- Email: [email protected]
Experimental characterization of a long counter for neutron fluence measurement
Radiation Measurements 119 (2018) 16–21
Contents lists available at ScienceDirect
Radiation Measurements journal homepage: www.elsevier.com/locate/radmeas
Experimental characterization of a long counter for neutron fluence measurement
T
Z.M. Hua, X.Y. Penga, Z.J. Chena, T.F. Dua, L.J. Gea, X. Yuana, Z.Q. Cuia, W.J. Zhua, Z.M. Wanga, X. Zhua, J.X. Chena, X.Q. Lia, G.H. Zhanga, J. Chenb, H. Zhangc, G. Gorinid, T.S. Fana,∗ a
School of Physics and State Key Laboratory of Nuclear Physics and Technology, Peking University, Beijing, 100871, China China Institute of Atomic Energy, Beijing, 102413, China c National Institute of Metrology, Beijing, 100029, China d Dipartimento di Fisica ‘G. Occhialini’, Università degli Studi di Milano-Bicocca, Milano, 20126, Italy b
A R T I C LE I N FO
A B S T R A C T
Keywords: Long counter Calibration Response function Effective center
A De Pangher-type long counter was constructed at Peking University for neutron fluence measurement. The responses and effective centers of the long counter were calculated using the Monte Carlo method. The variations with neutron energy of the position of the effective center, calculated by the Monte Carlo code, was experimentally validated with a 241AmBe neutron source and several mono-energetic neutron sources ranging from 100 keV to 6 MeV. Long counter calibration was performed using a radionuclide source from Peking University, and neutron fluence measured by the calibrated long counter was successfully compared to values determined with reference instruments (two recoil proton counters and a 238U fission chamber). The relative deviations were lower than 6% along the whole energy range. The calibrated long counter was thus successfully applied in the calibration of a Bonner sphere spectrometer.
1. Introduction Experimental validation of the response function of a Bonner sphere spectrometer was performed in the experimental hall of Peking University at six mono-energetic neutron fields ranging from 100 keV to 20 MeV provided by a 4.5-MV Van de Graaff accelerator (Hu et al., 2014b, 2017) at the State Key Laboratory of Nuclear Physics and Technology. Neutron fluence measurement devices should be employed to determine the neutron fluence value at the measurement position of the Bonner sphere spectrometer (Alevra and Thomas, 2003; Birattari et al., 2010; Klein and Thomas, 2003; Lacoste et al., 2004). Two types of reference devices, namely two recoil proton proportional counters and one 238U fission chamber (Hu et al., 2017; Zhang et al., 2011), were used to measure neutron fluence within energy ranges from 100 keV to 1 MeV and from 2 MeV to 20 MeV, respectively. They were designed by China Institute of Nuclear Energy and the two counters were traceable to international standards (Ryves, 1987). The uncertainty on the neutron fluence is approximately 2% with the counters (Chen et al., 2007) and 3% with the fission chamber. This chamber is made of a 2-cm diameter 238U sample attached to the inner face of a copper cylinder box. The 238U sample has a mass of (547.2 ± 1.3%) μg with enrichment greater than 99.997% (Zhang et al., 2011). To provide reference ∗
fluence values between 1 MeV and 2 MeV, a calibrated long counter was selected, and the range of use varied from a few eVs/keVs to 20 MeV. The long counter has been widely applied in neutron fluence measurement due to characteristics such as a relatively flat response in a wide neutron energy range between a few eVs or keVs and a few MeVs; insensitivity to gamma rays with an appropriate threshold; and good stability (Tagziria and Thomas, 2000). The Hanson and McKibben (1947) and De Pangher (De Pangher and Nichols, 1966) types of long counters have been employed as standard instruments in many laboratories (Gressier et al., 2014; Nolte and Thomas, 2011; Roberts et al., 2010; Tagziria and Thomas, 2000). They are based on a cylindrical moderator, usually made of paraffin or polythene materials with a thermal neutron tube in the center. The moderator is embedded in an additional shield to reduce sensitivity to room-scattered neutrons and endow the counter with high efficiency to neutron incidents from the front face along the axis. The response of long counters can be roughly independent of energy, up to 3 MeV–5 MeV depending on type. Recently, some newly developed long counters have also been reported in the literature (Gressier and Lacoste, 2014; Harano et al., 2011; Hu et al., 2014a; Kim et al., 2012; Lacoste, 2010; Lacoste and Gressier, 2010; Mazunga et al., 2017; Tanimura et al., 2014), using Monte Carlo simulations to either optimize the geometry or employ new materials.
Corresponding author. E-mail address: [email protected] (T.S. Fan).
https://doi.org/10.1016/j.radmeas.2018.08.017 Received 4 November 2017; Received in revised form 24 August 2018; Accepted 28 August 2018 Available online 29 August 2018 1350-4487/ © 2018 Published by Elsevier Ltd.
Radiation Measurements 119 (2018) 16–21
Z.M. Hu et al.
detector reading (number of 3He (n, p)3H reactions) to the incident neutron fluence. In response function calculations, a parallel neutron beam starting from a disk with a 40-cm diameter irradiated the front face of the long counter along its axis, and the F4 and FM4 tally cards of MCNP5 were combined to derive the response values (Hu et al., 2014b, 2017). The effective center, r0, is defined as the position of the calibration point. The long counter is regarded as a point detector in the fluence measurements from a point neutron source, such that the count rate varies as (r + r0)−2, where r is typically the distance from the point source to the front face of the long counter (Lacoste and Gressier, 2010; Roberts et al., 2004; Tagziria and Thomas, 2000). In calculating effective centers, the inverse of the square root of the count rate against distance r was linearly fitted to yield the effective centers. The counts of the long counter at 51 distances were simulated. The pseudorandom number for each distance was different and randomly generated as in reference (Roberts et al., 2004). This method can overcome correlations between calculations at each distance by using the same pseudorandom number. At each distance, the calculation was separated into six runs with different initial random numbers.
A long counter based on the De Pangher design was constructed at Peking University to calibrate a Bonner sphere spectrometer and for other applications. This long counter, which generally does not have a perfectly flat response over a wide energy range, needed to be experimentally calibrated because its efficiency cannot be derived from first principles like recoil proton telescopes (Roberts et al., 2010). For any feasible fluence measurement, the effective center of the long counter, depending on the counter design and neutron energy, should also be accurately determined by combining measurements and detailed Monte Carlo simulations (Roberts et al., 2010; Tagziria and Thomas, 2000). In this paper, Monte Carlo simulations were performed to determine the fluence responses and effective centers as a function of neutron energy. Simulations were then validated by measurements at eight mono-energetic neutron fields between 100 keV and 14.0 MeV as well as in front of a 241AmBe radioactive source. Good agreement was obtained between the experimental and simulated values. 2. Materials and methods 2.1. Construction of long counter The schematic geometry and image of the De Pangher-type long counter are presented in Fig. 1. The cylindrical long counter measures 43 cm in length and 39 cm in diameter and consists mainly of a thermal neutron counter, two annular moderators, an outer annular shield, and three thermal neutron absorbers. The thermal neutron counter, manufactured by CENTRONIC Ltd., is a proportional tube with a diameter of 3.8 cm and a length of 45 cm, filled with 3He gas at a pressure of 6 atm. The moderators and shield are made of polyethylene with a mass density of 0.93 g/cm3. Boron loaded plastic, measuring 1 cm thick, is installed between the moderators and outer shield. A 0.6-mm thick rear cadmium sheet is inserted between the front and rear polyethylene moderators, and another cadmium sheet is mounted on the front face of the long counter. An annular trough is dredged in the front moderator. All parts of the long counter are supported and fixed with aluminum materials.
2.3. Experimental setup The experimental determination of the response and effective center values, which aimed to verify the calculated values, was carried out using a 241AmBe source and eight mono-energetic neutron sources ranging from 100 keV to 14.0 MeV in the experimental hall (20 m long, 12 m wide, and 8 m high) at Peking University (Hu et al., 2017). The neutron emission rate of the 241AmBe source was determined to be (6.94 ± 0.75%) × 106 s−1 on July 21, 2006 based on manganese bath equipment from China Institute of Atomic Energy (Roberts et al., 2011). The manganese bath equipment was applied in neutron emission rate comparisons in 1984 (Axton, 1987) and 2005. The neutron spectrum and anisotropy factor were calculated using the Monte Carlo method (Li et al., 2013; Liu et al., 2007). The charged particle beams accelerated by a 4.5-MV Van de Graaff accelerator were sent to various solid targets to produce mono-energetic neutron fields of 0.144 MeV, 0.565 MeV, 1.2 MeV, 2.5 MeV, 4.0 MeV, 5.0 MeV, 6.0 MeV, and 14.0 MeV through the 7Li (p, n)7Be, T (p, n)3He, D (d, n)3He, and T (d, n)4He reactions (Hu et al., 2017; Zhang et al., 2011). The targets were positioned near the center of the hall and 1.8 m above the ground. A cylindrical hole, 3 m in diameter and 2 m in depth, was dug in the ground just below the target to decrease the scattered neutrons near the target. The energies and angle distributions of mono-energetic neutrons emitted from the targets were calculated by the PTB Monte Carlo code TARGET (Schlegel, 2005).
2.2. Monte Carlo simulations Monte Carlo simulations were performed using the MCNP5 code (X5 Monte Carlo Team, 2003). Detailed geometric parameters, including the dimensions and density of the materials, were described based on the design values. The standard cross-section data for most elements were taken from the ENDF/B-VII.1 library (Chadwick et al., 2011). All calculations were performed in a vacuum environment. The response to neutron fluence was defined as the ratio of the
Fig. 1. (a) Schematic geometry and (b) image of the long counter. 17
Radiation Measurements 119 (2018) 16–21
Z.M. Hu et al.
The shadow cone method was employed to subtract from the counter response the contribution of the room- and air-scattered neutrons in the response and effective center determination experiments (Birattari et al., 2010; Klein and Thomas., 2003) following the ISO 8529-2 guideline (International Standard Organization, 2000). Three shadow cones were used, consisting of a 20-cm-long conical front end made of iron and a 30-cm-long conical rear end, covering an angle from 6.8° to 14.2°. Materials of the rear end are paraffin and borated polyethylene, respectively, for the largest shadow cone and for the two others. The response due to neutrons scattered by the target was also subtracted. The TARGET code developed by PTB was employed to calculate energy distributions of the un-scattered and scattered neutrons from target materials and mounting. The folding of neutron energy distribution with a calculated response function of the long counter allowed for evaluation of the contribution of target-scattering neutrons to the total response. The influence of the source anisotropy on the response was also considered when processing experimental data. The fluence value was normalized to the target-to-counter direction after considering the relative sensitivity of the long counter to neutrons reaching its front face within different solid angles. The air attenuation of neutrons from the target to the detector by air was evaluated with the formula provided by ISO 8529-2. In accelerator experiments, the two recoil proton proportional counters and the 238U fission chamber described in Section 1 were used for neutron fluence measurements. A second long counter, with a BF3 tube as a central counter, and a beam integrator, were employed as neutron yield monitors for all neutron sources. In addition, two liquid scintillators (Xie et al., 2014; Yuan et al., 2013) were applied to monitor the yield of neutron sources above 2 MeV. The electronic system consisted of a charge sensitive preamplifier (ORTEC 142PC), a main amplifier (ORTEC 572A), and a multichannel pulse analyzer (designed by Sichuan University).
Fig. 2. Effective center of an 8-inch Bonner sphere obtained from linear fitting of the square root of the inverse count rate M after air-attenuation correction against distances from a 241AmBe source to the sphere center. The calculated spectrum obtained by Monte Carlo method was used as an input spectrum (Li et al., 2013; Liu et al., 2007).
3. Results and discussion 3.1. Validation of reliability of effective center determination in experimental hall An accurate determination of the effective center of the long counter is required for neutron fluence measurements, although experimental identification of the effective center with reasonable accuracy remains challenging (Tagziria and Thomas, 2000). It is therefore essential to confirm the reliability of the experimental determination of the effective center of the long counter from the start. The Bonner sphere is a good candidate because its effective center is at or close to its geometric center (Hunt, 1984). Thus, determination of the effective center of an 8inch Bonner sphere to a 241AmBe neutron source was carried out experimentally at 14 distances between the source and the sphere center ranging from 125 to 500 cm. The shadow cone method was employed to subtract from the sphere response the contribution of the room- and airscattered neutrons; the air attenuation of neutrons was also considered. As shown in Fig. 2, the effective center was determined to be (−0.16 ± 2.34) cm, reasonably close to zero. The calculation of the MCNP5 code yielded an effective center of (0.23 ± 0.98) cm, which agreed well with the experimental value. The reliability of the determined effective center of a detector in the experimental hall was therefore validated.
Fig. 3. Comparison of three calculation methods of response functions. The first two sets (263 points) were obtained using the “parallel source model” with default pseudorandom number (solid square dots and solid line) and from the average of six calculations with different pseudorandom numbers (solid circle dots and dashed line), respectively. The third set (149 points) was derived from effective center determination using the “point source model” (upward triangle dots and dashed dotted line).
calculated results from the six runs was considered. The relative uncertainty of each calculated response was less than 1% (k = 1). The calculated response functions are depicted in Fig. 3. Notably, a difference within 3% was found between responses calculated using the default pseudorandom number and the averaged number at approximate energies of 2.5 MeV and 5 MeV as shown in Fig. 4. Calculation results reveal that the standard deviations between the six calculations were not always less than or equal to the single calculation uncertainty, indicating that the deviation was not always due to statistical fluctuations. Response values could also be derived from effective center determination using the ‘’point source model’’ as shown in Fig. 3; values were numerically obtained by 4π divided by the square of the slope. The number of source neutrons in MCNP5 at each distance was 106, and the relative uncertainty of the calculated response at each distance was less than 1% (k = 1) for neutrons below 10 MeV and less than 1.3% (k = 1) for those between 10 MeV and 20 MeV. The response values
3.2. Response function 3.2.1. Simulated response function The response function of the long counter between 1 keV and 20 MeV was calculated at hundreds of neutron energies in two ways, with a parallel neutron source as explained in Section 2.2: either with a single default number as starting point of the calculation or with six different pseudorandom numbers. For the latter case, the average of the 18
Radiation Measurements 119 (2018) 16–21
Z.M. Hu et al.
Fig. 5. Pulse height spectra measured for the bare 3He counter in a cosmicinduced background neutron field and for the counter covered by its shield in a 565 keV mono-energetic neutron field. The threshold was at channel 70 (vertical line).
Fig. 4. Calculated response values for mono-energetic neutrons up to 7 MeV and for a 241AmBe source compared to experimental data from seven neutron sources (six mono-energetic neutron sources and one 241AmBe source).
derived from the “point source model” concurred well with the other two calculation methods using a parallel neutron source. Good agreement among the three sets of response values confirmed the consistency of these models in the response calculation. More than 90% of responses calculated using the default pseudorandom number agreed well with those calculated using the other two models (i.e., within 2%), with the largest deviation being less than 3%. The latter two response functions exhibited deviations of less than 2% and showed better consistency. The “parallel source model”, using six pseudorandom numbers, was therefore chosen for response calculations, as it could improve calculation accuracy compared to the default pseudorandom number. Additionally, the calculation time was far lower (about 50 times less) than when using the “point source model”. The three calculated response functions were also compared based on linear coordinates in the energy region up to 7 MeV, as shown in Fig. 4. The calculation results revealed that the long counter had a relatively flat response to neutrons with an upper limit energy reaching approximately 5 MeV. Some local variations near 1.12 × 10−3 MeV in the response curve resulted from resonance of the cadmium absorber in the front face (Lacoste et al., 2004). Other prominent structures in the MeV energy region were caused by resonances in the neutron total cross-section of carbon in polyethylene, the major element by weight in the long counter (Harano et al., 2011; Roberts et al., 2004; Tagziria and Thomas, 2000).
approximately 2.8% (k = 1) at 0.144 MeV and 0.565 MeV and 3.6% (k = 1) at other energy points. For the 241AmBe source measurement, the uncertainty came from the total emission rate of the source (0.75%) and the linear fitting procedure of effective center determination (2.2%); the total uncertainty was thus 2.3% (k = 1). The calibration factor of the counter was determined to be 0.995 ± 0.023 (k = 1) using the experimental to calculated response ratio for the 241AmBe source. The neutron emission rate was traceable to international standards. The calculation was made by the parallel source model using six different pseudorandom numbers. The number of 3He (n,p)3H reactions in the long counter was taken from the summation of counts above a threshold in the pulse height spectra, as indicated in Fig. 5. We used the calibrated long counter to measure neutron fluence at seven mono-energetic neutron sources and compared them with the fluence values measured by proportional counters and a 238U fission chamber. The ratios of neutron fluence measured by the long counter and other devices are given in Table 1. The uncertainties were higher for mono-energetic neutron fields above 2.5 MeV, as the fission chamber was used as a fluence measurement device with an approximate distance of 4 cm between the 238U sample in the fission chamber and the target. The small detector-to-target distances were measured repeatedly before and after measurement. The neutron fluence measured by the long counter agreed well with the reference results at all energies. The deviations were within the uncertainties, even at 4.0 MeV, as uncertainties were set at a coverage factor of k = 1. The comparison results demonstrate that the calibration was valid for the entire studied energy range. The relative uncertainty in the variation of the response with energy was estimated to be 3.5%, using the relative standard deviation of experimental data compared to the simulated
3.2.2. Experimental determination of response function Experimental responses of the long counter at several neutron sources are illustrated in Fig. 4. The long counter and fluence measurement devices were positioned 250 cm from the target and at 0° (i.e., in the incident ion beam direction) except for 14 MeV energy, where the long counter was placed at 101°. The 14 MeV neutrons were produced via the reaction between the 1.23 MeV deuterium beam with a T/Ti target at 101° (17.1 MeV neutrons were produced at 0°). The neutron fluence at 101° was derived from that measured at 0° using the calculated angular distribution of neutrons emitted from the target. Several sources of uncertainty were considered: (1) contributions of neutron fluence determined by the fluence measurement devices around approximately 2.0% and 3.0% for neutron sources below and above 2 MeV, respectively; (2) counts of the long counter at 250 cm distance were used, and the distance uncertainty was estimated to be 0.7 cm with an approximate contribution of 0.6%; and (3) the count statistics of the counter and monitor were less than 1% and contributed about 1.2% of the uncertainty. The total resulting uncertainties were therefore
Table 1 Ratios of neutron fluence measured by the long counter ΦLC and other devices Φ at seven mono-energetic neutron sources (uncertainties set at k = 1).
19
Neutron energy/MeV
ΦLC/Φ
0.144 0.565 2.5 4.0 5.0 6.0 14.0
0.994 0.992 1.003 1.053 1.041 0.983 1.036
Reference fluence measurement device ± ± ± ± ± ± ±
0.037 0.039 0.044 0.046 0.044 0.043 0.045
Recoil proton counter Recoil proton counter U fission chamber 238 U fission chamber 238 U fission chamber 238 U fission chamber 238 U fission chamber 238
Radiation Measurements 119 (2018) 16–21
Z.M. Hu et al.
Fig. 6. Variations of calculated effective centers as a function of neutron energy up to 20 MeV. For each of the six calculations at an energy point, the uncertainty came from the linear fitting procedure. The standard uncertainty at each energy point was derived from the standard deviation of the corresponding six calculated results.
Fig. 8. Comparison between calculated and experimentally determined effective centers at four mono-energetic neutron sources and one 241AmBe source. Standard uncertainties of experimental responses were obtained from the linear fitting procedure.
natural carbon σC,tot and the elastic scattering cross-section on 1H atoms σ 1H . These cross-sections were taken from the ENDF/B-VII.1 library. The higher the neutron energy, the larger the mean free path of the neutron; it could penetrate deeper into the polyethylene before being detected by the 3He counter, leading to a large effective center value. The mean free path is the reciprocal of the macroscopic cross-section, ρN represented by the formula Σ = MA (2σC , tot + 4σ 1H ) . Here, ρ is the mass density of polyethylene, NA is the Avogadro constant, and M = 28.0536 g/mol is the molar mass of the polyethylene molecule C2H4. Fig. 7 illustrates that the structural variation of the effective center resulted from the energy dependence of the total neutron crosssection of carbon in the polyethylene. The effective centers of the long counter were experimentally determined at four mono-energetic neutron fields and one 241AmBe source as illustrated in Fig. 8. With 0.565 MeV, 1.2 MeV, and 2.5 MeV neutron sources, measurements were performed at either 9 or 11 distances, between 225 and 500 cm, from the front face of the long counter to the sources. At 5 MeV, measurements were performed at 14 distances from 165 cm to 500 cm. For the 241AmBe source, measurements were taken at 17 distances between 135 cm and 500 cm. The effective center was determined by linearly fitting the inverse square root of the long counter count rate per monitor count against the distance (Tagziria and Thomas, 2000). The calculated values agreed well with the experimental results within the measured uncertainties with systematic underestimation. The uncertainty of variation in the effective center with energy was estimated to be 0.5 cm using the standard deviation of experimental values compared to the calculations. Fig. 7 shows the variation in energy of the effective center up to 7 MeV.
results (2.6%), to be quadratically added to the uncertainty of the calibration factor (2.3%).
3.3. Effective center Calculations of the effective center position were performed for 149 neutron energies from 1 keV to 20 MeV at 51 distances, between the front face of the long counter and the neutron sources, varying from 150 cm to 600 cm. As illustrated in Fig. 6, the variation of the effective center with neutron energy exhibited some sharp structures consistent with the total neutron cross-section of carbon, concurring with structures observed in the response function. Similar results have been reported for the De Pangher and McTaggart long counters (Roberts et al., 2004) and for a homemade long counter (Harano et al., 2011). On average, the effective center position from the long counter front face increased from 6 cm to 14 cm with the neutron energy up to approximately 10 MeV and then continued increasing slowly. For the long counter in our work, the moderators were made of polyethylene. The relationship of the mean free path against the neutron energy in the moderator is shown in Fig. 7, along with the total cross-section on
4. Conclusions A De Pangher-type long counter has been constructed and was calibrated in a 241AmBe neutron field at Peking University. The calibration factor was determined to be 0.995 ± 0.023 (k = 1) by combining measurements and Monte Carlo simulations. The response function of the long counter between 1 keV and 20 MeV was calculated at hundreds of neutron energies, with a “parallel source model’’ using six pseudorandom numbers. Neutron fluence measured by the calibrated long counter was successfully compared to values determined by reference instruments (two recoil proton counters and a 238U fission chamber) with relative deviations lower than 6% in the energy range between a few hundred of keVs and 14 MeV; hence, the long counter was well
Fig. 7. Mean free path of neutrons in polyethylene of 0.93 g/cm3 with total cross-sections on natural carbon and elastic scattering cross-sections on 1H atoms. 20
Radiation Measurements 119 (2018) 16–21
Z.M. Hu et al.
characterized in the whole energy range of the calculations. The relative uncertainty on the variation of the response with energy was estimated to be 3.5%. The calculation method and measurement protocol for the effective center determination were validated via their application to a Bonner sphere. The random method was employed to predict effective center values. The sharp structures in the variation of the effective center with energy matched the total neutron cross-section of carbon, present in the polyethylene shield of the long counter. The calculated effective centers agreed reasonably well with the experimental results within uncertainties even though systematic underestimation could not be excluded. Uncertainty on the variation of the effective center with energy was estimated to be 0.5 cm. Compared to other De Pangher long counters, such as the NPL version, the long counter exhibited higher response values due to greater sensitivity of the 3He tube to thermal neutrons; however, the long counters had comparable effective center positions because of their similar geometry. The long counter compensates for the lack of devices for neutron fluence measurement from 1 MeV to 2 MeV at the State Key Laboratory of Nuclear Physics and Technology. Consequently, the counter has been successfully applied to experimental characterization of a Bonner sphere spectrometer, which has been presented elsewhere (Hu et al., 2017).
Hu, Q.Y., Zhang, J.H., Zhang, D., Guo, H.S., Yang, G.Z., Li, B.J., Ye, F., Si, F.N., Liu, J., Fu, Y.C., Ning, J.M., Yang, J., Yang, H.H., Wang, W.C., 2014a. An improved long counter for neutron fluence measurement with a flat response over a wide energy range from 1 keV to 15 MeV. Nucl. Instrum. Methods Phys. Res., Sect. A 768, 43–45. Hu, Z.M., Xie, X.F., Chen, Z.J., Peng, X.Y., Du, T.F., Cui, Z.Q., Ge, L.J., Li, T., Yuan, X., Zhang, X., Hu, L.Q., Zhong, G.Q., Lin, S.Y., Wan, B.N., Gorini, G., Li, X.Q., Zhang, G.H., Chen, J.X., Fan, T.S., 2014b. Monte Carlo simulation of a Bonner sphere spectrometer for application to the determination of neutron field in the Experimental Advanced Superconducting Tokamak experimental hall. Rev. Sci. Instrum. 85 11E417. Hu, Z.M., Chen, Z.J., Peng, X.Y., Du, T.F., Cui, Z.Q., Ge, L.J., Zhu, W.J., Wang, Z.M., Zhu, X., Chen, J.X., Zhang, G.H., Li, X.Q., Chen, J., Zhang, H., Zhong, G.H., Hu, L.Q., Wan, B.N., Gorinie, G., Fan, T.S., 2017. Experimental determination of the response functions of a Bonner sphere spectrometer to monoenergetic neutrons. J. Instrum. 12, T06002. Hunt, J.B., 1984. The calibration of neutron sensitive spherical devices. Radiat. Protect. Dosim. 8, 239–251. International Standard Organization (ISO), 2000. Reference Neutron Radiations — Part 2: Calibration Fundamentals of Radiation protection Devices Related to the Basic Quantities Characterizing the Radiation Field, ISO 8529-2. Kim, G.D., Woo, H.J., Park, J.H., 2012. Characteristics of the KIGAM long-counter for the neutron energy range below 2.5 MeV. J. Kor. Phys. Soc. 61, 347. Klein, H., Thomas, D.J., 2003. 10 Quality assurance. Radiat. Protect. Dosim. 107, 37. Lacoste, V., Gressier, V., Pochat, J.L., Fernández, F., Bakali, M., Bouassoule, T., 2004. Characterization of Bonner sphere systems at monoenergetic and thermal neutron fields. Radiat. Protect. Dosim. 110, 529. Lacoste, V., 2010. Design of a new long counter for the determination of the neutron fluence reference values at the IRSN AMANDE facility. Radiat. Meas. 45, 1250. Lacoste, V., Gressier, V., 2010. Experimental characterization of the IRSN long counter for the determination of the neutron fluence reference values at the AMANDE facility. Radiat. Meas. 45, 1254. Li, Y.M., Chen, J.X., Zhang, G.H., Fan, T.S., 2013. Study of physical characteristics for compressed-mixture type 241AmO2-Be neutron source. Atomic Energy Sci. Technol. 47, 1–6 (in Chinese). Liu, Z.Z., Chen, J.X., Zhu, P., Li, Y.M., Zhang, G.H., 2007. The 4.438 MeV gamma to neutron ratio for the Am–Be neutron source. Appl. Radiat. Isot. 65, 1318–1321. Mazunga, M., Li, T.S., Li, Y.N., Hong, B., Wang, Y.F., Ji, X., 2017. Design of an extended range long counter using super Monte Carlo simulation. Radiat. Protect. Dosim. 175, 413. Nolte, R., Thomas, D.J., 2011. 8,Monoenergetic fast neutron reference fields: II. Field characterization. Metrologia 48, S274. Roberts, N.J., Jones, L.N., Wang, Z., Liu, Y., Wang, Q., Chen, X., Luo, H., Rong, C., Králik, M., Park, H., Choi, K.O., Pereira, W.W., da Fonseca, E.S., Cassette, P., Dewey, M.S., Moiseev, N.N., Kharitonov, I.A., 2011. International key comparison of measurements of neutron source emission rate (1999-2005) – CCRI(III)-K9.AmBe. Metrologia 48, 06018. Roberts, N.J., Tagziria, H., Thomas, D.J., 2004. Determination of the Effective Centres of the NPL Long Counters, NPL Report DQL RN004. Roberts, N.J., Thomas, D.J., Lacoste, V., Böttger, R., Loeb, S., 2010. Comparison of long counter measurements of monoenergetic and radionuclide source-based neutron fluence. Radiat. Meas. 45, 1151. Ryves, T.B., 1987. International fluence-rate intercomparison for 144 and 565 keV neutrons. Metrologia 24, 27. Schlegel, D., 2005. Target User's Manual, PTB Laboratory Report PTB-6.42-05-2, Braunschweig Germany. Tagziria, H., Thomas, D.J., 2000. Calibration and Monte Carlo modelling of neutron long counters. Nucl. Instrum. Methods Phys. Res., Sect. A 452, 470. Tanimura, Y., Tsutsumi, M., Yoshizawa, M., 2014. Development of portable long counter with two different moderator materials. Radiat. Protect. Dosim. 161, 144–148. X-5 Monte Carlo Team, 2003. MCNP — a General Monte Carlo Code N-particle Transport Code, Version 5, Manual LA-UR-03-1987. Los Alamos National Laboratory. Xie, X.F., Chen, Z.J., Peng, X.Y., Yuan, X., Zhang, X., Gorini, G., Cui, Z.Q., Du, T.F., Hu, Z.M., Li, T., Fan, T.S., Chen, J.X., Li, X.Q., Zhang, G.H., Yuan, G.L., Yang, J.W., Yang, Q.W., 2014. Neutron emission measurement at the HL-2A tokamak device with a liquid scintillation detector. Rev. Sci. Instrum. 85 103506. Yuan, X., Zhang, X., Xie, X.F., Gorini, G., Chen, Z., Peng, X., Chen, J., Zhang, G., Fan, T., Zhong, G., Hu, L., Wan, B., 2013. Neutron energy spectrum measurements with a compact liquid scintillation detector on EAST. J. Instrum. 8, P07016. Zhang, G.H., Gledenov, YuM., Khuukhenkhuu, G., Sedysheva, M.V., Szalanski, P.J., Koehler, P.E., Voronov, YuN., Liu, J.M., Liu, X., Han, J.H., Chen, J.X., 2011. 149Sm (n,α)146Nd cross sections in the MeV region. Phys. Rev. Lett. 107, 252502.
Acknowledgements This work was supported by the National Key Research and Development Program of China (No. SQ2016YFZG020067), the National Magnetic Confinement Fusion Science Program of China (Nos. 2013GB106004 and 2012GB101003), and the National Natural Science Foundation of China (No. 91226102). References Alevra, A.V., Thomas, D.J., 2003. Neutron spectrometry in mixed fields: multisphere spectrometers. Radiat. Protect. Dosim. 107, 37. Axton, E.J., 1987. Intercomparison of neutron-source emission rates (1979-1984). Metrologia 23, 129. Birattari, C., Dimovasili, E., Mitaroff, A., Silari, M., 2010. A Bonner Sphere Spectrometer with extended response matrix. Nucl. Instrum. Meth. Phys. Res. 620, 260. Chadwick, M.B., et al., 2011. ENDFB-VII.1 nuclear data for science and Technology cross sections, covariances, fission product yields and decay data. Nucl. Data Sheets 112, 2887–2996. Chen, J., Wang, Z., Rong, C., Lövestam, G., Plompen, A., Puglisi, N., Gilliam, D.M., Eisenhauer, C.M., Nico, J.S., Dewey, M.S., Kudo, K., Uritani, A., Harano, H., Takeda, N., Thomas, D.J., Roberts, N.J., Bennett, A., Kolkowski, P., Moisseev, N.N., Kharitonov, I.A., Guldbakke, S., Klein, H., Nolte, R., Schlegel, D., 2007. International key comparison of neutron fluence measurements in mono-energetic neutron fields CCRI(III)-K10. Metrologia 44, 06005. De Pangher, J., Nichols, L.L., 1966. A Precision Long Counter for Measuring Fast Neutron Flux Density, BNWL-260. Pacific Northwest Laboratory, Washington USA. Gressier, V., Bonaldi, A.C., Dewey, M.S., Gilliam, D.M., Harano, H., Masuda, A., Matsumoto, T., Moiseev, N., Nico, J.S., Nolte, R., Oberstedt, S., Roberts, N.J., Röttger, S., Thomas, D.J., 2014. International key comparison of neutron fluence measurements in monoenergetic neutron fields - CCRI(III) - K11. Metrologia 51, 06009. Gressier, V., Lacoste, V., 2014. A Martin and M Pepino, Characterization of a measurement reference standard and neutron fluence determination method in IRSN monoenergetic neutron fields. Metrologia 51, 431–440. Hanson, A.O., McKibben, J.L., 1947. A neutron detector having uniform sensitivity from 10 keV to 3 MeV. Phys. Rev. 72, 673. Harano, H., Matsumoto, T., Nishiyama, J., Masuda, A., Uritani, A., Kudo, K., 2011. Development of a compact flat response neutron detector. IEEE Trans. Nucl. Sci. 58, 2421.
21
Contents lists available at ScienceDirect
Radiation Measurements journal homepage: www.elsevier.com/locate/radmeas
Experimental characterization of a long counter for neutron fluence measurement
T
Z.M. Hua, X.Y. Penga, Z.J. Chena, T.F. Dua, L.J. Gea, X. Yuana, Z.Q. Cuia, W.J. Zhua, Z.M. Wanga, X. Zhua, J.X. Chena, X.Q. Lia, G.H. Zhanga, J. Chenb, H. Zhangc, G. Gorinid, T.S. Fana,∗ a
School of Physics and State Key Laboratory of Nuclear Physics and Technology, Peking University, Beijing, 100871, China China Institute of Atomic Energy, Beijing, 102413, China c National Institute of Metrology, Beijing, 100029, China d Dipartimento di Fisica ‘G. Occhialini’, Università degli Studi di Milano-Bicocca, Milano, 20126, Italy b
A R T I C LE I N FO
A B S T R A C T
Keywords: Long counter Calibration Response function Effective center
A De Pangher-type long counter was constructed at Peking University for neutron fluence measurement. The responses and effective centers of the long counter were calculated using the Monte Carlo method. The variations with neutron energy of the position of the effective center, calculated by the Monte Carlo code, was experimentally validated with a 241AmBe neutron source and several mono-energetic neutron sources ranging from 100 keV to 6 MeV. Long counter calibration was performed using a radionuclide source from Peking University, and neutron fluence measured by the calibrated long counter was successfully compared to values determined with reference instruments (two recoil proton counters and a 238U fission chamber). The relative deviations were lower than 6% along the whole energy range. The calibrated long counter was thus successfully applied in the calibration of a Bonner sphere spectrometer.
1. Introduction Experimental validation of the response function of a Bonner sphere spectrometer was performed in the experimental hall of Peking University at six mono-energetic neutron fields ranging from 100 keV to 20 MeV provided by a 4.5-MV Van de Graaff accelerator (Hu et al., 2014b, 2017) at the State Key Laboratory of Nuclear Physics and Technology. Neutron fluence measurement devices should be employed to determine the neutron fluence value at the measurement position of the Bonner sphere spectrometer (Alevra and Thomas, 2003; Birattari et al., 2010; Klein and Thomas, 2003; Lacoste et al., 2004). Two types of reference devices, namely two recoil proton proportional counters and one 238U fission chamber (Hu et al., 2017; Zhang et al., 2011), were used to measure neutron fluence within energy ranges from 100 keV to 1 MeV and from 2 MeV to 20 MeV, respectively. They were designed by China Institute of Nuclear Energy and the two counters were traceable to international standards (Ryves, 1987). The uncertainty on the neutron fluence is approximately 2% with the counters (Chen et al., 2007) and 3% with the fission chamber. This chamber is made of a 2-cm diameter 238U sample attached to the inner face of a copper cylinder box. The 238U sample has a mass of (547.2 ± 1.3%) μg with enrichment greater than 99.997% (Zhang et al., 2011). To provide reference ∗
fluence values between 1 MeV and 2 MeV, a calibrated long counter was selected, and the range of use varied from a few eVs/keVs to 20 MeV. The long counter has been widely applied in neutron fluence measurement due to characteristics such as a relatively flat response in a wide neutron energy range between a few eVs or keVs and a few MeVs; insensitivity to gamma rays with an appropriate threshold; and good stability (Tagziria and Thomas, 2000). The Hanson and McKibben (1947) and De Pangher (De Pangher and Nichols, 1966) types of long counters have been employed as standard instruments in many laboratories (Gressier et al., 2014; Nolte and Thomas, 2011; Roberts et al., 2010; Tagziria and Thomas, 2000). They are based on a cylindrical moderator, usually made of paraffin or polythene materials with a thermal neutron tube in the center. The moderator is embedded in an additional shield to reduce sensitivity to room-scattered neutrons and endow the counter with high efficiency to neutron incidents from the front face along the axis. The response of long counters can be roughly independent of energy, up to 3 MeV–5 MeV depending on type. Recently, some newly developed long counters have also been reported in the literature (Gressier and Lacoste, 2014; Harano et al., 2011; Hu et al., 2014a; Kim et al., 2012; Lacoste, 2010; Lacoste and Gressier, 2010; Mazunga et al., 2017; Tanimura et al., 2014), using Monte Carlo simulations to either optimize the geometry or employ new materials.
Corresponding author. E-mail address: [email protected] (T.S. Fan).
https://doi.org/10.1016/j.radmeas.2018.08.017 Received 4 November 2017; Received in revised form 24 August 2018; Accepted 28 August 2018 Available online 29 August 2018 1350-4487/ © 2018 Published by Elsevier Ltd.
Radiation Measurements 119 (2018) 16–21
Z.M. Hu et al.
detector reading (number of 3He (n, p)3H reactions) to the incident neutron fluence. In response function calculations, a parallel neutron beam starting from a disk with a 40-cm diameter irradiated the front face of the long counter along its axis, and the F4 and FM4 tally cards of MCNP5 were combined to derive the response values (Hu et al., 2014b, 2017). The effective center, r0, is defined as the position of the calibration point. The long counter is regarded as a point detector in the fluence measurements from a point neutron source, such that the count rate varies as (r + r0)−2, where r is typically the distance from the point source to the front face of the long counter (Lacoste and Gressier, 2010; Roberts et al., 2004; Tagziria and Thomas, 2000). In calculating effective centers, the inverse of the square root of the count rate against distance r was linearly fitted to yield the effective centers. The counts of the long counter at 51 distances were simulated. The pseudorandom number for each distance was different and randomly generated as in reference (Roberts et al., 2004). This method can overcome correlations between calculations at each distance by using the same pseudorandom number. At each distance, the calculation was separated into six runs with different initial random numbers.
A long counter based on the De Pangher design was constructed at Peking University to calibrate a Bonner sphere spectrometer and for other applications. This long counter, which generally does not have a perfectly flat response over a wide energy range, needed to be experimentally calibrated because its efficiency cannot be derived from first principles like recoil proton telescopes (Roberts et al., 2010). For any feasible fluence measurement, the effective center of the long counter, depending on the counter design and neutron energy, should also be accurately determined by combining measurements and detailed Monte Carlo simulations (Roberts et al., 2010; Tagziria and Thomas, 2000). In this paper, Monte Carlo simulations were performed to determine the fluence responses and effective centers as a function of neutron energy. Simulations were then validated by measurements at eight mono-energetic neutron fields between 100 keV and 14.0 MeV as well as in front of a 241AmBe radioactive source. Good agreement was obtained between the experimental and simulated values. 2. Materials and methods 2.1. Construction of long counter The schematic geometry and image of the De Pangher-type long counter are presented in Fig. 1. The cylindrical long counter measures 43 cm in length and 39 cm in diameter and consists mainly of a thermal neutron counter, two annular moderators, an outer annular shield, and three thermal neutron absorbers. The thermal neutron counter, manufactured by CENTRONIC Ltd., is a proportional tube with a diameter of 3.8 cm and a length of 45 cm, filled with 3He gas at a pressure of 6 atm. The moderators and shield are made of polyethylene with a mass density of 0.93 g/cm3. Boron loaded plastic, measuring 1 cm thick, is installed between the moderators and outer shield. A 0.6-mm thick rear cadmium sheet is inserted between the front and rear polyethylene moderators, and another cadmium sheet is mounted on the front face of the long counter. An annular trough is dredged in the front moderator. All parts of the long counter are supported and fixed with aluminum materials.
2.3. Experimental setup The experimental determination of the response and effective center values, which aimed to verify the calculated values, was carried out using a 241AmBe source and eight mono-energetic neutron sources ranging from 100 keV to 14.0 MeV in the experimental hall (20 m long, 12 m wide, and 8 m high) at Peking University (Hu et al., 2017). The neutron emission rate of the 241AmBe source was determined to be (6.94 ± 0.75%) × 106 s−1 on July 21, 2006 based on manganese bath equipment from China Institute of Atomic Energy (Roberts et al., 2011). The manganese bath equipment was applied in neutron emission rate comparisons in 1984 (Axton, 1987) and 2005. The neutron spectrum and anisotropy factor were calculated using the Monte Carlo method (Li et al., 2013; Liu et al., 2007). The charged particle beams accelerated by a 4.5-MV Van de Graaff accelerator were sent to various solid targets to produce mono-energetic neutron fields of 0.144 MeV, 0.565 MeV, 1.2 MeV, 2.5 MeV, 4.0 MeV, 5.0 MeV, 6.0 MeV, and 14.0 MeV through the 7Li (p, n)7Be, T (p, n)3He, D (d, n)3He, and T (d, n)4He reactions (Hu et al., 2017; Zhang et al., 2011). The targets were positioned near the center of the hall and 1.8 m above the ground. A cylindrical hole, 3 m in diameter and 2 m in depth, was dug in the ground just below the target to decrease the scattered neutrons near the target. The energies and angle distributions of mono-energetic neutrons emitted from the targets were calculated by the PTB Monte Carlo code TARGET (Schlegel, 2005).
2.2. Monte Carlo simulations Monte Carlo simulations were performed using the MCNP5 code (X5 Monte Carlo Team, 2003). Detailed geometric parameters, including the dimensions and density of the materials, were described based on the design values. The standard cross-section data for most elements were taken from the ENDF/B-VII.1 library (Chadwick et al., 2011). All calculations were performed in a vacuum environment. The response to neutron fluence was defined as the ratio of the
Fig. 1. (a) Schematic geometry and (b) image of the long counter. 17
Radiation Measurements 119 (2018) 16–21
Z.M. Hu et al.
The shadow cone method was employed to subtract from the counter response the contribution of the room- and air-scattered neutrons in the response and effective center determination experiments (Birattari et al., 2010; Klein and Thomas., 2003) following the ISO 8529-2 guideline (International Standard Organization, 2000). Three shadow cones were used, consisting of a 20-cm-long conical front end made of iron and a 30-cm-long conical rear end, covering an angle from 6.8° to 14.2°. Materials of the rear end are paraffin and borated polyethylene, respectively, for the largest shadow cone and for the two others. The response due to neutrons scattered by the target was also subtracted. The TARGET code developed by PTB was employed to calculate energy distributions of the un-scattered and scattered neutrons from target materials and mounting. The folding of neutron energy distribution with a calculated response function of the long counter allowed for evaluation of the contribution of target-scattering neutrons to the total response. The influence of the source anisotropy on the response was also considered when processing experimental data. The fluence value was normalized to the target-to-counter direction after considering the relative sensitivity of the long counter to neutrons reaching its front face within different solid angles. The air attenuation of neutrons from the target to the detector by air was evaluated with the formula provided by ISO 8529-2. In accelerator experiments, the two recoil proton proportional counters and the 238U fission chamber described in Section 1 were used for neutron fluence measurements. A second long counter, with a BF3 tube as a central counter, and a beam integrator, were employed as neutron yield monitors for all neutron sources. In addition, two liquid scintillators (Xie et al., 2014; Yuan et al., 2013) were applied to monitor the yield of neutron sources above 2 MeV. The electronic system consisted of a charge sensitive preamplifier (ORTEC 142PC), a main amplifier (ORTEC 572A), and a multichannel pulse analyzer (designed by Sichuan University).
Fig. 2. Effective center of an 8-inch Bonner sphere obtained from linear fitting of the square root of the inverse count rate M after air-attenuation correction against distances from a 241AmBe source to the sphere center. The calculated spectrum obtained by Monte Carlo method was used as an input spectrum (Li et al., 2013; Liu et al., 2007).
3. Results and discussion 3.1. Validation of reliability of effective center determination in experimental hall An accurate determination of the effective center of the long counter is required for neutron fluence measurements, although experimental identification of the effective center with reasonable accuracy remains challenging (Tagziria and Thomas, 2000). It is therefore essential to confirm the reliability of the experimental determination of the effective center of the long counter from the start. The Bonner sphere is a good candidate because its effective center is at or close to its geometric center (Hunt, 1984). Thus, determination of the effective center of an 8inch Bonner sphere to a 241AmBe neutron source was carried out experimentally at 14 distances between the source and the sphere center ranging from 125 to 500 cm. The shadow cone method was employed to subtract from the sphere response the contribution of the room- and airscattered neutrons; the air attenuation of neutrons was also considered. As shown in Fig. 2, the effective center was determined to be (−0.16 ± 2.34) cm, reasonably close to zero. The calculation of the MCNP5 code yielded an effective center of (0.23 ± 0.98) cm, which agreed well with the experimental value. The reliability of the determined effective center of a detector in the experimental hall was therefore validated.
Fig. 3. Comparison of three calculation methods of response functions. The first two sets (263 points) were obtained using the “parallel source model” with default pseudorandom number (solid square dots and solid line) and from the average of six calculations with different pseudorandom numbers (solid circle dots and dashed line), respectively. The third set (149 points) was derived from effective center determination using the “point source model” (upward triangle dots and dashed dotted line).
calculated results from the six runs was considered. The relative uncertainty of each calculated response was less than 1% (k = 1). The calculated response functions are depicted in Fig. 3. Notably, a difference within 3% was found between responses calculated using the default pseudorandom number and the averaged number at approximate energies of 2.5 MeV and 5 MeV as shown in Fig. 4. Calculation results reveal that the standard deviations between the six calculations were not always less than or equal to the single calculation uncertainty, indicating that the deviation was not always due to statistical fluctuations. Response values could also be derived from effective center determination using the ‘’point source model’’ as shown in Fig. 3; values were numerically obtained by 4π divided by the square of the slope. The number of source neutrons in MCNP5 at each distance was 106, and the relative uncertainty of the calculated response at each distance was less than 1% (k = 1) for neutrons below 10 MeV and less than 1.3% (k = 1) for those between 10 MeV and 20 MeV. The response values
3.2. Response function 3.2.1. Simulated response function The response function of the long counter between 1 keV and 20 MeV was calculated at hundreds of neutron energies in two ways, with a parallel neutron source as explained in Section 2.2: either with a single default number as starting point of the calculation or with six different pseudorandom numbers. For the latter case, the average of the 18
Radiation Measurements 119 (2018) 16–21
Z.M. Hu et al.
Fig. 5. Pulse height spectra measured for the bare 3He counter in a cosmicinduced background neutron field and for the counter covered by its shield in a 565 keV mono-energetic neutron field. The threshold was at channel 70 (vertical line).
Fig. 4. Calculated response values for mono-energetic neutrons up to 7 MeV and for a 241AmBe source compared to experimental data from seven neutron sources (six mono-energetic neutron sources and one 241AmBe source).
derived from the “point source model” concurred well with the other two calculation methods using a parallel neutron source. Good agreement among the three sets of response values confirmed the consistency of these models in the response calculation. More than 90% of responses calculated using the default pseudorandom number agreed well with those calculated using the other two models (i.e., within 2%), with the largest deviation being less than 3%. The latter two response functions exhibited deviations of less than 2% and showed better consistency. The “parallel source model”, using six pseudorandom numbers, was therefore chosen for response calculations, as it could improve calculation accuracy compared to the default pseudorandom number. Additionally, the calculation time was far lower (about 50 times less) than when using the “point source model”. The three calculated response functions were also compared based on linear coordinates in the energy region up to 7 MeV, as shown in Fig. 4. The calculation results revealed that the long counter had a relatively flat response to neutrons with an upper limit energy reaching approximately 5 MeV. Some local variations near 1.12 × 10−3 MeV in the response curve resulted from resonance of the cadmium absorber in the front face (Lacoste et al., 2004). Other prominent structures in the MeV energy region were caused by resonances in the neutron total cross-section of carbon in polyethylene, the major element by weight in the long counter (Harano et al., 2011; Roberts et al., 2004; Tagziria and Thomas, 2000).
approximately 2.8% (k = 1) at 0.144 MeV and 0.565 MeV and 3.6% (k = 1) at other energy points. For the 241AmBe source measurement, the uncertainty came from the total emission rate of the source (0.75%) and the linear fitting procedure of effective center determination (2.2%); the total uncertainty was thus 2.3% (k = 1). The calibration factor of the counter was determined to be 0.995 ± 0.023 (k = 1) using the experimental to calculated response ratio for the 241AmBe source. The neutron emission rate was traceable to international standards. The calculation was made by the parallel source model using six different pseudorandom numbers. The number of 3He (n,p)3H reactions in the long counter was taken from the summation of counts above a threshold in the pulse height spectra, as indicated in Fig. 5. We used the calibrated long counter to measure neutron fluence at seven mono-energetic neutron sources and compared them with the fluence values measured by proportional counters and a 238U fission chamber. The ratios of neutron fluence measured by the long counter and other devices are given in Table 1. The uncertainties were higher for mono-energetic neutron fields above 2.5 MeV, as the fission chamber was used as a fluence measurement device with an approximate distance of 4 cm between the 238U sample in the fission chamber and the target. The small detector-to-target distances were measured repeatedly before and after measurement. The neutron fluence measured by the long counter agreed well with the reference results at all energies. The deviations were within the uncertainties, even at 4.0 MeV, as uncertainties were set at a coverage factor of k = 1. The comparison results demonstrate that the calibration was valid for the entire studied energy range. The relative uncertainty in the variation of the response with energy was estimated to be 3.5%, using the relative standard deviation of experimental data compared to the simulated
3.2.2. Experimental determination of response function Experimental responses of the long counter at several neutron sources are illustrated in Fig. 4. The long counter and fluence measurement devices were positioned 250 cm from the target and at 0° (i.e., in the incident ion beam direction) except for 14 MeV energy, where the long counter was placed at 101°. The 14 MeV neutrons were produced via the reaction between the 1.23 MeV deuterium beam with a T/Ti target at 101° (17.1 MeV neutrons were produced at 0°). The neutron fluence at 101° was derived from that measured at 0° using the calculated angular distribution of neutrons emitted from the target. Several sources of uncertainty were considered: (1) contributions of neutron fluence determined by the fluence measurement devices around approximately 2.0% and 3.0% for neutron sources below and above 2 MeV, respectively; (2) counts of the long counter at 250 cm distance were used, and the distance uncertainty was estimated to be 0.7 cm with an approximate contribution of 0.6%; and (3) the count statistics of the counter and monitor were less than 1% and contributed about 1.2% of the uncertainty. The total resulting uncertainties were therefore
Table 1 Ratios of neutron fluence measured by the long counter ΦLC and other devices Φ at seven mono-energetic neutron sources (uncertainties set at k = 1).
19
Neutron energy/MeV
ΦLC/Φ
0.144 0.565 2.5 4.0 5.0 6.0 14.0
0.994 0.992 1.003 1.053 1.041 0.983 1.036
Reference fluence measurement device ± ± ± ± ± ± ±
0.037 0.039 0.044 0.046 0.044 0.043 0.045
Recoil proton counter Recoil proton counter U fission chamber 238 U fission chamber 238 U fission chamber 238 U fission chamber 238 U fission chamber 238
Radiation Measurements 119 (2018) 16–21
Z.M. Hu et al.
Fig. 6. Variations of calculated effective centers as a function of neutron energy up to 20 MeV. For each of the six calculations at an energy point, the uncertainty came from the linear fitting procedure. The standard uncertainty at each energy point was derived from the standard deviation of the corresponding six calculated results.
Fig. 8. Comparison between calculated and experimentally determined effective centers at four mono-energetic neutron sources and one 241AmBe source. Standard uncertainties of experimental responses were obtained from the linear fitting procedure.
natural carbon σC,tot and the elastic scattering cross-section on 1H atoms σ 1H . These cross-sections were taken from the ENDF/B-VII.1 library. The higher the neutron energy, the larger the mean free path of the neutron; it could penetrate deeper into the polyethylene before being detected by the 3He counter, leading to a large effective center value. The mean free path is the reciprocal of the macroscopic cross-section, ρN represented by the formula Σ = MA (2σC , tot + 4σ 1H ) . Here, ρ is the mass density of polyethylene, NA is the Avogadro constant, and M = 28.0536 g/mol is the molar mass of the polyethylene molecule C2H4. Fig. 7 illustrates that the structural variation of the effective center resulted from the energy dependence of the total neutron crosssection of carbon in the polyethylene. The effective centers of the long counter were experimentally determined at four mono-energetic neutron fields and one 241AmBe source as illustrated in Fig. 8. With 0.565 MeV, 1.2 MeV, and 2.5 MeV neutron sources, measurements were performed at either 9 or 11 distances, between 225 and 500 cm, from the front face of the long counter to the sources. At 5 MeV, measurements were performed at 14 distances from 165 cm to 500 cm. For the 241AmBe source, measurements were taken at 17 distances between 135 cm and 500 cm. The effective center was determined by linearly fitting the inverse square root of the long counter count rate per monitor count against the distance (Tagziria and Thomas, 2000). The calculated values agreed well with the experimental results within the measured uncertainties with systematic underestimation. The uncertainty of variation in the effective center with energy was estimated to be 0.5 cm using the standard deviation of experimental values compared to the calculations. Fig. 7 shows the variation in energy of the effective center up to 7 MeV.
results (2.6%), to be quadratically added to the uncertainty of the calibration factor (2.3%).
3.3. Effective center Calculations of the effective center position were performed for 149 neutron energies from 1 keV to 20 MeV at 51 distances, between the front face of the long counter and the neutron sources, varying from 150 cm to 600 cm. As illustrated in Fig. 6, the variation of the effective center with neutron energy exhibited some sharp structures consistent with the total neutron cross-section of carbon, concurring with structures observed in the response function. Similar results have been reported for the De Pangher and McTaggart long counters (Roberts et al., 2004) and for a homemade long counter (Harano et al., 2011). On average, the effective center position from the long counter front face increased from 6 cm to 14 cm with the neutron energy up to approximately 10 MeV and then continued increasing slowly. For the long counter in our work, the moderators were made of polyethylene. The relationship of the mean free path against the neutron energy in the moderator is shown in Fig. 7, along with the total cross-section on
4. Conclusions A De Pangher-type long counter has been constructed and was calibrated in a 241AmBe neutron field at Peking University. The calibration factor was determined to be 0.995 ± 0.023 (k = 1) by combining measurements and Monte Carlo simulations. The response function of the long counter between 1 keV and 20 MeV was calculated at hundreds of neutron energies, with a “parallel source model’’ using six pseudorandom numbers. Neutron fluence measured by the calibrated long counter was successfully compared to values determined by reference instruments (two recoil proton counters and a 238U fission chamber) with relative deviations lower than 6% in the energy range between a few hundred of keVs and 14 MeV; hence, the long counter was well
Fig. 7. Mean free path of neutrons in polyethylene of 0.93 g/cm3 with total cross-sections on natural carbon and elastic scattering cross-sections on 1H atoms. 20
Radiation Measurements 119 (2018) 16–21
Z.M. Hu et al.
characterized in the whole energy range of the calculations. The relative uncertainty on the variation of the response with energy was estimated to be 3.5%. The calculation method and measurement protocol for the effective center determination were validated via their application to a Bonner sphere. The random method was employed to predict effective center values. The sharp structures in the variation of the effective center with energy matched the total neutron cross-section of carbon, present in the polyethylene shield of the long counter. The calculated effective centers agreed reasonably well with the experimental results within uncertainties even though systematic underestimation could not be excluded. Uncertainty on the variation of the effective center with energy was estimated to be 0.5 cm. Compared to other De Pangher long counters, such as the NPL version, the long counter exhibited higher response values due to greater sensitivity of the 3He tube to thermal neutrons; however, the long counters had comparable effective center positions because of their similar geometry. The long counter compensates for the lack of devices for neutron fluence measurement from 1 MeV to 2 MeV at the State Key Laboratory of Nuclear Physics and Technology. Consequently, the counter has been successfully applied to experimental characterization of a Bonner sphere spectrometer, which has been presented elsewhere (Hu et al., 2017).
Hu, Q.Y., Zhang, J.H., Zhang, D., Guo, H.S., Yang, G.Z., Li, B.J., Ye, F., Si, F.N., Liu, J., Fu, Y.C., Ning, J.M., Yang, J., Yang, H.H., Wang, W.C., 2014a. An improved long counter for neutron fluence measurement with a flat response over a wide energy range from 1 keV to 15 MeV. Nucl. Instrum. Methods Phys. Res., Sect. A 768, 43–45. Hu, Z.M., Xie, X.F., Chen, Z.J., Peng, X.Y., Du, T.F., Cui, Z.Q., Ge, L.J., Li, T., Yuan, X., Zhang, X., Hu, L.Q., Zhong, G.Q., Lin, S.Y., Wan, B.N., Gorini, G., Li, X.Q., Zhang, G.H., Chen, J.X., Fan, T.S., 2014b. Monte Carlo simulation of a Bonner sphere spectrometer for application to the determination of neutron field in the Experimental Advanced Superconducting Tokamak experimental hall. Rev. Sci. Instrum. 85 11E417. Hu, Z.M., Chen, Z.J., Peng, X.Y., Du, T.F., Cui, Z.Q., Ge, L.J., Zhu, W.J., Wang, Z.M., Zhu, X., Chen, J.X., Zhang, G.H., Li, X.Q., Chen, J., Zhang, H., Zhong, G.H., Hu, L.Q., Wan, B.N., Gorinie, G., Fan, T.S., 2017. Experimental determination of the response functions of a Bonner sphere spectrometer to monoenergetic neutrons. J. Instrum. 12, T06002. Hunt, J.B., 1984. The calibration of neutron sensitive spherical devices. Radiat. Protect. Dosim. 8, 239–251. International Standard Organization (ISO), 2000. Reference Neutron Radiations — Part 2: Calibration Fundamentals of Radiation protection Devices Related to the Basic Quantities Characterizing the Radiation Field, ISO 8529-2. Kim, G.D., Woo, H.J., Park, J.H., 2012. Characteristics of the KIGAM long-counter for the neutron energy range below 2.5 MeV. J. Kor. Phys. Soc. 61, 347. Klein, H., Thomas, D.J., 2003. 10 Quality assurance. Radiat. Protect. Dosim. 107, 37. Lacoste, V., Gressier, V., Pochat, J.L., Fernández, F., Bakali, M., Bouassoule, T., 2004. Characterization of Bonner sphere systems at monoenergetic and thermal neutron fields. Radiat. Protect. Dosim. 110, 529. Lacoste, V., 2010. Design of a new long counter for the determination of the neutron fluence reference values at the IRSN AMANDE facility. Radiat. Meas. 45, 1250. Lacoste, V., Gressier, V., 2010. Experimental characterization of the IRSN long counter for the determination of the neutron fluence reference values at the AMANDE facility. Radiat. Meas. 45, 1254. Li, Y.M., Chen, J.X., Zhang, G.H., Fan, T.S., 2013. Study of physical characteristics for compressed-mixture type 241AmO2-Be neutron source. Atomic Energy Sci. Technol. 47, 1–6 (in Chinese). Liu, Z.Z., Chen, J.X., Zhu, P., Li, Y.M., Zhang, G.H., 2007. The 4.438 MeV gamma to neutron ratio for the Am–Be neutron source. Appl. Radiat. Isot. 65, 1318–1321. Mazunga, M., Li, T.S., Li, Y.N., Hong, B., Wang, Y.F., Ji, X., 2017. Design of an extended range long counter using super Monte Carlo simulation. Radiat. Protect. Dosim. 175, 413. Nolte, R., Thomas, D.J., 2011. 8,Monoenergetic fast neutron reference fields: II. Field characterization. Metrologia 48, S274. Roberts, N.J., Jones, L.N., Wang, Z., Liu, Y., Wang, Q., Chen, X., Luo, H., Rong, C., Králik, M., Park, H., Choi, K.O., Pereira, W.W., da Fonseca, E.S., Cassette, P., Dewey, M.S., Moiseev, N.N., Kharitonov, I.A., 2011. International key comparison of measurements of neutron source emission rate (1999-2005) – CCRI(III)-K9.AmBe. Metrologia 48, 06018. Roberts, N.J., Tagziria, H., Thomas, D.J., 2004. Determination of the Effective Centres of the NPL Long Counters, NPL Report DQL RN004. Roberts, N.J., Thomas, D.J., Lacoste, V., Böttger, R., Loeb, S., 2010. Comparison of long counter measurements of monoenergetic and radionuclide source-based neutron fluence. Radiat. Meas. 45, 1151. Ryves, T.B., 1987. International fluence-rate intercomparison for 144 and 565 keV neutrons. Metrologia 24, 27. Schlegel, D., 2005. Target User's Manual, PTB Laboratory Report PTB-6.42-05-2, Braunschweig Germany. Tagziria, H., Thomas, D.J., 2000. Calibration and Monte Carlo modelling of neutron long counters. Nucl. Instrum. Methods Phys. Res., Sect. A 452, 470. Tanimura, Y., Tsutsumi, M., Yoshizawa, M., 2014. Development of portable long counter with two different moderator materials. Radiat. Protect. Dosim. 161, 144–148. X-5 Monte Carlo Team, 2003. MCNP — a General Monte Carlo Code N-particle Transport Code, Version 5, Manual LA-UR-03-1987. Los Alamos National Laboratory. Xie, X.F., Chen, Z.J., Peng, X.Y., Yuan, X., Zhang, X., Gorini, G., Cui, Z.Q., Du, T.F., Hu, Z.M., Li, T., Fan, T.S., Chen, J.X., Li, X.Q., Zhang, G.H., Yuan, G.L., Yang, J.W., Yang, Q.W., 2014. Neutron emission measurement at the HL-2A tokamak device with a liquid scintillation detector. Rev. Sci. Instrum. 85 103506. Yuan, X., Zhang, X., Xie, X.F., Gorini, G., Chen, Z., Peng, X., Chen, J., Zhang, G., Fan, T., Zhong, G., Hu, L., Wan, B., 2013. Neutron energy spectrum measurements with a compact liquid scintillation detector on EAST. J. Instrum. 8, P07016. Zhang, G.H., Gledenov, YuM., Khuukhenkhuu, G., Sedysheva, M.V., Szalanski, P.J., Koehler, P.E., Voronov, YuN., Liu, J.M., Liu, X., Han, J.H., Chen, J.X., 2011. 149Sm (n,α)146Nd cross sections in the MeV region. Phys. Rev. Lett. 107, 252502.
Acknowledgements This work was supported by the National Key Research and Development Program of China (No. SQ2016YFZG020067), the National Magnetic Confinement Fusion Science Program of China (Nos. 2013GB106004 and 2012GB101003), and the National Natural Science Foundation of China (No. 91226102). References Alevra, A.V., Thomas, D.J., 2003. Neutron spectrometry in mixed fields: multisphere spectrometers. Radiat. Protect. Dosim. 107, 37. Axton, E.J., 1987. Intercomparison of neutron-source emission rates (1979-1984). Metrologia 23, 129. Birattari, C., Dimovasili, E., Mitaroff, A., Silari, M., 2010. A Bonner Sphere Spectrometer with extended response matrix. Nucl. Instrum. Meth. Phys. Res. 620, 260. Chadwick, M.B., et al., 2011. ENDFB-VII.1 nuclear data for science and Technology cross sections, covariances, fission product yields and decay data. Nucl. Data Sheets 112, 2887–2996. Chen, J., Wang, Z., Rong, C., Lövestam, G., Plompen, A., Puglisi, N., Gilliam, D.M., Eisenhauer, C.M., Nico, J.S., Dewey, M.S., Kudo, K., Uritani, A., Harano, H., Takeda, N., Thomas, D.J., Roberts, N.J., Bennett, A., Kolkowski, P., Moisseev, N.N., Kharitonov, I.A., Guldbakke, S., Klein, H., Nolte, R., Schlegel, D., 2007. International key comparison of neutron fluence measurements in mono-energetic neutron fields CCRI(III)-K10. Metrologia 44, 06005. De Pangher, J., Nichols, L.L., 1966. A Precision Long Counter for Measuring Fast Neutron Flux Density, BNWL-260. Pacific Northwest Laboratory, Washington USA. Gressier, V., Bonaldi, A.C., Dewey, M.S., Gilliam, D.M., Harano, H., Masuda, A., Matsumoto, T., Moiseev, N., Nico, J.S., Nolte, R., Oberstedt, S., Roberts, N.J., Röttger, S., Thomas, D.J., 2014. International key comparison of neutron fluence measurements in monoenergetic neutron fields - CCRI(III) - K11. Metrologia 51, 06009. Gressier, V., Lacoste, V., 2014. A Martin and M Pepino, Characterization of a measurement reference standard and neutron fluence determination method in IRSN monoenergetic neutron fields. Metrologia 51, 431–440. Hanson, A.O., McKibben, J.L., 1947. A neutron detector having uniform sensitivity from 10 keV to 3 MeV. Phys. Rev. 72, 673. Harano, H., Matsumoto, T., Nishiyama, J., Masuda, A., Uritani, A., Kudo, K., 2011. Development of a compact flat response neutron detector. IEEE Trans. Nucl. Sci. 58, 2421.
21