomography, using the projection data obtained by Monte Carlo technique, for the detection and localization of light materials such as those containing hydrogen, concealed by heavy materials such as iron and lead. For tomographic reconstructions of the samples simulated use was made of only six equal projection angles distributed between 0º and 180º, with reconstruction making use of an algorithm (ARIEM), based on the principle of maximum entropy. With the neutron tomography it was possible to detect and locate polyethylene and water hidden by lead and iron (with 1cm-thick). Thus, it is demonstrated that thermal neutrons tomography is a viable test method which can provide important interior information about test components, so, extremely useful in routine industrial applications.

Neutron tomography; nondestructive evaluation; Monte Carlo simulation

Neutron tomography using projection data obtained by Monte Carlo simulation for nondestructive evaluation

A. X. da Silva; V. R. Crispim

[PEN/COPPE-DNC/EE]CT/UFRJ, P.O. Box 68509, Ilha do Fundão, 21945-970 Rio de Janeiro, RJ. Brazil

ABSTRACT

This work present the application of a computer package for generating of projection data for neutron computerized tomography, and in second part, discusses an application of neutron tomography, using the projection data obtained by Monte Carlo technique, for the detection and localization of light materials such as those containing hydrogen, concealed by heavy materials such as iron and lead. For tomographic reconstructions of the samples simulated use was made of only six equal projection angles distributed between 0^o and 180^o, with reconstruction making use of an algorithm (ARIEM), based on the principle of maximum entropy. With the neutron tomography it was possible to detect and locate polyethylene and water hidden by lead and iron (with 1cm-thick). Thus, it is demonstrated that thermal neutrons tomography is a viable test method which can provide important interior information about test components, so, extremely useful in routine industrial applications.

Keywords: Neutron tomography, nondestructive evaluation, Monte Carlo simulation

Introduction

X-rays and gamma rays interact with the electrons of the atomic shell, and the attenuation thus increases with the atomic number in a regular way. Therefore, X-rays or gamma rays are used for viewing heavy objects (high atomic number materials) in light materials (low atomic number materials). On the other hand, absorption or scattering of thermal neutron (25 meV) is an interaction between nucleons showing a random dependence of the total cross section on the atomic mass (Domanus, 1992). For many technical applications, however, thermal neutrons can be used to great advantage. Therefore, the neutron radiography is a non-destructive testing method that has been applied in special cases of inspection, where it is difficult to take radiographs by X-rays or gamma rays (Mishima et al., 1999). In neutron radiography, by passing a uniform and collimed beam of neutrons through a sample, it is possible to record its internal structure on a detector. In this procedure, internal structure of one part of sample may mask another. Thus resolving the structure becomes difficult and sometimes impossible. To overcome this difficulty, Computerized Tomography (CT) could be employed to produce an image of a thin layer of a sample without any interference (Treimer et al., 1998, Schillinger et al., 1999). In this case, the source and detector array both translate and rotate around a scan circle in which the object is centered. The neutron beam penetrates the object and is detected on the opposite site. A set of projection data is obtained at a particular angle. The source and detector are then rotated at a small angle, and a new projection is obtained. Each neutron beam transmission measurement is converted into an electrical signal and computer processed. The reconstructed CT image is cross-sectional, and the computed attenuation coefficient at any point in the cross section is independent of overlying material (Herman, 1980).

The components of the tomographic inspection system, i.e. the characteristics of the source (size, shape and spectrum of the radiation source), the composition and configuration of the medium through which photons and neutrons are transported, the detector geometry and type (energy, flux detection, etc), as well as imaging process, can be examined individually by computer codes. The computer simulation gives the possibilities to optimize testing parameters, to make feasibility analysis for special testing problems, and to support the interpretation of testing results by making use of the forward model to eliminate a number of disturbances from the resulting radiography or tomography (Raine and Brenizer, 1997). In this context, the availability of a computer package for the simulation of the neutron imaging process in nondestructive evaluation is essential. The aim of this study were: (1) the obtaining, via Monte Carlo simulation, of projection data used for neutron computed tomography; (2) the reconstruction of a bi-dimensional image of the test sample executed by a reconstruction algorithmic using as projection data those generated on step (1); and (3) the comparison between the simulated tomographic images obtained either by a photon beam with energy of 662 keV or by a thermal neutron (energy of 25 meV) beam, aiming industrial applications. To this purpose the MCNP (Briesmeister, 1997) / ARIEM (Crispim, 1993) code systems were used, leading to a two steps calculation. The emphasis here is on obtaining an overall indication of the presence of the hydrogened materials hidden by heavy materials and not on determining exact physical densities.

Monte Carlo Simulation

Monte Carlo simulation have recently been used for various scientific and engineering applications (Gardner and Liu, 2000). There are a numbers of reasons for this, including: (1) the increased cost and inadequacy of experimentation for design and interpretation purposes; (2) the availability of low cost, large memory, and fast personal computers; and (3) general availability of general purpose Monte Carlo codes that are increasingly user-friendly, efficient, and accurate (Gardner and Liu, 2000).

In this work, the code used for the generating of projection data was the Monte Carlo N-Particle, MCNP, version 4B (Briesmeister, 1997). MCNP is a general purpose, three-dimensional general geometry, time-dependent, Monte Carlo N-Particle code that is used to calculate coupled neutron-photon-electron transport. The discrete cross section data used by the MCNP code are part of the Evaluated Nuclear Data File (ENDF) and Evaluated Nuclear Data Library (ENDL). The code allows for generalized three-dimensional object geometry and material content. A sketch of the geometry and the detector configuration used in MCNP plotting routine aiming to obtain the computed tomographic projection (lines) dates can be seen in Fig. 1. The source was defined to be monoenergetic with an energy of 25 meV (for thermal neutron beam) and 662 keV (for photons beam). Hence, the gamma source simulated was the ¹³⁷Cs, which emits gamma-rays of energy 662 keV and is widely used in industrial applications. The particles (neutrons and photons) were started from a single source plane, and were given a direction perpendicular to the imaging plane was which segmented into 103 subregions to represent the individual pixel locations with areas of 500mmx500mm. The cylindrical shape phantoms 10 mm in diameter and height of 30 mm consists of (1) polyethylene (CH₂) and (2) light water (H₂O), which were simulateds by using the elements: hydrogen-carbon and hydrogen-oxygen, respectively. The phantoms were concealed by iron and lead cylindrical tubes with diameters of 30 mm and wall thickness of 5 mm. Were obtained dates relatives to the six angular projections (0^o, 30^o, 60^o, 90^o, 120^o and 150^o) in each case.

Tomographic Reconstruction

Assuming monodirectional and monoenergetic neutron beam of intensity, I_o, incident on a homogeneous sample of thickness, X, the transmitted neutron intensity, I_, by matter obeys law :

where S_T is the total macroscopic cross section of neutron interactions in the material along path X. The projection of the attenuated neutron beam from this ray path onto a know detector localization is defined as (Herman, 1980):

Sets of projection data must be collected over many angles of rotation about the samples. By means of a reconstruction method, the combined sets of projection data can be transformed into a series of cross-sectional images of the sample. A detailed description of the principles of CT is given in Kak and Slaney (1988).

Since the time for the realization of a inspection in routine industrial is limited the use of algebraic reconstruction techniques (Kak and Slaney, 1988) will be considered that can provide reasonable results with a small number of projections which reduces the inspection time. For this we have used the Maximum Entropy Image Reconstruction Algorithm – ARIEM (Crispim, 1993). This code gives good results specially when the available number of projections is small. The merit of maximixing entropy lies in the factor the reconstructed image requires a minimum of information about the system configuration. Simulated projection data from different angles are processed together to obtain a reconstructed tomographic of samples. We used six projection angles starting at 0^o and uniformly distributed between 0^o and 180^º.

Results

Figure 2 shows the data related to the six projection profiles of transmitted neutrons ( ln I_o/I ) generated by the MCNP4B, of: (a) lead (Pb) and (b) iron (Fe) tubes with the water (H₂O) and polyethylene (CH₂) phantoms insert, respectively. It is show that the MCNP radiation transport code can successfully simulate neutron projection data. Aiming a comparative analysis between simulated tomography images obtained by thermal neutron and a gamma ray beam, the same modeling was adopted substituting the thermal neutron beam by a monoenergetic 662 keV photon beam.

The MCNP simulated projection data sets of the phantoms were used as input in CT reconstruction routines from the ARIEM code. To evaluate the difference between thermal neutron and gamma ray CT, the images obtained with the two beams are displayed together. Figures 3 and 4 shown the tomographic reconstructions of the lead and iron tubes with the water and polyethylene phantoms inserts, from the MCNP projections dates generated using: (a) 662 keV gamma ray and (b) 25 meV neutron beams. In gamma ray CT images, Figs. 3a (10 mm thick Pb) and 4a (10 mm thick Fe), it should be noted that the water and polyethylene detection was not possible, since the lead and iron owns high linear attenuation coefficients for photons of these energy levels. So, dense materials, like lead or iron, cannot be inspectioned using the tomography with gamma ray beam (of 662 keV) aiming the detection and localization of materials based on water and plastic hidden by these heavy materials. On the other hand, the images reconstructed with thermal neutrons, Figs. 3b and 4b, shown clearly the water and polyethylene samples concealed by the lead (10 mm in thickness) and iron (10 mm in thickness), respectively. This fact demonstrates that the computerized tomography technique using thermal neutron beams allows the detection and the localization of hydrogened materials such water and polyethylene when concealed by high atomic number materials.

Conclusions

The demand of industry non-destructive inspection is rising due to the increasing complexity and costs of technical components. Therefore more sophisticated and computerized systems will be applied for solving specific problems. In this context, investigations with neutrons will provide an unique technique complementary to other well established methods (X and g-rays, eddy-current, ultrasound).

This paper has shown that the MCNP transport code can successfully simulated projection data used for CT (Fig. 2), and in a second part, the advantages and complementarities of using thermal neutron tomography technique with a small number of projections over the one employing photons of 662 keV for the detection and localization of light materials hidden by heavy materials. With the thermal neutron CT was possible to detect and locate water and polyethylene concealed by lead and iron (with 1cm-thick). Neutron CT can view through materials that simply cannot be viewed by gamma ray CT, due to the high attenuation coefficients involved. Therefore, it has been recommend to employ the CT technique using thermal neutron beams for detecting and localizing of similar materials the water and polyethylene inside the materials with high atomic numbers, such as metallic objects. Thus, it is clearly demonstrated that thermal neutrons are extremely useful in CT applications industrial.

Acknowledgements

The authors would like to thank the Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) for their financial support.

Article received August, 2001

Technical Editor: Atila P. Silva Freire

References

• Briesmeister, J. F., 1997, "MCNP - A General Monte Carlo N-Praticle Transport Code, Version 4B". Los Alamos, NM: Los Alamos National Laboratory; LA-12625-M, March.
• Domanus, J.C., 1992, "Practical Neutron Radiography", Kluwer Academic Publishers, Dordrecht, Holland.
• Herman, G. T., 1980, "Image Reconstruction from Projections The Fundamentals of Computarized Tomography". New York. Academic Press.
• Gardner R. P., Liu L., 2000, "Monte Carlo simulation for IRRMA". Applied Radiation and Isotopes, Vol. 53, pp.837-855.
• Kak, A.C. and Slaney, M., 1988, "Principles of Computerized Tomographic Imaging" (New York: IEEE).
• Mishima, K., Hibiki, T., Saito, Y., Nakamura, H. Matsubayashi, M., 1999, "The Review of the Application of Neutron Radiography to Thermal Hydraulic Research". Nuclear Instruments and Methods in Physics Research, Vol. A424, pp.66-72.
• Raine, D. A., Brenizer, J.S., 1997, "The Analysis and Correction of Neutron Scattering Effects In Neutron Imaging". Materials Evaluation, pp. 1174-1178.
• Schillinger, B., Blumlhuber, W., Fent, A., Wegner, M., 1999, " 3D Neutron Tomography: Recent Developments and First Steps Towards Reverse Engineering". Nuclear Instruments and Methods in Physics Research, Vol. A424, pp.58-65.
• Treimer, W., Feye-Treimer, U., Herzig, C., 1998, " On Neutron Tomography " Physica B, vol. 241-243, pp. 1197-1203.

Publication Dates

• Publication in this collection
  08 Sept 2003
• Date of issue
  Nov 2002

History

• Accepted
  Aug 2001
• Received
  Aug 2001