A Monte Carlo transport code study of the space radiation environment using FLUKA and ROOT

Source of Acquisi tion NASA Johnson Space Center

A Monte Carlo Transport Code Study of the Space Radiation Environment Using FLUKA and ROOT

Thomas Wilson l , Lawrence Pinsky2, Federico Carrninati3a, Rene Brun3b,

Alfredo Ferrari4, Paola Sala4, A. Empl 1.2, and Jane MacGibbon l,2

I National Aeronautics and Space Administration, Johnson Space Center, Houston, TX 77058 2 Department of Physics, University of Houston, Houston, TX 77204

3uCERN - EP Division and 3bCERN - IP Division, CH-J 2 J J Geneva 23, Switzerland -iIns(ituto Nazionale di Fisisa Nucleare, Milan, Italy

twilson@ems-isc. nasa.gov

Abstract. We report on the progress of a current study ai med at developing a state-of-the-art Monte-Carlo computer si mulation of the space rad iation env ironment using advanced computer software techniques recently available at CERN. the European Laboratory fo r Panicle Physics in Geneva, Swi tzerland. By taking the next-generation computer software appearing at CERN and adapt ing it to known problems in the implementation of space exploration strategies. this research is identify ing changes necessary to bring these two advanced technologies together. The radiation transport tool being developed is tailored to the problem of taking measured space rad iation fluxes impingi ng on the geometry of any panicular spacecraft or planetary habitat and simulating the evolution of that flux through an accurate model of the spacecraft material. The simulation uses the latest known res ults in low-energy and high-energy physics. The output is a prediction of the detailed nature of the radiation environment experienced in space as we ll as the thermal neutron albedo and secondary part icle albedo created by the spacecraft material itself. Beyond do ing the physics transport of the incident fl ux using a Monte Carlo code called FLUKA, our software too l wi ll provide a self-contai ned stand-alone object-oriented analysis and visualization infrastructure. The latter is known as ROOT. We will also describe the method for defining spacecraft geometries by utilizing aerospace fi ni te element models (FEMs).

INTRODUCTION

One focus of NASA's Human Exploration and Development of Space (HEDS) enterprise is long-duration missions to the Moon and Mars, and a permanent human presence in space beg inning with the International Space Station (ISS) in low-Earth orbit (LEO). To accomplish such an ambitious task the latest progress in advanced technology must be assessed and applied to the basic problem of risk mitigation for human exposure to the complex space radiation environment.

In this regard, we are developing a Monte-Carlo-based tool designed to simulate the space radiation environment using CERN-derived computational technology. We have chosen FLUKA as the primary physics engine, and ROOT as the user interface and analysis infrastructure. These have both been described earlier (Pinsky et ai. , 1999a,b; 2000b) and this report describes progress in the continuation of an ongoing three-year study. Our goal is to provide a single user-friendly, GUI (Graphical User Interface)-based environment which is complete with all of the needed input and analysis capabilities built into a self-contained package for understanding and studying space radiation.

After explaining the reason for undergoing this investigation, we will briefly discuss FLUKA and ROOT, the virtual Monte Carlo interface known as ALIROOT, and fmally the geometry generators which ideally will interface with industry-standard CAD (Computer-Aided drawing) files and aerospace FEMs.

WHY ARE WE DOING THIS?

In the course of acquiring more and more radiation flight data (Badhwar et ai. , 2000), some anomalies in neutron flux measurements have appeared when compared w ith certain standard shielding calculations made by conventional methods. Figure I g ives the STS-57 data acquired in June 1993, along with such comparisons to SAIC and HZETRN math models. Concern regarding the actual source of the errors shown has prompted others

(Pinsky, Carminati, and Ferrari. 2000a) to look into this, and it has led to the present investigation of advanced Monte Carlo techniques as a poss ible means for reconciling these data with known physics.

\

I

SAle-SOLAR rvIAX 51.6

0 x 470 Ian

/

STS-57 (JUN'93) 28 .5° x 470 km

STS-57 (28 .50

x 470 Ian) (BRYNfRNIHZETRN)

E, Neutron Energy (MeV)

---.

FIGURE 1. Comparison of Space Shuttle neutron measurements with model calculations. Reprinted with permission (Badhwar, Keith, and Cleghorn 2000).

A second justification for this investigation is fact that most of the cosmic-ray (CR) astrophysics community is currently using an out-dated Monte Carlo program known as GEANT3.21 for the design of experimental detectors such as ACCESS (Advanced Cosmic-Ray Composition Experiment on the Space Station). ACCESS (Wilson and Wefel, 1999a,b,c) is one of three new starts in NASA's "Cosmic Journeys" program under the Structure and Evolution of the Universe enterprise. GEANT3 was frozen at CERN some eight years ago. One consequence of the research discussed here will be a much more advanced Monte Carlo tool for use by the high-energy astrophysics community at energies approaching 10 15 eV which is the "knee" in the CR spectrum.

FLUKA - THE PHYSICS ENGINE

The background, history, and rationale for selecting FLUKA as the particle physics transport code have already been described (Pinsky et aI. , 1999a). This code is arguably one of the most accurate integrated physics packages available today. It also includes the technology for simulating the transport of low-energy neutrons developed over 30 years by the nuclear reactor community, and it embeds the well known EGS (Electromagnetic Gamma-ray Shower) code for electromagnetic interactions. Another feature of FLUKA is its "plug-in" capability to add physics

that is not already integrated into it. An example is the full treatment of all possible heavy-ion interactions being undertaken in the present study.

Nevertheless, FLUKA 's current input formats are cumbersome for very complex geometries. Therefore, a part of the present exercise is to "civilize" FLUKA by making it more user-friendly while extending its physical range to include all of the energies of interest in the study of space radiation. The required FLUKA additions include: (a) addition of heavy-ion interactions (all ions from Fe down to He, including the full range of interactions); (b) conversion of FLUKA input and output to a ROOT interface (similar to ALI ROOT discussed in a later section); and (c) establishing the interchangeability of the geometry format.

ROOT - THE USER INTERFACE

Software computation packages are necessary to analyze and display data from Monte Carlo calculations. However, these are rapidly changing. In order to overcome many of the shortcomings of an earlier CERN product known as PA W (Physics Analysis Workstation) used in conjunction with the Monte Carlo code GEANT (Brun et a!. , 1987), one of us (Brun, 1995 ; Brun, 1997) has introduced ROOT. This is both an analysis and a visualization

FIGURE 2. Functional schematic of the ALIROOT architecture.

software program in one package. At its heart, ROOT is based upon Object-Oriented (00) data structures. Such use of 00 programming allows many difficult tasks with multiple applications to be done at once. Another feature of ROOT is that much of the manipulation is provided for via GUI menus, displays, and browsers. In addition, it uses C++ as a scripting language.

Thus a set of software tools with new data analysis infrastructure is evolving. These are Object Oriented and use open source code like Linux. Most operating systems are supported, such as Unix, Linux, MSWindows, Sun, and so forth . Many objects and classes are already available.

In summary, ROOT is both a tool and a framework. As a tool , one enters commands in C++ to manipulate data, and GUI-based displays exist for most common operations. As a framework , ROOT can be used as a library and a link in new user C++ code. The results can then be used as a tool again.

ALIROOT - THE VIRTUAL MONTE CARLO INTERFACE

The ALICE (A Large Ion Collider Experiment) project under study for the LHC (Large Hadron Collider) at CERN includes a software team which has already started the development of a unified framework for their collaboration team . This has led to the definition of a complete environment (ALI ROOT, or ALICE using ROOT) where the software being developed off-line by different experiment groups is integrated (Brun et aI. , 2000).

ALI ROOT is the ALICE software team ' s name for a virtual Monte Carlo interface. This is a ROOT-based software infrastructure that can employ any adapted radiation transport code as the physics engine. Examples could be GEANT3 .21, GEANT4, or FLUKA. Figure 1 shows a logical schematic of the ALIROOT architecture.

FIGURE 3. Space Shuttle example of a Finite Element Model (FEM) with seven classes of external material composition.

In a sense, ALIROOT can be thought of as getting all of the physics transport engines for the price of one, plus the GUI infrastructure of ROOT. A simulation session can start with an old-fashioned Monte Carlo like GEANT3, and its rough approximations can then be refmed using the more rigorous FLUKA engine. Obviously, such a capability

., - - --~'~ '- - --- ~

will be very advantageous for transitioning the old FORTRAN (Formula Transformation) users of GEANT onto the new ALI ROOT technology of tomorrow using C++ scripting language.

FROM FEMs TO CADs - GEOMETRY GENERA TORS

A prominent piece in the perplexing puzzle of space radiation analysis is the geometry of the spacecraft itself As with the design of experimental detectors and their instrumentation (e .g., ALICE) in the beamline of the LHC at CERN, a very sophisticated representation of the "target" structure is required. Not only is a rigorous depiction of the dimensions in the geometry necessary , but an elaborate definition of the isotopic elemental composition in the structural material is also mandatory in order to arrive at the correct neutron albedo, for example.

[t was realized during the course of the present study that most of the level of detail required in the definition of "target" structure. geometry, and composition is contained in what are called FEMs (Finite Element Models) by the aerospace engineering community. To qualify any spacecraft structure or payload for launch, safety reviews and flight readiness reviews are required. In order to pass these, mechanical engineers must demonstrate that the body­bending modes have resonance amplitudes below a certain magnitude. To accomplish this task, FEMs are defined and used in the design, development, and test phase of spacecraft evolution. Mass properties and inertia tensors are modified until the flight article is brought into compliance for said reviews. An additional feature of the FEMs is their three-dimensional nature which will help in exercising the full power of FLUKA as a physics engine.

FIGURE 4. Example of an ISS module FEM.

Examples of Shuttle and ISS FEMs are given in Figure 3 and Figure 4. We are presently studying the file transformation requirements for converting FEMs into a standard Auto-CAD (Computer Aided Drawing) fonnat. With the proper geometry database interface illustrated in Figure 2, a FEM-to-CAD conversion is envisioned as a means for defining the target geometry for FLUKA at a high level of accuracy. Supplemental work will have to be done to the FEMs since minor substances such as insulation are not included and these can prove important because of their elemental hydrogen content. Finally, a FEM strategy solves the additional problem of configuration management, that is, keeping track logistically of the flight-to-flight changes in each spacecraft.

SP ACE RADIATlON TOOL KIT

One objective of the present research, of course, is to produce a tool for space radiation analysis. An overview of such an eventual "tool kit" is summarized here.

• Inputs (via ROOT and GUI-based or file edit). • Mission parameter formats (dates, times, orientations, orbital parameters, etc.). • Built-in "standard' GCRs (galactic cosmic-rays) with solar modulation, trapped belt models (AP8 and

successors), and "standard" solar flare models. • SpacecraftlHabitat geometry (long-term goal of direct conversion from industry standard CAD fonnats) . • Special scoring (bui[t-in human body geometry models).

---~.--.- .-~ ----~~----

• Standard classes will be available. • Special classes can be created and saved for inclusion in future standard libraries. • Objectivity will greatly increase the ease with which new classes can be created.

• Outputs (direct GUI and file creation). • Online monitoring of progress during execution. • Histograms (one-dimensional and two-dimensional). • 3-D plots of parameters such as dose versus geometry. • History files (to speed up sub-geometry re-runs) . • Temporal evaluations (e.g., dose versus mission-elapsed-time).

• Run Times. • Execution times scalable with parallel processing (Beowulf). • 2x increase with 64-bit processors (initial goal of 32-to-128 nodes). • Goal- nominal full statistics runs in a few hours ofreal time. • Goal - one-hour sub-runs based on indexed history files. • Realistic near-term prospects : - Over-night.

CONCLUSIONS

The present status of our three-year investigation into a Monte-Carlo-based s imulation of the space radiation environment has been summarized. Progress is being made at civilizing the very sophisticated physics engine known as FLUKA by using ROOT as the user interface and analysis infrastructure. This is being accomplished in collaboration with the ALICE software team at CERN and their concept of a virtual Monte Carlo interface known as ALIROOT. Additional accomplishments include the identification of FEMs as a probable database source for preliminary definition of spacecraft and habitat geometry. We are much closer to our goal of providing a standard user-friendly, GUI-based simulation of space radiation which is well-supported, portable, free, and available as open source code to the user community.

ACKNOWLEDGMENTS

This research is primarily funded through the Marshall Space Flight Center under NASA grant NASG8-1658 . Additional support is provided by the State of Texas under the Joint JSC-University of Houston Post-Doctoral Aerospace Fellowship (PDAF) program. Further support from the Institute for Space Systems Operations at the University of Houston is also acknowledged. Figures 3 and 4 are adapted from the public website for NASA's Hypervelocity Impact Facility (HITF) available at http ://hitf.jsc .nasa.gov, and these are shown with permission.

REFERENCES

Badhwar, G., Keith, J., and Cleghorn, T., "Neutron Measurements Onboard the Space Shuttle," Rad. Measurements, to appear (October 2000).

Brun, R., et aI. , GEANT, Geneva, CERN, 1987. Brun, R. http://root.cern.ch/, 1995. Brun, R., "Root, An Interactive Object-Oriented Framework and its Application to NA49 Analysis," in proceedings

of Computing in High-Energy Physics (CHEP) - 1997, Elsevier, 1997. Brun, R., Carminati, F. , and Rademakers, F., "The ALICE Off-line Strategy: A Successful Migration to 00," in

proceedings of Computing in High-Energy Physics (CHEP) - 2000, to be published by Elsevier. Pinsky, L., Carminati, and Ferrari, A., "Simulation of Space Shuttle Neutron measurements with FLUKA," Rad. Measurements,

to appear (2000a). Pinsky, L., Wilson, T. , Ferrari, A. , Sala, P., Carrninati, F., and Brun, R., "A Monte-Carlo Simulation of Space Radiation using

the FLUKA and ROOT Codes," in proceedings of First International Workshop on Space Radiation Research, Arona, Italy (European Science Institute, May, 2000b), to be published in Physica Medica.

Pinsky, L., Carminati, F., Brun, R., Ferrari, A., Sala, P., Sill, A. , Badhwar, G., MacGibbon, 1., and Wilson, T., "Developing a Space Radiation Simulation and Analysis Package," in Space Technology and ApplicatiOns International Forum -1999, edited by M. El-Genk, AlP Conference Proceedings 458, New York, 1999a, pp. 328-433.

Pinsky, L., MacGibbon, 1., Badhwar, G. , and Wilson, T., "A Space Radiation Monte-Carlo Computer Simulation using the FLUKA Code," in proceedings of Space Radiation Environment Workshop, Farnborough, United Kingdom, 1999b (to be published in 1. Geophys. Res.).

Wefel, J. , and Wilson, T. , et aI. , "Advanced Cosmic-ray Composition Experiment for Space Station," in Space Technology and Applications International Forum - 1999, edited by M. El-Genk, AlP Conference Proceedings 458, New York, 1999a, pp. 245-250.

Wefel, J., and Wilson, T., "The ACCESS Mission: ISS Accommodation Study," Proc. 26th International Cosmic Ray Coni. Salt Lake City, 1999b.

Wilson, T., and Wefel, 1., "ISS Accommodation Study for the ACCESS Mission," Bull. American Physical Society, Vol. 44, 1999c, pp. 1294-1295.