Simulation of the Heavy Ion Contribution to the Radiation Environment onboard the International Space Station

(1)

Simulation of the Heavy Ion Contribution to the

Radiation Environment onboard the International

Space Station

Anton Andersson and Birger Vaksdal

antonand@kth.se, vaksdal@kth.se

SA104X Degree Project in Engineering Physics, First Level

Supervisor: Bengt Lund-Jensen

Department of Physics

School of Engineering Sciences

Royal Institute of Technology (KTH)

Stockholm, Sweden

(2)

Abstract

The health risks from exposure to the space radiation environment makes it important to have reliable simulation models for the radiation doses inside spacecrafts. Simulations of the heavy ion parts, especially iron, of the Galactic Cosmic Ray (GCR) radiation has been carried out using the simulation software Geant4 with geometry models from the DESIRE project. The total dose rate from iron ions is found to be 1.57 µGy/d and dose equivalent is found to be 17µSv/d. This means that iron contributes between 5% and 10% to the total dose equivalent from the GCR radiation, substantially higher than its relative abundance of 0.03 percent.

67% of the dose rate from iron comes from particles with energies in the range 1 000 -10 000 MeV/nucleon, and 26% comes from particles in the -10 000 - 50 000 MeV/nucleon range. The great majority of the dose is deposited by iron ions, opposed to secondary particles. This result was not expected. Calculated dose rates are found not to be significantly dependent of the ISS altitude.

(3)

1 Introduction

The atmosphere and magnetic field of the earth is acting as a shield against a lot of the cosmic radiation. In space however, the radiation levels are notably higher. This makes it important to be able to estimate the risk posed to the astronauts onboard the International Space Station (ISS), for future interplanetary spaceflights and longer visits to space.

1.1 High energetic space particle radiation

The high energetic radiation environment in space near Earth can be divided into three sources: trapped radiation in the Van Allen Belts, particles associated with solar proton events and galactic cosmic rays.

Van Allen trapped radiation is primarily electrons and protons, contained in two toroidal (Van Allen) belts around the earth at altitudes ranging from hundreds of kilo-meters to around 60 000 km. The energies of the electrons range from 50 keV to 5 MeV, and the energies of the protons range from 100 keV to several hundred MeV [1].

The solar protons events are fluxes of high energy protons from the sun, which vary greatly in time.

The Galactic Cosmic Rays (GCR) are isotropically distributed protons and heavier ions, with energies ranging from 1 MeV/nucleon to more than 100 000 MeV/nucleon [2]. Hydrogen constitutes about 85% of the total GCR flux, and iron about 0.03% [3]. For the relative abundance of some of the heavier elements in the GCR, see Table 1.1. The GCR particles are likely originating from supernova explosions of stars and are, apart from the solar condition limiting the amount reaching earth, constant in time [2]. The solar wind (the plasma flow of mostly hydrogen from the sun) influences the GCR flux; the flux is peaking during solar minimum and reaches a minimum during solar maximum [4]. Due to the earth’s magnetosphere, only particles with high energy can reach the surface of the earth. In the polar areas the particles can travel parallel to the magnetic fields, resulting in more particle radiation in these areas [1].

(5)

Table 1.1: Heavy cosmic ray element abundance, with oxygen as a reference level. Table adapted from [1].

Element Relative abundance

He 44,700 ± 500 Li 192 ± 4 Be 94 ± 2.5 B 329 ± 5 C 1130 ± 12 N 278 ± 5 O 1000 Mg 203 ± 3 Al 36 ± 1.5 Si 141 ± 3 Cl 9 ± 0.6 Ti 14.4 ± 0.9 Cr 15.1 ± 0.9 Mn 11.6 ± 1 Fe 103 ± 2.5 Ni 5.6 ± 0.6

1.2 Biological effects of the space radiation

The radiation exposure in the orbit of the ISS is considerably increasing the risk of cancer for the crew. Other medical problems that are increased in likelihood is cataract, temporal sterility and damage to the bone marrow effecting in decreased red blood cell formation. These effects may limit the possibilities of manned flights to other planets [5]. Heavy ion radiation induces high level of damage in the path of the particle, and the the area surrounding it due to secondary electrons (delta rays) produced in the ionizing process[2]. Astronauts observe visual flashes that are thought to be caused by heavy ions passing through the retina. The frequency of these flashes has been observed, together with the location of of the observations, giving a crude image of the radiation fields [6].

Due to the different amount of damage caused by different particle types and their rate of energy loss, the average dose is a poor way to assess the risks. One attempt to make a better risk assessment tool is the dose equivalent (see section 1.3) [4]. It should be noted that the dose equivalent does not take localized cell damage into account, and therefore is a far from perfect way to assess the biological effects [5].

(6)

1.3 Dose and dose equivalent

The dose is the total amount of energy deposited per mass. It is usually given in the unit Gray (Gy), equal to J/kg. The dose rate is the dose per time unit.

The dose equivalent can be calculated by adding the energy depositions per mass, weighed by the quality factor Q. The quality factor can be calculated as

Q(L) =      1 L < 10 keV /µm 0.32L − 2.2 L ∈ [10, 100] keV /µm 300/√L L > 100 keV /µm

where L is the unrestricted Linear Energy Transfer (the amount of energy lost by a particle per distance traveled) in water [7].

1.4 International Space Station

The International Space Station (ISS) began manned service in year 2000, and has since been expanded in several steps. It has been visited by over 200 individuals, and has in its orbit travelled the equivalent of over 8 round-trips to the sun since its launch. It spans 109 by 73 meters and currently weighs 419 metric tons. The angle between the plane of the ISS orbit and the equatorial plane is 51.6◦ [8]. During the last 5 years, the altitude of the ISS has been in the range 337 km to 418 km [9]. In 2008 the European module Columbus was attached to the ISS, adding 13 metric tons of mass and 75 m3 _{of space}

[10]. In this thesis, the radiation doses in Columbus will be examined.

The shielding properties of the walls and instruments of ISS is equivalent to 1-30 cm water, this can be compared to that of the earth’s atmosphere, 10 m of water [5].

(7)

Figure 1.1: The International Space Station, viewed from the Space Shuttle Endeavour. Image reprinted from NASA [11].

1.5 Related work

DESIRE (Dose Estimation by Simulation of the ISS Radiation Environment) is a project aimed at estimating the radiation dose received onboard the ISS, covered in a doctoral thesis made by Tore Ersmark ([4]). One of the findings in his thesis is that further work has to be done on simulating heavy ion radiation due to lacking support from the simulation software (Geant4, see section 2.1) at the time of his thesis work.

’Dosemetric Mapping’ is an experiment carried out by the science program of NASA’s Human Research Facility in 2001. The total absorbed doses from the GCR was measured in the US Lab in the ISS with a silicon dosimetry telescope (DOSTEL). Another study performed by the NASA Johnson Space Center used a tissue-equivalent proportional counter (TEPC) to measure the dose rates [12]. The results from these two studies can be seen in Table 1.2.

Table 1.2: Experimental GCR dose rates. Table adapted from [12].

Model µGy/d µSv/d

DOSTEL 92 409

(8)

1.6 Objective and scope

This thesis seeks to provide data on the heavy ion parts of the GCR radiation using an updated version of the simulation software.

The heavy ion species used in the simulations will be limited to iron and oxygen, since these two are some of the most abundant heavier elements in the GCR radiation (see Table 1.1). Primarily values for solar maximum will be used for particle spectra, since the solar is expected to reach maximum this year (2013). The position of the ISS in its orbit will not be considered; average values will be used. The shadow effect of the earth will influence the total amount of incoming particles, but the simulations will assume an isotropic particle flux.

(9)

2 Software for particle spectra,

simula-tion and analysis

To run the simulations, Geant4 is used with a geometry model from DESIRE. The data is given in ROOT tree files, and ROOT is used to process the data before exporting it to Matlab. The values obtained for different energy levels are then weighted with the GCR particle flux data obtained from CREME96.

2.1 Geant4

Geant4([13], [14]) is a Monte Carlo based toolkit for simulations of particle transport through matter. It is developed by a world-wide collaboration of software engineers and scientists, and is based on object-oriented technology, implemented in the C++ program-ming language. Geant4 can be used for various implementations such as electromagnetic, hadronic and optical processes and can handle complex geometries.

The earlier version of Geant4 used by Tore Ersmark et al. in earlier studies did not support particles with energies above 10 GeV/nucleon. Improvements and new physics models now allows Geant4 to handle these energies.

2.2 DESIRE ISS geometry model

One of the outputs from DESIRE (see section 1.5) is a detailed geometry model of the ISS and in even more detail the European Columbus module, that can be used for Geant4 simulations. Included in this model are 15 virtual detectors (ICRU-spheres) spread out in the Columbus module. The ICRU-spheres have a radius of 30 cm, and are consisting of (by weight) 76.2% oxygen, 10.1% hydrogen, 11.1% carbon and 2.6% nitrogen [15].

2.3 ROOT

ROOT provides an object-oriented framework for efficient analysis of large amounts of data. It is implemented in the C++ programming language. ROOT is an official project within the Physics Department at CERN. It was developed for the NA49 experiment at CERN, which generated about 10 terabytes of data per run and therefore needed efficient data analysis [16].

(10)

2.4 CREME96

CREME is an acronym for ”Cosmic Ray Effects on Micro-Electronics code”, of which we are using the version originating in 1996. CREME96 is a suite of programs for creating models of the radiation environment in space, close to the Earth. The flux histogram given has bin widths that are linearly increasing in size with the energy per particle. The bin width is about 1 MeV/nucleon for energies of 100 MeV/nucleon, and a bin width of 1000 MeV for energies of 100 GeV/nucleon. For background on CREME96, see [17] [18] [19].

(11)

3 Simulations and data analysis

3.1 Simulation set up

The simulations were carried out by generating particles on the surface of a spherical source with 75 m radius, centered on the ISS. Several simulations were made, with a varying number of particles with linear dependency between numbers of particles and simulation time. A flat spectrum (equal amount of flux for all studied energies) was used, and the output data was weighted to give results corresponding to the particle spectra given by CREME96. The simulations generate data in the ROOT tree format, containing information for each particle (including secondary particles from reactions with wall material) that deposited energy in the 15 ICRU-spheres in the Columbus module.

3.2 Generating particle spectra

In order to get particle spectra, a number of parameters has to be given to CREME96. The altitude used to get the particle spectrum for the majority of the calculations is 380 km, opposed to the current altitude of about 400 km (as of April 2013) [9]. The differences in spectra for these altitudes can be seen in Figure 3.1. The differences are only noticeable for energies below 10 MeV/nucleon. Virtually all particles at these lower energies are stopped in the aluminum hull of the ISS, and do not contribute to the dose inside the station (see section 4.4 for validation of this claim).

Figure 3.2 shows the difference between solar minimum and solar maximum. There is a clear difference for energies below 10 000 MeV/nucleon.

(12)

100 101 102 103 104 105 10−8 10−7 10−6 10−5 10−4 Energy [MeV/nucleon] Flux [particles/(m 2*s*sr*MeV/nucleon)] 330 km 380 km 430 km

Figure 3.1: Energy spectra from CREME96 at solar maximum for different altitudes.

100 101 102 103 104 105 10−8 10−7 10−6 10−5 10−4 Energy [MeV/nucleon] Flux [particles/(m 2*s*sr*MeV/nucleon)] Solar Maximum Solar Minimum

Figure 3.2: Energy spectra from CREME96 at 380 km for different solar conditions.

(13)

3.3 Hardware

The simulations have been carried out using multiple parallel processes on four computers, with four CPU cores per computer. The iron simulations have used about 700 hours of CPU time, and the oxygen simulations about 200 hours of CPU time.

3.4 Analysis

For the analysis, ROOT tree files from several simulations, and data from the 15 ICRU-spheres (within each ROOT tree file), have been merged to get more accurate results for the average dose and dose equivalent rates.

The main part of the analysis has been carried out in Matlab, including weighting of the flat spectrum from Geant4 and the final calculations to get the total dose and dose equivalent rates.

(14)

4 Results

Dose and dose equivalent rates for primarily iron ions, but also for hydrogen, helium and oxygen, have been calculated as described in chapter 3. The dose and dose equivalent rates are the average values for the 15 ICRU-spheres in the ISS.

4.1 Total dose rate contribution from different ion species

Figure 4.1 shows the dose rates and dose equivalent rates for hydrogen, helium, oxygen and iron. The average dose equivalent quality factor, Q, is for iron about 11 and for oxygen, helium, and hydrogen about 2.

H He O Fe 0 10 20 30 40 50 60 70 80 90 100 Dose rate [ µ

Gy/d], Dose eq. [

µ

Sv/d]

Dose Dose eq.

Figure 4.1: Dose rates for different incident elements at 380 km and solar maximum.

(15)

4.2 Dose and dose equivalent rates from primary and

secondary particles

Figure 4.2 and Figure 4.3 shows the dose and dose equivalent rates for primary and secondary particles that deposits energy in any of the 15 ICRU-spheres, with simulations for iron. Most of the dose and almost all of the dose equivalent rate is from the primary iron ions themselves, and not from secondary particles. As noted by Ersmark et al. in [20], iron should not constitute the great majority of the dose equivalent rate. This result suggests that the Geant4 physics model used does not handle heavy ions in a satisfactory way. 0 0.2 0.4 0.6 0.8 1 1.2 Photon e− e+ Pion−Pion+Muon+Muon−_{NeutronProton} Deuteron

TritonAlphaN14 N15 N17 C12 O16 Si28 Ar37 Sc44Cr52Fe56

Dose rate [

µ

Gy/d]

Figure 4.2: Dose rates from different depositing particle types at 380 km and solar maximum.

(16)

Figure 4.3: Dose equivalent rates from different depositing particle types at 380 km and solar maximum.

4.3 Dose rates at different altitudes and solar

condi-tions

Figure 4.4 shows the dose and dose equivalent rates from iron particles, at different altitudes and at solar minimum and maximum. The relative difference between solar minimum and maximum is about 30%, while the difference between different altitudes is much smaller. This was expected (see section 3.2), and is further explained in section 4.4.

350 km 380 km 450 km 0 5 10 15 20 25 30 Dose rate [ µ

Gy/d], Dose eq. [

µ

Sv/d]

Dose Sol max Dose eq. Sol max Dose Sol min Dose eq. Sol min

Figure 4.4: Dose and dose equivalent rates for different altitudes and solar conditions.

(17)

4.4 Dose rates for different energy intervals

Figure 4.5 shows the dose rates from iron particles in different energy intervals. 67% of the dose comes from particles with energies in the 1-10 GeV/nucleon range, and 26% from particles in the 10-50 GeV/nucleon range. The low contribution from particles with a maximum energy of 1000 MeV/nucleon is consistent with minimal differences at the different altitudes (see Figure 4.4 and Figure 3.1). Also, the low dose from particles in the 50 to 100 GeV/nucleon range is consistent with the assumption that particles with energies above 100 GeV/nucleon can be neglected. In Figure 4.6, this is even more apparent, with less of the dose equivalent rate from the high energy particles.

0−1000 1000−10 000 10 0000 − 50 000 50 000− 100 000 −0.2 0 0.2 0.4 0.6 0.8 1 1.2 Energy [MeV/nucleon] Dose rate [ µ Gy/d]

Figure 4.5: Dose rates at 380 km and solar maximum for iron particles, for different energy intervals.

(18)

0−1000 1000−10 000 10 0000 − 50 000 50 000− 100 000 −2 0 2 4 6 8 10 12 14 16 Energy [MeV/nucleon]

Dose equivalent rate [

µ

Sv/d]

Figure 4.6: Dose equivalent rates at 380 km and solar maximum for iron particles, for different energy intervals.

(19)

5 Summary and conclusions

The objective of this thesis was to simulate heavy ion radiation onboard the ISS. Depen-dency of variations in solar cycle and ISS altitude have been evaluated and simulations for hydrogen and helium have been carried out to serve as reference.

The results show that iron contributes 1.57 µGy/d to the total dose and 17 µSv/d to the total dose equivalent. This means that even though iron only constitutes 0.03% of the total GCR particle flux, it contributes between 5% and 10% to the total dose equivalent. Incident iron ions with energies in the range 1-50 GeV/nucleon contributes to most of the dose and dose equivalent. When it comes to depositing particle types, the great majority of the dose equivalent comes from the primary iron ions themselves, and not from secondary particles. This result indicates a need for further studies and improvements to the physics models of Geant4.

(20)

6 Acknowledgements

We would like to thank our supervisor Bengt Lund-Jensen for his substantial input and feedback on the work that has gone into this report, and for letting us do this thesis at the Particle and Astroparticle Physics group. We would also like to thank Oscar Larsson, Tore Ersmark and Christer Fuglesang for valuable help and input.

(21)

Bibliography

[1] D. Hastings, H. Garret, ”Spacecraft-Environment Interactions” Cambridge: Cam-bridge University Press, 1996

[2] D. Grahn, ”HZE Particle Effects in Manned Spaceflight” Washington, D.C.: Na-tional Academy of Sciences, NaNa-tional Academic Press, 1973.

[3] W. Schimmerling,”The Space Radiation Environment: An Introduction” Available at http://three.usra.edu/concepts/SpaceRadiationEnviron.pdf

[4] T. Ersmark, ”Geant4 Monte Carlo Simulations of the International Space Sta-tion RadiaSta-tion Environment” [PhD Thesis], Stockholm, Sweden: Royal Institute of Technology, 2006.

[5] G. Reitz, ”Biological Effects of Space Radiation”, K¨oln, Germany: DLR, Institut f¨ur Luft- und Raumfahrtmedizin, 1998.

[6] C. Fuglesang, "Using the human eye to image space radiation or the history and status of the light flash phenomena" , Nuclear Instruments and Methods in Physics Research A, 580(2007) 861-865

[7] J. Valentin, ”Relative biological effectiveness (RBE), quality factor (Q), and radia-tion weighting factor (wR): ICRP Publicaradia-tion 92”, Annals of the ICRP, 33 4 (2003) 1-121

[8] NASA, ”International Space Station: Facts and Figures”, [Webpage] [May 10, 2013] Available at

http://www.nasa.gov/mission_pages/station/main/onthestation/facts_and_figures.html [9] Heavens-Above GmbH, ”ISS-Orbit”, [Webpage] [April 12, 2013] Available at

http://www.heavens-above.com/orbit.aspx?satid=25544

[10] European Space Agency, ”European Columbus Laboratory: Facts and Figures”, [Webpage] [May 10, 2013] Available at

http://www.esa.int/Our_Activities/Human_Spaceflight/Columbus/ European_Columbus_laboratory

[11] STS-134 Shuttle Mission Imagery, ”S134-E-010137”, [Webpage] [May 10, 2013] Available at http://www.spaceflight.nasa.gov/gallery/images/shuttle/sts-134/html/s134e010137.html

(22)

[13] S. Agostinelli, et al., ”Geant4 - a simulation toolkit”, Nucl. Instrum. Methods A, 506 (2003) 250-303

[14] J. Allison, et al., ”Geant4 development and applications”, Nuclear Science, IEEE Transactions on, 53 (2006) 270-278

[15] D. Harder, K.-P. Hermann, ”Tissue-Equivalent Materials and the ICRU Sphere”, Radiation Protection Dosimetry, 12 2 (1985) 125-128

[16] The ROOT Team, ”About ROOT” [webpage] [April 10th, 2013] Available at root.cern.ch/drupal/content/about

[17] A.J. Tylka, et al., ”CREME96: A Revision of the Cosmic Ray Effects on Micro-Electronics Code”, IEEE Trans. Nucl. Sci. 44 6 (1997) 2150-2160

[18] R.A. Weller, et al., ”Monte carlo simulation of single event effects”, IEEE Trans. Nucl. Sci., 57 4 (2010) 1726-1746

[19] M.H. Mendenhall and R.A. Weller, ”A probability-conserving cross-section biasing mechanism for variance reduction in Monte Carlo particle transport calculations”, Nucl. Inst. & Meth A, 667 1 (2012) 38-43

[20] T. Ersmark, et al., ”Geant4 Monte Carlo Simulations of the Galactic Cosmic Ray Radiation EnvironmentOn-board the International Space Station/Columbus”, Nu-clear Science, IEEE Transactions on, 54 5 (2007) 1854-1862

Simulation of the Heavy Ion Contribution to the Radiation Environment onboard the International Space Station