General Information
Geant4 Space Users' Workshop –G4SUW– is focused on new results on space radiation interaction with components, sensors and shielding analysis, as well as on Geant4-based tools and developments applicable to space missions. The particular topics of interest for this workshop include:
Location
Silverroom, Space Center Houston
1601 E NASA Pkwy, Houston, TX 77058
URL : https://spacecenter.org/
Registration and Attendance Fee
The registration website is here.
Registration will start in September 2018.
Abstract Submission
The Indico website for abstract submission will open when registration starts.
The presentation will focus on the recent Geant4 related activities started and performed by ESA for the planning of space and planetary exploration missions.
New developments and improvements of some of the ESA Geant4 tools will be presented.
The ESA tools list includes: GRAS (3D radiation analysis), MULASSIS (1D radiation analysis), MAGNETOCOSMICS (cosmic ray rigidity) and PLANETOCOSMICS (cosmic ray showers) and SPENVIS (Space Environment Information System, www.spenvis.oma.be).
Geant4 has been an invaluable tool for understanding the data of MAVEN's Solar Energetic Particle (SEP) instrument. Raw data from SEP consists of two sets of counts across 256 bins. Each bin is essentially a counter corresponding to an event type and energy deposit (to be further explained in the presentation). Simulations were done in Geant4 in order to understand what differential energy flux spectra could produce observed results.
One of the primary goals of The Mars Science Laboratory (MSL) rover is to understand current and historical water abundance on Mars (Grotzinger et al., 2012). A key instrument on MSL for accomplishing this goal is the Dynamic Albedo of Neutrons (DAN; Mitrofanov et al., 2012). The DAN instrument is sensitive to hydrogen (H) from the surface down to roughly 60 cm depth in the regolith. DAN cannot discriminate between different host molecules for the H it detects, but for convenience, all H abundances are provided in terms of Water Equivalent Hydrogen (WEH), which is the amount of water that would exist if all of the H was bound in water.
DAN has two modes of operation: passive and active. In passive mode, spallated neutrons in the regolith are sourced by galactic cosmic rays and by the rover’s Multi Mission Radio-isotope Thermoelectric Generator (MMRTG). In active mode, a Pulsing Neutron Generator (PNG) is employed to produce roughly 107 high-energy neutrons per pulse at a frequency of 10 Hz (Sanin et al., 2015). In both modes neutrons are detected by two 3He proportional detectors, one that records neutrons with energies below ~ 1keV (Counter of Total Neutron: CTN), and another that records neutrons with energies between 0.4 keV and 1 keV (Counter of Epithermal Neutrons: CETN). After the PNG emits fast neutrons in active mode, the detector portion of DAN counts epithermal and thermal neutrons in 64 logarithmic time-dependent bins. The resultant plot of neutron counts versus time is called a ‘die-away’ curve. By analyzing the shape of a die-away curve, we determine the WEH abundance and depth distribution in the regolith directly beneath the rover (Sanin et al., 2015). It has been shown that varying chlorine (Cl) abundances are the biggest source of variability in neutron absorption for the Martian regolith (Hardgrove et al., 2011), however there are other absorbers (e.g., iron). Neutron absorbers are collectively referred to as Absorption Equivalent Chlorine (AEC), following the convention of (Tate et al., 2015).
By comparing measurements to simulations, hydrogen abundance is determined. Simulations of the DAN instrument are typically done with MCNPX/MCNP6 (McKinney et al., 2006), but this nuclear transport code subject to export control, making it difficult to run on open-architecture computing facilities. Thus, our goal is to be the first to use Geant4 (Agostinelli et al., 2003) for DAN modeling. We will use MCNPX/MCNP6 simulations, as well as DAN-active measurements to determine the viability of our Geant4 application.
We will present our ongoing work to use Geant4 as a replacement for MCNPX/MCNP6. Our Geant4 application is designed to mimic the DAN instrument and the Martian environment. We have included a PNG as the primary event, and the detectors are the CTN and CETN. The neutronHP package processes used include neutronElastic, neutronInelastic, nCapture, and nFission. With the use of a .g4mac macro simulations will be run for a variety of WEH and AEC values, and regolith geometries. Simulations are run with the University of Tennessee’s Advanced Computing Facility. We hope to implement the multithread function (G4Threading) to take full advantage of running on a computer cluster.
References
Agostinelli et al. (2003) Nuc Instruments and Methods in Phys. Res. A, 506(3), 250–303. Grotzinger et al. (2012) Space Sci. Rev., 170, 5-56. Hardgrove et al. (2011) Nucl. Instruments Methods Phys. Res. A, 659, 442-455. McSween, et al. (2010) J. Geophys. Res., 115. Mitrofanov et al. (2012) Space Sci. Rev., 170, 559-582. Sanin et al. (2015) Nucl. Instruments Methods Phys. Res. A, 789, 114-127. Tate et al. (2015) Icarus, 262, 102-123.
At the core of the LISA gravitational wave observatory are isolated test masses that act as mirrors for a 2.5 million km baseline interferometer. The test masses must be kept in near pure free-fall: isolated from all unwanted force disturbances above the femto-Newton level. Electrical charging of the test masses produced by cosmic ray impacts gives rise to electrostatic forces that can reduce the sensitivity of the observatory to gravitational waves. In previous work, GEANT4 simulations have been used to predict the test-mass charging rate in orbit. We revisit these simulations, comparing the results with measured test-mass charging rates in the LISA Pathfinder technology demonstration mission during 2016 and 2017. Combining charge measurements with the results of in situ measurements with a dedicated particle detector allow us to probe fluctuations in the charging rate driven by cosmic ray variations. The observations made in orbit confirm the importance to this seemingly simple problem of physical processes from eV to TeV energy scales.
The space elevator is a means for moving large amounts of mass into space at relatively low cost. Modern concepts of this structure involve a gravitationally stabilized thin tether made of strong materials such as carbon nanotubes or single-crystal graphene, extending to an altitude of 100,000 km. Along the full length of the tether are many hazards, such as space debris, solar storms, radiation damage on tether materials and electromagnetic de-stabilzation of tether motion, which must be studied. The planned use of Geant4 for each of these studies will be discussed after a brief introduction to the space elevator concept.
The concept of active magnetic shielding is to use high-temperature superconducting coils to induce very high magnetic fields around a human spacecraft. The induced magnetic field will deflect incoming charged particles (solar particles and galactic cosmic rays), thereby reducing the particle flux and radiation dose to astronauts behind the shield.
The goal of this project is to create a model for determining the value of active magnetic shielding in reducing radiation dose to astronauts on an interplanetary mission. Over 100 mission scenarios will be simulated in Geant4, varying parameters such as solar cycle, shielding configuration, magnetic field strength, crew gender, and phantom type. A sensitivity analysis on the effect of varying each parameter will highlight scenarios that minimize astronaut radiation dose for a given mission profile.
Science missions are demanding progressively more detailed simulations increasing the computational time due to the complexity of the 3D geometrical models. Generally, in space dosimetry calculations, the sensitive part of the electronic components is significantly small compared to spacecraft size. In such cases the reverse Monte Carlo (RMC) method, also known as the adjoint Monte Carlo, can be used to speed up these calculations.
This presentation will provide a status report on the Geant4 Reverse Monte Carlo (RMC) implementation in the GRAS tool. An evaluation of the method is provided for some realistic cases with a particular attention to some challenging environments like the Jovian radiation belt that will be encountered by the ESA JUICE (Jupiter Icy Moons Explorer) mission.
Simulations were performed with simple and complex geometries and a realistic spacecraft 3D model has been designed for the tests.
Potential issues have been identified and possible solutions are under development. Some comparisons with other tools and different methods, like ray tracing, will be presented as well.
SoftWare for Optimization of Radiation Detectors (SWORD) is a CAD-like development tool and front end for several radiation transport codes. By default, SWORD ships with Geant4. SWORD also supports MCNP (radiation transport Monte Carlo developed at Los Alamos National Laboratory) and Denovo (deterministic discrete solver developed at Oak Ridge National Laboratory), both of which must be obtained separately. SWORD also provides a library of prebuilt objects, including detectors, various crafts, and buildings as well as a library of common spectra. There is an ongoing effort to include more spacecraft and space-relevant objects in this library.
SWORD was designed to allow for easy use of these radiation transport codes without the extensive learning curve and to quickly make changes to the geometry. SWORD also allows the creation of complex sources including multiple distinct sources and sources that have several different spectra. SWORD also allows the construction of scenarios where objects can move and simulations can be run at various time intervals very easily. Work is currently being performed to improve integration with existing space weather software like AE9 and AP9 to improve SWORD usability for space applications. This would allow more detailed analysis to be conducted simply.
SWORD was developed with funding from the Department of Homeland Security/Countering Weapons of Mass Destruction, the Defense Threat Reduction Agency, and the Office of Naval Research, but is available to the community at large. SWORD is available in the USA from the Radiation Safety Information Computational Center at Oak Ridge National Laboratory and in Europe from the Organisation for Economic Co-operation and Development Nuclear Energy Agency. SWORD has been used for both dosimetry and background modeling for several space instruments, including the SoloHI instrument on the ESA Solar Orbiter and a future gamma-ray burst instrument currently being developed.
This research was supported by the Chief of Naval Research.
As China's first X-ray astronomical satellite, Insight-HXMT (Hard X-ray Modulation Telescope) successfully launched on Jun 15, 2017. HXMT carries three main payloads onboard: the High Energy X-ray telescope (HE, 20-250 keV, NaI(Tl)/CsI(Na)), the Medium Energy X-ray Telescope (ME, 5-30 keV, Si-Pin) and the Low Energy X-ray telescope (LE, 1-15 keV, SCD). The response function and efficient areas of HXMT payloads can be derived by using the Geant4.
An X-ray spectral model for the clumpy torus was constructed in an active galactic nucleus (AGN) using Geant4, which includes the physical processes of the photoelectric effect, Compton scattering, Rayleigh scattering, gamma conversion, fluorescence line, and Auger process. Since the electrons in the torus are expected to be bounded instead of free, the deviation of the scattering cross section from the Klein–Nishina cross section has also been included, which changes the X-ray spectra by up to 25% below 10 keV. We have investigated the effect of the clumpiness parameters on the reflection spectra and the strength of the fluorescent line Fe Kalpha.
Monte Carlo methods to model radiation transport are now extensively used to model detector efficiency and response for space radiation sensors. Often, the modeling efforts focus on the response of the sensor to isotropic or unidirectional fluxes at low to moderate counting rates. This is acceptable for situations where the design requirements of the sensor allow it to be operated in these near ideal circumstances. With the advent of space radiation sensors being developed for a variety of applications as well as platforms this simplified approach to assessing the performance of a detector on orbit may be not sufficient. This is because almost all sensors rely upon sensitive analog and digital electronics to take and interpret data from these sensors. These signal chains can behave in non-ideal manners due to effects such as dead-time, pulse pile-up, and accidental coincidences. Those behaviors are often analyzed for single sensors and then effects are approximated for complex systems such as multi-element particle telescopes. However, with the advent of large memory depth signal generators it has become possible to develop pulse sequences that can stimulate these signal chains to better understand departures from ideal behavior. Our current efforts underway to go from Monte-Carlo simulation of the detector interaction all the way to pulse sequences for stimulation of the signal chains for multi-element coincidence based particle telescopes are described.
On the Zugspitze mountain (2650 m a.s.l.), Germany, the Institute of Radiation Protection operates a Bonner Sphere Spectrometer (BSS), running continuously since 2004, to measure the neutron energy distribution of secondary cosmic rays. These measurements are highly affected by various environmental parameters, and in particular by the amount of snow in the near environment. To quantify this effect, Geant4 is used to simulate the neutron radiation field in this environment.
In a first step, the radiation field was simulated at 5 km a.s.l.. For this purpose the Earth atmosphere was adopted as a cylinder with a height of 100 km and a diameter of 6,000 km, consisting of several layers with temperature, density and air composition according to the 1976 U. S. Standard Atmosphere. At the top of this cylinder a surface source was placed covering the whole cylinder cross section and emitting primary cosmic rays (protons and alphas). Energy and angular distribution was implemented using a modulated local interstellar (LIS) spectrum with local cutoff rigidity and a cosine law angular distribution.
In a second step, the cylinder was reduced to 200 km in diameter to enhance the particle fluence, and the radiation field obtained in the first-step simulation was used as an input field at 5 km a.s.l. For this step, the ground level environment (mountain, housing of measurement location, several snow depths) will also be implemented.
This presentation shows the first results of this project, and further plans will be discussed.
NASA Shields-1 will be launched in the upcoming NASA CubeSat Launch Initiative (CSLI) ELaNaXIX Mission in December 2018. Shields-1 hosts a research payload experiment with 8 μdosimeters behind radiation shielding in slab geometries to make dose-depth curves. Shields-1 hosts atomic number (Z)-graded radiation shielding materials as part of the research payload with baseline aluminum materials for the comparison of dose-depth curves. This radiation shielding experiment has been prepared from radiation shielding computational modeling of a geotransfer orbit (GTO) with slab geometries using GEANT4 MUlti-LAyered Shielding SImulation Software (MULASSIS) from The Space Environment Information System (SPENVIS). Z-graded radiation shielding materials show a ~30% increased shielding effectiveness for electrons at half the thickness in comparison to aluminum of the same areal density The Shields-1 electronics enclosure (vault) has been developed using the LaRC Z-Shielding materials for reducing total ionizing dose on the electronic cards in order to increase the mission lifetime of the commercial parts.
GEANT4-dna-chemistry is a relatively new part of GEANT4 which allows a particle-in-cell (PIC) like simulation of water electrolysis products in liquid. By applying this to a microscopic geometric model of nuclear DNA we can estimate genetic damage caused by particle radiation. To apply this to macroscopic bodies we have furthered the PIC-like methodology by implementing repeated geometry. With certain limiting assumptions, this can be a resource-friendly alternative to pure PIC methods of simulation. However, in the process we have run across certain difficulties in the current state of GEANT4-dna-chemistry. We will summarize our models, the difficulties encountered, and our work-arounds.
Intensity modulated proton therapy (IMPT) is an advance form of proton therapy, in which tumor is irradiated by proton beamlet spot-by-spot and layer-by-layer through controlling the trajectory and energy of a focus beam of protons.1,2 While proton radiobiological effects depend primarily on their physical dose distribution, studies have showed that linear energy transfer (LET) plays an important role too.3 LET itself can be used as indicator for radiobiological outcome at the microscopic level, which would justify the use of purely LET-based objectives in treatment plan optimization.4
We developed a method using a hybrid approach to calculate proton linear energy transfer (LET) for intensity modulated proton therapy (IMPT) based on the data pre-computed by Geant4 Monte Carlo (MC) simulations. The hybrid method was incorporated into our in-house developed treatment planning system (TPS), as an extension to calculate LET in voxelized patient geometries. First we commissioned the Geant4 MC code to model three proton treatment nozzles installed in our hospital without range shifter (VAC machine), with range shifter at 42.5 cm away from iso-center (RS machine), and with range shifter at 30 cm away from iso-center (ERS machine). The code was used to generate pencil beam type of LET kernels for all 97 proton energies used clinically. Second, the LET kernels were incorporated into the in-house developed TPS using the ray-casting algorithm. The inhomogeneities were taken into account using water-equivalence-thickness (WET). Since the LET kernels were pre-calculated, the calculation of LET distribution in patient geometries takes much less time. It is found that the LET distribution calculated by the in-house developed TPS agrees well with the MC calculation. The calculated LET distributions were used to evaluate potential clinical benefit and toxicity for various tumor sites including lung, head and neck, esophageal, and brain etc. The LET calculation code has also been used in the IMPT treatment planning, allowing for radiobiological optimization by including LET-weighted constraints in the inverse treatment planning process.
The MPEXS series is yet another software toolkit to simulate interactions between particles and matter. MPEXS was designed and built from scratch for GPU in the CUDA language, but still, algorithms and data necessary are taken from Geant4. MPEXS, the major part of the MPEXS series including handlers of geometry, material, incident particles, particle transportation, Electro-Magnetic physics, and so on, was developed by the collaboration among Stanford University, SLAC and KEK, as a fully parallel computing software. MPEXS-h and MPEXS-DNA were started to be developed by KEK later. MPEXS-h is for hadronic interactions, and MPEXS-DNA is a CUDA version of Geant4-DNA respectively. The MPEXS series was already confirmed to provide the same level of reproducibility of physics processes as Geant4 in many benchmark programs.
The development of the MPEXS series is continuing to provide more functionalities and better computing performance. As of today, MPEXS-DNA has achieved the speedup factor of 2,800 on an NVIDIA TITAN-V for a benchmark program against Geant4-DNA running on one core of Intel Xeon CPU. In other words, this means just one TITAN-V has the equivalent computing power of 2,800 core of Intel Xeon CPU. Any other groups who tried to develop a CUDA version of Geant4 never succeeded to have the similar achivements.
We will report the status and plan for the MPEXS series in this talk.
The Fortran code system Penelope performs Monte Carlo simulation of coupled electron-photon transport in arbitrary materials for a wide energy range. Penelope implements the most reliable interaction models that are currently available, limited only by the required generality, and gives results in good agreement with a variety of experimental data. We present a translation of the latest Penelope 2018 physics routines to C++ that is designed to be used with the Geant4 toolkit with multithreading capabilities. This program effectively uses the available computational resources, giving results equivalent to those from the original Penelope programs with a much shorter running time, which is roughly inversely proportional to number of threads used.
In view of manned missions targeted to the Moon and Mars, for which radiation exposure is one of the greatest challenges, it is of fundamental importance to have tools, which allows the determination of the particle flux and spectra at any time during the mission from transit trajectory to planetary surface operations.
The environment at the surface is, apart from occasional solar energetic particle events, dominated by galactic cosmic radiation, secondary particles produced in their interaction with the atmosphere and albedo particles from the regolith. The highly energetic primary cosmic radiation consists mainly of fully ionized nuclei creating a complex radiation field at the surface. This complex field, its formation and its potential health risk posed to astronauts on future manned missions can only be fully understood using a combination of measurements and model calculations. We will present our ongoing effort in modeling RAD detector response as part of the effort of understanding this complex environment.
Another important aspect of the problem, modeling how vehicle shielding mitigates the dose accumulated by astronauts. This is an essential step toward reducing these risk. In order to do that it is necessary to handle complex object geometries. Standard Monte Carlo programs use mainly simple geometrical forms such as parallelepipeds, ellipsoids, or planes to construct more complex geometries. Vehicle/habitat analysis can be substantially improved by implementing better methods of directly infusing CAD architecture into Monte-Carlo transport codes. In this context, we will present results obtained with the DAGMC package