A Charged Particle Spectrometer for Crewed Space Missions

1
A Charged Particle Spectrometer for Crewed Space
Missions D. H. Kaplan1 and J. H. Adams, Jr.2
1Department of Physics, Southern Illinois
University Edwardsville, Edwardsville, Illinois
62026 2 NASA Marshall Space Flight Center,
Huntsville, Alabama 35805 
Abstract It is necessary to measure the
radiation environment within crewed space systems
to check the accuracy of computer models used to
determine the radiation exposure of the crew. All
radiologically significant components of the
radiation field need to be measured.These include
neutrons, protons and the nuclei of the nuclei of
all the abundant elements from H to Fe over the
energy range where their fluxes are high enough
to be radiologically significant.The
instrumentation to make these measurements
includes various types of dosimeters, a Tissue
Equivalent Proportional Counter (TEPC), a
directional charged particle spectrometer for
both internal and external use and an external
electron detector. This poster will present a
concept for a directional charged particle
spectrometer to monitor the nuclei of the
elements C though Fe from 20 MeV to 10 GeV within
the crew quarters and externally.   Introduction
While computer codes aimed at predicting
the radiation environment inside spacecraft on
deep space missions exist, their predictions need
to be validated or corrected. It is therefore
important that all radiologically significant
components of the radiation field be measured in
crew quarters on at least one deep space mission.
Subsequently, radiation exposures can be
determined by using calculations normalized to
dosimetry measurements.   Among the
significant components will be neutrons, protons,
and nuclei of elements up to and including Fe.
Due to fragmentation reactions in the spacecraft
walls, the abundances of different elements will
be enhanced or diminished relative to the
progenitor galactic cosmic ray (GCR) background.
The enhancement will vary with location within
the spacecraft according to the wall thickness
distribution and other factors. Necessary
corrections to computer code predictions will
most likely be due to errors in knowledge of
shielding thickness distribution and/or in
relevant cross sections used in the modeling. The
effects of both of these types of error will be
most pronounced at relatively low energies
(roughly 100 MeV per nucleon (MeV/n) to 500
MeV/n.) For this reason, measurement in this low
energy range is particularly important.  
While current and planned future instrumentation
includes various passive dosimeters and a Tissue
Equivalent Proportional Counter, it is necessary
to separately measure and monitor the abundance
and energy of each elemental component at various
locations inside the craft over a range from
about 100 MeV per nucleon (MeV/n) to about 10 GeV
per nucleon (GeV/n). Existing instruments that
have flown on the International Space Station
(ISS) and/or on the Space Shuttle (SS) are
inadequate to this task, due to intrinsic
limitations of capability, insufficient
geometrical acceptance, or both. For these and
other reasons, a recent NASA report has
criticized existing internal active area
dosimetry, calling for substantial updating.
A suitable, active charged particle spectrometer
should should be capable, with daily
measurements, of providing statistical accuracy
of about 20 on the abundance of each element
over the 100MeV/n 10GeV/n energy range, with
binning emphasizing the 100 500 MeV/n region.
The daily sampling frequency of updating is
minimum frequency for a lunar mission For
deployment within a spacecraft, the instrument
volume should be limited to about 0.05 m3, it
should be operable on standard spacecraft
voltages, should draw no more than 15-20 W power,
and should weigh no more than about 10 kg. Due
to the wide range of nuclear charges that must be
identified and measured, it is necessary to
deploy two instruments. One of these will measure
H and He, and a second, more challenging to
design, is to identify and measure nuclei the
range . This paper is concerned with the design
of this second instrument. Adaptation to
Specific Requirements   The broad energy
range to be covered suggests a vertically
segmented detector with stacked, separated
threshold cherenkov counters. The minimum
necessary sizes and the optimal ordering of the
components within the device are dependent on the
rates to be encountered. For this purpose, a
computer code, CRÈME96, was run to estimate flux
rates through an aluminum spacecraft wall of
thickness 10g/cm2, assuming galactic cosmic ray
background under Solar minimum (cosmic ray
maximum) conditions. It was found that about 2/3
of the flux behind the shielding is above 500
MeV/n in energy. Further, as transmitted
particles with energies above about 800 MeV/n are
much more susceptible to fragmentation than to
stopping in tissue, and because the fragmentation
cross sections are less dependent on energy,
finer energy binning is necessary in the range
below 500 MeV/n than above it. Thus, the
identities and energies of the lower energy
particles should be determined in the upper
layers of a multi-layer detector, as this
maximizes the geometrical acceptance factors for
such particles. Within this design paradigm,
the primary dE/dx signal can be provided by
planes of relatively coarsely pixellated
instrumented Silicon. However, in order to
minimize power consumption and allow for dual use
as a trigger, instead of Silicon microstrip or
mm-scale pixellated Silicon, a scintillating
fiber hodoscope with all fibers read out at both
ends by multianode PMT and ASIC IC signal
processing is proposed. With this scheme, a fast
dynode signal can then be used to replace the
traditional trigger scintillator layer, thus
making a thinner front end. (In addition to its
higher power draw, due to its relatively slow
readout time, a Silicon-based plane is
inappropriate for a trigger.)
For measurement of energies above 2.5 GeV/n, an
aerogel-silica (n 1.04) cherenkov radiator is
proposed. For the range 300 2500 MeV/n, an
acrylic plastic (n 1.49) cherenkov radiator can
be used. Simple calculations show that, for
energies in the range 100 MeV/n 300 MeV/n, the
rangeout technique is not appropriate, because of
the additional depth, weight and grammage that
would be required in the upper detector layers.
Therefore, we propose the use of a layer of
CVD-diamond (n 2.4) or of Moissanite (n 2.64)
as a separate cherenkov radiator for this
purpose. Due to the high index of refraction,
such a layer can be quite thin (see below) or
even in the form of a solution or paste of small
crystals in solid or liquid solution. In this
report, we assume a 1mm thick solid layer (15 cm
x 15 cm x 0.1 cm). In order to homogenize the
cherenkov signals against variations in yield
with lateral position of the track and such
factors as chance proximity to a readout
phototube, in this design each radiator will
deposit photons into a separate light collection
box positioned underneath, the internal walls of
which are coated with diffuse reflector, on which
cherenkov photons will undergo repetitive
bounces. For the lowest energy segment, the 1 mm
CVD Diamond or Moissanite layer would be adhered
directly to the upper inside surface of the light
box. While the ideal box shape would be cubical,
the success of the TIGER mission has demonstrated
that an aspect ratio (depth/length or width) of
20 provides an adequate photoelectron yield for
15 photocathode efficiency. Thus, for 15 cm x
15 cm planar area, the depth of each light box
need only be 3 cm. Light boxes can thus be read
out with 1 diameter photomultiplier tubes placed
side-by side around the box end faces. The PMTs
are to be followed by ASIC ICs including
discrimination and pulse height analysis. The
expected photoelectron yields of a 1mm thick
layer of CVD diamond for incident Carbon and
Oxygen nuclei are shown these assume a
normalization factor similar to that found for
cherenkov detectors on the TIGER mission with
similar light box aspect ratio. In line with
the principle of early low-energy identification,
the Diamond or Moissanite detector should be
placed as close to the top of the detector stack
as possible, above the acrylic and aerogel
detectors. The thinness of the radiator layer and
light box (lt 4cm) is consistent with this. The
overall fractional energy resolution of the
3-layer composite cherenkov detector for Z6 (the
worst case) is better than adequate for the bin
widths of the design.   For assessing rate
capability requirements for a dosimeter, Iron
provides a useful benchmark due both to its high
quality factor and the fact that Fe rates behind
shielding are somewhat lower than in the
background GCRs. Using, for example, the CREME96
result of 0.15 Fe per cm2-sr-day for the range
100 200 MeV/n behind 10 g/cm2 of Al shielding,
requires a geometrical factor (S) of at least 167
cm2-sr to achieve 25 Fe (i.e., ) per day in this
range. For a detector surface area of 15 cm x 15
cm (a reasonable transverse area choice for
spacecraft deployment), for this S, a detector
segment thickness of less than 13 cm is required,
the less, the better. In our design conception
the thickness of the portion of the device
measuring energies and charge in the 100-300
Mev/n range is about 6 cm, which implies S 340
cm2 sr or about 50 Fe/day in this range. The
flux of Fe for all other 100 MeV/n wide bins at
higher mean energy than this is greater.
The success of the NASA TIGER mission has
demonstrated that a ratio of hodoscope pitch to
plane separation of 1-2 is adequate for the
required 0.25 cu resolution. Using a hodoscope
plane separation of 6 cm for the low energy
portion of the detector and 1mm square cross
section scintillating fiber elements (1 mm pitch)
would then provide an adequate pitch to
separation ratio of 1.7. Assuming that the
fibers are read out at both ends, the number of
readout channels per hodoscope plane is then 300.
For redundancy of position measurement, the
proposed device incorporates 6 such hodoscope
planes instead of four thus the total number of
scintillating fiber channels is 1800. The fibers
would be read out with Hamamatsu 64-anode compact
multianode photomultipliers (MAPMTs) 6 per
plane are then required, bringing the total to 36
MAPMTs.  Conclusions   We have presented a
concept for an active on board heavy element
monitor capable of identifying and measuring the
energies of charged nuclei from Carbon through
Iron in the range 100 MeV/n 10 GeV/n within a
crewed spacecraft. The device is compact, light
weight and draws less than 15 W. Further
work needs to be done in verifying and completing
the design, including Monte Carlo simulations and
especially beam tests of the proposed Cerenkov
system. Also, the design presented above does
not address the need for monitoring of nuclei
with Zlt6. For these nuclei, a special attachment
needs to be designed as the photon yields will
be low in the relatively thin Cherenkov radiators
mentioned above. Due to the higher rates expected
at low Z, however, such an attachment, with
thicker Cherenkov radiators, can have a smaller
transverse surface area, and so the total volume
would still be manageable.
CRÈME 96 results for flux of elements 14 (Si)
26 (Fe) behind 10 g/cm2 of Al
shielding.
Expected Cherenkov photoelectron yields for 1mm
thick CVD Diamond layer with incident (a) Carbon
and (b) Oxygen nuclei.
(Left) Schematic of a scintillating fiber. Fiber
has additives which emit photons upon traversal
by nuclei. Photons are transmitted through fiber
via repeated total internal reflection. Fibers
are fabricated at Washington University.
(Right) A prototype plane of parallel running
optical scintillating fibers constructed in the
Washington University Laboratory for Experimental
Astrophysics (J.T. Link et al., (2003), 28th
International Cosmic Ray Conference 4, 1781).