Journal of the Korean Physical Society, Vol. 59, No. 2, August 2011, pp. 1195∼1198
Computational Tools and Nuclear Data for Radioprotection Studies in Low Energy Light Ions Accelerators P. Sauvan,∗ A. Mayoral, J. Sanz, F. Ogando, M. Garc´ıa and D. Lopez Departamento de Ingenier´ıa Energ´etica, E.T.S. Ingenieros Industriales, UNED, Madrid, 28040, Spain and Instituto de Fusi´ on Nuclear, Madrid, 28006, Spain
A. J. Koning Nuclear Research and Consultancy Group NRG, Petten, 1755 ZG, The Netherlands
A. Ibarra CIEMAT, Madrid, 28040, Spain (Received 26 April 2010) Some of the current accelerator programs such as TechnoFusion, Spiral2 or EVEDA/IFMIF will use low energy particles such as proton, deuteron and alpha. For these projects, the radioprotection studies require a more reliable prediction of the neutron and photon generated by interaction of these charged particles. In this paper, some developments to address this issue are presented. Except for proton, Monte Carlo codes such MCNPX or PHITS use built-in semi-empirical nuclear models to deal with charged particles interactions. Such models, applied to the above mentioned accelerators beam characteristics (particle type and energy), lead to unreliable prediction of secondary particle production. Consequently, there is a need to both generate accurate evaluated data libraries for light ion nuclear reactions and extend MCNPX to handle the evaluated charged particle nuclear data. The MCNPX code has been extended by the authors to handle proton, deuteron, triton and alpha nuclear data libraries. This tool is here presented and applied to analyse the reliability of available evaluated nuclear data for incident deuterons of energies those used in the abovementioned facilities (up to 40 MeV). Deuteron-induced reactions have been chosen because this particle is expected to be used in all these facilities. Concerning target material, copper has been selected as it is one the main constituents of accelerating components and beam dumps. The evaluated nuclear data are provided by the TENDL library, which is the only one available for deuterons with a wide range of target elements, including those of concern for the accelerators here considered. The testing of the TENDL nuclear data is carried out by comparing existing experimental data on thick target neutron yields for the given materials with those computed by the modified MCNPX code using TENDL cross sections. As a result, the assessment of its applicability to radioprotection studies of those accelerators is discussed. PACS numbers: 29.85.+c, 07.05.Tp Keywords: Experimental nuclear data, Extended MCNPX, MCUNED, TENDL library DOI: 10.3938/jkps.59.1195
I. INTRODUCTION Several important research programs based on using a deuteron accelerator facility are currently under development. Some of the most important are SARAF [1], SPIRAL2 [2] and IFMIF/EVEDA [3,4], in which deuterons of relatively low energy (less than 40 MeV) interact with a dedicated target to produce neutrons for different applications. Other facilities addressing also the simulation of fusion reactor materials follow a complementary approach by using triple ion beams, in which one of them can be a deuteron or alpha beam. A recent proposal at ∗ E-mail:
this respect is the Spanish Technofusi´on Project [5]. Radioprotection studies of such facilities require reliable nuclear data for the production and transport of neutrons and photons produced by deuteron and alpha nuclear reactions. Up to now, due to the lack of such nuclear data libraries, Monte Carlo transport codes use built-in semi empirical nuclear models to deal with charged particle interactions. Those models at low incident particle energy up to 9 MeV on Copper [6], and up to 20 Mev for different elements of accelerator equipments [7], were found to predict deficiently the secondary particle production. The release of the TENDL libraries [8] allows to use tabulated data produced by the nuclear code TALYS [9] instead of semi-empirical model in trans-
[email protected]
-1195-
-1196-
Journal of the Korean Physical Society, Vol. 59, No. 2, August 2011
port calculations. The use of such tabulated data introduces the flexibility to change easily nuclear data used in transport code when they need to be improved. Unfortunately, current Monte Carlo transport codes, such as MCNPX, PHITS or FLUKA, are not able to handle light ion libraries except for protons. To overcome this drawback the MCNPX code has been extended to handle deuteron, triton, and alpha nuclear data libraries. In this new extended MCNPX called MCUNED [10], a new variance reduction technique has also been implemented for the production of secondary particles induced by light ions nuclear reactions, allowing a drastic reduction in the computing time in many applications. The objective of this paper is to test the TENDL library against available experiments for deuteron induced reactions on thick copper target, for incident energies up to 40 MeV. The assessment of its applicability to radioprotection studies of those accelerators is discussed. II. MCUNED: AN EXTENDED MCNPX CODE In order to handle deuteron libraries, the MCNPX code has been modified. The MCNPX subroutines devoted to reading and processing proton nuclear cross sections from external libraries have been modified allowing the use of deuteron libraries. Subroutines dedicated to sample nuclear reactions have been modified as well, in order to be able to use tables instead of models during the deuteron transport. The particle transport processes remain the original from MCNPX. In the MCUNED code a new variance, reduction technique has also been included for the production of secondary particles. When the interest of the simulation is emphasized in the transport of secondary particles, like radioprotection studies in accelerators, it is necessary to perform the calculation with a large number of secondary particle histories in order to achieve an acceptable statistical error. In such calculations, if the production of secondary particles per primary is very low, a large amount of source particles is necessary to produce enough secondary particles to obtain the desired accuracy. The main idea of the variance reduction technique implemented in MCUNED, for the production of secondary particles from light ions induced nuclear reaction, is to always force the nuclear reaction for each charged particle history. The weight of the secondary particle produced is then multiplied by the probability of such nuclear absorption, in order to keep the yield of secondary particle production. III. TENDL DEUTERON LIBRARY COMPARISON AGAINST EXPERIMENTAL DATA In this section, experimental data on deuteron nuclear reactions with copper are compared with simulations
Table 1. Neutron production (over 4π). Ref 11 12 13 13
Ed (MeV) 10 15 33 33
En > (MeV) 0 1 0 4
Exp (n/d) 8.81E-4 4.65E-4 1.81E-2 7.88E-3
TENDL (n/d) 6.44E-4 1.51E-3 1.50E-2 4.09E-3
Yield frac. 1 0.64 1 0.27
Table 2. Neutron yield in forward direction. Ref
Ed (MeV)
En > (MeV)
Exp (n/d/Sr)
TENDL (n/d/Sr)
Yield frac.
12 14 13 14
15 16 33 33
1 4 4 4
7.69E-5 2.76E-4 6.16E-3 6.68E-3
1.71E-4 5.94E-5 7.49E-4 7.49E-4
0.70 0.20 0.39 0.39
performed with the MCUNED code using TENDL 2009 library. In the first section, the total neutron production is considered. Then, neutron spectrum and neutron angular distribution are compared with available experiments. Finally the results of the comparison the evaluated data with experimental is discussed. 1. Total neutron production
Experiments [11-14] on deuteron nuclear reaction on copper have been performed for various incident deuteron energies. In these works the neutron yield is measured in thick target experiments. In order to simulate these experiments, the MCUNED code has been used with the TENDL 2009 library for deuteron interaction data. The experimental and simulated results are summarised in Tables 1 and 2. For most of the experiments listed above, the yield is measured only for neutrons with energy higher than 1 or 4 MeV. This amount of fast neutrons represents only part of the total neutron yield. In Tables 1 and 2, the yield fraction (last column) represents the proportion of neutron with energy above En in the spectra calculated with MCUNED + TENDL. The measured total yield including all neutron energies is in good agreement with the simulations. The major discrepancy between experiments and simulations affects the neutron yield in the forward direction. In all those comparisons the difference between experimental and simulated values are never higher than one order of magnitude. 2. Neutron spectrum and angular distribution A. Neutron spectrum
Figure 1 show the neutron spectrum emitted in the forward direction for two incident deuteron energies. It is seen that the neutron spectrum in the forward direction
Computational Tools and Nuclear Data for Radioprotection Studies in Low Energy Light Ions· · · – P. Sauvan et al.
MCUNED+TENDL
MCUNED+TENDL 0,75
Exp. [Ref. 12,13]
1E-4 0,50
E =33 MeV
1/sr
n/d / Sr / MeV
Meulders et al.
-1197-
1E-5
d
E
n
> 4 MeV
0,25
E =33 MeV d
1E-6
10
20
30 0,15
E =15 MeV d
1E-5
1/sr
n/d / Sr / MeV
0,00
1E-4
E
n
0,10
> 1 MeV
E =16 MeV
1E-6
d
0,05
1E-7
0
5
10
E
n
20
40
60
80
100
120
140
15 Angle (Deg)
(MeV)
Fig. 1. (Color online) Neutron Spectra emitted in forward direction.
MCUNED-Tendl09 K.Shin etal.
Fig. 3. (Color online) Neutron angular distribution. The angular neutron yield is normalized by total neutron yield form Table 1.
spectra don’t match each other. From 45 degrees and higher the simulated and experimental spectra are in good agreement.
-4
10
B. Neutron angular distribution =0º
-6
10
-4
10
=15º
n/d/MeV/sr
-6
10
-4
10
=45º
In order to complete the study of the neutron emission from deuteron Copper nuclear reaction, the angular distribution of produced neutron has been plotted in Fig. 3. For both incident deuteron energies, the experimental and simulated angular distributions are similar for angle higher than 45 degrees, while in the forward direction the experimental distribution has a stronger contribution.
-6
10
3. Discussion
-4
10
=75º
-6
10
-4
10
=135º
-6
10
0
10
20
30
E (MeV) n
Fig. 2. (Color online) Neutron spectra produced by 33 MeV incident deuteron on thick.
is not correctly reproduced by TENDL, in agreement with values of Table 2, and also it seems to reveal that the differences between both spectra tend to be lower as the deuteron incident energy decreases. Figure 2 shows the neutron spectra at various emission angles for 33 MeV incident deuterons. This figure shows that up to 45 degrees the simulated and experimental
The difference between experimental and simulated spectra can be attributed to the deuteron breakup process during the nuclear reaction. The bump at 13 MeV in Fig. 2 at 0 degree is the typical signature of the breakup process [15], which can be defined as having a projectile fragment emerging from the reaction in a relatively narrow peak centred close to the beam velocity and strongly directed toward forward angles. Since breakup process is includes in TALYS with a simple model [16] the TENDL deuteron library should reproduce this behaviour. Angular integrated neutron spectra produced by TALYS exhibit the breakup bump feature, but the code is not able to reproduce correctly the angular spectra. In fact, to produce the angular neutron spectrum, TALYS use the Kalbach [17] systematic. It is known that this systematic cannot reproduce satisfactorily the peaked forward angular distribution, and instead of producing a pronounced breakup contribution to the forward angle, this systematic smooth the breakup contribution over a large angular aperture. Very recently a new phenomenological model for projectile breakup reaction has been proposed
-1198-
Journal of the Korean Physical Society, Vol. 59, No. 2, August 2011
[18]. In this model, a peaked forward angular distribution is given to reproduce the angular distribution of the emitted fragment of the breakup reaction. IV. TENDL RELIABILITY ENERGY RANGE FOR RADIOPROTECION STUDIES From section III, some conclusions on the reliability of the TENDL library can be assessed. In 10 – 40 MeV incident deuteron energy range, the neutron yield and backward neutron spectrum could be used for radioprotection purpose whereas neutron emitted is the forward direction would be underestimated. Figures 1 and 3 show that the difference between experiments and simulations decrease with the deuteron energy. This behaviour can be explained by the reduction of the breakup contribution to the neutron production [19]. Therefore, it might be expected that below 10 MeV the deuteron TENDL library could be used with a relative reliability. V. CONCLUSION For the purpose of radioprotection studies, the actual version of the deuteron TENDL library reproduce satisfactorily the angular integrated total neutron yield. Unfortunately, the angular neutron spectrum does not reproduce reasonably the experimental spectra. As experimental angular distribution exhibit a peaked forward neutron emission, the simulation gives a flatter neutron angular distribution. This bad estimation of neutron spectra has a very bad consequence on the reliability of dose rate calculations produced by secondary neutrons generated by a target interaction with an incident deuteron beam. In this condition, the dose in the beam direction would be underestimated while it would be overestimated in the other directions. The resulting shielding design of a facility would be then very different considering one or other neutron angular distribution. However, for low incident deuteron energy, a correct behaviour of the TENDL library could be expected.
To solve this problem in the TALYS calculations, a new model for projectile breakup is under implementation. This new model should be available for the release of TENDL 2010 deuteron libraries. ACKNOWLEDGMENTS Work supported by Association EURATOM/CIEMAT for Fusion (AEC) and Plan Nacional de I + D + I (2008 – 2011) Fusion Nuclear, ENE2008-06403-C06-02, MCIN, Spain. REFERENCES [1] I. Mardor et al., Particle Accelerator Conference (Vancouver, Canada, 2009). [2] M. Lewitowicz, Acta Phys. Pol. B 40, 811 (2009). [3] IFMIF Comprehensive Design Report, IFMIF International Team, 2004. [4] P. Garin, M. Sugimoto, Fusion Eng. Des. 84 259 (2009). [5] ISBN: 978-84-7834-628-8, 2009. [6] J. Sanz et al., Fusion Sci. Technol. 56 273 (2009). [7] V. Blideanu, J.Sanz et al., in Proceeding 14th ICFRM (Sapporo, Japan, 2009). [8] A. J. Koning, S. Hilaire, M. C. Duijvestijn, http://www.talys.eu/tendl-2009/. [9] A. J. Koning, S. Hilaire, M. C. Duijvestijn, in Proceedings of Inter. Conf. on Nucl. Data for Sci. and Techn. (Nice, France, 2007), p. 211. [10] P. Sauvan, J. Sanz, F. Ogando, Nucl. Instrum. Methods Phys. Res. Sect. A 614, 323 (2010). [11] L. W. Smith and P. G. Kruger, Phys. Rev. 83, 1137 (1951). [12] A. J. Allen et al., Phys. Rev. 81, 536 (1951). [13] K. Shin et al., Phys. Rev. C 29, 1307 (1984). [14] J. P. Meulders, P. Leleux, P. C. Macq and P. Pirart, Phys. Med. Biol. 2, 235 (1975). [15] C. Kalbach, Phys. Rev. C 71, 034606 (2005). [16] A. J. Koning, S. Hilaire, M. C. Duijvestijn, Talys 1.2. User Manual (2009). [17] C. Kalbach, Phys. Rev. C 37, 2350 (1988). [18] C. Kalbach, http://www-nds.iaea.org/fendl3/docs/dBreakupRCM2.pdf. [19] P. Bem et al., Phys. Rev. C 79, 044610 (2009).
