Available online at www.sciencedirect.com Physics Procedia 26 (2012 ) 88 96 Union of Compact Accelerator-Driven Neutron Sources I & II Neutronic studies on a pulsed thermal neutron source based on the Be(p,n) reaction by using a compact proton accelerator Hiroyuki Hasemi*, Fujio Hiraga, Yoshiaki Kiyanagi Graduate Schoool of Engineering, Hokkaido University, Kita-13 Nishi-8, Kita-ku, Sapporo 060-8628, Hokkaido, Japan Abstract A neutron source based on a compact proton accelerator can be used for neutron scattering and imaging experiments. To promote the neutron utilization in various fields of sciences and technologies, it is necessary to develop a highperformance compact accelerator based neutron source for efficient neutron beam experiments. For this purpose, we have performed design studies for an optimal moderator system for thermal neutrons aiming at the imaging. Therefore, we have carried out the neutronic studies by using the MCNPX code. The Be(p,n) reaction was assumed (proton energy 11 MeV), which gives a neutron yield of 2.15x10 13 n/sec at 1 ma. Firstly, we studied neutronic characteristics of the intensity, the brightness and the pulse width to choose a moderator material for imaging between light water and heavy water. Secondly, we optimized moderator size and reflector thickness. It was indicated that a thermal neutron flux at 10 m from the moderator was 9.43x10 5 n/cm 2 /sec (L/D is 70.7) in case of a slab type. Finally, we investigated the effect of a target cooling system, a moderator vessel and a gap between the Be target and the moderator on a neutron flux. In such a practical model, the thermal neutron flux reduced to about 76 % of a simple model. 2011 2012 Published by Elsevier Ltd. B.V. Selection and/or peer-review under responsibility of of UCANS Open access under CC BY-NC-ND license. Keywords: Thermal neutron source, Compact proton accelerator, Be(p,n) reaction, MCNPX, Neutron imaging ; 1. Introduction A compact accelerator-driven neutron source has several advantages over a large accelerator-driven neutron source in terms of low cost for construction and maintenance, flexible machine time. Therefore, utilization of neutron is promoted by increasing such a compact neutron source. Neutron imaging is one of * Corresponding author. Tel.: +81-11-706-6653; fax: +81-11-706-7896. E-mail address: hasemi@eng.hokudai.ac.jp. 1875-3892 2012 Published by Elsevier B.V. Selection and/or peer-review under responsibility of UCANS Open access under CC BY-NC-ND license. doi:10.1016/j.phpro.2012.03.013
Hiroyuki Hasemi et al. / Physics Procedia 26 ( 2012 ) 88 96 89 the main fields for the compact neutron sources. So far, reactor based neutron radiography facilities have been used. However, the accelerator based neutron sources are very useful for the industrial use and also the on-demand use. We have aimed to develop a pulsed thermal neutron source by using a compact proton accelerator for imaging. Because a produced neutron intensity of 10 13 n/sec or over is desired for efficient neutron beam experiments, we have assumed the Be(p,n) reaction with 11 MeV protons, which gives a neutron yield of 2.15x10 13 n/sec/ma[1]. To develop a high-performance compact accelerator-driven pulsed thermal neutron source, we have carried out simulation calculations for neutronic studies. First, we have chosen a moderator material for thermal neutron between light water and heavy water in terms of intensity, brightness and characteristics of pulse. Secondly, we have optimized a target-moderator-reflector assembly (TMRA). Finally, we have studied effects of some design parameters considered in a practical TMRA on thermal neutron flux. 2. Choice of a moderator material It is well known that light water is much better than heavy water as a pulsed moderator. However, the intensity is expected to be higher than that of light water in the case of a wider moderator. For the choice of the moderator material for the imaging we have compared the neutronic characteristics of light water with those of heavy water by simulation. Fig. 1 shows the calculation model. Here, we assumed a thickness of 2 mm and a radius of 2 cm for a Be target. The thickness of Be reflector, which is well known to have good characteristics as a reflector, was chosen as 30cm tentatively, although the thickness is little bit smaller than the optimal one as shown later. In this study, the MCNPX[2] code of version 2.6, and ENDF/B-V,-VI [3][4]and -VII[5] were used. Be target Moderator (H 2 O or D 2 O) Tally at 5 m Proton Neutron Be reflector Fig. 1. Calculation model for choice of a moderator material 2.1. Intensity Firstly, we calculated thermal neutron flux less than 0.5 ev at 5 m from the moderator as a function of side length of moderator surface by point tally. Here, we tallied the neutrons from the moderator and the flight path length was chosen only for check the intensity. Fig. 2 shows thermal neutron flux depending on area of the moderator surface. The light water moderator has a thickness of 5 cm and the heavy water 20 cm. The light water moderator with an area of 30 x 30 cm 2 gives the maximum thermal neutron flux, and in the heavy water moderator case even at an area of 50 x 50 cm 2 the intensity almost levels off. Next,
90 Hiroyuki Hasemi et al. / Physics Procedia 26 ( 2012 ) 88 96 we fixed the side length of moderator surface 30 cm and 50 cm, respectively, and studied thickness dependence. Fig. 3 shows the thermal neutron flux depending on the thickness of the moderator. The moderator thickness which gives maximum thermal neutron flux is 4 cm in case of the light water moderator and 18 cm in case of the heavy water moderator. The thermal neutron flux of the heavy water moderator is 1.7 times as large as that of light water moderator. However, compared with at same L/D, intensity of light water moderator is 1.6 times as large as that of the heavy water moderator since the intensity is proportional to (D/L) 2. Here, L is the flight path length of neutrons and D the moderator side length. Here, we assumed direct use of the neutrons emitted from the whole surface of the moderator. No collimators are used. 7.0E-07 6.0E-07 5.0E-07 4.0E-07 3.0E-07 2.0E-07 1.0E-07 light water heavy water 20 30 40 50 Side length of moderator surface [cm] Fig. 2. depending on area of moderator surface. Thickness of light water moderator is 5 cm and that of heavy water moderator 20 cm. Thickness of heavy water moderator [cm] 14 16 18 20 22 24 26 7.0E-07 6.0E-07 5.0E-07 4.0E-07 3.0E-07 2.0E-07 light water 1.0E-07 heavy water 1 3 5 7 9 11 Thickness of light water moderator [cm] Fig. 3. depending on thickness of moderator.
Hiroyuki Hasemi et al. / Physics Procedia 26 ( 2012 ) 88 96 91 2.2. Brightness Although the integrated intensity over the moderator surface has been discussed, the spatial distribution of the neutron intensity at the moderator surface is important since it defines the real viewed area of the moderator. Therefore, we calculated distribution of thermal neutron flux at the emission surface of the moderator, i.e. brightness. In this calculation, light water moderator has a size of 30 x 30 x 4 cm 3, and heavy water moderator 50 x 50 x 18 cm 3. Fig. 4 shows the spatial distribution. The neutron flux at the centre of the light water moderator is 3.0 times as large as that of heavy water moderator, indicating that effectively the better L/D will be obtained by the light water moderator. Fig. 4. Spatial distribution of neutrons at the moderator surface. Position x is distance from the centre of moderator 2.3. Characteristics of pulse 1.4E-03 1.2E-03 1.0E-03 8.0E-04 6.0E-04 4.0E-04 2.0E-04 light water heavy water -30-20 -10 0 10 20 30 position x [cm] A moderator giving sharper pulses gives higher energy resolution, so such a moderator is preferable for the pulsed neutron experiments. We simulated the emission time distribution for each moderator thickness, and then compared peak intensities and FWHMs depending on neutron energy. In this calculation, light water moderator has the area of 30 x 30 x 4 cm 3, and heavy water moderator 50 x 50 x14 cm 3 as before. Fig. 5 shows the pulse peak intensities and Fig. 6 shows FWHMs of emission time distribution depending on the neutron energy. The peak intensity of the light water moderator is 3.2 times as large as that of heavy water moderator at the energy of 27.7 mev. FWHM of light water moderator is 11 % of that of the heavy water moderator at 14 cm at the energy of 27.7meV. These results indicate that the heavy water moderator does not have advantage for the imaging experiments.
92 Hiroyuki Hasemi et al. / Physics Procedia 26 ( 2012 ) 88 96 Peak intensity [n/cm 2 /MeV/sec/source] 3.0E+04 2.5E+04 2.0E+04 1.5E+04 1.0E+04 5.0E+03 light water 4cm heavy water 14cm 0 100 200 300 400 500 Neutron energy [mev] Fig. 5. Pulse peak intensities of emission time distribution depending on neutron energy. Light water moderator has thickness of 4 cm and heavy water moderator 14 cm. FWHM [μsec] Fig. 6. FWHMs of emission time distribution depending on neutron energy. Light water moderator has thickness of 4 cm and heavy water moderator 14 cm. 3. Optimization of TMRA 800 700 600 500 400 300 200 100 0 light water 4cm heavy water 14cm 0 100 200 300 400 500 Neutron energy [mev] We studied two types of TMRA, a slab type and a lateral type. Fig. 7 shows the simple calculation models of the slab type and the lateral type. The slab type gives higher neutron beam intensity while the lateral type can reduce the fast neutrons from the target. We obtained thermal neutron flux in an energy region of less than 0.5eV at 10 m from the moderator as a function of the side length of moderator surface, the thickness of moderator and reflector. The reason we calculated thermal neutron flux at 10 m from the moderator is to make the L/D higher than before, 5 m case. Here, we have optimized moderator having a side length of 10 cm because moderators around 10 cm are usually used. Fig. 8 shows thermal neutron flux depending on thickness of the moderator. The optimal thicknesses obtaining the highest
Hiroyuki Hasemi et al. / Physics Procedia 26 ( 2012 ) 88 96 93 thermal intensity are 4 cm for the slab type and 4.5 cm for the lateral type. Fig. 9 shows thermal neutron flux depending on the thickness of the reflector. At a thickness of 50 cm the increase of the thermal neutron intensity almost levels off. The optimal thickness of the reflector is 50 cm in both types. In this study we used a reflector with a thickness of 50 cm. The intensities obtained under these conditions with a proton current 1 ma are 9.43x10 5 n/cm 2 /sec for the slab type and 8.85x10 5 n/cm 2 /sec for the lateral type at 10 m from the moderator. The L/D in this case is 70.7. Fig. 10 shows thermal neutron flux and normalized intensity depending on the side length of the moderator surface in the slab type. The normalized intensity is the thermal neutron flux divided by the area of moderator surface, which means the intensity at the same L/D. The moderator having side length of 30 cm gives the maximum thermal neutron flux, while the moderator having side length of 4 cm has the maximum normalized intensity. Compared at the same L/D, a moderator having a side length of 4 cm gives the highest intensity. Therefore, the optimal side length of moderator surface is 4 cm in this study. Be reflector H 2 O moderator Be reflector Proton Neutron Neutron (a) Slab type Be target Proton (b) Lateral type Fig. 7. Simple type calculation models for optimizing TMRA. (a) Slab type, (b) Lateral type. 4.5E-08 4.4E-08 4.3E-08 4.2E-08 4.1E-08 Slab Lateral 4.0E-08 2 3 4 5 6 7 Thickness of moderator [cm] Fig. 8. depending on the moderator thickness (side length: 10 cm, reflector thickness: 50 cm).
94 Hiroyuki Hasemi et al. / Physics Procedia 26 ( 2012 ) 88 96 5.0E-08 4.0E-08 3.0E-08 2.0E-08 1.0E-08 Slab Lateral 20 30 40 50 60 Thickness of reflector [cm] Fig. 9. depending on the reflector thickness (moderator side length: 10 cm, moderator thickness: optimal thicknesses for both types). 1.0E-07 1.0 9.0E-08 0.9 8.0E-08 0.8 7.0E-08 0.7 6.0E-08 (Time- and areaintegrated intensity) 0.6 5.0E-08 0.5 Normalized intensity 4.0E-08 (Time-integrated 0.4 3.0E-08 intensity) 0.3 2.0E-08 0.2 1.0E-08 0.1 0.0 0 10 20 30 40 50 Side length of moderator surface [cm] Fig. 10. and normalized intensity depending on side length of moderator surface in Slab type. 4. Effect of some design parameters It is necessary to consider effect of design parameters of a realistic model of the moderator system on the thermal neutron flux. The parameters considered here are moderator vessel, cooling system and distance from Be target to moderator vessel. Fig. 11 shows the practical calculation model of the slab type. The moderator size is 10 x 10 cm 2 and the thickness is optimal ones for both cases, and the reflector thickness is 50 cm. A moderator vessel is assumed to be made with 5 mm aluminium, and a cooling system of the target consists of a can of copper and cooling water. Here, we chose copper simply by the reason of thermal conductance although we should consider the total performance of the target system including heat removal, activity and so on. The practical model of the lateral type has the same design parameters. We calculated thermal neutron flux as a function of the thicknesses of copper can and cooling water, and distance from Be target to moderator vessel. Fig. 12 and Fig. 13 show the ratio of thermal Nortmalized intensity [a.u.]
Hiroyuki Hasemi et al. / Physics Procedia 26 ( 2012 ) 88 96 95 neutron flux depending on distance from Be target to moderator vessel in case of the slab type and the lateral type. In the practical model of the slab type, the thermal neutron flux is less than 76 % of the simple model and decreases at a rate of about 2.8 % per 1 cm distance. In the practical model of the lateral type, the thermal neutron flux is less than 75 % of the simple model, and decreases at a rate of about 2.5 % per 1 cm distance. in the practical slab model will be less than 7.17x10 5 n/cm 2 /sec, and that of the lateral model will be less than 6.66x10 5 n/cm 2 /sec at 10 m from the moderator under the condition of proton current 1 ma. Be target Cu can Proton Neutron φ 4 cm Cooling water H 2 O Moderator vessel Be reflector Cooling system moderator Al, t=5mm Fig. 11. A practical calculation model of Slab type. Ratio of thermal neutron flux (practical model / simple model) 0.80 0.78 0.76 0.74 0.72 0.70 0.68 0.66 0.64 0.62 0.60 0 1 2 3 4 5 6 Distance from Be target to moderator vessel [cm] Cu(3mm), water(4mm) Cu(3mm), water(2mm) Cu(1mm), water(4mm) Cu(1mm), water(2mm) Fig. 12. Ratio of thermal neutron flux depending on distance from Be target to moderator vessel in a practical model of the slab type.
96 Hiroyuki Hasemi et al. / Physics Procedia 26 ( 2012 ) 88 96 Fig. 13. Ratio of thermal neutron flux depending on distance from Be target to moderator vessel in a practical model of the lateral type. 5. Conclusion Ratio of thermal neutron flux (practical model / simple model) 0.80 0.78 0.76 0.74 0.72 0.70 0.68 0.66 0.64 0.62 0.60 0 1 2 3 4 5 6 Distance from Be target to moderator vessel [cm] Cu(3mm), water(4mm) Cu(3mm), water(2mm) Cu(1mm), water(4mm) Cu(1mm), water(2mm) As a compact neutron source based on an accelerator a moderator system using the Be(p,n) reaction was studied for imaging experiments. The neutronics of a heavy water moderator was studied since it gives higher intensity than the light water moderator if we use a large surface moderator. However, light water is much better than heavy water as a moderator for imaging experiments since light water moderator gives higher intensity than heavy water moderator under the same L/D condition. The thermal neutron flux in a practical model of the slab type is 7.17x10 5 n/cm 2 /sec at 1 ma when L/D is 70.7. This intensity is acceptable for the imaging experiments. Therefore, a pulsed thermal neutron source based on the Be(p,n) reaction by using a compact proton accelerator can be applied for imaging experiments. However, we did not consider the detailed design of the target system here and the practical design of the target should be done in the near future. References [1] S. Kamada, et al., Preprints 2006 Spring Meeting of At. Energy Soc. Jpn., K42, (2006), [in Japanese] [2] J. F. Briesmeister, ed., MCNP A General Monte Carlo N-Particle Transport Code, Version 4C, Los Alamos National Laboratory report LA-13709-M (March 2000) [3] P.F. Rose (editor), ENDF-201: ENDF/B-VI Summary Documentation, Tech. Rep. BNL-NCS-17541, National Nuclear Data Center, BNL, (October 1991) [4] R. E. MacFarlane, Los Alamos National Laboratory Report LA-12639-MS (ENDF356) (1994) [5] M. B. Chadwick et al., ENDF/B-VII.0 Next Generation Evaluated Nuclear Data Library for Nuclear Science and Technology, Nuclear Data Sheets, 107(12), 2931-3059 (2006)