- Email: [email protected]
Bremsstrahlung source term estimation for high energy electron accelerators
Radiation Physics and Chemistry 113 (2015) 1–5
Contents lists available at ScienceDirect
Radiation Physics and Chemistry journal homepage: www.elsevier.com/locate/radphyschem
Bremsstrahlung source term estimation for high energy electron accelerators M.K. Nayak a,n, T.K. Sahu a, H.G. Nair a, R.V. Nandedkar b, Tapas Bandyopadhyay a, R.M Tripathi a, P.R. Hannurkar c, D.N Sharma d a
Health Physics Division, Bhabha Atomic Research Centre, Mumbai 400085, India Raja Ramanna Centre for Advanced Technology, Indore 452013, India c Indus Operations and Accelerator Physics Design Division, Raja Ramanna Centre for Advanced Technology, Indore 452013, India d Health, Safety & Environment Group, Bhabha Atomic Research Centre, Mumbai 400085, India b
H I G H L I G H T S
Experimental determination of bremsstrahlung source term at 450 and 550 MeV electrons. Monte Carlo calculations performed for validation of experimental data. Thick and thin target bremsstrahlung source term is studied. Brensstrahlung Source term is determined up to 3 GeV electron energies.
art ic l e i nf o
a b s t r a c t
Article history: Received 28 January 2015 Received in revised form 9 April 2015 Accepted 11 April 2015 Available online 16 April 2015
Thick target bremsstrahlung source term for 450 MeV and 550 MeV electrons are experimentally determined using booster synchrotron of Indus facility at Raja Ramanna Centre for Advanced Technology, Indore, India. The source term is also simulated using EGSnrc Monte Carlo code. Results from experiment and simulation are found to be in very good agreement. Based on the agreement between experimental and simulated data, the source term is determined up to 3000 MeV by simulation. The paper also describes the studies carried out on the variation of source term when a thin target is considered in place of a thick target, used in earlier studies. & 2015 Elsevier Ltd. All rights reserved.
Keywords: Source term Depth dose Bremsstrahlung Thick target Thin target
1. Introduction Bremsstrahlung source term for accelerators is usually expressed as dose equivalent rate or absorbed dose rate per unit beam power at 1 m distance from a high atomic number (Z) thick target (Ferrari et al., 1993). Estimation of source term is important for evaluation of shielding requirement for particle accelerator facilities. The most widely used source terms for electron accelerators are those suggested by Swanson (1979). It is based on experimental data obtained from the measurement of absorbed dose rate in a 30 cm tissue equivalent phantom. The bremsstrahlung radiation for these experiments were generated by bombarding high energy electrons on high Z thick target. For thick targets, the target thickness considered is equivalent to the range n
Corresponding author. E-mail address: [email protected] (M.K. Nayak).
http://dx.doi.org/10.1016/j.radphyschem.2015.04.004 0969-806X/& 2015 Elsevier Ltd. All rights reserved.
of electrons in the medium (Wyckoff et al., 1971). Experimentally obtained maximum absorbed dose data for 20, 30 and 100 MeV electron energies in a polymethyl methacrylate (PMMA) phantom generated by Wyckoff et al. was used by Swanson in formulating a semi-empirical relation for thick target bremsstrahlung source term. The empirical relations suggested for forward and lateral source terms are available up to 100 MeV. For energies in the range 20–100 MeV, the suggested relations for thick target source term in the forward direction are found to be linear in energy. Whereas, for higher energies the relations are assumed to be linear (Swanson,1979; NCRP, 2003). When an accelerated electron beam moving in vacuum envelope is lost, it encounters a thin target of few millimeter thickness, instead of a thick target, which has been studied earlier. For thin targets, generated bremsstrahlung photons in the forward direction are of higher energies compared with those of thick targets (Haridas et al., 2006). Bremsstrahlung photons generated by thin
2
M.K. Nayak et al. / Radiation Physics and Chemistry 113 (2015) 1–5
targets further generate electromagnetic shower (ICRU, 1978) when incident on shield structures. Moreover as the thickness of vacuum chamber is less, substantial electrons are also transmitted along with bremsstrahlung photons and the source term is different from that obtained from the empirical relations for the thick target case. The absorbed dose rate from tungsten target at 200 MeV electrons incident as a function of angle of emission of bremsstrahlung and thickness calculated by Fasso. et al. (1984), shows that as the thickness of the target is reduced, the absorbed dose rate increases at low emission angles (forward direction). For larger emission angles, the dose rate is higher from thick targets. Thus the target thickness plays an important role in source term determination. The forward source term for 100 MeV–1000 MeV electrons have been calculated using Monte Carlo simulation code FLUKA by Ferrari. et al. (1993). It has been found that the forward source term is higher by 7 times at 100 MeV and 22 times at 1000 MeV when compared with the values obtained by Swanson’s empirical relations. Tromba et al. (1990) also proposed empirical relation for the thick target source term using Monte Carlo calculations performed with EGS4 code for electron energies from 100 MeV up to 10 GeV. In the present work, thick target source term is experimentally determined from the depth dose data for 450 MeV and 550 MeV electrons in the forward direction on booster synchrotron at Indus synchrotron facility of Raja Ramanna Centre for Advanced Technology (RRCAT), Indore. The experimental data is also simulated using EGSnrc code and is found to be in excellent agreement with the experimental results. Subsequently, the source term data is simulated for higher energies up to 3000 MeV. Source term simulations are also performed for thin targets at two electron energies, 450 MeV1 and 2500 MeV1. The simulated data for 450 MeV is experimentally verified at the booster synchrotron. The thin target source terms are found to be higher than the thick target source term obtained from empirical relations by Swanson.
2. Experimental details 2.1. Materials and methods CaSO4: Dy TLD disks of size 6 mm diameter and 1 mm thickness used for the experiments are annealed for 1 h at 400 °C in air. The sensitivity of these TLDs is checked by exposing the disks uniformly to a known dose of 5 mGy from a Co-60 source. TLDs within 7 3% sensitivity are selected and used for these experiments. A cuboid water phantom of size 30 cm 30 cm 30 cm made from 10 mm thick Perspex sheet (with top open) is used for measuring the depth dose in water. The phantom has been placed 1 m away from a lead target in the forward direction with TLDs placed at different depths at the mid plane of the phantom. It is ensured that the geometrical center of the target, the center of the water phantom and the axis of the TL disks within the phantom are in a straight line. The schematic diagram of the experimental setup for the measurement of forward source term is shown in Fig. 1.
Fig. 1. Experimental set up and geometry used for simulation.
aluminum) window. The extracted beam is allowed to fall on the center of the lead target of 25 mm2 (1 CSDA range of 450 MeV electrons) (ICRU,1984). The position of beam at the center of the target is ensured by observing the fluorescence, emitted from a fluorescence screen placed before the target, with the help of a charge-coupled device (CCD) camera. The TLDs are exposed for a period of 900 s (the time of exposure was decided based on a preliminary dose rate measurement with ion chamber). The circulating beam current in the booster synchrotron is 1.5 mA. During the exposure period, the beam pulse from the booster synchrotron is measured with the help of a Wall Current Monitor (WCM) which gave a pulse height of 25 mV/beam pulse. Two pulses per second are fired on the target. From the pulse height data, the extracted electrons per second is calculated to be 3.26 108. Similar experiments are repeated for 550 MeV electrons with a 26.3 mm (1CSDA range) lead target and dose is measured with a new set of TLDs. The extracted electrons per second obtained from the WCM data was 1.66 108. The experiments for thin target source term for 450 MeV electrons are also carried out by allowing the electrons to be incident on a 2 mm Stainless Steel (SS) target close to the vacuum chamber thickness of Indus-1 storage ring. TLDs are exposed within the water phantom for 600 s. The exposed TLDs for each set of experiments are analyzed using a TLD reader (Laboratory Reader-Analyzer RA'94, RADPRO, POLAND). From the TL counts the relative dose per unit beam power with respect to Co-60 is found out.
3. Monte Carlo simulations
The electron beam at 450 MeV from the booster synchrotron is extracted to a transport line via an extraction septum and the electron beam is allowed to come out through a thin (1 mm
The experimental results obtained are simulated using EGSnrc Monte Carlo code. The geometry for the experimental setup (Fig. 1) is made using DOSERZ user code from EGSnrc. In the simulation of depth dose curve, a parallel beam of 3.5 mm radius (which is the electron beam radius) of 450 MeV is allowed to incident on 25 mm thick lead target and absorbed dose per incidence fluence was scored in the TLD disks (scoring area 0.363 cm2), placed in different depths of the water phantom. For radiation transport, an electron cut off (Ecut) of 521 keV and a photon cut-off (Pcut) of 10 keV are used in the simulation with a history of 106. Simulation for 550 MeV electron case is carried out by changing the lead target thickness to 26.3 mm. The statistical accuracy obtained in the present simulation is within 71%. The simulations of depth dose in phantom with thin targets (2 mm Stainless Steel and 30 mm Aluminum) are also performed for calculating the source term for Indus-1 and Indus-2 storage rings (vacuum chamber of Indus-1 and Indus-2 are made of 2 mm
1 The 450 MeV and 2500 MeV are the electron beam energies in the storage rings, Indus-1 and Indus-2 respectively at RRCAT, Indore.
2 It is equivalent to 4.5 radiation length, which is more than enough to develop electromagnetic cascade fully in the lead.
2.2. Experiments
M.K. Nayak et al. / Radiation Physics and Chemistry 113 (2015) 1–5
3
Fig. 2. Measured and simulated absorbed dose rate for thick target due to 450 MeV electron beam (Gy/h-kW).
Fig. 4. Thick target Source term up to 3000 MeV electrons.(solid and dash line represent the Swanson's curve, data points are from present work).
Fig. 3. Measured and simulated absorbed dose rate for thick target due to 550 MeV electron beam (Gy/h-kW).
Table 1 Thick target source term obtained for various electron energies from EGSnrc code and empirical relation. Electron energy (MeV)
450 550 1000 2000 3000
Absorbed dose rate at 1 m(Gy- m2/h-kW)
Ratio
EGSnrc
Empirical relation (Swanson W.P.)
EGSnrc/empirical relation
1.07 105 1.32 105 1.63 105 2.73 105 3.61 105
1.35 1005 1.65 1005 3.00 1005 6.00 1005 9.00 1005
0.80 0.80 0.54 0.45 0.40
thick Stainless Steel and 30 mm thick Aluminum, respectively). For thin target depth dose simulations, only change made in the geometry is in the target material and thickness. 4. Results and discussion The depth dose curves per unit beam power obtained from
experiments with 450 MeV and 550 MeV electron energies and the simulated data are shown in Figs. 2 and 3, respectively. The thick target source term in forward direction is calculated from the measured depth dose curve as the maximum dose within water phantom per unit electron beam power incident on the target and normalized to 1 m distance. The measured and simulated maximum dose rate for 450 MeV electrons are 1.29 105 Gy/h-kW and 1.32 105 Gy/h-kW respectively at 5 cm depth. The values are in agreement within 72.3%. For 550 MeV, the dose rate obtained from both experiment and simulation is 1.52 105 Gy/h-kW at depth of 9 cm, in excellent agreement. Experimentally observed thick target source term (dose rate normalized to 1 m per unit beam power) is 1.05 105 Gy-m2/h-kW for 450 MeV and 1.32 105 Gy-m2/h-kW for 550 MeV. The source terms vary approximately as the ratio of the electron energies. Since the results of experiment and simulation are found to be in very good agreement at 450 MeV and 550 MeV, source term is simulated up to 3000 MeV electrons using similar geometry (shown in Fig. 1). The results are given in Table 1. The thick target source term data obtained from the present study is plotted in Fig. 4 and compared with Swanson’s suggested curve. The solid line shows the source Table 2 Thin target source term for Indus-1 and Indus-2 and compared with thick target source term. Electron energy (MeV)
Indus-1 (450 MeV) Indus-2 (2500 MeV)
Absorbed dose rate at 1 m(Gy- m2/h-kW) Thin target (EGSnrc) Thick target (Swanson W.P.)
EGSnrc/empirical relation (Swanson)
2.71 106
9.45 104
28.73
6
5
14.61
5.48 10
3.75 10
4
M.K. Nayak et al. / Radiation Physics and Chemistry 113 (2015) 1–5
Fig. 5. Measured and simulated absorbed dose rate from thin target (2 mm stainless steel target) for 450 MeV electron beam (Gy/h-kW).
in the case of thin target in comparison to a thick target (29 times in case of 450 MeV and 15 times in case of 2500 MeV approximately). This is due to the fact that in the case of a thin target, substantial electrons are transmitted out with a small reduction in energy along with high energy photons (Haridas et al., 2006). To explain the higher source term in the case of thin target, the simulated photon fluence coming out from thick and thin target is shown in Fig. 6. The fluence of high energy photons are found to be more for the thin target (0.2 cm). These photons and transmitted electrons incident on the phantom will travel deeper inside the water phantom and generate more cascade electrons and positrons. This will subsequently give rise to higher source term, in comparison to thick target. Thus in the case of estimation of shielding requirements of high energy electron accelerators, Swanson's empirical relations underestimate the source term and hence additional shield needs to be provided to counter the increased source term on account of thin target actually encountered. Otherwise, the source has to be made thick by providing additional thickness of suitable material close to the vacuum chamber. As the energies used for the experiments (450 MeV and 550 MeV) are above the photo-neutron threshold, neutron production is possible. Majority of photo-neutrons emitted are fast neutrons (Swanson, 1979) and hence its effect on the dose recorded in CaSO4: Dy TLD is evaluated. The response of CaSO4: Dy for fast neutrons is reported to be negligible (Weng and Chen, 1974; Pradhan et al. 2002) and for thermal neutrons it is 3.8 mGy (equivalant to Co-60 gamma ray) for a neutron fluence of 1010 n/cm2 (Vohra et al., 1980; Pradhan et al., 2002). Since the response of CaSO4: Dy TLD to fast neutron is negligible and thermal neutrons is very low, the contribution of neutron dose in the experimental TLD data is neglected.
5. Conclusion
Fig. 6. Photon fluence from a thick (25 mm lead) and a thin (2 mm stainless steel) target when 450 MeV electrons is incident on it (EGSnrc data).
term data compiled by Swanson based on experimental data up to 100 MeV and dashed line represents extrapolation beyond 100 MeV. In the figure unfilled circles indicate the data on thick target source term based on the present experiments at 450 MeV and 550 MeV. The solid circles show the simulated source term. The present results indicate that beyond 450 MeV, source term is not following the trend as suggested by Swanson. It shows a decreasing trend in comparison with Swanson’s suggested source term, as the energy is increased beyond 450 MeV–3000 MeV. An empirical relation obtained from the fit of the source term data from the present work in the energy range, 450 MeV–3000 MeV is given in Eq. (1).
Source term ⎡⎣Gy−m2 /h/kW⎤⎦ = 98.3E + C
Thick target source term in forward direction is experimentally determined for 450 and 550 MeV electrons at booster synchrotron of Indus synchrotron facility, RRCAT, India. These are also simulated using EGSnrc code and are found to be in very good agreement with experimental results. Subsequently thick target source term data is extended up to 3000 MeV. An empirical relation is obtained from the fit of the data for thick target source term beyond 450 MeV. Source term for thin target has also been determined experimentally for 450 MeV and compared with the simulated data. On agreement of the data at 450 MeV, simulation for thin target was carried out for 2500 MeV and the source term is determined. From the studies it is concluded that The thick target source terms is found to be not following the trend as suggested by Swanson’s empirical relations beyond about 450 MeV. The shielding thickness for electron accelerator based on thick target bremsstrahlung source term, may underestimate the thickness as the target is not actually thick in reality and give rise to higher source term. Care has to be taken in avoiding under shielding of such high energy electron accelerator facilities on account of the thin targets actually encountered.
(1)
where C is a constant (6.95 104) and E is the electron energy in MeV Table 2 shows the calculated thin target source term for Indus-1 (450 MeV) and Indus-2 (2500 MeV). Comparison of the simulated data for Indus-1 with experiment is found to match very well (Fig. 5). From the results we see that the source term is higher
Acknowledgments We are highly thankful to Dr. P.D.Gupta, Director, Raja Ramanna Centre for Advance Technology (RRCAT), for providing all necessary support for carrying out the present investigations. The authors are thankful to Indus Operation Crew for providing us the
M.K. Nayak et al. / Radiation Physics and Chemistry 113 (2015) 1–5
beam at target for experiments. We are indebted to Dr. A.K.Sinha, ISUD, RRCAT for critical review of the manuscript. We are also thankful to Shri K.K.Thakkar for encouragement in carrying out the work. Thanks are due to Dr. Palani Selvam, RP&AD, BARC for providing help in the use of EGSnrc code. The authors are also thankful to Shri Lokesh Babbar, Beam Diagnostic Section, RRCAT for providing help in measurement.
References Fasso, A., Goebel, K., et.al., 1984. Radiation problems in the design of the large electron-positron collider (LEP). CERN 84-02. Ferrari, A., Pelliccioni, M., Sala, P.R., 1993. Bremsstrahlung source terms for intermediate energy electron accelerators. Nucl. Instrum. Methods B 82 (1), 32–38. Haridas, G., Vipin, Dev, M.K., Nayak, et al., 2006. Determination of dose build-up thickness for absorbed dose measurement in high energy electron-photon radiation at electron storage rings. Radiat. Prot. Dosim. 121 (2), 92–98. ICRU, 1978. Basic Aspects of High Energy Particle Interactions and Radiation Dosimetry, ICRU Report No. 28.
5
ICRU, 1984. Stopping Powers for Electrons and Positrons, ICRU Report No. 37. National Council on Radiation Protection and Measurements, 2003. Radiation Protection for Particle Accelerator Facilities, NCRP Report No. 144. Pradhan, A.S., Adtani, M.M., Varadharajan, G., Bakshi, A.K., 2002. Hand book on the Use of TLD Badge Based on CaSO4: Dy Teflon TLD Discs for Individual Monitoring, BARC Report No. BARC/2002/E/025. Swanson, W.P., 1979. Radiological Safety Aspects of the Operation of Electron Linear Accelerators, IAEA Technical Report No.188 . Tromba, G., Rindi, A., Fabretto, M., 1990. The shilelding of electron accelerators: Monte Carlo evaluation of gamma source terms. In: Proceedings of European Particle Accelerator Conférence, Nice, France . Vohra, K.G., Bhatt, R.C., Chandra, Bhuwan, Pradhan, A.S., Lakshmanan, A.R., Shastry, S.S., 1980. A Personnel Dosimetry TLD Badge based on CaSO4: Dy Teflon TLD Discs. Health Phys. 38, 193–196. Weng, P.S., Chen, K.M., 1974. Response of CaSO4(Dy) phosphor to neutrons. Nucl. Instrum. Methods 117, 89–92. Wyckoff, J.M., Pruitt, J.S., Svensson, G., 1971. Dose vs. angle and depth produced by 20 to 100 MeV electrons incident on thick targets. In: Proceedings of the International Congress on Protection Against Accelerators and Space Radiation, Geneva, 26–30 April 1971, Report CERN 71-16, Vol. 2, European Organization for Nuclear Research, Geneva, 773.
Contents lists available at ScienceDirect
Radiation Physics and Chemistry journal homepage: www.elsevier.com/locate/radphyschem
Bremsstrahlung source term estimation for high energy electron accelerators M.K. Nayak a,n, T.K. Sahu a, H.G. Nair a, R.V. Nandedkar b, Tapas Bandyopadhyay a, R.M Tripathi a, P.R. Hannurkar c, D.N Sharma d a
Health Physics Division, Bhabha Atomic Research Centre, Mumbai 400085, India Raja Ramanna Centre for Advanced Technology, Indore 452013, India c Indus Operations and Accelerator Physics Design Division, Raja Ramanna Centre for Advanced Technology, Indore 452013, India d Health, Safety & Environment Group, Bhabha Atomic Research Centre, Mumbai 400085, India b
H I G H L I G H T S
Experimental determination of bremsstrahlung source term at 450 and 550 MeV electrons. Monte Carlo calculations performed for validation of experimental data. Thick and thin target bremsstrahlung source term is studied. Brensstrahlung Source term is determined up to 3 GeV electron energies.
art ic l e i nf o
a b s t r a c t
Article history: Received 28 January 2015 Received in revised form 9 April 2015 Accepted 11 April 2015 Available online 16 April 2015
Thick target bremsstrahlung source term for 450 MeV and 550 MeV electrons are experimentally determined using booster synchrotron of Indus facility at Raja Ramanna Centre for Advanced Technology, Indore, India. The source term is also simulated using EGSnrc Monte Carlo code. Results from experiment and simulation are found to be in very good agreement. Based on the agreement between experimental and simulated data, the source term is determined up to 3000 MeV by simulation. The paper also describes the studies carried out on the variation of source term when a thin target is considered in place of a thick target, used in earlier studies. & 2015 Elsevier Ltd. All rights reserved.
Keywords: Source term Depth dose Bremsstrahlung Thick target Thin target
1. Introduction Bremsstrahlung source term for accelerators is usually expressed as dose equivalent rate or absorbed dose rate per unit beam power at 1 m distance from a high atomic number (Z) thick target (Ferrari et al., 1993). Estimation of source term is important for evaluation of shielding requirement for particle accelerator facilities. The most widely used source terms for electron accelerators are those suggested by Swanson (1979). It is based on experimental data obtained from the measurement of absorbed dose rate in a 30 cm tissue equivalent phantom. The bremsstrahlung radiation for these experiments were generated by bombarding high energy electrons on high Z thick target. For thick targets, the target thickness considered is equivalent to the range n
Corresponding author. E-mail address: [email protected] (M.K. Nayak).
http://dx.doi.org/10.1016/j.radphyschem.2015.04.004 0969-806X/& 2015 Elsevier Ltd. All rights reserved.
of electrons in the medium (Wyckoff et al., 1971). Experimentally obtained maximum absorbed dose data for 20, 30 and 100 MeV electron energies in a polymethyl methacrylate (PMMA) phantom generated by Wyckoff et al. was used by Swanson in formulating a semi-empirical relation for thick target bremsstrahlung source term. The empirical relations suggested for forward and lateral source terms are available up to 100 MeV. For energies in the range 20–100 MeV, the suggested relations for thick target source term in the forward direction are found to be linear in energy. Whereas, for higher energies the relations are assumed to be linear (Swanson,1979; NCRP, 2003). When an accelerated electron beam moving in vacuum envelope is lost, it encounters a thin target of few millimeter thickness, instead of a thick target, which has been studied earlier. For thin targets, generated bremsstrahlung photons in the forward direction are of higher energies compared with those of thick targets (Haridas et al., 2006). Bremsstrahlung photons generated by thin
2
M.K. Nayak et al. / Radiation Physics and Chemistry 113 (2015) 1–5
targets further generate electromagnetic shower (ICRU, 1978) when incident on shield structures. Moreover as the thickness of vacuum chamber is less, substantial electrons are also transmitted along with bremsstrahlung photons and the source term is different from that obtained from the empirical relations for the thick target case. The absorbed dose rate from tungsten target at 200 MeV electrons incident as a function of angle of emission of bremsstrahlung and thickness calculated by Fasso. et al. (1984), shows that as the thickness of the target is reduced, the absorbed dose rate increases at low emission angles (forward direction). For larger emission angles, the dose rate is higher from thick targets. Thus the target thickness plays an important role in source term determination. The forward source term for 100 MeV–1000 MeV electrons have been calculated using Monte Carlo simulation code FLUKA by Ferrari. et al. (1993). It has been found that the forward source term is higher by 7 times at 100 MeV and 22 times at 1000 MeV when compared with the values obtained by Swanson’s empirical relations. Tromba et al. (1990) also proposed empirical relation for the thick target source term using Monte Carlo calculations performed with EGS4 code for electron energies from 100 MeV up to 10 GeV. In the present work, thick target source term is experimentally determined from the depth dose data for 450 MeV and 550 MeV electrons in the forward direction on booster synchrotron at Indus synchrotron facility of Raja Ramanna Centre for Advanced Technology (RRCAT), Indore. The experimental data is also simulated using EGSnrc code and is found to be in excellent agreement with the experimental results. Subsequently, the source term data is simulated for higher energies up to 3000 MeV. Source term simulations are also performed for thin targets at two electron energies, 450 MeV1 and 2500 MeV1. The simulated data for 450 MeV is experimentally verified at the booster synchrotron. The thin target source terms are found to be higher than the thick target source term obtained from empirical relations by Swanson.
2. Experimental details 2.1. Materials and methods CaSO4: Dy TLD disks of size 6 mm diameter and 1 mm thickness used for the experiments are annealed for 1 h at 400 °C in air. The sensitivity of these TLDs is checked by exposing the disks uniformly to a known dose of 5 mGy from a Co-60 source. TLDs within 7 3% sensitivity are selected and used for these experiments. A cuboid water phantom of size 30 cm 30 cm 30 cm made from 10 mm thick Perspex sheet (with top open) is used for measuring the depth dose in water. The phantom has been placed 1 m away from a lead target in the forward direction with TLDs placed at different depths at the mid plane of the phantom. It is ensured that the geometrical center of the target, the center of the water phantom and the axis of the TL disks within the phantom are in a straight line. The schematic diagram of the experimental setup for the measurement of forward source term is shown in Fig. 1.
Fig. 1. Experimental set up and geometry used for simulation.
aluminum) window. The extracted beam is allowed to fall on the center of the lead target of 25 mm2 (1 CSDA range of 450 MeV electrons) (ICRU,1984). The position of beam at the center of the target is ensured by observing the fluorescence, emitted from a fluorescence screen placed before the target, with the help of a charge-coupled device (CCD) camera. The TLDs are exposed for a period of 900 s (the time of exposure was decided based on a preliminary dose rate measurement with ion chamber). The circulating beam current in the booster synchrotron is 1.5 mA. During the exposure period, the beam pulse from the booster synchrotron is measured with the help of a Wall Current Monitor (WCM) which gave a pulse height of 25 mV/beam pulse. Two pulses per second are fired on the target. From the pulse height data, the extracted electrons per second is calculated to be 3.26 108. Similar experiments are repeated for 550 MeV electrons with a 26.3 mm (1CSDA range) lead target and dose is measured with a new set of TLDs. The extracted electrons per second obtained from the WCM data was 1.66 108. The experiments for thin target source term for 450 MeV electrons are also carried out by allowing the electrons to be incident on a 2 mm Stainless Steel (SS) target close to the vacuum chamber thickness of Indus-1 storage ring. TLDs are exposed within the water phantom for 600 s. The exposed TLDs for each set of experiments are analyzed using a TLD reader (Laboratory Reader-Analyzer RA'94, RADPRO, POLAND). From the TL counts the relative dose per unit beam power with respect to Co-60 is found out.
3. Monte Carlo simulations
The electron beam at 450 MeV from the booster synchrotron is extracted to a transport line via an extraction septum and the electron beam is allowed to come out through a thin (1 mm
The experimental results obtained are simulated using EGSnrc Monte Carlo code. The geometry for the experimental setup (Fig. 1) is made using DOSERZ user code from EGSnrc. In the simulation of depth dose curve, a parallel beam of 3.5 mm radius (which is the electron beam radius) of 450 MeV is allowed to incident on 25 mm thick lead target and absorbed dose per incidence fluence was scored in the TLD disks (scoring area 0.363 cm2), placed in different depths of the water phantom. For radiation transport, an electron cut off (Ecut) of 521 keV and a photon cut-off (Pcut) of 10 keV are used in the simulation with a history of 106. Simulation for 550 MeV electron case is carried out by changing the lead target thickness to 26.3 mm. The statistical accuracy obtained in the present simulation is within 71%. The simulations of depth dose in phantom with thin targets (2 mm Stainless Steel and 30 mm Aluminum) are also performed for calculating the source term for Indus-1 and Indus-2 storage rings (vacuum chamber of Indus-1 and Indus-2 are made of 2 mm
1 The 450 MeV and 2500 MeV are the electron beam energies in the storage rings, Indus-1 and Indus-2 respectively at RRCAT, Indore.
2 It is equivalent to 4.5 radiation length, which is more than enough to develop electromagnetic cascade fully in the lead.
2.2. Experiments
M.K. Nayak et al. / Radiation Physics and Chemistry 113 (2015) 1–5
3
Fig. 2. Measured and simulated absorbed dose rate for thick target due to 450 MeV electron beam (Gy/h-kW).
Fig. 4. Thick target Source term up to 3000 MeV electrons.(solid and dash line represent the Swanson's curve, data points are from present work).
Fig. 3. Measured and simulated absorbed dose rate for thick target due to 550 MeV electron beam (Gy/h-kW).
Table 1 Thick target source term obtained for various electron energies from EGSnrc code and empirical relation. Electron energy (MeV)
450 550 1000 2000 3000
Absorbed dose rate at 1 m(Gy- m2/h-kW)
Ratio
EGSnrc
Empirical relation (Swanson W.P.)
EGSnrc/empirical relation
1.07 105 1.32 105 1.63 105 2.73 105 3.61 105
1.35 1005 1.65 1005 3.00 1005 6.00 1005 9.00 1005
0.80 0.80 0.54 0.45 0.40
thick Stainless Steel and 30 mm thick Aluminum, respectively). For thin target depth dose simulations, only change made in the geometry is in the target material and thickness. 4. Results and discussion The depth dose curves per unit beam power obtained from
experiments with 450 MeV and 550 MeV electron energies and the simulated data are shown in Figs. 2 and 3, respectively. The thick target source term in forward direction is calculated from the measured depth dose curve as the maximum dose within water phantom per unit electron beam power incident on the target and normalized to 1 m distance. The measured and simulated maximum dose rate for 450 MeV electrons are 1.29 105 Gy/h-kW and 1.32 105 Gy/h-kW respectively at 5 cm depth. The values are in agreement within 72.3%. For 550 MeV, the dose rate obtained from both experiment and simulation is 1.52 105 Gy/h-kW at depth of 9 cm, in excellent agreement. Experimentally observed thick target source term (dose rate normalized to 1 m per unit beam power) is 1.05 105 Gy-m2/h-kW for 450 MeV and 1.32 105 Gy-m2/h-kW for 550 MeV. The source terms vary approximately as the ratio of the electron energies. Since the results of experiment and simulation are found to be in very good agreement at 450 MeV and 550 MeV, source term is simulated up to 3000 MeV electrons using similar geometry (shown in Fig. 1). The results are given in Table 1. The thick target source term data obtained from the present study is plotted in Fig. 4 and compared with Swanson’s suggested curve. The solid line shows the source Table 2 Thin target source term for Indus-1 and Indus-2 and compared with thick target source term. Electron energy (MeV)
Indus-1 (450 MeV) Indus-2 (2500 MeV)
Absorbed dose rate at 1 m(Gy- m2/h-kW) Thin target (EGSnrc) Thick target (Swanson W.P.)
EGSnrc/empirical relation (Swanson)
2.71 106
9.45 104
28.73
6
5
14.61
5.48 10
3.75 10
4
M.K. Nayak et al. / Radiation Physics and Chemistry 113 (2015) 1–5
Fig. 5. Measured and simulated absorbed dose rate from thin target (2 mm stainless steel target) for 450 MeV electron beam (Gy/h-kW).
in the case of thin target in comparison to a thick target (29 times in case of 450 MeV and 15 times in case of 2500 MeV approximately). This is due to the fact that in the case of a thin target, substantial electrons are transmitted out with a small reduction in energy along with high energy photons (Haridas et al., 2006). To explain the higher source term in the case of thin target, the simulated photon fluence coming out from thick and thin target is shown in Fig. 6. The fluence of high energy photons are found to be more for the thin target (0.2 cm). These photons and transmitted electrons incident on the phantom will travel deeper inside the water phantom and generate more cascade electrons and positrons. This will subsequently give rise to higher source term, in comparison to thick target. Thus in the case of estimation of shielding requirements of high energy electron accelerators, Swanson's empirical relations underestimate the source term and hence additional shield needs to be provided to counter the increased source term on account of thin target actually encountered. Otherwise, the source has to be made thick by providing additional thickness of suitable material close to the vacuum chamber. As the energies used for the experiments (450 MeV and 550 MeV) are above the photo-neutron threshold, neutron production is possible. Majority of photo-neutrons emitted are fast neutrons (Swanson, 1979) and hence its effect on the dose recorded in CaSO4: Dy TLD is evaluated. The response of CaSO4: Dy for fast neutrons is reported to be negligible (Weng and Chen, 1974; Pradhan et al. 2002) and for thermal neutrons it is 3.8 mGy (equivalant to Co-60 gamma ray) for a neutron fluence of 1010 n/cm2 (Vohra et al., 1980; Pradhan et al., 2002). Since the response of CaSO4: Dy TLD to fast neutron is negligible and thermal neutrons is very low, the contribution of neutron dose in the experimental TLD data is neglected.
5. Conclusion
Fig. 6. Photon fluence from a thick (25 mm lead) and a thin (2 mm stainless steel) target when 450 MeV electrons is incident on it (EGSnrc data).
term data compiled by Swanson based on experimental data up to 100 MeV and dashed line represents extrapolation beyond 100 MeV. In the figure unfilled circles indicate the data on thick target source term based on the present experiments at 450 MeV and 550 MeV. The solid circles show the simulated source term. The present results indicate that beyond 450 MeV, source term is not following the trend as suggested by Swanson. It shows a decreasing trend in comparison with Swanson’s suggested source term, as the energy is increased beyond 450 MeV–3000 MeV. An empirical relation obtained from the fit of the source term data from the present work in the energy range, 450 MeV–3000 MeV is given in Eq. (1).
Source term ⎡⎣Gy−m2 /h/kW⎤⎦ = 98.3E + C
Thick target source term in forward direction is experimentally determined for 450 and 550 MeV electrons at booster synchrotron of Indus synchrotron facility, RRCAT, India. These are also simulated using EGSnrc code and are found to be in very good agreement with experimental results. Subsequently thick target source term data is extended up to 3000 MeV. An empirical relation is obtained from the fit of the data for thick target source term beyond 450 MeV. Source term for thin target has also been determined experimentally for 450 MeV and compared with the simulated data. On agreement of the data at 450 MeV, simulation for thin target was carried out for 2500 MeV and the source term is determined. From the studies it is concluded that The thick target source terms is found to be not following the trend as suggested by Swanson’s empirical relations beyond about 450 MeV. The shielding thickness for electron accelerator based on thick target bremsstrahlung source term, may underestimate the thickness as the target is not actually thick in reality and give rise to higher source term. Care has to be taken in avoiding under shielding of such high energy electron accelerator facilities on account of the thin targets actually encountered.
(1)
where C is a constant (6.95 104) and E is the electron energy in MeV Table 2 shows the calculated thin target source term for Indus-1 (450 MeV) and Indus-2 (2500 MeV). Comparison of the simulated data for Indus-1 with experiment is found to match very well (Fig. 5). From the results we see that the source term is higher
Acknowledgments We are highly thankful to Dr. P.D.Gupta, Director, Raja Ramanna Centre for Advance Technology (RRCAT), for providing all necessary support for carrying out the present investigations. The authors are thankful to Indus Operation Crew for providing us the
M.K. Nayak et al. / Radiation Physics and Chemistry 113 (2015) 1–5
beam at target for experiments. We are indebted to Dr. A.K.Sinha, ISUD, RRCAT for critical review of the manuscript. We are also thankful to Shri K.K.Thakkar for encouragement in carrying out the work. Thanks are due to Dr. Palani Selvam, RP&AD, BARC for providing help in the use of EGSnrc code. The authors are also thankful to Shri Lokesh Babbar, Beam Diagnostic Section, RRCAT for providing help in measurement.
References Fasso, A., Goebel, K., et.al., 1984. Radiation problems in the design of the large electron-positron collider (LEP). CERN 84-02. Ferrari, A., Pelliccioni, M., Sala, P.R., 1993. Bremsstrahlung source terms for intermediate energy electron accelerators. Nucl. Instrum. Methods B 82 (1), 32–38. Haridas, G., Vipin, Dev, M.K., Nayak, et al., 2006. Determination of dose build-up thickness for absorbed dose measurement in high energy electron-photon radiation at electron storage rings. Radiat. Prot. Dosim. 121 (2), 92–98. ICRU, 1978. Basic Aspects of High Energy Particle Interactions and Radiation Dosimetry, ICRU Report No. 28.
5
ICRU, 1984. Stopping Powers for Electrons and Positrons, ICRU Report No. 37. National Council on Radiation Protection and Measurements, 2003. Radiation Protection for Particle Accelerator Facilities, NCRP Report No. 144. Pradhan, A.S., Adtani, M.M., Varadharajan, G., Bakshi, A.K., 2002. Hand book on the Use of TLD Badge Based on CaSO4: Dy Teflon TLD Discs for Individual Monitoring, BARC Report No. BARC/2002/E/025. Swanson, W.P., 1979. Radiological Safety Aspects of the Operation of Electron Linear Accelerators, IAEA Technical Report No.188 . Tromba, G., Rindi, A., Fabretto, M., 1990. The shilelding of electron accelerators: Monte Carlo evaluation of gamma source terms. In: Proceedings of European Particle Accelerator Conférence, Nice, France . Vohra, K.G., Bhatt, R.C., Chandra, Bhuwan, Pradhan, A.S., Lakshmanan, A.R., Shastry, S.S., 1980. A Personnel Dosimetry TLD Badge based on CaSO4: Dy Teflon TLD Discs. Health Phys. 38, 193–196. Weng, P.S., Chen, K.M., 1974. Response of CaSO4(Dy) phosphor to neutrons. Nucl. Instrum. Methods 117, 89–92. Wyckoff, J.M., Pruitt, J.S., Svensson, G., 1971. Dose vs. angle and depth produced by 20 to 100 MeV electrons incident on thick targets. In: Proceedings of the International Congress on Protection Against Accelerators and Space Radiation, Geneva, 26–30 April 1971, Report CERN 71-16, Vol. 2, European Organization for Nuclear Research, Geneva, 773.