·Home ·Table of Contents ·Workshop - Neutron workshop

Status and Prospects of neutron Tomography in Europe E. H. Lehmann[CH-5232 Villigen PSI, Switzerland], P. Vontobel, Paul Scherrer Institut, Switzerland B. Schillinger, T. Bücherl, Technical University Munich, Germany S. Baechler, J. Jolie, University of Fribourg, Switzerland W. Treimer, Hahn Meitner Institut and University of Applied Sciences Berlin, Germany R. Rosa, ENEA Roma, P. Chirco, University of Bologna, Italy G. Bayon, S. Legoupil, CEA Saclay, France S. Körner, H. Böck, Atominstitut Vienna, Austria V. Micherov, Lebedev Physics Institute Moscov, Russia M. Balasko, A. Kuba, KFKI Buadapest, Hungary Contact

Abstract

The situation of computed neutron tomography (NT) is described from the point of view of the COST action 524 members (DE, IT, FR, UK, SL, CH, HU, SP, RU, CZ, AT). This European collaboration is dealing with "Neutron radiography techniques for the detection of defects in materials" and established a Working Group "Neutron Tomography and Computation Analysis" in 1998.
Neutron tomography set-ups for the inspection of macroscopic samples have been built in several countries with different performances concerning the neutron beam properties, the specific detection devices and methods used for image reconstruction.
This article explains the principles of tomography with neutrons, makes a comparison of the individual solutions in the member countries and gives an impression about presently obtained results of selected examples.
Prospects regarding higher spatial resolution and faster operation cycles (image generation, reconstruction and visualisation) are presented.

1. Introduction

Neutron radiography has a tradition since neutron sources and imaging devices (mainly film in the beginning) became available [1,2]. Non-destructive testing with neutrons has been an established method in many countries for several years [3,4]. However, this method cannot compete directly with X-ray methods on a broad scale because of to few strong neutron sources (which are mainly stationary ones). Therefore, only some sites in different countries are able to perform radiography with neutrons [5].
The specific advantage of neutrons compared to X-rays is their high interaction probability with hydrogen and the lower attenuation in several heavy elements which are "black" for X-rays (e.g. lead, bismuth, uranium). This gives the justification for using neutrons for special applications where X-rays must fail even if neutron imaging is more expensive and demanding. The investigation of moisture and corrosion, the detection of explosives and adhesive connections and the inspection of defects in nuclear fuel or in thick metallic samples are examples where neutron can be utilised favourably.
In the past few years, several techniques in digital imaging were successfully be performed providing high sensitive detectors with important performances regarding their dynamic range and linearity. This holds mainly for imaging plates [6] and detectors based on CCD-cameras with a scintillator screen as primary neutron-to-light converter [7]. Other methods are not yet completely optimised for neutron imaging (amorphous silicon arrays, micro-strip gas counters [8]), but have a high potential as detectors in radiography and in neutron scattering experiments as well.
For tomography it is required to have many two-dimensional projections in digital form of the observed object rotated over 180° angle range related to its central axis. To obtain precise voxel information about the sample, the detector position and the sample manipulation must be adjusted as accurate in space as the size of one pixel in the projections (mm to mm).
The investigation of the third dimension of samples became achievable by a set-up with a stationary digital neutron detector and a turntable, rotating the sample in the beam. First attempts were done for such kind of arrangements by [9,10] and improvements towards a routinely operating system were described in [11]. Today, several groups have a tomography equipment available. The situation in the COST-524 member countries regarding neutron tomography will be presented on the next pages.

2. Tomography with neutrons - boundary conditions and solutions

Compared to X-ray computed tomography, which is in routine operation since several years for medical purposes and in industry for technical inspections, the neutron source must be considered as stationary. This is caused by the demands for a good beam collimation (high L/D ratio) for a high spatial resolution and a reasonable short duration of a single projection measurements, which can only be provided by strong sources. Therefore, such neutron sources are reactor or accelerator based. Neutron tomography with mobile sources seems to be not possible with the presently available neutron detection systems, but some attempts are on the way (see 3.8.).
Therefore, the common arrangement for neutron tomography is such as shown in Fig. 1 with a fixed beam line and a stationary detector, whereas the sample is rotated between them on a turntable. In this way, as many projections of the object as needed are obtained (of the order of few hundred).

Fig 1: Arrangement for tomography investigations using neutrons from a stationary source (assumed as illuminating from the right in the figure)

The beam geometry (parallel beam or conical beam) has to be considered as well for the applied reconstruction method as the obtainable spatial resolution. In the X-ray field, the use of micro-focus tubes provides the great advantage of object enlargement during the investigations by factors. A similar approach [23] with an aperture of some micrometers only (given by a diaphragm in front of the primary source or by using of capillary optics) would reduce the intensity of the neutron beam dramatically and a useful tomography device is a challenge. Nevertheless, this approach needs to be tested in practice (see 3.5).
A better option seems to be a quasi-parallel beam with a L/D-ratio as high as achievable. In this way, no geometrical distortion occurs and the spatial resolution should be limited only by the properties of the detector device.
At most facilities dealing with tomography (more details are presented in chapter 3), the beam properties regarding collimation are far away from being parallel (L/D of the order of 100 only) with apertures of some cm. This is caused by the individual layout of the radiography station which are not specially devoted to tomography with highest resolution but reasonably low exposure time per image. Corrections for beam divergence and inherent blurring of the images have to be applied in this case.

Fig 2: Example of investigations of a small combustion engine by 2D-radiography (middle - PSI) and 3D-tomography (right - TU Munich)

3. Tomography facilities in European COST-524 countries

A summary of the most important properties of set-ups in the different institutions dealing with neutron tomography is given in Table 1. Besides the use of mainly thermal neutrons for tomography there are some options with cold neutrons and fast (fission) neutrons as well.
Some detailed descriptions of the individual solutions, the experimental conditions and the obtained performances are presented in the following sub-sections.

3. 1. Technical University Munich and the FRM-1 set-up
At the FRM-1 reactor in Garching, two set-ups for cold and thermal and for fast neutrons have been installed.
The thermal/cold system was installed at a long neutron guide of 2.5 cm width and 25 cm height. The spectrum was cold or sub-thermal (cut-off at 2 Å by the curvature of the guide), depending on if the cold source was in operation or not, with fluxes of 1·10⁸ cm^-²s^-1 and 3·10⁷ cm^-²s^-1. Due to the low guide width, the whole system consisting of scintillator, mirror, custom-built cooled 512 x 512 pixel CCD camera and rotation table was scanned across the guide to illuminate larger area from 12 x 12 cm² to 28 x 28 cm². The inherent divergence of the neutron guide limited the achievable resolution for large samples below the effective pixel resolution. The measured effective L/D ratio was 71 and 115 [12]. Examinations were done on several technical and archaeological objects with fluences of about 2.5 10⁸ cm^-² and scan times according to the object width.
The neutron guide and beam line have recently been dismantled due to the construction of the new reactor FRM-II. The tomography equipment has now been installed at a new thermal beam tube in the reactor hall with a flux of about 1·10⁶ cm^-²s^-1 and an L/D ration of 500. The beam geometry is greatly improved, but the reduced flux and increased gamma background require long exposure times and allow only for lower dynamics in the images.
The fast neutron facility is located at the thermal-to-fast neutron converter beam line, mainly used for medical treatment. The detection system built by IKE Stuttgart consists of a slit collimator/scintillator/photo multiplier combination. Pulse shape discrimination allows the simultaneous collection of gamma and neutron tomography images. Due to the strong small-angle scattering of fast neutrons in the sample, a long slit collimator between sample and detector is required, which enables only for single point scanning. Several experiments with a scintillator-CCD Camera system have been conducted in Garching by the Lebedev Physiscs Institute, Moscow (see 3.8.).

3. 2. The facility NEUTRA at SINQ of the Paul Scherrer Institute
A tomography setup is available at the thermal neutron beamline NEUTRA [13] of the PSI spallation source SINQ. This beamline provides a highly collimated thermal neutron beam with a large area (400 mm diameter) at its outermost detector position. There the main characteristics are a L/D ratio of 550, a neutron flux of 3 10⁶ [cm^-2 s^-1 mA^-1] (i.e. neutron flux per mA proton beam onto the spallation target, with an actually available proton beam of about 1.2 mA and a target with lead rods). A higher neutron flux is available at the innermost detector position closest to the spallation target (5 times higher flux, L/D ratio 250).
The components of the neutron tomography set-up at NEUTRA are a nitrogen cooled CCD camera system (Astromed 3200 by Perkin Elmer Life Sciences, Cambridge, U.K.) with a 512 x 512 CCD chip providing 16 bit greyscale images. It looks via a mirror onto a 0.4mm thick ZnS(Ag) + ⁶LiF converter/scintillator screen (NE 427), fixed in a light tight box.
Secondly, there is a rotating table (Franke, Aalen, Germany) driven by a 2-phase stepping motor and supplemented by an additional goniometer. A LabView (National Instruments, Austin, TX, USA) based software interface implemented on a PC controls the rotation sequence, checks for sufficient neutron beam intensity and sends the triggering signal to the CCD camera.
Depending on the object size, different camera lenses are in use allowing a varying size of the field of view from 20 x 20mm to 250 x 250 mm, and correspondingly pixel sizes from 50 mm to 500 m m. Exposure times from 10 - 60 seconds per picture are imposed by the neutron flux level, the objects neutron transmission given by its thickness and isotopic composition and the camera lens in use. Typically 200 projections with an 0.9° angle increment are taken from 0° to 180°.
Due to the slow readout rate of our CCD camera, this leads to a total acquisition time of 1.5 to 3 hours. In order to reduce the total acquisition time to 15 - 20 minutes a 10 times more efficient detector system with fast readout electronics is valuated.

Fig 3: Image reconstructions of samples investigated at the PSI station NEUTRA by the camera set-up. From the left: a relais with about 4 cm diameter, a aircraft turbine blade (about 10 cm high), a sparking plug

Some results from tomography investigations are shown in the Fig. 3 and are reported in [14].

3. 3. Hahn Meitner Institute and the BER-2
The first experiments on neutron tomography at the Hahn-Meitner-Institute have been performed in 1995, using the monochromatic and cold neutron beam of a double crystal diffractometer (DCD) [15]. These test experiments were dedicated to fundamental aspects such as the imaging of simple objects by absorption contrast Fig. 4, Fig. 5), but also to the investigation of additional effects such as refraction and small angle scattering as imaging signals (Fig. 6) and the feasibility of the use of polarised neutrons to image magnetic fields (Fig. 7). Refraction and small angle contrast in neutron tomographies can only be detected with the help of a high resolution DCD, the detection of magnetic fields requires a polarised neutron beam and a polarisation analysis of the transmitted neutrons [16].

Fig 5: Views of a ball bearing, snapping cage(3) and its inner structure (5)

Fig 4: "Classical" neutron tomography: Reconstructed slice of a chicken leg, diameter ~ 7mm

Fig 6: Aluminium cylinder with hole and glass fibre bundle: Enhanced contrast at the edges of the cylinder and the hole and by the glass fibres due to refraction; note that this contrast is much higher than the usual absorption contrast in the bulk of the sample.

Fig 7: 2-dimensional reconstruction of a small permanent magnet and its magnetic fields by means of spin polarisation analysis of the transmitted neutron beam, diameter of the rod = 20mm

The neutron tomographies in Figs 4 and 5 have been registered in the "conventional mode" (l = 0.52 nm, parallel object scanning with a slit system, object rotation and an 1-dimensional neutron detection), the one in Fig. 6 was performed with the double crystal diffractometer (DCD), detecting absorption + refraction signals, using the option of the high angular correlation between two perfect crystal of a DCD, the neutron tomography of Fig. 7 used polarised neutrons and a spin-analysing system to detect the depolarisation of the neutron spin by the magnetic field of the sample. The first 3-dimensional reconstructions from 2-dimensional projections (Fig. 5) were performed by a film-converter-system due to the lack of an appropriate 2-dimensional detector. The results, however, have been surprisingly good, keeping in mind the rather low number of projections (40) due to the extreme low neutron flux.
Other investigations on neutron tomography dealt with reconstruction techniques, esp. with ART (algebraic reconstruction techniques) in order to compare the results of reconstructions (of so-called ill posed problems) yielded with standard techniques with those of (modified) ART.
Now the V12b instrument (DCD) at HMI will be equipped with a graphite crystal as monochromator for 0.52 nm neutron wave length, which delivers an app. 50 times higher neutron flux than the perfect Silicon crystal and with a neutron low flux CCD camera (512x512 pixel for a circle with 25mm diameter). In the next phase of development the DCD together with the CCD camera shall be used as well for refraction contrast tomography.

3. 4. ENEA and the TRIGA facility in Rome
A neutron tomography facility has been developed at the thermal column of the 1 MW TRIGA Mark II reactor running at the ENEA Casaccia Research Centre.
The R&D activities are carried out in a cooperation of ENEA with the University of Roma La Sapienza Nuclear Science Dept. and the University of Bologna Physics Dept. These activities concern the imaging of industrial samples as well as the detection of hydrogen in metallic objects. In previous work, the possibility of establishing a linear correlation between the reconstructed CT value and the local hydrogen concentration was demonstrated.
An improved neutron collimator for the radial channel of the thermal column has been recently replaced in order to enhance the beam quality for radiographic purposes. The collimator aperture is now circular with a diameter of 4 cm and an estimated L/D ratio of 54.
Neutrons are detected by a typical chain of ⁶LiF/ZnS:Ag screen (r) Image intensifier (r)CCD camera placed in a light tight box. The group acquired a good experience in the use of self-developed CCD cameras and pioneered the use of back-illuminated sensors for neutron tomography applications.
Tomographic projections are usually acquired with steps of 0.9° by using a turntable placed on a horizontal and vertical micrometric translation system that ensures an accurate centering of the detector.
The system acquires 200 projections in about 30 min. The filtered back projection reconstruction algorithm includes proprietary filters for the correction of spatial and temporal inhomogeneities of the neutron field corrections and filters. The software generates raw data sets that can be visualized with commercial tools and it is able to handle images from the CCD presently in use (with a 192x165 pixel matrix) up to 512x512 pixels matrix.
In the near future the design and implementation of a new near parallel collimator to be used at the tangential channel having a circular aperture of about 10 cm is planned. This would greatly enhance the testing capabilities of the entire system, now limited to small size objects.

3. 5. Cold beam line PGA at SINQ operated by University of Fribourg
At SINQ (PSI, Switzerland), neutron tomography has been performed exploiting a high intense cold neutron beam. The detection system based on a converter and a CCD camera was located at the end of the beamline 14, mainly used for PGAA. The neutron to visible light converter consisted of a 420 mm-thick ZnS(Ag) + ⁶LiF layer. The converter was imaged using an optical lens, a 45-degree mirror and a CCD camera. An active area of 2.5 cm² is projected on the CCD. The mirror is used to keep the camera out of the direct beam. In addition, a lead glass was placed between the mirror and the camera to protect the CCD from emitted photons. The Sensicam CCD (PCO, Munich, Germany) has a sensitive area of 8.6x6.9 mm² covered by 1280x1024 pixels, a dynamic range of 12 bits (4096 gray-levels per pixel) and a fast read-out of 125 ms. The spectral sensitivity of the camera matched with the blue emission spectrum of the scintillator. At the sample position, a 50 mm high and 20 mm wide neutron beam has an intensity of the order of 10⁸ s^-1×cm^-2. The detection system was protected by neutron/gamma-shields to avoid neutron damage on the one hand, and to suppress the interference from scattering gamma rays and neutrons on the other. Materials used for this purpose are ⁶LiF and B₄C plates for neutrons and lead bricks for gamma-rays.

Fig 8: The example of a tooth sample demonstrates the obtainable resolution (but also the limitations) of presently available tomography systems with cold neutrons. The metallic inlay can easily be distinguished from the basis tooth material.

Due to the high neutron flux, an exposure time smaller than one second per image was necessary. It took about 20 minutes to perform a 400 projections tomography scan. Thus, the new setup at SINQ eliminates most of limitations in time resolution. Neutron tomography limitations in spatial resolution are given by the detector system and the properties of the neutron source; the resolution was about 220 mm. In order to improve the spatial resolution, a neutron tomography setup using a neutron focusing capillary lens is under study.

3. 6. The CEA Saclay approach
The development of a neutron tomography system has started at CEA Saclay. The neutron NDT activities are mostly dedicated to the control of pyrotechnic material for the aerospatial industry (diameter<50 mm). Traditionally, tested objects are controlled by radiographic techniques [17]. As a second stage, industrial partners would like that defects are visualized and measured in the 3^rd dimension. In this context, the objective is to reach the finest spatial resolution as possible while the acquisition time must not be longer than 12 hours.

Fig 9: In order to increase the spatial resolution of the CEA set up at the guide for cold neutrons an optimum has to find between the achievable L/D ratio and the detector efficiency which is limited by the thickness of the neutron sensitive layer

The selection of the detector should take benefit of the long acquisition time. However, an effort has to be carried on the adaptation of the neutron guide to reach the objective. The small L/D ratio (not yet estimated) needs to be enlarged. On one hand, a good spatial resolution can be achieved if the sensitive area of the detector is very thin. As a consequence, the neutron flux must high enough to satisfy the reconstruction process (signal to noise ratio on projections). On the other hand, the enhancement of the L/D ratio implies a reduction of the available neutron flux on the detector. As illustrated below (Fig. 9), in order to keep constant the number of incident neutrons, the work consists in the optimisation of the L/D ratio versus the efficiency of the available 2D neutrons detectors. To get a small spatial resolution, the L/D ratio must be large enough while the detector thickness (and efficiency) is decreased in order to maximise its spatial resolution.
During the development of the NT device at CEA Saclay, one pays a particular attention to the detectors developed by physics particle teams.

3. 7. Atominstitut Vienna and its TRIGA facility
At the 250 kW TRIGA Mark II reactor in Vienna, two beam lines are dedicated to neutron radiography. Station 1 has a large beam diameter of 40 cm suitable for industrial applications and station 2 is used for examinations with high demands concerning the resolution (beam diameter 8 cm). The detectors mainly used for radiography in the past consist of X-ray films and a Gd converter enclosed in a vacuum cassette [18].
Recently, a neutron tomography set-up has been developed and brought into operation. Therefore, the first step was the design and optimisation of a suitable detector, based on a nitrogen cooled slow scan CCD camera and a neutron sensitive scintillator screen (Levy Hill: ZnS(Ag)-⁶LiF) [19]. The CCD driver electronics provides 16-bit digitization. Samples are placed on a rotary table to allow taking images from different view angles. Camera control software and the motion control software have been connected to automate measurements. Special thanks are devoted to Dr. B. Schillinger, (TU München, Germany) for his ongoing advice and his support with the motion control system as well as for the 3D reconstruction on the basis of our first tomography data set in parallel with P. Vontobel, (PSI, Switzerland). The whole assembly was brought into operation at station 2, in front of the thermal column. Despite the low neutron flux, exposition times can get as low as 20 s per image, due to the high sensitivity of the new detector. This means that a tomography can be made within one working day.
Most important application in the near future will be the investigation of bonding of fusion relevant materials. Next step will be the design of a tomography set-up at station 1 to allow investigation of larger objects.

3. 8. P.N. Lebedev Physics Institute (LPI, Moscow, Russia)
Thermal and fast neutron tomography systems have been developed at the LPI. To optimise the measuring conditions and to meet various specific demands they use a whole set of CCD-detectors differing mostly by sensitive area, spatial resolution, composition of the luminescent screen and optical schemes [20,21].
Two kinds of optics are used in the detector for elements junction: taper and lens system. A magnifying taper can be used to improve spatial resolution of the system. The same detector unit can be used for thermal and fast neutrons simply by changing the luminescent screen. For thermal neutrons the screen is made of Gd₂O₂S:Tb(Eu). Screens with different phosphor loading are manufactured and used with respect to different thermal neutron spectrum of the source to optimise efficiency of registration and spatial resolution. A composite screens made of a silicon organic resin and embedded Gd₂O₂S:Tb(Eu) is used for fast neutron imaging.
Tomographic reconstruction is performed by a software combining Convolution and Back Projection (BP) method with the Maximum Likelihood (ML) one [22]. The ML method is appropriate first of all in the case of instability of instrumental function and comparatively low luminosity of the neutron source. Another important advantage of the ML reconstruction is its ability to take into account neutron scattering in a sample. This procedure is slower than BP but it can process incomplete or even ill-consistent data. Besides the main processing part the developed software contains some extra facilities to visualise experimental data, images and the process of reconstruction.
At present, tomographic measurements by the LPI are carried out at various neutron sources in collaboration with a number of Institutions: with thermal neutrons - at the nuclear reactor of Moscow Engineering Physics Institute (LPI's neutron facility DNM-I); with fast neutrons - at the converter facility of the Munich Technical University. Feasibility of fast neutron radiography and tomography on non-reactor neutron sources is now being investigated at: the stationary fast neutron generator of the Institute of Nuclear Research (Moscow) and portable neutron generators of the All-Russian Research Institute of Automation.

3.9. Neutron tomography at the Budapest reactor at KFKI
The realisation of neutron tomography (NT) was decided in 1999. A portable neutron tomography adapter (NTA) was designed which is available to work on both neutron radiography stations of the reactor. It has capability at the N^o1 channel to make neutron-, gamma-, and X-ray tomography with high intensities at the same beam position. The NTA allows neutron tomography with low intensity of the filtered neutron beam at the N^o2 channel. A Peltier cooled CCD camera (working in TV-mode) is installed as the detector connected to a PC. A photo of the NTA is given in Fig. 10.

Fig 10: Device for sample positioning and rotation at the KFKI installation in Budapest, suitable for different beam ports at the reactor.

The reconstruction of the exposed NT data will be performed by a team of Department of Applied Informatics, University of Szeged. An image reconstruction and visualisation system was developed and tested by them for creating and presenting NT data.

Table 1: Comparison of the different approaches to neutron tomography within the COST-524 member countries


After the necessary pre-processing steps (homogeneity correction, background subtraction and centre of rotation correction) the images of the cross-sections of the objects were reconstructed. Several reconstruction techniques were tested (algebraic reconstruction techniques, filtered back-projection of the software package SNARK93 [1]). In order to present 3D images special visualisation software tool should be developed.

4. Results of investigations by neutron tomography

As shown in the previous chapter, the existing tomography systems are based on a CCD-camera set-up with a neutron sensitive scintillator as detector. Depending on the neutron field and the number of pixel in the CCD-chip (presently between 512 and 1024) objects from some cm to about 30 cm in diameter can be investigated. Additionally to the limitations by the size of the neutron beam for some materials, a limited neutron transmission gives inherent restrictions for investigations.
The limitation for higher spatial resolution with camera based systems is finally given by the properties of the scintillator. For most devices it was optimised to highest light output during neutron exposure. The spread of light in the scintillator limits the detector blurring to about 0.2 mm which might to be not sufficient for some future applications.
Fig. 2 to 8 gives some examples for results of tomography measurements. It is important to mention that, compared to simple radiography imaging, there are much more difficulties for the visualisation of the tomography results due to the amount of data and the adequate presentation of the third dimension. Today, there are some software tools available [22,23] which help to solve this problems professionally.

5. Main tasks of the Working Group

A collaboration in neutron tomography between different small groups in Europe should help to exchange information about the progress in research, to improve the performance of the method, and to attract partners and clients from industry.

In this way, two tasks were started for all participants:

• to compare the software tools applied for volume reconstruction by delivering a complete set of projections of one observed object (see Fig. 12),
• to compare the detector (and beam) performance by a set of test specimen with well defined geometry and material composition (see Fig. 11).

Fig 11: Test object for neutron tomography performance evaluation, proposed and provided by the COST-524 Working Group 5 "Tomography and Computation Analysis". The specimen are available in aluminium, lead and steel; the holes can be filled with cadmium, polyethylene and steel rods.

Fig 12: Test object (floppy disk drive) used for comparison of the performances of reconstruction algorithm applied in the different groups. The raw data were measured at the NEUTRA beam line at PSI.

This tasks are not yet completed but first results demonstrated that a good knowledge is available in the participating laboratories.
In the future it is important to increase the performance of tomography facilities in two respects:

• to be able to lower the spatial resolution to the limits given by the beam line conditions and
• to improve the performance for image taking and reconstruction to be able to present tomography results in a time scale less than 1 hour.

Both this activities will be considered in the participating countries in relation to the existing systems.

6. Prospects for neutron tomography

Recent developments on the detector side (amorphous silicon arrays, CCD-cameras with much more pixels, fast readout imaging plate systems) should help to enhance the performance in neutron tomography too. This is important in order to study samples in much deeper geometric details or to investigate much bigger samples (up to the limits for transmission).
However, with an increase of the field of view in the images there is also a dramatic higher demand in data acquisition and data handling performances (by the 3^rd order of pixel number increase). Therefore, reconstruction and visualisation of tomography data remains a demanding task even if the hard and software will improve its capacity as it has been in the past years.
From the application side, it is necessary to find those application fields where the solution of the dedicated problem will justify the much higher expense in tomography compared to single transmission analysis. Today this holds for reverse engineering for components which are not very transparent for X-ray but for neutrons, or those which have inner structures with higher contrast to neutrons.
Furthermore, the study of hydrogen in its different occurrence (moisture, steam, hydrides, chemical compositions) and its spatial (3-dimensional) and time dependent distribution will be important for some practical or fundamental problems. This holds for transport phenomena in porous materials as well as thermo-hydraulic processes. The higher rate for tomography image acquisition will help to increase the activity in these fields.

7. Summary

This article intended to describe the situation in the field of tomography with neutrons in countries in Europe taking part in the COST-524 action. All potential partners outside this collaboration are friendly invited for participation in the exchange of knowledge and experience in the future.
Due to a strong development on the detector and computer side it is expected, that the potential for this technique will be heavily increased in the near future. Therefore, it is very important to attract industrial partners to match their demands for tomography studies with the development programs in the individual institutes. Because of the relative high investments for tomography devices it seems reasonable to look for industrial contributions for dedicated installations.

References

1. P. von der Hardt, H. Röttger (ed.), Neutron Radiography Handbook Domanus, D. Reidel Publ. Comp., 1981, p. 1-2
2. H. Kallman, Research 1 (1947-48), p. 254
3. H. Berger, Radiography, Methods, Capabilities and Applications, Elsevier Publ. Co. (1965)
4. J. P. Barton, Contrast Sensitivity in Neutron Radiography, Appl. Materials Res., April 1965, p. 90
5. E. Lehmann, Facilities for neutron radiography in Europe: performance, application and future use, to be published
6. K. Takahashi et al., Nucl. Instr. & Meth. A 377 (1996) 119-122
7. Y. Matsubayashi et al., High resolution static imaging system in JRR-3M neutron radiography facility, Proc. 5^th World Conf. on Neutron Radiography, Berlin 1996
8. V. Dangendorf et al., Thermal neutron imaging detector combining novel composite foil converters and gaseous electronic multipliers, Nucl. Instr. and Meth. in Physics Reasearch A 350 (1994) 503-510
9. E.W. McFarland, R. C. Lanza, G.W. Poulos: "Multi-dimensional neutron-computed tomography using cooled charge-coupled devices", IEEE Trans. on Nucl. Sci. Vol. 38-2, 1991
10. T. Bücherl, E. Steichele: "Contrast variation in neutron computer tomography by change of the neutron wavelength", Proc. 3^rd World Conf. Neutron Radiography Osaka 1989, printed 1990, pp. 821-825
11. B. Schillinger: " Improved Radiography and 3D Tomography due to better beam geometry " presented at the 6^th World Conf. Neutron Radiography, Osaka 1999
12. B. Schillinger, Dissertation TU Munich, Institute E21, 1999
13. E. Lehmann et al., Properties of the Radiography Facility NEUTRA at SINQ and its Potential for Use as European Reference Facility, 6^th World Conf. on Neutron Radiography, Osaka, 1999
14. P. Vontobel et al., Neutron tomography setup at SINQ: Status and first results, PSI-Annual Report, Volume VI: Large Research Facilities
15. Treimer, W. Schaper, J., BENSC Exp. Rep. (1996) p 319
16. Treimer, W., Feye-Treimer, U., Herzig, C., Physica B 241-243 (1998), 1197-1203
17. G. Bayon, Present applications of neutron radiography in France, Nucl. Instr. and Meth. in Physics Research A 424 (1999) 92-97
18. S. Körner, H. Pleinert, H. Böck, Review of Neutronradiography Activities at the Atominstitut of the Austrian Universities, Proceedings of the 5th World Conference on Neutron Radiography, Berlin 1996
19. S. Körner, E. Lehmann, P. Vontobel, Design and Optimisation of a CCD - Neutron Radiography Detector, accepted for publication in Nuclear Instruments and Methods A.
20. Mikerov V., Waschkowski W.: A Two-dimensional detector for fast neutron fields imaging, Nuclear Instruments and Methods in Physics Research A 424 (1999) 48-52.
21. Podolyak E.: Maximum Likelihood Algorithm in Neutron Tomography, Proceedings of the 5^th World Conference on Neutron Radiography, pp. 155-160 (June 17-20, 1996, Berlin, Germany).
22. Mikerov V. et al.: High Resolution Neutron Tomography // Physica Scripta 57 190-195 (1995)
23. K.M. Podurets et al., Application of capillary optics to neutron radiography, Review of Scientific Instruments, Vol. 69, No. 10, p. 3541 (Oct. 1998)

© AIPnD , created by NDT.net

|Home|    |Top|