Document Sample


Test of Spectrometers and Dosemeters for the Investigation of the
Radiation Environment onboard Spacecraft and around High-
Energy Accelerators

DATE      2006-10-03

Günther Reitz, Deutsches Zentrum für Luft- und Raumfahrt (DLR), Cologne,
Germany, e-mail:

Thomas Berger, Deutsches Zentrum für Luft- und Raumfahrt (DLR), Cologne,
Les Bennet, Michelle Boudreau, Royal Military College of Canada, Kingston, Canada
Martin Smith, Bubble Technologies, Chalk River, Canada
Burkhard Wiegel, Marlies Luszik-Bhadra, Stefan Röttger, Ralf Nolte,
Physikalisch-Technische Bundesansalt (PTB), Braunschweig, Germany
Torsten Radon, Gesellschaft für Schwerionenforschung (GSI) Darmstadt, Germany
F.D. Smit, Zinhle Buthelezi, TLABS, Cape Town, South Africa
A. Buffler, UCT, Cape Town, South Africa

        Neutron fields encountered around high-energy accelerators and on board spacecraft
are both characterised by a broad spectral distribution extending from thermal energies to
several hundred MeV. Therefore they pose a considerable challenge for appropriate
monitoring of this component for radiation protection purposes on ground and in orbit. This is
of particular importance for new accelerators like the upcoming FAIR project in Germany,
which are designed to deliver much higher currents at higher energies and for the
International Space Station (ISS) where, due to the higher mass, the neutron exposure of the
astronauts increased compared to former missions. In addition, the duration of missions was
recently increased to more than 180 days.
        The proposed project focuses on the experimental characterisation of instruments
which are either designed for the investigation of the neutron energy distribution in these
fields or used for routine monitoring of the exposure of personnel. Due to the deficiencies of
nuclear models in the medium energy range around and above 200 MeV, an experimental
verification of calculated response functions and detection efficiencies using quasi-
monoenergetic neutrons is still mandatory for reliable measurements. Energies around
100 MeV are of special interest for space applications since the neutron field present in the
ISS has two peaks in the evaporation region and at 100 MeV. The unique features of the
TLABS neutron beam are therefore ideal to achieve both goals.


(a)     Pure basic research                                                               30----------------- %

(b)     Directed basic research                                                           30----------------- %

(c)     Applied research                                                                  30----------------- %

(d)     Experimental development, services                                                10------------------ %

        Around accelerators and in spacecraft, complex radiation fields are produced by the
interaction of primary charged particles with material structures. In the case of high-energy
accelerators, the shielding around experimental areas usually prevents charged particles to
reach positions where they could cause a direct radiation hazard to personnel. However,
when interacting with structural parts like beam tubes, dumps or shielding walls, the primary
charged particles create secondary neutrons with energies of up to a few hundred MeV. To
keep the radiation exposure of personnel below acceptable limits even in case of beam
losses, shielding of considerable thickness is required, in particular since new accelerator
projects like FAIR at GSI aim at much higher beam currents than older machines. This is why
optimisation of shielding is a question of radiation protection as well as of cost effectiveness
which is a key issue for such projects.

                                                           FLUKA prediction
                                              70           NEMUS solution

                ΦE (En) ⋅ En / (cm 10 ions)






                                                -9    -8     -7    -6    -5    -4    -3     -2    -1    0    1     2    3
                                              10     10    10     10    10    10    10    10     10    10   10    10   10
                                                                              En / MeV ⎯→

Fig. 1 Results of a shielding experiment carried out at GSI. The primary source is a 400 MeV/u 12C beam incident
on a thick C target. The neutron spectrum behind a thick concrete shield was measured with the PTB Bonner
Sphere Spectrometer NEMUS and calculated with the Monte Carlo code FLUKA. Strong deviations are observed
in the so-called ‘spallation peak’ around 100 MeV.

        Monte Carlo codes such as FLUKA, MCNPX or PHITS (presently under
development) are used as much as possible for this task. However, as demonstrated in
Fig. 1, significant deviations still exist between Monte Carlo calculations and the results of
shielding experiments. Therefore shielding is generally designed very conservative with
safety factors of up to two. This demonstrates that benchmark experiments for thick shielding

is still necessary to improve the existing codes. Such experiments require the measurement
of secondary neutron spectra extending over about nine orders of magnitude from thermal
energies up to several hundred MeV. The instrument best suited for such a task is the
Bonner sphere spectrometer (see below).
        In contrast to accelerators, the radiation fields encountered by crew members
onboard spacecraft has a dominating component of charged particles. In low earth orbits
(LEO), galactic cosmic rays (97% protons, 2% alpha particles and 1% heavier ions), protons
and electrons trapped in the Van Allen radiation belts as well as protons ejected during solar
particle events (SPE) are the major components of the radiation field. Neutrons encountered
in LEO are produced in interactions of these particles with the spacecraft structure and in
interactions of the galactic cosmic rays with the earth’s atmosphere (albedo neutrons).
        The radiation field around the International Space Station (ISS) is changing
dynamically depending on the position of the space station, the solar cycle and sporadic
solar particle events. Measurements inside the ISS were mostly carried out for the charged
particle component, resulting in dose equivalent values around 500 µSv/d depending on the
location inside the ISS. However, insufficient attention has been given in the past to the
contribution of neutrons to the dose equivalent. Only results of a few measurements with
neutron spectrometers are available. Measurements with Bonner spheres equipped with
activation foil detectors made near solar maximum in low inclination with high altitude
resulted in a dose equivalents ranging from 45 μSv/d to 345 μSv/d depending on the orbit
inclination and altitude [1]. Comparison of these measurements with calculations showed that
an additional dose equivalent (about 50 percent) had not been detected because the energy
range of the Bonner sphere spectrometer was limited to about 15 MeV by the maximum size
of the moderator spheres used. In addition, the spectral distribution measured by Bonner
spheres and the calculated ones showed considerable deviations.
        Recently, new measurements were performed by a Japanese group onboard the ISS
in 1998 [2] and in 2001 with a Bonner sphere spectrometer using active detectors but with
the same limitations in energy range. The neutron dose equivalent for neutrons with energies
below 15 MeV resulted in 85 μSv/d and changed by about 30% by a small relocation of the
device to 109 μSv/d. Addition of the part at higher neutron energies which still has not been
measured adequately results in a neutron component which considerably contributes to the
dose received by astronauts and which can change considerably with the amount of
shielding inside the space station. Neutron measurements using the electrochemical etching
technique for the nuclear track detector (CR39) revealed similar results. Calculations of the
neutron dose show some improvements, but suffer from a still limited knowledge of cross
sections and approximate shielding distributions for selected locations inside the ISS.
        This deficiency in neutron spectrometry motivated several groups from space agency
laboratories world-wide to develop neutron spectrometers for application in space. Because
of the more compact design, most instruments are based on scintillation detectors.
        As indicated above, astronauts in long term missions in LEO by far exceed the annual
dose limits applied for radiation workers on earth and can reach or even exceed dose limits
specified in the ‘Radiation protection guidance for activities in low-earth orbit’ [3]. The
assessment of the exposure therefore requires a higher accuracy than usually required for
radiation protection on earth. This can be achieved by measurements done at special
locations inside the ISS with passive and active detectors indicating the radiation quality (CR-
39, TLD, TEPC). These measurements can than be used to benchmark model calculations.
It was found in these studies that the details of the vehicle geometry and materials as well as
the detector response had to be accurately modelled to relate the measured data to
computed instrument responses based on calcuated neutron fluence rates at the detector
location within the vehicle. Over-simplifications of the details resulted in poor agreement.
This resulted in several comparison programs, where devices were irradiated together (partly
also combined with radiobiological investigations) with good documentation of their positions
and shielding conditions.

       The present proposal therefore asks for beam time to study the performance of
several instruments which were constructed for the characterisation of neutron fields either
onboard spacecraft or around high-energy accelerators using the quasi-monoenergetic 100
MeV and 200 MeV neutron beams at TLABS. Details on the instrument as well as on the
proposed measurements are given below.


Dosemeters for the MATROSHKA project at the ISS
        For the assessment of the effective dose, which is the quantity related to radiation
risk, organ doses have to be determined. This is the objective of the MATROSHKA
experiments. MATROSHKA is a facility of the European Space Agency (ESA). The scientific
programme was defined as a bilateral undertaking between ESA and IMBP, Moscow, with
German Aerospace Center (DLR) as scientific lead [4]. Seventeen international research
organisations, space agencies and universities from Japan, USA and Europe participate with
active and passive dosimetry systems in the scientific payload.

Fig. 2 The MATROSHKA experiment on ISS. The first picture from the left shows the head and in the lower body
region of the phantom with the inserted detectors. The TLD’s are inserted in the tubes. The rectangular packages
are active sensors (with cable) and a passive detector packages. The second picture shows the phantom dressed
with a poncho and a hood carrying also TLD’s and passive detector packages for skin measurements. In front of
the phantom the TEPC can be seen. The third figure represents the closed facility and the right figure shows the
facility dressed with multilayer insulation and equipped with cables for power supply and data transfer.

        The MATROSHKA facility consists of a human phantom torso, mounted on a base
structure and covered by a carbon fibre container. Passive detector packages (plastic
nuclear track detectors (PNTD), thermoluminescence detectors (TLD)) and active sensors
(silicon detectors and scintillator cubes with anticoincidence and a tissue equivalent
proportional counter (TEPC)) are placed into and around the phantom to provide the required
data, such as particle and LET spectra, dose and dose rates and neutron fluence rates and
dose rates. Fig. 2 gives an inside view and a description of the facility.
        MATROSHKA was exposed for 18 month outside ISS at the hull of the Russian
module from February 2004 until August 2005. Since January 2006, MATROSHKA is
operated inside ISS with a new passive detector set. The set is planned to come back at the
end of 2006. The first results were presented at the 11th Workshop of Radiation Monitoring
on the ISS (WRMISS) [5].
       An extensive ground-based programme aimed at the investigation of active and
passive radiation detectors in a simulated space radiation environment is an indispensable
complement of measurements in space. Experiments using charged particle beams were
performed at HIMAC of the National Institute of Radiological Sciences, Chiba, Japan as well

as at NSRL, Brookhaven (heavy ions) and Loma Linda (protons) USA [6,7]. In ten irradiation
campaigns between 2002 and 2005, the responses of all active and passive detectors used
in the MATROSHKA project were determined for a variety of ion beams ranging from 70 MeV
protons to 1 GeV/u 26Fe ions.
        The experiments with charged-particle beams have to be supplemented by
measurements using neutron beams since, as outlined above, a considerable part of the
total dose equivalent is produced by neutrons. In this respect, the quasi-monoenergetic
neutron beams available at TLABS will be important for a determination of the response of
the MATROSHKA instruments, in particular the energy range from 100 MeV to 200 MeV
which was not assessed properly so far. In a first stage, it is planed to investigate the
response of passive detectors used on the MATROSHKA phantom to 100 MeV and 200 MeV
neutrons at TLABS. Later also the active devices and/or the whole phantom should be
investigated in the high energy neutron fields available at TLABS.

Spectrometer for Low-Earth Orbits (CHENSS)
         The Canadian high-energy neutron spectrometry system (CHENSS) was developed
for the Canadian Space Agency (CSA) to measure the neutron flux and energy spectrum in
LEO from a few MeV up to 100 MeV [8].Traditionally, neutron detectors are made of liquid
compounds that are considered to be hazardous cargo for space flights. Therefore, the core
of CHENSS is a detector made of a novel visco-elastic polymer, which will not spill or run if
its container is opened. This material has pulse-shape n/γ discrimination properties similar to
those of the widely-used NE-213 and BC-501A liquid scintillators. The gelled scintillator is
housed in a pressure vessel, 12.7 cm high and 12.7 cm in diameter, providing enough
material to reliably detect 100 MeV neutrons. In order to reject events due to GCR and
trapped protons, which exhibit the same signature as pulses due to neutron-induced recoil
protons, the elastomer is completely surrounded by a box of eight 1-cm-thick plastic
scintillators. Protons from the background interact in both the plastic and gelled scintillators
and can thus be discriminated from neutron-induced events with veto efficiency close to
        Detection of neutrons in the required range of a few MeV to 100 MeV requires a
factor of 300 in light output dynamic range. In order to achieve this, each event is recorded at
three different gain settings (referred to as low-, mid- and high-gain). For each of the three
outputs, the signal amplitude and a shape signal are recorded so that pulse-shape analysis
can be performed. The amplitude is integrated over two ranges (0 - 30 ns and 0 - 300 ns)
and the shape signal is proportional to the difference of the two time integrals. An internal
energy calibration is provided by a 22Na source, mounted in the centre of the scintillating
vessel and two pulsed light-emitting diodes (LED’s) provide signals for relative gain
calibration of the three dynode signals.
        The CHENSS was recently taken to PTB by RMC and Bubble Technologies (BTI) and
irradiated by monoenergetic neutron beams with four incident energies: 2.5 MeV, 5 MeV,
14.8 MeV and 19-MeV. Energy-shape spectra, with background components subtracted, are
presented in Fig. 3. A polygonal coincidence window has been applied to the energy-shape
spectra to isolate neutron events from the γ-ray background. The resulting neutron spectrum
has been unfolded using the response matrix determined for the cylindrical liquid scintillator
discussed in [9] which has the same size as the gelled scintillator used in the CHENSS.
        The calculation of response matrices of organic scintillation detectors becomes more
and more difficult with increasing neutron energy because the contribution of 12C(n,x)
reactions becomes dominant. For the simulation of a detector, the correlation of all charged
particles emitted in an individual reactions has to be modelled since this determines the total
amount of scintillation light produced. It has to be noted that the nuclear data base of
general-purpose transport code like MCNP(X) does not contain these correlation. In the
energy range up to 100 MeV, this task is only partly solved by dedicated codes like
SCINFUL, in particular with respect to the prediction of the full pulse-height response which

is required for CHENSS. Above 100 MeV, the situation is even worse since only very few
semi-empirical extensions of codes originally developed for lower energies are available. The
energy range between 100 MeV and 200 MeV is particularly difficult since here the
intranuclear cascade (INC) model implemented in codes like MCNPX is not yet applicable.
Moreover, this model can only reliably predict emission spectra for A=1 and A=2 while the
more complex particles and the correlation of multiple products are treated insufficiently. This
why an experimental characterisation of the response of CHENSS using the quasi-
monoenergetic 100 MeV and 200 MeV neutron beams available at TLABS is mandatory for
a use of the instrument for neutron spectrometry onboard spacecraft.

Fig. 3 Total energy-shape spectrum, after subtraction of background components, following irradiation with (a) 2.5-
MeV, (b) 5-MeV, (c) 14.8-MeV and (d) 19-MeV neutron beams. It should be noted that panels (a) and (b) show
high-gain spectra whereas panels (c) and (d) show mid-gain spectra.

The PTB Bonner Sphere System NEMUS for Spectrometry in broad high-energy
neutron fields
        Before commercial survey monitors can be used for routine dose measurements in
the complex radiation fields encountered around high-energy accelerators, the spectral
neutron fluence, i.e. the neutron energy distribution has to measured, from which the ambient
dose equivalent is calculated by applying the fluence-to-ambient dose equivalent conversion
coefficients recommended by ICRU. With this method, site-specific corrections factors for the
readings of the survey monitors can be provided.
       The most suitable instrument to measure the neutron spectrum from thermal energies
up to several hundred MeV is the Bonner sphere spectrometer. It consists of a series of
moderator sphere made of polyethylene with a detector (active or passive) sensitive to
thermal neutrons placed in the centre of the spheres. The spectrometer available at PTB
Braunschweig consists of additional spheres with lead and copper shells imbedded in the
polyethylene sphere to increase the response to neutrons with energies above 20 MeV by
neutron multiplication via (n,xn) reactions. This system is known as NEMUS [10] (neutron
multisphere spectrometer), see Fig. 4.

Fig. 4: Components of NEMUS: In the back are five out of 10 moderator spheres made of polyethylene (PE): left
the largest with 30.48 cm diameter, right the smallest with 7.62 cm. In the mid front the 3He gas filled proportional
counter, single and mounted in a PE cylinder. The foreground shows on the left side one of the lead modified
sphere and on the right side the copper modified sphere.

                                 9            3
                                          bare He proportional counter
                                          3" PE sphere to       12" PE sphere
                                          3P5_7 (lead)      4C5_7 (copper)
                                 7        4P5_7 (lead)      4P6_8 (lead)
                Rd(En) / cm ⎯→


                                 0 -9 -8 -7 -6 -5 -4   -3  -2 -1 0 1 2 3  4
                                 10 10 10 10 10 10 10 10 10 10 10 10 10 10
                                                  En / MeV ⎯→

Fig 5: NEMUS response functions. The neutron energy of the maximum response increases with increasing
diameter of the PE spheres. There is an increase in response of the four modified spheres at high neutron
energies due (n, xn) reactions in the lead and copper shells.

       The spectrometer is characterised by the response Rd(En) of spheres with diameter d
as function of neutron energy En, see Fig. 5. The response functions are needed to
determine the neutron spectrum from the readings of each sphere by means of
deconvolution or better known as ‘unfolding’. The response functions were calculated using
the Monte Carlo transport code MCNPX and the calculations were experimentally verified
using radionuclide sources, a thermal field at a nuclear research reactor and the mono-
energetic neutron fields available at PTB (1.2 keV up to 19 MeV) and at the Université
Catholique de Louvain (UCL) at 60 MeV.

        For neutron energies above 60 MeV, the response functions are only based on the
evaluated cross section tables (up to 150 MeV) and on reaction models implemented in the
MCNPX code. The experimental verification of the calculated response with 150 MeV and
200 MeV neutron beams at TLABS is therefore of fundamental interest. A better knowledge
of the response function will reduce the uncertainties in the unfolded neutron spectrum in the
energy range above 60 MeV and as a consequence reduce the uncertainties in the neutron
dose rates calculated from the neutron spectrum.

Beam characterisation for the experiments at TLABS
         The same techniques as used earlier by the PTB group for measuring the neutron
spectrum of the TLABS beams will be employed for the present measurements. A liquid
scintillation detector is used to measure the relative neutron spectrum at low beam intensities
with high TOF resolution and a 238U fission ionisation chamber is used for an ‘absolute’
measurement with lower TOF resolution at the beam currents used for the experiments with
the instruments to be studied. A combination of these two measurements yields a high-
resolution spectral neutron distribution measured relative to the 238U(n,f) reference cross
section. A monitor system consisting of two transmission detectors will be used together with
the integrated beam charge to relate the beam specification and the measurements with the
detectors under study.
       A particular problem arises from the fact that several instruments like most of the
Bonner spheres as well as the CHENSS spectrometer will be larger than the beam size of
about 10 cm x 10 cm. To simulate an homogeneous irradiation, scanning procedures will
have to be applied. For this purpose, the PTB group will use a remotely controlled scanning
machine which can be used for step-by-step as well for continuous Pseudo-Lissajous
scanning schemes.

  [1] J.E. Keith, G.D. Badhwar, D.J. Lindstrom., Neutron spectrum and dose-equivalent in
      shuttle flights during solar maximum, Nucl. Tracks. Radiat. Meas. 20, 41 - 48
  [2] H. Matsumoto, T. Goka, K. Koga, S. Iwai, T. Uehara, O. Sato and S. Takagi Real-time
      measurement of low-energy-range neutron spectra on board the space shuttle STS-89
      (S/MM-8) • Radiation Measurements, Volume 33, Issue 3, June 2001,321-333
  [3] NCRP Report No. 132, Radiation Protection Guidance for Activities in Low-Earth Orbit,
      December 2000, Bethesda, Maryland
  [4] G. Reitz, T. Berger, The MATROSHAK Facility – Dose determination during an EVA,
      Radiat Prot Dosimetry, September 2006; 120: 442 - 445
  [5] Workshop on Radiation Monitoring for the International Space Station
  [6] ICCHIBAN project
 [7] Y.Uchihori, E.R. Benton (Ed). Results from the first Two InterComparison of Dosimetric
      Instruments for Cosmic Radiation with Heavy Ions Beams at NIRS (ICCHIBAN 1 & 2)
      Experiments, NIRS Report HIMAC-078, 2004
 [8] M.B. Smith et. al., Canadian high-energy neutron spectrometry system (CHENSS), in
      proceedings of "International Workshop on Fast Neutron Detectors and Applications,
      University of Cape Town, Cape Town, South Africa, April 2006 PoS(FNDA2006)006,
 [9] N. Nakao et. al., Measurements of response function of organic liquid scintillator for
      neutron energy range up to 135 MeV, Nucl. Instrum. Meth. Phys. Res. A 362, 454 (1995).
[10] B. Wiegel and A.V. Alevra, NEMUS—the PTB Neutron Multisphere Spectrometer:
      Bonner spheres and more Nucl. Instrum. Meth. A 476 36-41 (2002)
[11] H. Fehrenbacher, B. Wiegel, H. Iwase, T. Radon, D. Schardt, H. Schuhmacher,
      J. Wittstock, Spectrometry behind concrete shielding for neutrons produced by 400
      MeV/u 12C ions impinging on a thick graphite target, in: Proceedings of the 11th Intern.
      Congress of the International Radiation Protection Association (IRPA11), Madrid, Mai
      2004, contrib. 5i4 on CD or on


No additional costs affecting TLABS are anticipated.


Operation of the pulse selector device is required for adjusting the repetition
frequency to the requirements of TOF measurements.


14 shifts (2 weekends)


March 2007 (preferred) or later


No safety problems are anticipated.