Neutronyields[n/sr/proton]
Thickness of shield [cm]
Figure 5.10: Attenuation profiles using NE213 scintillator from neutron yields for concrete and steel
121.6±4.5 g/cm2and 121.9±4.2 g/cm2 from calculated results of PHITS and Fluka, respectively. These values are just between those given by reference
(1). PHITS and Fluka reproduces the spectral shape well and attenuation length was in good agreement within factor of 1.01 and 1.01. The neutron attenuation lengths for steel correspond to those for neutrons above 10 MeV are 121.4 ±8.7 g/cm2 from experimental results and 146.0 ±3.9 g/cm2 and 142.7 ±6.7 g/cm2 from calculated results of PHITS and Fluka, respectively.
These values are just between those given by reference(1). PHITS and Fluka reproduces the spectral shape well and attenuation length was in good
agree-steel, both calculations overestimated measured data by 20 %. To check this problems for steel, the C/E ratio were extracted over the indicated energy ranges, as shown in Fig 5.11. The figure shows same tendency according to the energy ranges.
0 10 20 30 40 50 60
0.6 0.8 1 1.2 1.4
C/E
40cm 60cm 80cm 100cm 120cm
0.8 1 1.2 1.4 1.6
C/E
40cm 60cm 80cm 100cm 120cm
0.6 0.8 1 1.2
C/E
40cm 60cm 80cm 100cm 120cm
1) 10 ~ 350 MeV
2) 50 ~ 350 MeV
3) 100 ~ 350 MeV
Figure 5.11: Neutron yields in steel for each measurement points over the indicated energy ranges
Fig 5.12 shows the neutron attenuation profiles using BSS for concrete removable blocks from 80 to 360 cm. The neutron attenuation lengths for concrete correspond to those for neutrons above 10 MeV are 103.3±4.9 g/cm2 from experimental results and 113.9 ±4.8 g/cm2 from calculated results.
These values are just between those given by reference (1). PHITS repro-duces the spectral shape well and attenuation length was in good agreement within factor of 1.10; however, its values were overestimated.
Neutronyields[n/sr/proton]
Thickness of concrete [cm]
Figure 5.12: Attenuation profiles using BSS from neutron yields for concrete and steel
6 Conclusion
This thesis aimed to obtain experimental data on neutron energy spectra for various thicknesses of common shielding materials to reveal neutron at-tenuation mechanism for energies up to GeV. It helps for shielding design of the high energy hadron accelerator. For this present study, two experiments were done to obtain data for transition of neutron spectrum.
In the first experiment, the author evaluated uncertainty in the neutron spectra measurement experimentally. The experiment was done at HIMAC, NIRS. The 100 and 290 MeV/nucleon 28Si beam was irradiated on a silicon target to generate high energy neutron spectrum of which are similar to one for high energy hadron accelerator facility. The neutrons were measured using a 12.7 cm in diameter, 12.7 cm in length liquid organic scintillator, NE213, coupled with a photo multiplier tube, at 75 degree with respect to the beam axis, 2.474 m away from the Si target. The neutron spectra were deduced using two different data analysis technique, TOF and unfolding. The neutron events were extracted from the data measured by the detector using pulse shape discrimination method.
On the TOF method, kinetic energies of neutrons were determined from time difference between beam signal and the NE213 scintillator. Neutron detection efficiency was obtained from SCINFUL-QMD code.
On the unfolding method, QDC light output from the NE213 was elabo-rately calibrated using Compton edges of gamma-rays from 60Co, 241Am-Be, cosmic-ray muons, and recoil proton edge from the TOF data because the unfolded energy spectrum is sensitive to the calibration of light output. Re-sponse function was calculated by SCINFUL-QMD code. Light output spec-tra were unfolded by RooUnfold code on Bayesian algorithm.
The difference between the TOF and the unfolded neutron spectra were 17 and 8 % for 100 and 290 MeV/nucleon 28Si beam at maximum. The ef-fect from uncertainty of calibration was evaluated to be several % on neutron
yield.
In the second experiment, the author obtained complete data set of neu-tron spectra for concrete and steel shields with various thicknesses. The experiment was done at CHARM facility of CERN. The proton beam of 24 GeV/c momentum was irradiated on the copper target with cylindrical shape of 8 cm in diameter and 50 cm in length. The neutron energy spectra were measured on the top roof of CHARM (CSBF). The thickness and material of a part of the roof, 80 cm×80 cm square area, were changed to see transition of the spectra.
The liquid organic scintillator used for the first experiment and two veto detectors were used for high energy neutron measurement. The response function of the scintillator was obtained using SCINFUL-QMD code. Neu-tron energy spectra were obtained by analyzing light output data of the scin-tillator with the unfolding method that was studied in the first experiment.
The spectrum data were obtained for concrete with thicknesses of 0, 40, 80, 120, 160, 200, 240 and 360 cm, and steel with thicknesses of 0, 20, 40, 60 and 80 cm.
For low energy neutron measurement, a Bonner sphere spectrometer with five moderator conditions was used for concrete with thicknesses of 80, 120, 200, 240, 360 cm. With combining these data and previously obtained acti-vation data of Bi and Al plates for concrete with thicknesses of 0, 80 and 160 cm, data sets for three different methods were obtained under the same con-dition. This is quite unique data set since no one provides neutron spectrum data with such variety of thickness and methods until now.
The neutron spectra of the data set were simulated by using Monte Carlo codes with high energy particle interaction models, PHITS and Fluka. To obtain the results with reasonable calculation time, simplified geometry was
shielding thickness condition.
On the measured spectra, prominent peaks were observed around 100 MeV for 0, 40, 80 and 120 cm thick concrete and 0 and 20 cm thick steel cases. The peaks disappear for the thicknesses of more than 160 cm of concrete and 40 cm of steel. The spectrum shapes are quite similar for the thicknesses that means attenuation of neutron follows one for high energy neutrons. The data for both materials shows three orders of magnitude attenuation, approxi-mately. In contrast, low energy neutron measured by the Bonner sphere spectrometer shows only one order of attenuation. The spectra show peak at 1 MeV which is from evaporation process. The spectra also show edge at around 100 MeV due to initial guess spectra. The spectra measured by the scintillator and Bonner are in agreement within a factor. The spectrum data are to complementary each other since the scintillator and BSS lack sensitiv-ity for low and high energy neutrons, respectively.
The results of Monte Carlo simulation successfully reproduce experimental data for all the cases, especially for thicker shielding. For thinner cases, the results underestimate neutrons having energies below the prominent peaks around 100 MeV. The differences are within a factor of two at maximum, which means the code results underestimate neutron up to 120 cm concrete and 20 cm iron shields.
To obtain attenuation length, the neutron energy spectrum for each thick-ness was used integrated for above 10 MeV, then the integrated neutron yields were fitted by an exponential function using a maximum likelihood es-timation. The attenuation lengths, which are important to estimate fluence and dose behind the shield, for concrete and steel, were obtained. Attenua-tion length determined by experiments was determined to be 119.9 g/cm2 for concrete, which is coincident with calculation within 1 %. On the other hand, that for steel was 121.4 g/cm2, which is lower than that of calculation within 20 %.
The data set and cross comparison described above provide more detail understanding of high energy neutron attenuation since it gives not only attenuation through detector output but also energy spectrum of neutrons.
The spectrum enables us to calculate any of quantities due to neutron, such as dose, energy deposition and induced activities etc, to compare previous data. It also enables us to direct comparison with calculation results based on various physics models without effects of complex response functions due to various nuclear reactions of high energy neutron that is difficult to evalu-ate experimentally. For these reasons, the author believes the data set will help a lot for shielding design of future high energy hadron accelerators.
References
(1) D. Filges et al.: "Handbook of Spallation Research: Theory, Experiments and Applications", Wiley, Berlin, (2010), 233.
(2) G.F. Knoll: "Radiation Detection and Measurement", John Wiley Sons, New York, (2000) 514.
(3) N. Nakao et al.: Nucl. Instrum. Meth., B266(2008), 93.
(4) D. Filges et al.: "Handbook of Spallation Research: Theory, Experiments and Applications", Wiley, Berlin, (2010), 19.
(5) D. Filges et al.: "Handbook of Spallation Research: Theory, Experiments and Applications", Wiley, Berlin, (2010), 44.
(6) B. J. Moyer: LBNL Rep. UCRL-9796, (1961).
(7) H. Iwase et al.: Prog. Nucl. Sci. Tech.,1 (2011), 40.
(8) S. Ban et al.: IAEA Technical Report Series No. 283, (1988).
(9) T. Sato et al.: J. Nucl. Sci. Technol.,55 (2018), 684.
(10) A. Ferrari et al.: CERN-2005-010 (2005).
(11) S. Agostinelli et al.: Nucl. Instrum. Meth.,A506 (2003), 250.
(12) N.V. Mokhov: "MARS Code System, Version 15", (2009), www-ap.fnal.gov/MARS.
(13) D.B. Pelowitz: LA-CP-05-0369 (2005).
(14) H. Hirayama et al.: Proc. SATIF-13, NEA/NSC/R 2 , (2018), 274.
(15) S. Ban et al.: Nucl. Instrum. Meth., 174 (1980), 271.
(16) H. Nakashima et al.: Nucl. Sci. Technol., 2 (2002), 1155.
(17) S. Agosteo et al.: Nucl. Instrum. Meth., A459(2001), 58.
(18) S. Agosteo et al.: Nucl. Instrum. Meth., B312 (2013), 36.
(19) M. Hagiwara et al.: Prog. Nucl. Sci. Tech.,1 (2011), 52.
(20) G. R. Stevenson et al.: Health Phys., 43(1982), 13.
(21) A. V. Alevra et al.: Radiat. Prot. Dosim., 107 (2003), 37.
(22) D. J. Thomas: Radiat. Prot. Dosim., 110 (2004), 141.
(23) E. Lee et al.: Nucl. Instr. Meth., B445(2019) 26.
(24) D. Satoh et al.: Nucl. Instr. Meth.,A644 (2011) 59.
(25) H. Bethe et al.: "Experimental Nuclear Physics", John Wiley Sons, New York, (1953) 253.
(26) T. Nakamura et al.: "Handbook on secondary particle production and Transport by High-energy heavy ions", World scientific, Singapore, (2006) 15.
(27) U.S. NIM Committee, Standard NIM Instrumentation System, DOE/ER-0457T (1990).
(28) D. Satoh et al.: Nucl. Instr. Meth.,B387 (2016), 10.
(29) D. Satoh et al.: J. Nucl. Sci. Tech., 43 (2006), 714.
(30) T. Kajimoto et al.: Nucl. Instr. Meth., A665(2011), 80.
(33) G. Dietze et al.: Nucl. Instr. Meth., 193 (1982), 549.
(34) T. Adye: Proc. of the PHYSTAT 2011 workshop on statistical issues re-lated to discovery claims in search experiments and unfolding (2011), 313.
(35) G. Dietze et al.: Nucl. Instr. Meth., 193 (1982), 549.
(36) T. Kajimoto et al.: Nucl. Instr. Meth., A (2018), 53.
(37) J.F. Ziegler et al.: Nucl. Instr. Meth.,B268 (2010), 1818.
(38) www.cern.ch (39) www.cern.ch/r2e/.
(40) M. Brugger: Proc. IPAC-2012-THPPP006 (2012), 3734.
(41) M. Brugger: AESJ in Prog. in Nucl. Sci. Tech., (2011), 948.
(42) R. Froeschl et al.: Proc. SATIF-12, NEA/NSC/R, 3 (2015), 14.
(43) E. Iliopoulou et al.: Nucl. Instrum. Meth., A855 (2018), 79.
(44) A. Thorton: Tech. Rep. CERN-ACC-NOTE-2016-12345 (2016).
(45) Y. Uwamino et al.: Nucl. Instrum. Meth., A 239 (1985), 299.
(46) T. Nunomiya et al.: KEK Internal 2003-8 (2003).
(47) https://acqiris.com/
(48) M.B. Chadwick et al.: Nucl. Sci. Eng., 131 (1999), 293.
(49) K. Shibata et al.: J. Nucl. Sci. Technol., 48(2011), 1.
(50) Y. Nara et al.: Phys. Rev. C., 61 (2000).
(51) A. Boudard et al.: Phys. Rev. C., 87(2013).
(52) Y. Sawada et al.: Nucl. Instrum. Meth.,B291 (2012), 38.
(53) H. Takada et al.: JAERI-Data/Code 98-005, (1998).
(54) K. Niita et al.: Phys. Rev. C., 52 (1995), 2620.
(55) S. Furihata et al.: Nucl. Instrum. Meth., B171(2000), 251.
(56) T.W. Armstrong et al.: ORNL-4869, (1973).
(57) H. Geissel, C. Scheidenberger, P. Malzacher, J. Kunzendorf, H. Weick, ATIMA, GSI.
(58) T. Furuta et al.: IFMBE Proc., 39(2013).
(59) H. Osnes: ISSN 0808-2839 (1997).
(60) I. Kawrakow et al.: Phys. Med. Biol., 45(8) (2000), 2163.
(61) LA-10363-MS DE86 004380.
Appendix A
(1) Input parameter for PHITS calculation
Some sections as input parameters in PHITS are listed in Table A.1.
Table 6.1: Sections as input parameters in PHITS code
Name Description
[title] Title
[parameters] Various type of parameters
[source] Source definition
[material] Material definition
[cell] Cell definition by geometry system [surface] Surface definition by geometry system [importance] Region importance definition
[t-track] Track length tally definition
[t-gshow] Region surface display definition for graphical plot
[end] End of input file
Number of histories, energy cut-off, nuclear data library and physics model settings, and other general parameters are defined in the [parameters] sec-tion. In the [source] section, the source types, energy and position are fined. The number of materials, composition and particle densities are de-fined in the [material] section. The surface and cell, which represent calcu-lation geometry, are defined in the [surface] and [cell] section, respectively.
The [importance] sets importance sampling (Russian Roulette /splitting) at boundary crossings. The [t-track] is used to obtain track length fluence in any specified region. The [t-gshow] is for two dimensional geometry plots.
(2) Input parameter for Fluka calculation
Some sections as input parameters in Fluka are listed inTable A.2.
Table 6.2: Sections as input parameters in Fluka code
Name Description
TITLE Title and comments for documentation purposes BEAM Beam characteristics definition
BEAMPOS Primary particle source definition GEOBEGIN Start geometry definition
GEOEND Stop geometry definition
MATERIAL Material definition
ASSIGANMAT Assigning materials to regions
SCORE Definition of particles to be scored in region BIASING Region biasing definition
USRTRACK Track length fluence estimator
RANDOMIZE Initialization of the random number sequence START Start signal and number of requested histories
STOP End of input file
The beam characteristics such as energy, profile, divergence and particle type are defined in the BEAM section. In the BEAMPOS section, the starting point of beam particles and the beam direction are defined. The GEOBEGIN and GEOEND define to start and stop the combinational geometry. The MA-TERIAL defines a materials and its properties. The ASSIGANMAT defines the correspondence between region and material indices and defines regions