Radiation Protection Dosimetry Advance Access published May 9, 2015 Radiation Protection Dosimetry (2015), pp. 1–5
doi:10.1093/rpd/ncv291
NEUTRON FLUX CHARACTERISATION OF THE PAVIA TRIGA MARK II RESEARCH REACTOR FOR RADIOBIOLOGICAL AND MICRODOSIMETRIC APPLICATIONS D. Alloni1,2,3,*, M. Prata1,3, A. Salvini1,3 and A. Ottolenghi2,3 1 LENA, Laboratory of Applied Nuclear Energy, University of Pavia, Via Aselli 41, Pavia, Italy 2 Department of Physics, University of Pavia, Via Bassi 6, Pavia, Italy 3 INFN National Institute of Nuclear Physics, Pavia Section, Via Bassi 6, Pavia, Italy
Nowadays the Pavia TRIGA reactor is available for national and international collaboration in various research fields. The TRIGA Mark II nuclear research reactor of the Pavia University offers different in- and out-core neutron irradiation channels, each characterised by different neutron spectra. In the last two years a campaign of measurements and simulations has been performed in order to guarantee a better characterisation of these different fluxes and to meet the demands of irradiations that require precise information on these spectra in particular for radiobiological and microdosimetric studies. Experimental data on neutron fluxes have been collected analysing and measuring the gamma activity induced in thin target foils of different materials irradiated in different TRIGA experimental channels. The data on the induced gamma activities have been processed with the SAND II deconvolution code and finally compared with the spectra obtained with Monte Carlo simulations. The comparison between simulated and measured spectra showed a good agreement allowing a more precise characterisation of the neutron spectra and a validation of the adopted method.
INTRODUCTION This work has the purpose to illustrate the results of the characterisation of the neutron fields in the irradiation facilities of the TRIGA Mark II Research Reactor of the University of Pavia which offers different in- and out-core neutron irradiation channels, each characterised by different neutron spectra. A campaign of measurements and simulations has been performed in order to guarantee a better characterisation of irradiation channels and to meet the demands of future irradiations that requires a precise information on these neutron spectra, such as radiobiological and microdosimetric studies. For the different reactor irradiation channels, the measurements of the neutron fluxes have been performed by means of the foils activation, which allow one to determine the spectrum and the absolute magnitude of a neutron flux starting from rate reaction values measured on activated samples, and the spectrum de-convolution technique based on the SAND II code(1 – 5). The results have been also used extensively to benchmark the MCNP(6, 7) Monte Carlo model of the reactor with the code. MATERIAL AND METHODS w
The Pavia TRIGA reactor The TRIGAw Mark II Research Reactor, located at the Laboratory of Applied Nuclear Energy (LENA; Pavia, Italy), is a steady-state 250 kW power and water moderated reactor with a core geometry
characterised by cylindrical symmetry. At the present time, the reactor core is a lattice (see Figure 1) of 90 cylindrical elements located on five concentric rings around the Central Thimble (CT) channel. Among these elements, 81 are moderator-fuel elements (uranium enriched at 20 % with 235U mixed with zirconium hydride), three are control rods (REGULATING, TRANSIENT and SHIM to regulate power level and ignition/shut down of the reactor) and one radium–beryllium source to trigger off chain reaction during reactor ignition; three are vertical irradiation channels (CT, F Thimble, Pneumatic Transfer System Thimble) which penetrate inside core allowing irradiation of small samples into the maximum flux region; the remaining elements are graphite elements (dummy elements). Fuel elements on the first, second and third rings have stainless steel cladding, while those on the fourth and fifth rings have aluminium cladding. All these features have been preserved during the simulation by MCNP code for flux calculations. Furthermore, the neutron source was simulated with a Watt fission energy spectrum defined by the equation E sinh ðbEÞ1=2 ; ð1Þ pðEÞ ¼ C exp a where C is a normalisation constant, a and b are parameters appropriate to neutron-induced fission in various materials and for spontaneous fission. Considering the reactor fuel elements, the neutron-induced fission isotope is 235U, the a and b parameter’s values (from code ENDF/B-V library) are a ¼ 0.988 and b ¼ 2.249, respectively. Each fuel element is an independent
# The Author 2015. Published by Oxford University Press. All rights reserved. For Permissions, please email: [email protected]
Downloaded from http://rpd.oxfordjournals.org/ at University of Cambridge on September 16, 2015
*Corresponding author: [email protected]
D. ALLONI ET AL.
isotropic source with a probability of neutron generation uniformly distributed over the fuel volume.
Table 1. Target materials used in this work. Material and reaction
Neutron activation analysis
23
Since the purpose of this work was to develop a standard technique to evaluate the energy spectrum and the absolute magnitude of a neutron flux in every irradiation positions of TRIGA reactor, the authors decided to begin with the characterisation of neutron flux in CT, that is the channel with the greatest available flux (reactor spectrum) composed of the three main energetic neutron components: thermal, epithermal and fast. To measure neutron spectra, the neutron activation analysis (NAA) technique has been used(8, 9). Because there are many probe substances whose neutron activation cross sections depend in different ways on the neutron energy, it is possible to derive information about the neutron field energy distribution from probe-induced activity measurements. The samples used in the measurement campaign have been chosen to cover the whole neutron energy spectrum up to 20 MeV for activation analysis (see Table 1). Furthermore, mono-elemental thin target foils or infinite diluted element samples have been chosen to avoid self-absorption effects and corrections to the measured induced activity. For the analysis of experimental activation data, the Spectrum Analysis by Neutrons Detector II (SAND II) deconvolution program has been used(1, 2). This code allows one to determine energetic spectrum and absolute magnitude of the neutron flux starting from measured induced activity values from samples which underwent that neutron field. After each irradiation, measurements of the induced gamma activity
Na(n, g)24Na Sc(n, g)46Sc 55 Mn(n, g)56Mn 59 Co(n, g)60Co 63 Cu(n, g)64Cu 65 Cu(n, g)66Cu 75 As(n, g)76As 81 Br(n, g)82Br 139 La(n, g)140La 197 Au(n, g)198Au 23 Na(n, g)24Na in Cadmium 45 Sc(n, g)46Sc in Cadmium 63 Cu(n, g)64Cu in Cadmium 65 Cu(n, g)66Cu in Cadmium 75 As(n, g)76As in Cadmium 81 Br(n, g)82Br in Cadmium 139 La(n, g)140La in Cadmium 197 Au(n, g)198Au in Cadmium 24 Mg(n, p)24Na 27 Al(n, p)27Mg 47 Ti(n, p)47Sc 48 Ti(n, p)48Sc 54 Fe(n, p)54Mn 115 In(n, n’)115In* in Cadmium 45
‘In Cadmium’ means that the same target foil was positioned in a small cadmium container in order to cut the thermal part of the neutron energy spectrum. *Excited indium state.
in the target foils have been evaluated by means of high resolution gamma-ray spectrometry (low background HPGe coaxial, vertical dip-stick detector EG&G ORTEC), which has a relative efficiency of about 25 –30 % and a resolution of 1.95 keV FWHM
Page 2 of 5
Downloaded from http://rpd.oxfordjournals.org/ at University of Cambridge on September 16, 2015
Figure 1. Top view of the TRIGA reactor core. The irradiation channels characterised in this work are indicated in the core map on top left of the figure. Horizontal channels dedicated for out-core irradiation facilities are also indicated.
NEUTRON FLUX CHARACTERISATION AT PAVIA TRIGA REACTOR
RESULTS Calculation of the specific activity at saturation Once known the irradiation time Tirr in the specific TRIGA irradiation channel and for the specific target, the time interval DT between the end of irradiation and the beginning of the counting, and the counting time Tcount, it was possible to calculate the induced activity AEOI at the end of irradiation (EOI) and then the specific activity at saturation Aspec – sat, for each element by using correction factors to take into account the radioactive decay (for further details,
see for example Alloni et al.(11)). All the values of the measured specific activities at saturation, for each reaction, constitute the input data for the SAND II code. Experimental data processing Sand II provides a ‘best fit’ neutron flux spectrum for a specific input set of infinitely dilute foil activities. The calculation procedure involves the selection of a known flux spectrum as the initial approximation (a guess spectrum taken from MCNP calculations) for the subsequent iterations and the final solution. This multiple foil iterative method can be expected to give integral neutron flux results over the energy range from 10210 to 18 MeV. Over this energy range the code uses a discrete interval description represented by 620 intervals (45 per decade up to 1 MeV, and 170 between 1 and 18 MeV). The solution spectrum is thus presented in tabular form at 621 points. The problem is essentially to solve for 621 unknowns (solution differential flux values) in a system of n linear equations, where n is the number of foils used in the experiment (for further details, see Mc Elroy et al.(4)). Furthermore, the output of SAND II can be lumped in different energy groups in order to better compare the experimental data with simulation outputs. In this case a subdivision in 135 energy groups has been chosen. A set of 11 reactions has been selected in order to cover all neutron energy spectrum (the ideal condition for the code input is a set of a number between 10 and 20 activity values), and after running the code with 15 iterations, a best fit for neutron spectrum in the CT
Figure 2. Comparison between the experimental neutron spectrum obtained by SAND II and the calculated neutron spectrum produced by MCNP (guess spectrum) for the CT as lethargy plot in 135 energy groups. Errors are standard deviations for both curves and are limited to the symbols.
Page 3 of 5
Downloaded from http://rpd.oxfordjournals.org/ at University of Cambridge on September 16, 2015
at 1.332 MeV (with a peak-to-Compton ratio of 55/1). The calibration efficiency of an HPGe spectrometer is defined as the ratio between the number of gamma rays detected and the number of gamma rays emitted by a radioactive source with certified activity. The calibration efficiency is a function of gamma energy and a curve of efficiency vs. energy has been obtained (the best fit of experimental points). The uncertainty on efficiency is one of the contributions to the total experimental uncertainty, and it is mainly due to the uncertainty on the certified activity of the source employed in this operation (usually 3– 5 %). The gamma-ray acquisition system consists of MAESTROw multichannel Analyzer (MCA) emulation software card, coupled with the detector through electronic modules, all manufactured by EG&G ORTEC. The multi-purpose gamma-ray analysis software GammaVisionw was used for peaks identification and evaluation.
D. ALLONI ET AL.
CONCLUSIONS The aim of this work was to evaluate, by means of direct measurements, the neutron flux distribution in different irradiation channels of the Pavia TRIGA reactor and to benchmark the results of the measurements with the data obtained through the simulation of the reactor preformed by means of the Monte Carlo code MCNP. The fluxes have been measured using different reaction channels through the neutron activation technique coupled with spectrum de-convolution based on SAND II code. The MCNP code allowed to one reproduce the reactor complex structure in detail, with regard to the reactor geometry, material composition, source and the evaluation of neutron energy spectrum and its absolute magnitude. The unfolding technique allowed one to determine neutron energetic spectrum and absolute magnitude
Figure 3. Comparison between the experimental neutron spectrum obtained by SAND II and the calculated neutron spectrum produced by MCNP (guess spectrum) for the RT as lethargy plot in 135 energy groups. Errors are standard deviations for both curves and are limited to the symbols.
Table 2. Comparison between processed experimental data (SAND) and calculated (MCNP) thermal, epithermal and fast integral flux in the CT and in the RT. Energy
Thermal (0–0.21 eV) Epithermal (0.21 eV–9.2 keV) Fast (E . 9.2 keV) Total
CT flux (SAND) (`1012 cm22 s21)
CT flux (MCNP) (`1012 cm22 s21)
RT flux (SAND) (`1012 cm22 s21)
RT flux (MCNP) (`1012 cm22 s21)
3.00 + 0.15 4.64 + 0.21 5.89 + 0.18 13.9 + 0.31
3.19 + 0.09 3.72 + 0.11 6.20 + 0.14 13.1 + 0.19
3.25 + 0.16 3.20 + 0.12 1.43 + 0.14 8.30 + 0.15
3.43 + 0.10 3.23 + 0.09 1.51 + 0.11 8.37 + 0.12
The reported experimental and simulated uncertainties are standard deviations.
Page 4 of 5
Downloaded from http://rpd.oxfordjournals.org/ at University of Cambridge on September 16, 2015
and in the Rabbit Thimble (RT) of the reactor has been obtained. This result is shown as lethargy plots (dw/d(log E) vs. log E) represented in Figures 2 and 3. The guess spectrum for iterative process was the neutron spectrum calculated, for each channel, with MCNP for the Pavia TRIGA reactor configuration (i.e. core configuration). The dependence of SAND results upon different guess spectra (selected between fluxes from MCNP and taking into account small flux variations in the associated error intervals for each energy bin) has been also tested showing no appreciable variations (within 2 %). In Table 2, as an example, a comparison between integral flux values obtained from MCNP calculation and those obtained from the measured activities and the unfolding method corresponding to the thermal, epithermal and fast energy region is shown both for CT and RT channels.
NEUTRON FLUX CHARACTERISATION AT PAVIA TRIGA REACTOR
2. 3. 4. 5. 6. 7. 8. 9. 10.
11.
REFERENCES 1. Berg, S. and McElroy, W. N. A computer-automated iterative method for neutron flux spectra determination by
foil activation. In: SAND II (Spectrum Analysis by Neutron Detectors II) and Associated Codes AFWLTR-67-41, vol II (1967). Griffin, P. J. SNL RML recommended dosimetry cross section compendium SAND92-0094 (1993). Griffin, P. J. et al. User’s Manual for SNL-SAND-II Code SAND-93-3957 (1994). Mc Elroy, W. N. et al. A computer-automated iterative method for neutron flux spectra determination by foil activation. In: AFWL-TR 67-41, vol. 1 (1967). Oak Ridge National Laboratory. SAND II Manual. Radiation Safety Information Computational Center, Oak Ridge National Laboratory (1965). Harmon, C. D., Busch, R. D., Briesmeister, J. F. and Forster, R. A. Criticality Calculations with MCNP: a Primer. Los Alamos, LA-12827 (1994). RSICC Computer Code Collection. MCNP4C—Monte Carlo N-Particle Transport Code System User Manual, Los Alamos (2001). Alfassi, Z. B. Activation Analysis, vols. I and II. CRC Press (1990). Alfassi, Z. B. Chemical Analysis by Nuclear Methods. John Wiley and Sons, Inc. (1994). Colautti, P., Moro, D., Chiriotti, S., Conte, V., Evangelista, L., Bortolussi, S., Protti, N., Postuma, I. and Altieri, S. Microdosimetric measurements in the thermal neutron irradiation facility of LENA reactor. Appl. Radiat. Isot. 88, 147–152 (2014). Alloni, D., Borio di Tigliole, A., Bruni, J., Cagnazzo, M., Cremonesi, R., Magrotti, G., Oddone, M., Panza, F., Prata, M. and Salvini, A. Neutron flux characterization of the SM1 sub-critical multiplying complex of the Pavia University. Prog. Nucl. Energ. 67, 98– 103 (2013).
Page 5 of 5
Downloaded from http://rpd.oxfordjournals.org/ at University of Cambridge on September 16, 2015
flux in Central Thimble and Rabbit Thimble starting from measured induced activity values for targets that have been irradiated in the reactor neutron field. The method described above together with the good agreement between MCNP simulations and experimental data will allow the same method to be adopted for experimental feasibility studies, in particular, for irradiation positions dedicated to radiobiology and microdosimetric study(10) where detailed information on neutron spectra are requested. The same method has been successfully applied in the past for the characterisation of the neutron field inside the subcritical facility(11) of the Pavia University. Currently, the Pavia TRIGA reactor facility is available for irradiation offering different fully characterised experimental channels. Accurate measurements of the neutron flux are also under study for the horizontal channels where neutron beams can be extracted from reactor core with different energy components depending on the channel design (e.g. the presence of core graphite reflector). Furthermore, the presence of neutron beams with different energy components will allow one, for example, to perform research activities on relative biological effectiveness (RBE) of neutrons.
doi:10.1093/rpd/ncv291
NEUTRON FLUX CHARACTERISATION OF THE PAVIA TRIGA MARK II RESEARCH REACTOR FOR RADIOBIOLOGICAL AND MICRODOSIMETRIC APPLICATIONS D. Alloni1,2,3,*, M. Prata1,3, A. Salvini1,3 and A. Ottolenghi2,3 1 LENA, Laboratory of Applied Nuclear Energy, University of Pavia, Via Aselli 41, Pavia, Italy 2 Department of Physics, University of Pavia, Via Bassi 6, Pavia, Italy 3 INFN National Institute of Nuclear Physics, Pavia Section, Via Bassi 6, Pavia, Italy
Nowadays the Pavia TRIGA reactor is available for national and international collaboration in various research fields. The TRIGA Mark II nuclear research reactor of the Pavia University offers different in- and out-core neutron irradiation channels, each characterised by different neutron spectra. In the last two years a campaign of measurements and simulations has been performed in order to guarantee a better characterisation of these different fluxes and to meet the demands of irradiations that require precise information on these spectra in particular for radiobiological and microdosimetric studies. Experimental data on neutron fluxes have been collected analysing and measuring the gamma activity induced in thin target foils of different materials irradiated in different TRIGA experimental channels. The data on the induced gamma activities have been processed with the SAND II deconvolution code and finally compared with the spectra obtained with Monte Carlo simulations. The comparison between simulated and measured spectra showed a good agreement allowing a more precise characterisation of the neutron spectra and a validation of the adopted method.
INTRODUCTION This work has the purpose to illustrate the results of the characterisation of the neutron fields in the irradiation facilities of the TRIGA Mark II Research Reactor of the University of Pavia which offers different in- and out-core neutron irradiation channels, each characterised by different neutron spectra. A campaign of measurements and simulations has been performed in order to guarantee a better characterisation of irradiation channels and to meet the demands of future irradiations that requires a precise information on these neutron spectra, such as radiobiological and microdosimetric studies. For the different reactor irradiation channels, the measurements of the neutron fluxes have been performed by means of the foils activation, which allow one to determine the spectrum and the absolute magnitude of a neutron flux starting from rate reaction values measured on activated samples, and the spectrum de-convolution technique based on the SAND II code(1 – 5). The results have been also used extensively to benchmark the MCNP(6, 7) Monte Carlo model of the reactor with the code. MATERIAL AND METHODS w
The Pavia TRIGA reactor The TRIGAw Mark II Research Reactor, located at the Laboratory of Applied Nuclear Energy (LENA; Pavia, Italy), is a steady-state 250 kW power and water moderated reactor with a core geometry
characterised by cylindrical symmetry. At the present time, the reactor core is a lattice (see Figure 1) of 90 cylindrical elements located on five concentric rings around the Central Thimble (CT) channel. Among these elements, 81 are moderator-fuel elements (uranium enriched at 20 % with 235U mixed with zirconium hydride), three are control rods (REGULATING, TRANSIENT and SHIM to regulate power level and ignition/shut down of the reactor) and one radium–beryllium source to trigger off chain reaction during reactor ignition; three are vertical irradiation channels (CT, F Thimble, Pneumatic Transfer System Thimble) which penetrate inside core allowing irradiation of small samples into the maximum flux region; the remaining elements are graphite elements (dummy elements). Fuel elements on the first, second and third rings have stainless steel cladding, while those on the fourth and fifth rings have aluminium cladding. All these features have been preserved during the simulation by MCNP code for flux calculations. Furthermore, the neutron source was simulated with a Watt fission energy spectrum defined by the equation E sinh ðbEÞ1=2 ; ð1Þ pðEÞ ¼ C exp a where C is a normalisation constant, a and b are parameters appropriate to neutron-induced fission in various materials and for spontaneous fission. Considering the reactor fuel elements, the neutron-induced fission isotope is 235U, the a and b parameter’s values (from code ENDF/B-V library) are a ¼ 0.988 and b ¼ 2.249, respectively. Each fuel element is an independent
# The Author 2015. Published by Oxford University Press. All rights reserved. For Permissions, please email: [email protected]
Downloaded from http://rpd.oxfordjournals.org/ at University of Cambridge on September 16, 2015
*Corresponding author: [email protected]
D. ALLONI ET AL.
isotropic source with a probability of neutron generation uniformly distributed over the fuel volume.
Table 1. Target materials used in this work. Material and reaction
Neutron activation analysis
23
Since the purpose of this work was to develop a standard technique to evaluate the energy spectrum and the absolute magnitude of a neutron flux in every irradiation positions of TRIGA reactor, the authors decided to begin with the characterisation of neutron flux in CT, that is the channel with the greatest available flux (reactor spectrum) composed of the three main energetic neutron components: thermal, epithermal and fast. To measure neutron spectra, the neutron activation analysis (NAA) technique has been used(8, 9). Because there are many probe substances whose neutron activation cross sections depend in different ways on the neutron energy, it is possible to derive information about the neutron field energy distribution from probe-induced activity measurements. The samples used in the measurement campaign have been chosen to cover the whole neutron energy spectrum up to 20 MeV for activation analysis (see Table 1). Furthermore, mono-elemental thin target foils or infinite diluted element samples have been chosen to avoid self-absorption effects and corrections to the measured induced activity. For the analysis of experimental activation data, the Spectrum Analysis by Neutrons Detector II (SAND II) deconvolution program has been used(1, 2). This code allows one to determine energetic spectrum and absolute magnitude of the neutron flux starting from measured induced activity values from samples which underwent that neutron field. After each irradiation, measurements of the induced gamma activity
Na(n, g)24Na Sc(n, g)46Sc 55 Mn(n, g)56Mn 59 Co(n, g)60Co 63 Cu(n, g)64Cu 65 Cu(n, g)66Cu 75 As(n, g)76As 81 Br(n, g)82Br 139 La(n, g)140La 197 Au(n, g)198Au 23 Na(n, g)24Na in Cadmium 45 Sc(n, g)46Sc in Cadmium 63 Cu(n, g)64Cu in Cadmium 65 Cu(n, g)66Cu in Cadmium 75 As(n, g)76As in Cadmium 81 Br(n, g)82Br in Cadmium 139 La(n, g)140La in Cadmium 197 Au(n, g)198Au in Cadmium 24 Mg(n, p)24Na 27 Al(n, p)27Mg 47 Ti(n, p)47Sc 48 Ti(n, p)48Sc 54 Fe(n, p)54Mn 115 In(n, n’)115In* in Cadmium 45
‘In Cadmium’ means that the same target foil was positioned in a small cadmium container in order to cut the thermal part of the neutron energy spectrum. *Excited indium state.
in the target foils have been evaluated by means of high resolution gamma-ray spectrometry (low background HPGe coaxial, vertical dip-stick detector EG&G ORTEC), which has a relative efficiency of about 25 –30 % and a resolution of 1.95 keV FWHM
Page 2 of 5
Downloaded from http://rpd.oxfordjournals.org/ at University of Cambridge on September 16, 2015
Figure 1. Top view of the TRIGA reactor core. The irradiation channels characterised in this work are indicated in the core map on top left of the figure. Horizontal channels dedicated for out-core irradiation facilities are also indicated.
NEUTRON FLUX CHARACTERISATION AT PAVIA TRIGA REACTOR
RESULTS Calculation of the specific activity at saturation Once known the irradiation time Tirr in the specific TRIGA irradiation channel and for the specific target, the time interval DT between the end of irradiation and the beginning of the counting, and the counting time Tcount, it was possible to calculate the induced activity AEOI at the end of irradiation (EOI) and then the specific activity at saturation Aspec – sat, for each element by using correction factors to take into account the radioactive decay (for further details,
see for example Alloni et al.(11)). All the values of the measured specific activities at saturation, for each reaction, constitute the input data for the SAND II code. Experimental data processing Sand II provides a ‘best fit’ neutron flux spectrum for a specific input set of infinitely dilute foil activities. The calculation procedure involves the selection of a known flux spectrum as the initial approximation (a guess spectrum taken from MCNP calculations) for the subsequent iterations and the final solution. This multiple foil iterative method can be expected to give integral neutron flux results over the energy range from 10210 to 18 MeV. Over this energy range the code uses a discrete interval description represented by 620 intervals (45 per decade up to 1 MeV, and 170 between 1 and 18 MeV). The solution spectrum is thus presented in tabular form at 621 points. The problem is essentially to solve for 621 unknowns (solution differential flux values) in a system of n linear equations, where n is the number of foils used in the experiment (for further details, see Mc Elroy et al.(4)). Furthermore, the output of SAND II can be lumped in different energy groups in order to better compare the experimental data with simulation outputs. In this case a subdivision in 135 energy groups has been chosen. A set of 11 reactions has been selected in order to cover all neutron energy spectrum (the ideal condition for the code input is a set of a number between 10 and 20 activity values), and after running the code with 15 iterations, a best fit for neutron spectrum in the CT
Figure 2. Comparison between the experimental neutron spectrum obtained by SAND II and the calculated neutron spectrum produced by MCNP (guess spectrum) for the CT as lethargy plot in 135 energy groups. Errors are standard deviations for both curves and are limited to the symbols.
Page 3 of 5
Downloaded from http://rpd.oxfordjournals.org/ at University of Cambridge on September 16, 2015
at 1.332 MeV (with a peak-to-Compton ratio of 55/1). The calibration efficiency of an HPGe spectrometer is defined as the ratio between the number of gamma rays detected and the number of gamma rays emitted by a radioactive source with certified activity. The calibration efficiency is a function of gamma energy and a curve of efficiency vs. energy has been obtained (the best fit of experimental points). The uncertainty on efficiency is one of the contributions to the total experimental uncertainty, and it is mainly due to the uncertainty on the certified activity of the source employed in this operation (usually 3– 5 %). The gamma-ray acquisition system consists of MAESTROw multichannel Analyzer (MCA) emulation software card, coupled with the detector through electronic modules, all manufactured by EG&G ORTEC. The multi-purpose gamma-ray analysis software GammaVisionw was used for peaks identification and evaluation.
D. ALLONI ET AL.
CONCLUSIONS The aim of this work was to evaluate, by means of direct measurements, the neutron flux distribution in different irradiation channels of the Pavia TRIGA reactor and to benchmark the results of the measurements with the data obtained through the simulation of the reactor preformed by means of the Monte Carlo code MCNP. The fluxes have been measured using different reaction channels through the neutron activation technique coupled with spectrum de-convolution based on SAND II code. The MCNP code allowed to one reproduce the reactor complex structure in detail, with regard to the reactor geometry, material composition, source and the evaluation of neutron energy spectrum and its absolute magnitude. The unfolding technique allowed one to determine neutron energetic spectrum and absolute magnitude
Figure 3. Comparison between the experimental neutron spectrum obtained by SAND II and the calculated neutron spectrum produced by MCNP (guess spectrum) for the RT as lethargy plot in 135 energy groups. Errors are standard deviations for both curves and are limited to the symbols.
Table 2. Comparison between processed experimental data (SAND) and calculated (MCNP) thermal, epithermal and fast integral flux in the CT and in the RT. Energy
Thermal (0–0.21 eV) Epithermal (0.21 eV–9.2 keV) Fast (E . 9.2 keV) Total
CT flux (SAND) (`1012 cm22 s21)
CT flux (MCNP) (`1012 cm22 s21)
RT flux (SAND) (`1012 cm22 s21)
RT flux (MCNP) (`1012 cm22 s21)
3.00 + 0.15 4.64 + 0.21 5.89 + 0.18 13.9 + 0.31
3.19 + 0.09 3.72 + 0.11 6.20 + 0.14 13.1 + 0.19
3.25 + 0.16 3.20 + 0.12 1.43 + 0.14 8.30 + 0.15
3.43 + 0.10 3.23 + 0.09 1.51 + 0.11 8.37 + 0.12
The reported experimental and simulated uncertainties are standard deviations.
Page 4 of 5
Downloaded from http://rpd.oxfordjournals.org/ at University of Cambridge on September 16, 2015
and in the Rabbit Thimble (RT) of the reactor has been obtained. This result is shown as lethargy plots (dw/d(log E) vs. log E) represented in Figures 2 and 3. The guess spectrum for iterative process was the neutron spectrum calculated, for each channel, with MCNP for the Pavia TRIGA reactor configuration (i.e. core configuration). The dependence of SAND results upon different guess spectra (selected between fluxes from MCNP and taking into account small flux variations in the associated error intervals for each energy bin) has been also tested showing no appreciable variations (within 2 %). In Table 2, as an example, a comparison between integral flux values obtained from MCNP calculation and those obtained from the measured activities and the unfolding method corresponding to the thermal, epithermal and fast energy region is shown both for CT and RT channels.
NEUTRON FLUX CHARACTERISATION AT PAVIA TRIGA REACTOR
2. 3. 4. 5. 6. 7. 8. 9. 10.
11.
REFERENCES 1. Berg, S. and McElroy, W. N. A computer-automated iterative method for neutron flux spectra determination by
foil activation. In: SAND II (Spectrum Analysis by Neutron Detectors II) and Associated Codes AFWLTR-67-41, vol II (1967). Griffin, P. J. SNL RML recommended dosimetry cross section compendium SAND92-0094 (1993). Griffin, P. J. et al. User’s Manual for SNL-SAND-II Code SAND-93-3957 (1994). Mc Elroy, W. N. et al. A computer-automated iterative method for neutron flux spectra determination by foil activation. In: AFWL-TR 67-41, vol. 1 (1967). Oak Ridge National Laboratory. SAND II Manual. Radiation Safety Information Computational Center, Oak Ridge National Laboratory (1965). Harmon, C. D., Busch, R. D., Briesmeister, J. F. and Forster, R. A. Criticality Calculations with MCNP: a Primer. Los Alamos, LA-12827 (1994). RSICC Computer Code Collection. MCNP4C—Monte Carlo N-Particle Transport Code System User Manual, Los Alamos (2001). Alfassi, Z. B. Activation Analysis, vols. I and II. CRC Press (1990). Alfassi, Z. B. Chemical Analysis by Nuclear Methods. John Wiley and Sons, Inc. (1994). Colautti, P., Moro, D., Chiriotti, S., Conte, V., Evangelista, L., Bortolussi, S., Protti, N., Postuma, I. and Altieri, S. Microdosimetric measurements in the thermal neutron irradiation facility of LENA reactor. Appl. Radiat. Isot. 88, 147–152 (2014). Alloni, D., Borio di Tigliole, A., Bruni, J., Cagnazzo, M., Cremonesi, R., Magrotti, G., Oddone, M., Panza, F., Prata, M. and Salvini, A. Neutron flux characterization of the SM1 sub-critical multiplying complex of the Pavia University. Prog. Nucl. Energ. 67, 98– 103 (2013).
Page 5 of 5
Downloaded from http://rpd.oxfordjournals.org/ at University of Cambridge on September 16, 2015
flux in Central Thimble and Rabbit Thimble starting from measured induced activity values for targets that have been irradiated in the reactor neutron field. The method described above together with the good agreement between MCNP simulations and experimental data will allow the same method to be adopted for experimental feasibility studies, in particular, for irradiation positions dedicated to radiobiology and microdosimetric study(10) where detailed information on neutron spectra are requested. The same method has been successfully applied in the past for the characterisation of the neutron field inside the subcritical facility(11) of the Pavia University. Currently, the Pavia TRIGA reactor facility is available for irradiation offering different fully characterised experimental channels. Accurate measurements of the neutron flux are also under study for the horizontal channels where neutron beams can be extracted from reactor core with different energy components depending on the channel design (e.g. the presence of core graphite reflector). Furthermore, the presence of neutron beams with different energy components will allow one, for example, to perform research activities on relative biological effectiveness (RBE) of neutrons.
