Loading...

Study of the Space Radiation Environment in GEO

Examination Thesis 2009 98 Pages

Physics - Electrodynamics

Excerpt

TABLE OF CONTENTS

Word Count

Dedication

Acknowledgement

Abstract

Table of Contents

List of Figures

LIST OF TABLES

1 Introduction
1.1 Background and Context
1.2 Scope and Objective
1.3 Achievements
1.4 Organization of the report

2 Space Radiation Environment in Geostationary Orbit
2.1 Introduction
2.2 History of the Space Environment Research
2.3 Space Environment Elements
2.3.1 Space Plasma Environment
2.3.2 Meteoroids/Orbital Debris Environment
2.3.3 Space Radiation Environment

3 SPACE RADIATION Environment MODELS
3.1 Trapped Radiation Models
3.2 Solar Proton Event Models
3.3 GCR Flux Models

4 Space Radiation Effects on Spacecraft, Design mitigation techniques and design guidelines
4.1 Space Radiation Effects
4.1.1 Total Ionization Dose (TID) effects
4.1.2 Single Event Effects (SEE)
4.1.3 Displacement Damage Effects
4.1.4 Surface Charging
4.1.5 Deep Dielectric Charging
4.2 Geostationary Spacecraft Anomalies
4.3 Spacecraft Design Mitigation Techniques
4.3.1 Mitigation of Single Event Effects
4.3.2 Mitigation of Total Ionizing Dose, Displacement Damage and Surface Charging
4.3.3 Spacecraft Deep Dielectric Charging Mitigation Techniques
4.4 Spacecraft Radiation Environment Design Guidelines

5 Modelling the Space Radiation Environment in Geostationary Orbit
5.1 Variation of GEO Radiation Environment
5.1.1 Variation of GEO Radiation Environment with Longitude
5.1.2 Variation of GEO Radiation Environment with Solar Cycle
5.1.3 Diurnal Variations of GEO Radiation Environment
5.2 Comparison between AE-8 Max and IGE-2006 Trapped Electron Models
5.3 Comparison of Solar Flare Proton Models

6 PREDICTION of the Space Radiation Environment for Paksat-1R Geostationary Communication Satellite
6.1 Step 1: Creating a Project in SPENVIS
6.2 Step 2: Simulation of Trapped Radiation Environment
6.3 Step 3: Simulation of Solar Energetic Protons (SEPs) for Paksat-1R
6.4 Step 4: Calculation of Dose-Thickness Curve
6.5 Step 5: Estimating solar cell degradation and cover glass thickness
6.6 Step 6: Simulation of Galactic Cosmic Rays (GCR) Environment for Paksat-1R
6.7 Step 7: Paksat-1R Single Event Upset Rates Calculation

7 Conclusion
7.1 Summary
7.2 Conclusion
7.3 Evaluation
7.4 Future Work

References

Appendix I - Work plan

Appendix II - Trapped Radiation Belt Models-I

Appendix III - Trapped Radiation Belt Models-ii

Appendix IV - Solar Proton Event Models

Appendix V - SPACE ENVIRONMENT INFORMATION SYSTEM (SPENVIS)

WORD COUNT

Number of Pages: 98

Number of Words: 24802

DEDICATION

I dedicate this work to my father Ghulam Mustafa, mother and brother Ghulam Murtaza (Pakistan Air Force) whose prayers and encouragement helped me complete this work.

ACKNOWLEDGEMENT

All praise is to ALLAH Almighty than whom there is no god. I convey my modest gratitude to Him for the courage, determination, intellect and innumerable blessings He bestowed upon me to complete my work. Glories and Durood-o-Salam upon Prophet Muhammad (SAW) whose love enabled me achieve my targets.

It’s my pleasure to acknowledge the cooperation, guidance and suggestions from all those who made this work precious and valuable. Special thanks to my Supervisor Dr. Craig I. Underwood for his motivation, encouragement, valuable comments and remarks to improve my work and timely finish of the project.

I am cordially thankful to Dr. Phil Palmer for his significant and appreciable comments at the time of interim interview of the project.

Last but not least, I am thankful to my family, all the colleagues and friends for their incalculable support and encouragement.

ABSTRACT

The Space radiation environment in GEO has always been a severe challenge to the spacecraft industry. The Spacecraft-Environment interaction has been the topic of deep investigation since 1970s to onwards. Very harsh space environment affects the spacecraft in various ways. The current project presents an overview of the characteristics of space radiation environment, its effects on spacecraft electronics and spacecraft operations. The elements of the space radiation environment such as Galactic Cosmic Rays (GCRs), Solar flare protons and trapped electron belt in GEO are explained comprehensively. The effects of hazardous space radiation environment on a GEO spacecraft including spacecraft charging, Total Ionizing Dose (TID), internal charging and Single Event Effects (SEE) are introduced with necessary details. The space radiation environment models currently available are critically analysed and explained in the light of the work of different space researchers. The limitations and risks involved with these models are briefly introduced. The spacecraft design mitigation techniques and design guidelines are presented to help the spacecraft community build the spacecraft capable of surviving in hazardous radiation environment. Then some case studies of GEO satellite anomalies are also briefly explained. The ESA based Space En- vironment Information System (SPENVIS) software package is utilized for analyzing the temporal, spatial and diurnal variations of radiation environment in geostationary orbit and the simulation results are compared with GOES data. A detailed space radiation environment analysis for a Paki- stani geostationary communication satellite Paksat-1R has been undertaken including the trapped electron flux estimation, solar proton flux estimation, Solar cell degradations and cover glass re- quirements, dose-depth curve for estimating the shielding level required for spacecraft protection and SEU rates calculations of different devices. The necessary comments and suggestions are also included with the analysis.

Keywords: Space Radiation Environment, Spacecraft effects, GCRs, SEU, SPENVIS

LIST OF FIGURES

Figure 2.1: K.E of particles versus particles diameter

Figure 2.2: Surface damage due to MMOD impacts on the LDEF

Figure 2.3: Composite motion of particles trapped in the magnetic field lines

Figure 2.4: Van Allan Radiation belts

Figure 2.5: Van Allan belt equatorial trapped particle flux versus altitude in earth radii

Figure 2.6: L-shell parameter

Figure 2.7: NOAA/NASA Solar Cycle 24 prediction panel results

Figure 2.8: GCR composition

Figure 2.9: Differential solar energy spectra at solar minimum and solar maximum

Figure 3.1: The sunspot number from 1960-2007. The blue line represents solar maximum while the green line represents solar minimum defined by JPL model

Figure 3.2: Comparison of CREME-86 and CREME-96 solar heavy ions models for GEO

Figure 4.1: Voltage shift caused by irradiation

Figure 5.1: L-values are plotted with respect to corresponding Longitude position of GEO satellite using AE-8 Max model

Figure 5.2: Change in L-parameter with respect to the longitude of a GEO satellite using CRRESELE model

Figure 5.3: The comparison of the L-parameter calculated by AE-8 Max model and CRRESELE model

Figure 5.4: Flux variations as a function of Longitude of geostationary satellite

Figure 5.5: Variations of solar proton flux of energy 200 MeV as a function of satellite year of operation

Figure 5.6: The solar proton flux data using ESP model

Figure 5.7: The solar proton data measured by the GOES satellite

Figure 5.8: The GOES data of grouped solar flares from 1976-

Figure 5.9: GOES satellite GCR Plot

Figure 5.10: GOES 3-hour prediction electron data plot

Figure 5.11: Comparison between the electron flux deduced from AE-8Max model and electron flux from IGE-2006 model in the lower, mean and upper cases

Figure 5.12: Comparison of the SPE fluences estimated by King’s, JPL 91and ESP models

Figure 6.1: Energy spectra of trapped protons for Paksat-1R using AP-8 MIN model

Figure 6.2: Energy spectra of trapped protons for Paksat-1R using AP-8 MAX model

Figure 6.3: The trapped electron spectra for Paksat-1R

Figure 6.4: Trapped electron spectra by using AE-8 MAX model calculated for Paksat-1R

Figure 6.5: Solar proton spectra predicted for Paksat-1R mission

Figure 6.6: Dose vs. thickness curve for Aluminium material over 15 years of satellite mission life time

Figure 6.7: Paksat-1R shielded LET spectra

LIST OF TABLES

Table 2.1: Space Debris Population

Table 2.2 : Solar Proton Events Classification with respect to their Energy

Table 2.3: Number of Observed Events by Solar Years

Table 6.1 Trapped Electron Peak Flux predicted for Paksat-1R

Table 6.2: Solar Energetic Protons Predicted for Paksat-1R Mission

Table 6.3 Solar Cell Displacement Damage Prediction for Paksat-1R

Table 6.4 LET vs. Fluence Level for Paksat-1R

Table 6.5 SEU Rates Calculated for Paksat-1R

Table 6.6: SEU rates of various devices

1 INTRODUCTION

1.1 Background and Context

The Space radiation environment in geostationary orbit has always been an indispensable challenge to the spacecraft industry. The Spacecraft-Environment interaction has been the topic of deep investigation since 1960s to onwards. The adverse effects of spacecraft-environment interactions include radiation damage, degradation of materials, Single event effects, thermal changes, contamination, excitation, spacecraft glow, surface charging and Internal charging etc. Space researchers have developed many codes and models to analyze these severe effects and have prepared standard guidelines and procedures for designing space systems.

1.2 Scope and Objective

This project intends to present the space environment in geostationary orbit (GEO) explaining the space environment parameters such as space radiation, space plasma, space debris and micrometeoroid and then space radiation analysis of a real GEO satellite.

The objective of this project is to generate a guide explaining;

1) Space radiation environment at GEO; explaining trapped electron belt, Solar Proton Events (SPEs) and Galactic Cosmic Rays (GCRs), their effects on GEO spacecraft alongwith the mitigation techniques and the critical analysis of the models currently available.
2) Analysis of the spatial and temporal variations of space radiation environment in GEO and comparing these variations with GOES satellite data.
3) As a case study; the space radiation analysis of a Pakistani geostationary communication satellite Paksat-1R including trapped radiation flux estimations, solar proton flux calcula- tions, Dose-Depth curve and SEU rates of onboard electronics.

The outcome of the project is a generic guide useful for the fresh engineers and scientists in the field of space mission planning and designing.

1.3 Achievements

Having a good eye on the literature review, major elements of Space environment in GEO have been studied and documented in this report alongwith their effects on spacecraft system design and operations. The spacecraft design mitigation techniques and Design guidelines for the spacecraft are presented according to the recommendations of NASA and ESA etc.

The variations of GEO space radiation environment with the spatial and temporal variations of the GEO spacecraft are analyzed successfully and the results are presented in an understandable way with comparison to the GOES data. Then the core objective of this project i.e. estimating the expected radiation environment for a Pakistani geostationary communication satellite Paksat-1R is carried out. The expected trapped particle flux, solar protons and GCRs heavy ions are simulated for Paksat-1R.All this analysis is carried out according to the recommendations of European Cooperation for Space Standardisation (ECSS) standard document.

SPENVIS online software is used for estimating radiation environment for Paksat-1R. The SEU rates are calculated and the dose- depth curve is plotted for Paksat-1R. The shielding level required to protect the spacecraft from hazardous radiation environment is calculated. The feasibility of the COTS devices to be used in Paksat-1R design is studied technically. Hence, completing the radiation calculations for Paksat-1R.

All the goals set for the project are achieved in a true spirit and within the time scale.

1.4 Organization of the report

This report is divided into seven chapters including all the necessary explanations. Chapter 1 presents the introduction of the dissertation and report.

Chapter 2 of this report presents an overview of the space environment at GEO discussing the space plasma, space debris, micrometeoroid and space radiation environment. The effects of each of the space environment element are acknowledged along with the satellite anomalies due to the respective element. Then the GEO radiation environment is discussed in detail. All the parameters of the radiation such as Trapped Radiation particles, Solar flare protons and Galactic Cosmic Rays (GCRs) are explained.

Chapter 3 explains the models developed for analyzing the radiation environment in detail. The limitations and risks involved with these models are also presented.

Chapter4 presents the effects of the radiation environment on spacecraft such as Total Ionizing Dose, Displacement Damage, Spacecraft charging and Single Event effects. Then Spacecraft design mitigation techniques and Spacecraft Design guidelines are presented.

Chapter 5 deals with modelling the space radiation environment in GEO. First, the variation of the space radiation environment with the variations in longitude of the spacecraft is analyzed by internal and external magnetic field consideration. Second, the variations of space radiation environment with the variation of solar activity are analyzed and the results are compared with GOES data. Then the radiation models are compared with each other.

Chapter 6 deals with estimation of expected space radiation environment for Pakistani geostationary communication satellite Paksat-1R.

Chapter 7 presents the Summary, Conclusion, evaluation and future work of the project.

2 SPACE RADIATION ENVIRONMENT IN GEOSTATION- ARY ORBIT

2.1 Introduction

Geostationary orbit (GEO) is a circular orbit at an altitude of 35786 km. The spacecraft placed in this orbit moves with same speed as that of earth hence appearing stationary with respect to earth. Solar flares and solar wind have unobstructed access to this specific orbit and the solar parti- cles emitted by sun affect the space systems in terms of total radiation dose and Single event phenomena. The spacecraft orbit drives the space environment effects encountering the system.

The space radiation environment is very dynamic; it varies by the spatial (longitude) and temporal (solar cycle and diurnal) variations of the satellite. These variations have their corresponding impacts on the satellite. Therefore, it is very important to analyze the space radiation environment of the satellite in the mission planning stage.

2.2 History of the Space Environment Research

Since the launch of the first satellite Sputnik in 1957, the space environment research takes its roots from the discovery of the existence of toroidal belts of energetic protons and electrons by Van Allan in 1962.This discovery gave rise to the race of studying the space radiation, as a result many space missions were flown with onboard radiation monitors. The researchers introduced many books on space environment meanwhile the effects of outgassing and particulate contamination on some spacecrafts were observed. For example, on a mariner mission star tracker lock was lost that was traced to particulate contaminants in the spacecraft near field. On the Gemini missions, the windows became contaminated due to the use of Silicones in the gaskets. In the Apollo missions, mass spectrometers observed high background signals.

In the light of above discussions, it can be noted that spacecraft community had got awareness of different elements of space environment like radiation effects, micrometeoroids plasma interactions and contaminations.

In 1970’s, the demand of commercial spacecrafts in geostationary orbits got increased because of their capability of being the best source for communication. Unfortunately, some of the geosta- tionary satellites got victimised of the harsh space environment. As a result, the researchers put their attention towards Geo orbit space environment. The initiative towards simulation codes was taken by NASA & US Air force by developing the NASCAP (NASA charging and analysis pro- gram).NASA launched a fully dedicated satellites called SCATHA in 1979.USSR developed two simulation codes akin to NASCAP .

In 1980’s, the SCATHA results were published (Koons, 1983).These results were simulated in NASCAP Geo charging Program. The researchers developed a geo spacecraft anomalies database. NASA introduced a comprehensive design guidelines document which is still a great source in satellite communication (Purvis et al, 1984).

In 1990s, the emphasis has been given on modelling the space environment. Models are being developed and further improved. Moreover, the space researchers are working on the environment effects on space mission design.

2000s is also seeing the improvements in understanding of the effects and mitigation tech- niques. Especially the Single event effects and charging effects are the areas of research in current years .New techniques are being developed for better functioning of the spacecraft components and systems.

2.3 Space Environment Elements

Three important elements of the space environment are briefly explained here.

1) Space plasma environment
2) Meteoroids and Space debris environment
3) Space radiation environment

The first two elements are presented in a short description and the radiation environment, being the theme of the current project, is explained in detail.

2.3.1 Space Plasma Environment

Plasma is produced when an atomic electron receives enough energy to escape the electrical at- tractions to the nucleus (A.C. Tribble, 1995). It results in a mixture of negative and positive ions. Mostly, in plasma, electron density ne is equal to the ion density ni and is simply referred to as plasma density n0. Over 99% of the universe, the sun and the stars, is plasma (A.C. Tribble, 1995). A spacecraft at GEO is on the edge of the plasmapause (Hastings, 1996). A GEO spacecraft mostly will be in tenuous, relatively cool plasma characteristic of the outer magnetosphere. This region, however, is characterized by sudden injections of high substorms. These events are thought to be the main cause of surface charging on GEO spacecraft (Hastings, et al. 1996).

Space Plasma Effects on Spacecraft

The plasma environment that a spacecraft will encounter depends upon its orbit (Hastings, et al. 1996). For a spacecraft in GEO, the charging environment is worse than in LEO because the plasma is much more active. Major plasma effects follow from the slow accumulation of the charges on surfaces (Hastings, et al. 1996). This accumulation of surface charging produces electrostatic fields that extend from surfaces into space and can result in a number of adverse interactions (Hastings, et al. 1996) such as;

i. Surface arc discharges that generate electromagnetic interference cause surface damage, in- duce currents in electronic systems, produce optical emissions and enhance the local plasma density;
ii. Enhanced contamination leading to changes in surface, thermal and optical properties,
iii. A shift of the spacecraft electrical ground, leading to problems with detectors collecting charged particles from the environment;
iv. Coulomb forces on the spacecraft components and materials as well as modifications of the drag coefficients and electromagnetic torques on the spacecraft.

Representative Cases

Intelsat K: Intelsat K suffered a disastrous electrostatic discharge. Due to this discharge, Momentum wheel control circuitry got disabled causing the spacecraft to wobble and produced antenna coverage fluctuations (K.L. Bedingfield, et al. 1996).

BS-3A: This satellite had to face severe electrostatic discharge on February 22, 1994 causing 60 minute telemetry outage (K.L. Bedingfield, et al. 1996).

GMS-4: This is a Geostationary Meteorological Satellite owned by Japan. It also faced electrostatic discharge in January and July 1991. Due to this discharge, the VIS-IR Spin Scan radiometer entered an unscheduled gain setting (K.L. Bedingfield, et al. 1996).

2.3.2 Meteoroids/Orbital Debris Environment

Our solar system comes with its own naturally occurring background of dust that results from the backup of comets, asteroids etc (A.C Tribble, 1995). These naturally occurring backgrounds are called micrometeoroids (MM). Since the launch of the Sputnik I, mankind has been creating a new cloud of orbiting particles, which are the left over pieces of nonoperational spacecrafts, boost stages, solid fuel particles, etc. The artificial environment is referred to as orbital debris (OD).MM and OD carry large kinetic energy associated with impacts at orbital velocities (A.C Tribble, 1995).

Assuming a density of 1g/ cm3, particulate matter impacting at this speed will carry the kinetic energies shown in Figure 2.1. Figure 2.2 shows the impact of MMOD upon surface of LDEF. The features of a hypervelocity collision depend upon impact velocity depending upon the state of the impacting particle; the physical processes responsible for transferring KE to the target may also differ (A.C Tribble, 1995).

illustration not visible in this excerpt

Figure 2.1: K.E of particles versus particles diameter (adapted from A.C. Tribble, 1995)

illustration not visible in this excerpt

Figure 2.2: Surface damage due to MMOD impacts on the LDEF (source, NASA, 2007)

Upto 2007, the state of the space debris is shown in Table 2.1.

Table 2.1: Space Debris Population (Source, NASA, 2007)

illustration not visible in this excerpt

2.3.3 Space Radiation Environment

Space radiation environment is a very important consideration for the space mission design. A lot of research has been made in this area.

It consists mainly of three broad elements;

1) Trapped radiation environment
2) Solar Proton Events (SPEs)
3) Galactic Cosmic Rays (GCRs)

These elements are explained in detail alongwith the models development for analysing the respective elements.

I. Trapped Radiation Environment

A charged particle is constrained to gyrate around magnetic field lines (A.C Tribble, 1995). In the earth’s polar region the magnetic field lines converge, increasing the local magnetic field strength. At the point where the magnetic field strength is

illustration not visible in this excerpt

A particle of mass m, velocity v, and magnetic moment μ will be reflected back in the initial di- rection of travel in order to conserve energy. For this reason particles travelling along the magnetic field lines are trapped to gyrate back and fourth along the earth’s magnetic field lines, as shown in Figure 2.3 .There are three types of motion which the trapped particles exhibit, gyration about mag- netic filed lines, motion of the guiding centre up and down the magnetic filed lines and the guiding centre drifting longitudinally with the earth (SPENVIS help). Theoretically, a particle may remain trapped forever in the magnetic field (A.C Tribble, 1995). However, by scattering, a particle may eventually be moved to higher or lower orbits or may be deflected along the magnetic field lines.

illustration not visible in this excerpt

Figure 2.3: Composite motion of particles trapped in the magnetic field lines (adapted from SPENVIS.oma.be/ help)

The inner (proton) belt is centred about an altitude of approximately 6000 miles and was dis- covered first. It contains mainly protons, some electrons and heavy particle ions whose energies are greater than 30MeV (D.L. Chenette, et al. 1984).The protons, especially the high energy compo- nent, are produced mainly by the decay of high energy neutrons leaking up out of the atmosphere (Messenger et al. 1997).

The source of these neutrons is the result of collision of cosmic ray ion with the oxygen and nitrogen nuclei of the atmosphere. The low energy neutrons are produced principally in the polar atmosphere by the same reaction. However, here the source neutrons are produced by solar proton interaction with upper atmosphere O2 and N2, analogous to cosmic ray neutrons producers for the high energy neutrons (NASA, 1996). The geomagnetic axis is offset and tilted with respect to the earth’s rotation axis, this tilt give rise to a phenomenon of higher radiation called South Atlantic Anomaly (SAA) as shown in Figure 2.4. Since, magnetic field is weak in SAA; particles penetrate deeper into the earth’s atmosphere in this region (Rami Vainio et al. 2009).

illustration not visible in this excerpt

Figure 2.4: Van Allan Radiation belts (adapted from

http://www.esa.int/TEC/Space_Environment/SEMEF3T4LZE_0.html)

However, in the outer Van Allen belt, centred at about 20,000 miles altitude, electrons are the major particles, and their levels fluctuate greatly with time, especially with solar wind induced geomagnetic storms (Messenger et al. 1997).

The electron flux seems to be in synchrony with geomagnetic storm activity, reflecting the in- fluence of the storms on the outer belt electrons. Also the mean life time for electrons in the outer belt is about 2-3 days, whereas their corresponding life time in the inner belt is about 400 days. At higher altitudes in the inner belt, the particle losses are mainly due to interaction with the ambient electromagnetic field, ultimately suffering the same fate as those in the lower part of the belt (Mes- senger et al. 1997).

As an aside, the boundary between inner and outer belt is somewhat arbitrary, as shown in Fig- ure 2.5.

illustration not visible in this excerpt

Figure 2.5: Van Allan belt equatorial trapped particle flux versus altitude in earth radii. (Adapted from E.G. Stassinopoulos, et al. 1983)

It is well known that the source of the outer belt electrons is not from the decay of the neutrons as in the inner belt (Messenger et al. 1997). It is generally claimed that electrons are injected into the outer belt from the ambient electron density in and around the solar wind shock wave outside the magnetopause. This is especially the case during a magnetic storm. As mentioned, the latter causes the atmosphere field to fluctuate greatly thereby producing corresponding electric fields (faraday’s law), which accelerate the electrons from beyond the magnetopause into the magnetosphere across its magnetic field lines (Messenger et al. 1997).

That the outer belt density fluctuates greatly during magnetospheric storm lends credence to this description of outer belt electron sources (Messenger et al. 1997). The electron energy spectrum (5MeV) is such that the electrons can be considered as fast (relativistic) particles. They present a serious radiation hazard for spacecraft orbits in these regions and may constrain the deployment of new semiconductor technologies in spacecraft.

The trapped radiation environment is generally represented in the form of B, L coordinates. So, it is very important to know the B, L coordinates. These are explained next.

B, L Coordinates

McIlwain (McIlwain, 1961) introduced the concept of the magnetic shell parameter, L. Given a certain field line, it is possible to imagine the particles associated with this line tracing out a shell as they drift around in longitude. This shell can then be labelled by an L number which (in a perfect dipolar field) is simply the radial distance at which it crosses the magnetic equator (McIlwain, 1961). Thus the L=2 shell crosses the magnetic equator at 2 earth radii. The second coordinate is the magnetic field strength B, which varies with position within the shell.

The L shells can be used to label the belts directly. Thus the heart of inner zone electron population is centred on the L=1.4 shell, whilst the outer zone electron population is centred on L=4.5.The L=2.3 shell marks the boundary between the inner and outer zones. The proton population forms a single zone centred on the L=1.7 shell.

Surfaces of constant B (magnetic field intensity) are concentric, roughly ellipsoidal shells encircling the Earth, while surfaces of constant L approximate the concentric shells generated by dipole field lines rotating with the Earth (Martin V. Zombick, 2007) as shown in Figure 2.6.

illustration not visible in this excerpt

Figure 2.6: L-shell parameter (adapted from Knecht et al. 1985)

B and L can be approximately mapped into polar coordinates by means of the following trans- formation:

B=M [4-(3R/L)]1 /2 /R3 R = Lcos 2 L

(Where M is the magnetic dipole moment of the earth). Thus a radial distance R and a "latitude” L may be computed (Knecht et al. 1985).

II. Solar Proton Events (SPEs)

A solar flare is a giant eruption of gases and plasma on the surface of the sun, whose arcs can extend to altitudes of more than 100,000 miles above the sun (Messenger et al., 1997). They are also characterized by photon bursts including X-rays, line spectra and microwave emission upto hundreds of MHz. As a result of this event, magnetic storms are produced on earth which cause solar plasma shock waves incident on the earth’s magnetosphere. These events cause extreme distortion of the geomagnetic field lines so that at high altitudes, the magnetic rigidity cut off can drop to negligible values, this means that the energy spectrum, and so the flux, of allowed cosmic ray ions is greatly increased (Messenger et al., 1997).

“A solar cycle is an 11 year period in which the sun is in “active phase” and “quite phases”. The solar active phase is a 7 years of period in which the solar proton fluence is enhanced to hazardous level, the two years before the peak fluence year and 4 years after it. The quite solar phase is the 4 years period in which the solar proton fluence is low enough to be neglected” (Feynman et al., 1990).

Energetic particles are accelerated in intermittent events at the sun associated with solar activity. SPEs occur occasionally throughout the solar cycle. A large scale research carried out in 1980s and 1990s (NRC, 2008).On the basis of a research work by Reames, et al. 1990; Cliver and Cane, 2002, Tylka and Lee, 2006, solar proton events were classified into two major categories, i) grad- ual and ii) impulsive.

In gradual SPEs, shocks driven by fast Coronal Mass Ejections (CMEs) are the dominant accelerators (NRC, 2008).

In Impulsive SPEs, particle acceleration is due to magnetic reconnection process. These SPEs are small in intensities, have short duration, energy lower upto the value that is unable to penetrate the typical shielding and are observed over a narrow range of solar longitudes. Impulsive SPEs are also characterized by distinctive patterns of enhancements in heavy ions (Reames, 2000a; Reames and Ng, 2004; Mason et al., 2004). Impulsive SPEs have low particle flux.

Usually, more than 90 percent of the energetic ions produced in an SPE are protons (NRC, 2008). That’s why protons are the major concern while evaluating SPE radiation hazard.

The solar activity from 1996-2020 covering two solar cycles is shown in the Figure 2.7.

illustration not visible in this excerpt

Figure 2.7: NOAA/NASA Solar Cycle 24 prediction panel results (Adapted from http://www.swpc.noaa.gov/SolarCycle/SC24/index.html)

When the solar activity starts, the intensity of the penetrating particles increases dramatically within minutes. At this earlier times, the particle is said to be anisotropic i.e. more particles from one direction can arrive than from other direction. The peak direction is generally towards inter- planetary magnetic field. Then after a short time, the particle becomes isotropic i.e. there is no preferred direction of flux. This depends largely on the energy of the particles (NRC, 2008).

Generally, there is no hard limit to the energy of SPEs; it varies from event to event.

The energy distribution also varies substantially from event to event. Normally, the most considerable energy range is from a few tens of MeV to a few hundred MeV. “Soft” events have a larger proportion of particles with lower energy. “Hard” events have more than the average proportion of high-energy particles.

E.G.Stassinopoulos et al. (1996) studied and analyzed the data taken from different spacecrafts in geostationary orbit for the solar cycle 20, 21 and 22 and presented his results in an excellent way. He classified the solar flare events in five major categories which are listed and defined in the table 2.2 and the number of these events occurred during solar cycle 20, 21 and 22 are shown in table 2.3.

Table 2.2 : Solar Proton Events Classification with respect to their Energy (Adapted from E.G.Stassinopoulos et al. 1996)

illustration not visible in this excerpt

Table 2.3: Number of Observed Events by Solar Years (adapted from

E.G.Stassinopoulos et al. 1996)

illustration not visible in this excerpt

III. Galactic Cosmic Rays (GCRs)

Galactic Cosmic Rays (GCRs) are continually and isotropically incident on the solar system (Kamide, 2007). These are of concern because it is well known that these heavy ions are certainly incident on integrated circuits in spacecraft (Messenger et al. 1997).They supply charge by producing ionization tracks within these microcircuits (Messenger et al. 1997).GCRs compose of almost all of the naturally occurring elements: 90% of GCRs are protons (Hydrogen), nearly 10 % are Helium and the remaining are the atoms heavier than helium (NRC, 2008).

This composition is shown in the Figure 2.8:

illustration not visible in this excerpt

Figure 2.8: GCR composition. (Adapted from NASA Goddard space flight centre, http://helios.gsfc.nasa.gov/ace/abund_plot).

In 1912, Hess discovered that cosmic rays are incident upon earth from space (Kamide, 2007). In 1954, Forbush showed that his ion chamber measurements between 1937 and 1952 were negatively correlated with the sunspot number over a solar cycle (Kamide, 2007). In 1955, J. Simpson interpreted this negative correlation in terms of the GCR modulation by the solar cycle, and suggested that the interplanetary magnetic fields may prevent GCRs from entering the solar system near the Earth’s orbit (Kamide, 2007).

Outside of the solar system, The GCR flux is considered to be constant (Lieberman and Me- lott, 2007).However, GCR had to pass through the heliosphere and space plasma in order to reach earth (Messenger et al. 1997). The Interplanetary Magnetic Field (IMF) changes with solar activity cycle; it is strongest near the solar maximum, hence GCR flux is at low point near solar maximum and at solar minimum, The GCR flux is at its highest peak (Cane et al., 1999). The IMF strength near the solar maximum is due to variations in the Coronal Mass Ejection (CME), (Cliver and Ling, 2001; Owens and Crooker, 2006).

The variations in the intensity of GCR with time depend upon variation in the GCR particle en- ergy and with solar activity. In the energy range of <a few GeV per nucleon, the flux decreases by 30 % -40 % from solar minimum to solar maximum (NRC, 2008) as shown in Figure 2.9.

illustration not visible in this excerpt

Figure 2.9: Differential solar energy spectra at solar minimum and solar maximum. SOURCE: Provided by P. O’Neill, NASA (Adapted from NRC, 2008).

3 SPACE RADIATION ENVIRONMENT MODELS

Introduction

The space radiation researchers have developed different models for modelling the elements of radiation environment. These models are based on the datasets of different space missions launched in different orbits. For analyzing the radiation environment for a space mission, the analysis is car- ried out using these models and then subsequent protection techniques are applied. Therefore, it is very necessary to understand the correct usage of these models and to understand the risks involved and limitations of these models.

This chapter introduces the critical analysis of all the three elements of space radiation models described as;

1) Trapped Radiation Belt Models
2) Solar Proton Event Models
3) GCR flux models

3.1 Trapped Radiation Models

NASA issued a series of radiation models denoting AE for electron and AP for protons in 1960s by Collecting the data from 43 satellites, (J. L. Barth, et al 2003). Since now eight models of this series have been released and the current version is AE8 and AP8 (J. L. Barth, et al 2003).These models are for both Solar Maxima and Minima conditions, termed as AE8 Max, AE8 Min, AP8 Max and AP8 Min. During these 27 years of continuous work, NASA released another model named as Starfish decay model (J. L. Barth, et al 2003). The current version of AP model, i. e. AP8 was released in 1976, keeping in view the need to include all the previous data and the data after 1970s (J. L. Barth, et al. 2003).The AE8 model was released in 1993 because of the fact that NASA radiation modelling budget was reduced in late 1970s (J. L. Barth, et al. 2003).Although AE8 and AP8 models are more than two decade old, These are still considered to be the preferred engineering approach for radiation modelling (Rami Vainio et al. 2009). However, several researchers have pointed out the serious problems in these models (Daly, et al. 1996).

Then NASA launched another mission CRRES to study the after effects of US and Soviet nu- clear detonations (J. L. Barth, et al. 2003). This mission measured the second peak of solar cycle 22 and measured a great magnetic activity during this phase (J. L. Barth, et al. 2003). At this time, the space community realized the need to develop the models with improved time resolution. NASA’s AP8 and AE8 models were good for long duration mission but for short term mission, a new and improved model was needed (J. L. Barth, et al. 2003).

Realizing this need, several space researchers worked in this context and on the basis of the data gathered by CRRESE mission, three empirical models were developed named as the CRRESPRO (M. S. Gussenhoven et al., 1993), CRRESELE (D. H. Brautigam et al., 1992), and CRRESRAD (M. S. Gussenhoven, et al.1992).

The CRRESPRO simulates both the active and quite conditions of the magnetosphere and is used for estimating proton levels before and after March 1991 (J. L. Barth, et al. 2003).

The CRRESELE model offers larger variation of magnetic activity giving eight conditions of magnetic activity, including worst case and average case (J. L. Barth, et al. 2003).

CRRESRAD model was developed from the data obtained by a dosimetry onboard CRRES mission (J. L. Barth, et al. 2003).

The major limitation of these models was that they were based on the data collected by CRRES mission which remained operational only for 14 months and actually covered the solar maximum. So these models are considered to be good for modelling solar maxima conditions and the larger magnetic activity (J. L. Barth, et al. 2003).

To extend the CRRESRAD model further to lower altitudes and higher latitudes, a mission named as APEX was launched; the data collected through this mission gave birth to a new model APEXRAD (M. S. Gussenhoven, 1997). APEXRAD estimates the outer zone electrons and their dynamics as a function of the magnetic activity (M. S. Gussenhoven, 1997).

Pfitzer worked to analyse the flux values for low inclination orbits that vary with the atmos- pheric density and was succeeded to plot the curves of predicted flux against the atmospheric density (K. A. Pfitzer, 1990). Later, he along with Huston analyzed the proton data collected by the CRRES and TIROS missions (D. Heynderickx, 1997). In 1998, Huston presented the first ever model that estimated the trapped protons during the whole solar cycle in lower altitude orbits (S. L. Huston, et al.1998).

In early 2000s, the researchers paid attention to model the geostationary electron flux (S. Bourdarie, et al. 2008). For this purpose, some French space researchers under the umbrella of ONERA and French space agency built the models POLE V1 (Boscher, et al. 2003) and POLE V2 (S.Bourdarie, et al. 2008).

The last version of these models is IGE-2006 (A. Sicard-Piet et al. 2007). This has been rec- ommended by the ECSS for modelling the geostationary orbit electron environment (ECSS -E-ST- 10-04C, 2008).

The tables summarizing the facts about almost all of the radiation belt models are provided as Appendix II and Appendix III.

The models which are widely used in space industry for Trapped radiation particle modelling are explained as under.

I. AE8, AP8 models

AE8 and AP8 are currently the mostly used radiation belt electron and proton models and maintain the fame as the accepted space industry standards. These models are distributed in the world by NSSDC (National Space Science Data Centre).AP8 model and AE8 model were released in 1976 and 1991 respectively (J. L. Barth, et al. 2003). They are equally applicable for low and high earth orbits. The energy ranges of AE8 and AP8 models are 40 KeV to 7 MeV and 40 KeV to 500 MeV respectively (Insoo Jun, et al. 2005). These models are available in two versions, i.e. MAX and MIN specifying the solar activity phases. In these models, the electron and proton flux distribution varies with McIlwain L parameter, particle energy E and B/Bo, where

Bo=M/L[illustration not visible in this excerpt] (Sawyer et al. 1976),

Bo is an adiabatic invariant; M is magnetic moment (D. Heynderickx, et al. 1996).

The AE8 MAX, AE8 MIN and AP8 MIN incorporate Jensen and Cain 1962 model (Jensen and Cain, 1962) for internal magnetic field and have no model for external magnetic field (SPENVIS help). The AP8 MAX incorporates GSFC 12/66, updated to 1970 (Cain et al., 1967).The data set for these models is collected by the satellites in late 1960’s and 1970’s such as Azur satellite, 1969, ATS 1969 and ATS 1974.

These models have been the first choice for the space mission planners. But the research has proved some serious problems in these models.

Limitations of AE8 and AP8 Models:

1) The data mostly used for the development of AE8 and AP8 model is based on the German AZUR satellite data (Hovestadt et al. 1972) which incorporates some problems with B, L coordinates calculations resulting in problems with AE-8 and AP-8 models. AP8 MAX model should be used in association with GSFC 12/66 model updated to 1970 (Cain et al., 1967) while AE8 can be used with Jensen and Cain 1962 model (Jensen and Cain, 1962) (D. Heynderickx et al., 1996).
2) In 1986, McCormack (1988) pointed out that there are great differences in dose estimations with NASA models.
3) Konradi et al. (1987) stated that if the (B,L) coordinates are calculated by using IGRF 1975 geomagnetic model, the AP-8 MIN model gives the ridiculous results of non zero proton flux below sea level.
4) D Heynderickx et al., (1995) reported that the Azur Satellite data which was used for developing AP-8 Max model covers only 3 months and not the whole six months as as- sumed in NSSDC and Daly et al. 1996 pointed out the errors in AP-8 MIN source code.
5) Konradi et al (1992) reported that the geographical location of South Atlantic Anomaly (SAA) is changed causing variation in geomagnetic field. SAA has been shifted westward almost 6.5o compared to the model predictions (Konradi et al 1992).
6) Daly (et al 1996) reported that the standard NASA models don’t include the explicit directionality.
7) Abel et al (1994) analyzed the data from six satellites including the CRRES, and saw the evidence that the AE-8 model was contaminated due to US nuclear detonations termed as Starfish and therefore gives the over-prediction of the electron flux.

In spite of the above mentioned limitations of the AE8 and AP8 models, these models still are used as a de facto engineering tool for analyzing the radiation belt particles. Success of these models is largely due to the fact that the radiation belt is very complex and it is a very hard task to model it. Many engineering models have been developed after these models but they lack the energy and spatial range as compared to these NASA models (Rami Vainio, 2009).

II. CRRESELE Model

The description given here for CRRESELE is taken from CRRESELE documentation by Donald H.Brautigam, Jabin T.Bell, 1Lt, USAF, Dec 1995.

CRRES was launched in July, 1990 in geosynchronous orbit; carrying High Energy Electron Fluxmeter (HEEF). HEEF measures the outer radiation belt flux. Using the data, a new radiation belt model was created named as CRRESELE.CRRESELE actually includes eight data models. Six out of eight CRRESELE models are parameterized by the geomagnetic activity index Ap. CRRES returned the data for approximately 14 months during solar maximum. The remaining two models, CRRES_AVE and CRRES_MAX compute the total mission average and worst case respectively.

The CRRESELE electron model is constructed using the full pitch distribution. The IGRF85 model specifies the magnetic field caused by currents inside earth.This field is nearly dipolar with an 11.5o tilt and 436 km offset with respect to earth’s spin axis and is slowly changing with time (Knecht and Shuman, 1985).The IGRF model accounts for the slow time variation of the field as well as the tilt and non dipole contributions to the field.

The Olson Pfitzer static magnetic field (W.P. Olson and Pfitzer, 1974) models the external con- tributions to the earth’s total magnetic field from the magnetopause current, the ring current and the magnetosphere tail current. The model also accounts for the tilt of earth’s dipole with respect to the earth-sun line.

The external currents that contribute to the magnetic field change with variation in the solar wind and magnetospheric activity. However, the external static model used in CRRESELE assumes these currents do not change and uses values from average quite conditions. The Energy range of CRRESELE is 0.5 to 6.69 MeV.

Limitation of CRRESELE: The lower energy limit of CRRESELE is 0.123 MeV, so CRRESELE may ignore some electrons with the energy lower than 0.123 MeV but possessing the capability to penetrate the thin shielding of the spacecraft (spenvis.oma.be/help).

III. POLE V1, POLE V2 and IGE-2006 Models

POLE is a trapped electron prediction model for the geostationary orbit (D.M. Boscher et al. 2003). POLE V1 assumes that the solar activity phase varies the flux at all energies. POLE V1 model is developed by ONERA/DESP, France and covers the energy range from 30 KeV to 2.5 MeV (D.M. Boscher et al. 2003).

The data set used for the development of POLE V1 is taken from the geostationary satellites of Los Alamos National Laboratory (LANL) from 1976-2001.The detectors onboard LANL GEO satellites are SOPA and CPA (Charged Particle Analyzer). (D.M. Boscher et al. 2003).

The CPA detector covers the range of 30 keV-2 MeV while SOPA covers the energy range of 50 keV-1.5 MeV. Through measurements and comparison with the existing GOES data set, it was ob- served that these detectors suffer contamination problem during solar proton events. Afterwards, during processing of data, the contaminating protons were separated from the usable electron data (D.M. Boscher et al. 2003).

The number of channels in both the detectors is different; the channels of one detector were in- terpolated to the other detector. The channels of CPA are transformed to the new channels of SOPA (D.M. Boscher et al. 2003).

POLE V2 and IGE-2006 models are extended from POLE V1 and are specifically used for computing the outer belt electron flux in geostationary orbit (S. Bourdarie et al., 2008). Since the energy of the POLE V1 is lower, it was required to extend the energy limit upto 5.2 MeV for mod- elling highly energetic Bremsstrahlung particles and for deep dose calculations (S. Bourdarie et al., 2008). To accomplish this purpose, the measurements from another detector, ESP were taken into account. To use the data of ESP, some data processing techniques were implemented such as analyzing the data in terms of contamination, background and saturation. This produced a clean and processed ESP data set. The data from ESP detector is available from 1996. The data from SOPA is extrapolated in energy between 1994 and 2004, i.e. the same years in which ESP was also opera- tional (S. Bourdarie et al., 2008).

For IGE-2006 model, the dataset since October 2002, from SDOM (Standard Dose Monitor) onboard the Japanese satellite DRTS is also included to increase the reliability of POLE V2 model. Before including this data, it was fully analyzed and cross calibrated. The low energy data from MPA (Magnetospheric Plasma Analyzer) onboard LANL geostationary satellite is also included in IGE-2006 so as to increase the energy range down to lower energies (Bame, S. J, 1993).

3.2 Solar Proton Event Models

Overview:

The space designers and researchers, since long, have been putting their efforts to develop a good understanding of solar proton events and to be able to model it as precisely as possible so that the expected proton fluence on the spacecrafts could be computed. Various missions were launched to measure the proton fluence at different altitudes. As a result, some models were developed such as King’s model (E.G. Stassinopoulos 1974), JPL-85 model (Feynman et al. 1990), JPL-91 (Feynman et al. 1993) and ESP (Emission of Solar Protons) model (Xapsos, 1999, 2000).Later on, ESP model was improved by Xapsos et al. (2004) by using the data set of IMP/GOES extending the proton energy >327 MeV. This model is applicable for both maximum and minimum solar ac- tivity phases.

Nymmik (1999a, 1999b) developed a model at Moscow State University (MSU), Russia. This model uses the data set by IMP from solar cycles 20-22 and gives the SEP fluence as a function of solar activity level.

Jun et al. (2007a) developed a model by using the IMP-8 data from 1973-1997 obtained during larger solar activity phase.

Barcelona university research team developed another model, SOLPENCO (Solar Particle Engineering Code) for predicting proton flux of SEP events (Aran et al. 2006). The summary of these models is given in Appendix IV.

Explanation of some of these widely used models is given as.

I. King’s Solar Proton Event Model

J. King (1974) has the honour of developing the first ever complete empirical statistical model for solar proton events (SPENVIS help). This model is based on the data collected by IMP 4, 5 and 6 satellites in the period of solar active years from 1966-1972 (SPENVIS help).But during the solar cycle 20, the solar activity was found different in two aspects as compared to the activity of the solar cycle 19.Firstly, the largest annual mean sunspot number of solar cycle 20 is much lower than solar cycle 19, this is shown in Figure 3.1.

Secondly, the intensity and event frequency observed during the solar cycle 19 is much greater than noted in solar cycle 20 (ISO 15391, 2004).King used the data base of 25 individual events. This data base also included the August, 1972 SPE. As this event was contributing largest part of over all solar protons fluence, King classified this event as “anomalously large (AL)” event and separated it from 24 other “Ordinary” events. He determined the occurrence frequencies of SPEs, separately, of both the classes of events during solar active years (SPENVIS help).

illustration not visible in this excerpt

Figure 3.1: The sunspot number from 1960-2007. The blue line represents solar maximum while the green line represents solar minimum defined by JPL model (Feynman et al. 1993) (Adapted from spenvis.oma.be/help, 2009).

Problems with King’s SPE Model

The current research proves that there are some problems with the King’s model. Some of these are described in ISO 15391, 2004 as;

1) King totally ignored the data set of solar cycle 19 and only considered the measurements of solar cycle 20 for representing the predictions of solar cycle 21 but his assumption that the annual integrated solar proton fluence is linearly related with the sunspot number proved to be wrong by the later observations of solar cycle 21
2) King’s assumption that SEPs happen only during 7-years active sun period and there is negligible SEP event in the rest 4 “quite” years has been proved to be wrong.
3) King’s classification of the solar proton events into “ordinary” and “Anomalously Large AL” events is wrong because, on the bases of later solar event research and observations, it has been established now that Solar proton event sizes follow a uniform fluence distribution (Nymmik, 1999) with the rollover for large size events (Nymmik, 1999) or (Xapsos et al. 1999).
4) Proton event fluences are independent of the size of the event and can best be presented by power law function of proton rigidity with a spectral droop at low energies (E<30 MeV) (Nymmik, 1993) and also the proton fluence energy spectra are not exponent.

Having a critical look of the above discussions, it can be concluded that King’s model predictions of solar proton fluence don’t agree with the latest experimental data and observations.

II. JPL Solar Proton Model

The first version of JPL model (JPL-1985) was presented by Feynman et al. in 1990. JPL-85 is based on the data set collected from 1956-1963 with detectors onboard rockets and balloons and then from 1963 onwards by satellite observations.

The overall data set used for the development of JPL-85 model covers three solar cycles (ISO 15391, 2004).

After a detailed analysis, Feynman and his partner researchers demonstrated that in a solar cycle of 11years, the 7 years are of solar active phase (ISO15391, 2004). They selected the exact solar maximum dates as zero reference point and showed that in a solar cycle, the larger proton fluence begins 2.5 years prior to the zero reference (solar maximum peak) and lasts until 4.5 years after this reference date.

JPL-85 model takes into account the solar fluences in the seven active years of a solar cycle and neglects the rest four quite years (SPENVIS help).

Later on JPL-91 model replaced JPL-85 model (Feynman et al., 1993) with an extended range of energy. As compared to the JPL-85, the dataset of JPL-91 presents almost continuous record (SPENVIS help).

JPL-91 model is organized by selecting a threshold fluence level and all the proton events taken into account in JPL-91 exceed this threshold level in terms of proton fluences. These thresholds are at 10,5,1,1 and 1 (in cm-2 s-1 sr-1 ) (SPENVIS help).

But even in JPL-91 model, the classification and definition of solar active and quite years remains the same as in JPL-85. JPL-91 model also assumes that no considerable solar event takes place in quite years (ISO15391, 2004).

Problems with JPL-91 models

The space researchers have compared the results of JPL models with the experimental data and have noted some imperfections in these models. These are described in ISO15391, 2004 as;

1) JPL model’s assumption that SEPs occur only during 7-years active sun period and there is negligible SEP event in the rest 4 “quite” years has been proved to be wrong.
2) JPL models ignore the SEP fluence at lower solar activity levels which results in underestimation of particle flux with energy below 100 MeV.
3) JPL models distribute the SEP event according to log normal function which can lead to overestimation of the proton flux.
4) JPL models provide only five values of integral proton fluence and not the differential energy spectra.

III. Emission of Solar Proton (ESP) Model for SEP Events

ESP model is developed by NASA and NRL based on the GOES and IMP satellites data set of solar cycles 20, 21 and 22 from 1963-1996. This model is used for predicting the worst case and cumulative solar proton fluences. (M.A. Xapsos et al. 1999).

ESP model is also based on the assumption of 7 years active sun period and assumes that the proton fluence is the same for all years and for all solar cycles. The events begin and end with a threshold proton flux and it is possible that the longer events may consist of many peaks and dips in flux (ISO 15391, 2004).

For worst case analysis, nineteen fluence energy thresholds have been analyzed and the distribution function follows the power law, not the log normal (ISO 15391, 2004).

Problems with ESP model

1) ESP model can be used for only the sun active years and not for the sun quite years (ISO 15391, 2004).
2) It ignores the lower solar activity SEPs (ISO 15391, 2004).

3.3 GCR Flux Models

Overview

Modelling the GCRs is essential for predicting the SEU rates in spacecraft systems. The first attempt towards modeling the GCR flux was made by a team of researchers under the leadership of Adams in Naval Research Laboratory (NRL), (Barth et al. 2003).As a result of continuous efforts, a code was developed naming Cosmic Ray Effects on Microelectronics (CREME).

I. CREME-86 Model

This model is the first ever model for calculating the radiation environment in all the near earth regions and for calculating the effects of radiation environment on different microelectronic de- vices (J.H. Adams et al. 1981). In this model, the variations in the GCR flux due to the solar activity cycle are representative by a simple sine wave (Barth et al., 2003). For upto ten years in particular and still in general, this model is considered as a standard model for calculating heavy ion environment.

Improvements in CREME-86 Model

The data set of space environment continued to add up and the need for an improvement was re- alized. A revolution in this context happened when Reams et al., 1990 on the basis of ISEE 3 data, noted an inverse correlation between the heavy ion abundance ratio and the proton intensity (Reames et al, 1990). Reams et al. contradicted the assumption by Adams that all SPEs are He rich (Reames et al, 1990). This meant that the CREME-86 model was over predicting the radiation en- vironment (Barth et al. 2003).

Later on, Sims compared the CREDO measurements with CREME-86 and found that all of the CREME models overpredict the flux (Barth et al. 2003).

The data form IMP-8 was analyzed at University of Chicago that proved to be very comprehensive especially for modeling fluences at very high energies (Tylka et al. 1997).

Tylka et al. 1997 analyzed the October 1989 event along with 100 other solar proton events and presented a “worst case model” on the basis of his results (Tylka et al., 1997). With this improve- ment, Tylka presented CREME-96 model in 1997 (Tylka et al., 1997).The estimations of heavy ion made by CREME-96 are significantly lower than CREME-86 model (Tylka et al., 1997). The Fig- ure 3.2 shows the comparison of CREME-96 and CREME-86 flux predictions for a geostationary orbit.

illustration not visible in this excerpt

Figure 3.2: Comparison of CREME-86 and CREME-96 solar heavy ions models for GEO (adapted from J Barth et al. 2003).

Later on, the measurements of Anomalous Cosmic Rays (ACRs) from SAMPEX were also included in the GCR model (Barth et al., 2003). With this improvement, four models in CREME-86 were replaced by one GCR model in CREME-96.These improvements in the CREME model significantly reduced the SEE rate calculations in low earth orbits (Tylka, et al. 1997).

II. ISO 15390 GCR Model

This model is developed by technical committee ISO/TC20, Space systems and operations, In- ternational Organization for Standardization (ISO). It is developed for calculating GCR radiation effects on hardware in space. In this model, The GCR flux varies as a function of the solar cycle phase. The GCR flux angular distribution is isotropic in the earth’s orbit beyond the magnetosphere (ISO 15390, 2003).

This model is recommended by European Cooperation for Space Standardisation (ECSS) (ECSS-E-ST-10-04C, 2008) for modelling GCR flux in the spacecraft orbits. Space Environment Information System (SPENVIS) online system developed by ESA incorporates this model.

4 SPACE RADIATION EFFECTS ON SPACECRAFT, DE- SIGN MITIGATION TECHNIQUES AND DESIGN GUIDELINES

The space environment causes severe effects on spacecraft, especially in geostationary orbit where the spacecraft is directly exposed to the hazardous environment and there is no stronger geomagnetic field to protect it. Therefore, it is very important for mission planning and designing to have a good understanding of these effects on spacecraft so that the required protection could be provided.

4.1 Space Radiation Effects

The space radiation environment effects are classified into five broad categories;

1) Total Ionization Dose (TID) Effects
2) Single Event Effects (SEE)
3) Displacement Damage (DD) effects
4) Surface charging
5) Deep dielectric charging

These are explained as under;

4.1.1 Total Ionization Dose (TID) effects

A charged particle when strikes at the surface of the spacecraft, it deposits a specific amount of energy (dose) to the electronics of the spacecraft through ionization (http://parts.jpl.nasa.gov/asic/Sect.3.4.html, 2009). This deposited energy is usually measured in “rad” which is equal to 100 ergs/g of the material. The spacecraft normally experiences 10-100 krad (Si) total dose level (http://parts.jpl.nasa.gov/asic/Sect.3.4.html, 2009). When the electron in the surface receives this energy; it jumps to some higher energy level and becomes available for conduction. These electrons produced in the ionization process are the major cause of the total dose effects.

In CMOS devices, the charge created inside the oxide becomes trapped at the interface (http://parts.jpl.nasa.gov/asic/Sect.3.4.html, 2009). This trapped charge changes potential of the interface structure and can increase leakage current or can change the characteristics of the device using this structure (if this interface exists at a biasing point).

The Total Dose degradation is largely affected by the bias conditions (http://parts.jpl.nasa.gov/asic/Sect.3.4.html, 2009). A positive gate-SiO2 field produces holes which are then transported to the Silicon-Silicon dioxide interface, a fraction of these holes is trapped here. A negative field has almost opposite effect i.e. the holes are trapped to the gate result- ing in the recombination of the holes with electrons and not creating the hole traps. For n-channel devices, the positive voltage is required for the hole transport and for p-channel devices, negative voltage is required at gate to source. The threshold voltage at the silicon-silicon dioxide interface is changed by these trapped holes. This threshold shift is negative for the n-channel devices which causes the transistor shifting gradually towards depletion mode with the increase in total dose. This threshold shift is opposite in p-channel devices.

These effects of threshold shifts are shown in Figure 4.1 both for p- channel and n-channel de- vices.

illustration not visible in this excerpt

Figure 4.1: Voltage shift caused by irradiation (adapted from http://parts.jpl.nasa.gov/asic/Sect.3.4.html#A0)

Total Ionization dose depends upon the total trapped charge and the rate at which the particles are coming.

4.1.2 Single Event Effects (SEE)

Many electronic devices have dimensions so small that the dose rate currents arising from the passage of a single energetic particle can be sufficient to change the operating characteristics of the device in question. The resulting disruptions are called Single Event Phenomena (SEP) or Single Event Effects (SEE). An effect is classified as “soft” if the damage is transitory and the device can recover (parts.jpl.nasa.gov, 2009). An example of the soft error is the reversible flipping of a memory bit. An effect is classified as “hard” if the damage is permanent and the device is lost, such as an irreversible bit flip (A.C Tribble, 1995).

For a geostationary satellite (GEO), the energy of the trapped protons is not sufficient to cause SEE (SEECA, 1996). Since, GEOs are fully exposed to GCR and solar flares, the protons with en- ergy 40-50 MeV are attenuated due to geomagnetic activity but this attenuation decreases largely during SPEs (SEECA, 1996). During these severe events, the field lines crossing the equator at 7 Re (Radius of earth) can be compacted to 4 Re. As a possible consequence of this shift, the parti- cles previously deflected can now reach much lower altitudes and latitudes (SEECA, 1996).

An ionizing particle interacting with the p-n junction causes electron-hole pairs. At the junction contact, a fraction of the charge is collected while a fraction of the charge recombines (SEECA, 1996) resulting in a short current pulse at the node.

SEEs are of different types with respect to their effects on the spacecraft systems. These are stated as;

1) Single Event Upset (SEU): An SEU refers to a change of state in a device (analog or digital) which is induced by a GCR or SPE ionizing particle. An SEU is termed as a soft error because a device after reset or rewritten behaves as normal (SEECA, 1996).

2) Single Hard Error (SHE): An SHE refers an eternal or non recoverable change of the device operation (SEECA, 1996).

3) Single Event Functional Interrupt (SEFI): An SEFI refers the conditions when the equip- ment stops its standard function and reset of power is needed for restoring its standard functions (SEECA, 1996).

4) Single Event Latchup (SEL): An SEL is a pathological state of a device which, under cer- tain conditions, can operate as a parasitic silicon- controlled rectifier, often leading to a device burnout if not powered down (Messenger et al. et. al. 1997). SEU induced Latchup pulse is one to two orders of magnitude shorter (10-100 ps) than that for transient dose rate induced latch up (10-100 ns).

SEU induced Latchup usually affects only a single component function such as a parasitic npnp chain in an inverter cell, where as dose rate induced Latchup can affect a complete circuit or an entire system. In the latter case, a single Latchup can expand to include latching the entire chip because of heat produced in neighborhood of original Latchup. With the advent of submicron feature size, the incident Latchup-producing particle track can encompass a goodly fraction of the areas of a group of information storage cells (Messenger et al. et. al. 1997).

5) Single Event Gate Rupture (SEGR): In MOSFET devices, a heavy ion piercing the gate insulator can cause damage through its perforation and consequent rupture (Messenger et al. et. al. 1997). This is called SEGR (Single Event Gate Rupture).It usually happens when the gate dielectric is being stressed by a high internal electric field, simultaneous with the heavy ion strike. Such fields occur during a write or clear operations in a nonvolatile SRAM or EEPROM (Electrically Erasable, Programmable, Read Only Memory) (F.W.Sexton et al. 1992).

The heavy ion in its transit through the gate produces a highly conductive plasma track through ionization of the dielectric. This forms a very low path resistance between the gate and substrate. If sufficient energy was stored in this gate capacitor prior to the ion strike, the plasma track becomes the discharge path for the capacitor. This discharge can cause excessive heating of the dielectricenough to melt or otherwise degrade it. Device SEGR tendencies include dependence on the LET (Linear Energy Transfer) of the incident particle, dielectric properties, corresponding electric field and heavy ion incidence angle (Messenger et al., et al. 1997).

6) Single Event Burnout (SEB): An SEB is a destructive condition in which the gate insulator of the device is burnout (SEECA, 1996).

7) Single Event Multiple Bit Error (SEMBE): An SEMBE refers to the event when a single ion causes more than one logical states change (J. Howard, et al. 1999).

4.1.3 Displacement Damage Effects

When the space radiation interacts with the material of spacecraft, it imparts a certain amount of energy to the atom as a whole (NASA report, J. Howard et.al. 1999). This energy is taken as heat by the increasing vibrational motion of the atoms. If this energy is sufficiently high, the energy im- parted to the atom can overcome the binding energy of the atom in the crystalline lattice of the material. As a result, the atom is displaced from its normal position to various end locations. The disturbance caused by the displacement damage phenomenon changes the device operation. This change may add a current path that previously did not exist (allowing increased leakage current) or make conduction more difficult in regions designed for flow. For example, diodes become less ef- fective (Two ways current flow becomes easier) and the inherent amplification capabilities of transistors are lost. Displacement damage is a cumulative effect. (NASA report, J. Howard et al. 1999).

4.1.4 Surface Charging

Surface charging occurs when the spacecraft interacts with the low energy electrons (Moldwin, 2008). A spacecraft is bombarded by both positively and negatively charged particles. If the net transfer of positive or negative charge is not equal, net charging takes place. Moreover, the sunlight with sufficient energy interacting with the spacecraft librates electrons from any conducting surface called the photoelectric effect. These processes electrically charge the spacecraft. Generally, the spacecraft is composed of the structures of different materials, each material has different charac- teristics. This phenomenon causes an electric discharge (spark) which may prove to be catastrophic (Moldwin, 2008). Any sensitive optical instrument onboard the spacecraft can be overloaded and hence degraded by the spark. If this electric discharge is on a piece of sensitive electronics, the component can be damaged.

4.1.5 Deep Dielectric Charging

Deep dielectric charging and discharging are very hazardous phenomenon for the spacecraft onboard electronics (Moldwin, 2008). At geostationary orbit, the Van Allen belt electrons have the energy high enough to penetrate the spacecraft (Moldwin, 2008). This incident charge is deposited on the dielectric material of the circuit boards in the memory devices. This electric charge gradually increases upto a level that the dielectric material breaks down, hence creating the pathway for the charge to flow through causing the electrical shorts.

4.2 Geostationary Spacecraft Anomalies

Many Geo spacecrafts have faced destruction due to hazardous space environment. National Geophysical Data Center (NGDC) has maintained a good database of the Geo spacecraft anomalies (D. C. Wilkinson, et. al. 1994). Some of those are explained by David P. Love et al. (2000) as under;

1) TELSTAR-401 (97 W): TELSTAR-1 Geo Communication Satellite suffered an operational problem on January 11, 1997 (J. C. Anselmo, 1997). On January 6, a massive solar outburst was reported causing the ultimate failure of the spacecraft after a huge electrical discharge was emitted from the spacecraft.
2) ANIK-E1 (110 W): ANIK-E1 was reported to lose its solar panel power on March, 26, 1996 (C. S. Powell, 1996).
3) GOES-8 (75.9 W): On February 14, 1995, the ADC subsystem of GOES-8 satellite failed (D. Dooling, 1995).It was reported that “solar eruption” was a major source of problem which resulted in electric discharges. This electric discharge induced 4 or 5 bit flips in the RAM operating the attitude control of the satellite.
4) ANIK-E1 AND E2 ( 110 W): ANIK-E1 caused a failure in its solar panel power on Jan 20, 1994 (J. C. Anselmo, 1997) and after one hour, the ANIK-E2 spacecraft suffered the same system failure (Anon, 1994).
5) GALAXY-4 (99 W): On May, 1998, the Galaxy-4 satellite suffered a communication problem due to hazardous electron fluence (David P. Love, et al. 2000).
6) GOES-9 (112.4 W): This satellite suffered a failure in data transmission on July 28, 1998 (D. C. Wilkinson, et al. 1998).
7) GOES-8 (75.9 W): This satellite faced a problem in its ADCS on Oct. 27, 1998 (D. C. Wilkinson, et al. 1998).

4.3 Spacecraft Design Mitigation Techniques

Spacecraft designers have faced the challenges of catastrophic space radiation environment and they have developed techniques to mitigate the effects of radiation environment. Many research studies have been conducted by major space agencies such as NASA and ESA etc. in this context. Having a close look on these techniques explained by (SEECA, NASA Report, 1996), some de- sign mitigation techniques are stated to counter SEEs, internal charging, dielectric charging and displacement damage effects. The brief discussion of these techniques is presented here, although the details are given in SEECA (NASA Report, 1996) and other references provided.

4.3.1 Mitigation of Single Event Effects

The space professionals generally divide the SEU risk devices into two classes: The memory devices and control devices. But this is not the strict classification because there can be some overlapping between these two categories. The potential SEE mitigation techniques call for additional software or hardware added to the system. These additions create complications and overheads in the system that is linear with power of mitigation schemes.

The SEE-hard devices may prove to be an appropriate technique for meeting SEE requirements. Radiation hardened devices are better to use but the volume, power, cost, performance and their availability restrict their use. Software or hardware design is an effective mitigation technique but the end design becomes very complex. Therefore, a good combination of the two systems is most efficient and effective option. The SEE mitigation techniques are explained for both types of de- vices as below.

I. Design Mitigation Techniques for Memory and Data Related Devices

One of the most important techniques for mitigating single event effects in memory or data de- vices is EDAC (Error Detection and Coding), not only at the system level but also at the part (microcircuit) level. A good EDAC scheme is one that can detect many errors by its intrinsic na- ture. However, it is important to have a sensibly realizable EDAC scheme for practical purposes.

At the part (microcircuit) level, SEU errors can be decreased by;

a) Special lithographic processing techniques during its manufacturing, such as a highly doped buried layer to curtain funneling
b) Chip circuit design to include feedback resistors in the memory cell to impose a low pass filter characteristic
c) SEU EDAC schemes on chip or off-chip
Different EDAC schemes are applied in the spacecraft systems such as Parity checks, Cyclic Redundancy Check (CRC), Hamming codes, Reed-Solomon codes and Convolution encoding etc.

The use of parity checks is the easiest and the simplest way of SEE mitigation in memory or data related devices. In this technique, the logic ones occurring in a data stream are counted. The parity is an additional single bit added in the original data stream. Parity defines, in even or odd, the total number of ones in the data stream. This number is then compared with the output data and the error, if occurred, can be detected. The parity check system is only for detecting the errors but not for correcting the errors.

Cyclic Redundancy Check (CRC) coding is also an error detection only technique (K. L Short, 1987). In this scheme, arithmetic operations are performed on the data structure and the results are interpreted as polynomials. Different CRC codes such as CRC-16 are used for spacecraft onboard storage devices.

Hamming codes are used for detecting the single and multiple bit errors. In this scheme, a check code encodes the entire data block (A.B. Carlson, 1975). These codes are capable of correcting all single errors or detecting all combinations of two or fewer errors within a block (Bernard, 2000). Syndrome decoding is especially suited for Hamming codes. This technique is widely applied in the solid state recorders onboard the spacecrafts. This scheme is also called memory scrubbing.

Reed-Solomon (R-S) Coding is also a block error coding scheme increasingly applied in modern space systems (NASA, 1994). R-S codes are used for the multiple and consecutive errors detection and correction. One of the example of R-S codes is (255,223) code. 223 bytes in this code are the data bytes and the rest 32 bytes are the overheads. This code can correct upto 16 consecutive bytes of error (NASA, 1994).

II. SEE Mitigation in Control Devices

The techniques described earlier for SEE mitigation in memory devices can also be applied spe- cifically to the control devices like microprocessor program memory. The devices such as microprocessors and VLSI may suffer the hazards like erroneous command to a spacecraft subsys- tem or functional disruption of the systems. Microprocessors have many hidden registers. These registers are externally inaccessible to the device but the control is provided internally and SEUs occurred can change the device operations. The space designers imply different techniques such as Redundancy, Lockstep systems and voting scheme for SEE mitigation in the control devices.

Redundancy is one of the major techniques used to cope with electronic system reliability. It is well known that for standby maintained systems, high level redundancy produces high reliability. On a system level, the redundancy is a good way to recover the device from the single event effect. One system is declared as prime system and when it is affected by the single event, a spare redun- dant system for the same functions is switched ON through a ground control or autonomous means. But redundancy has its own limitations of the power, weight, cost or complexities in the spacecraft design.

In lockstep system, two identical circuits are operated with synchronized clocking. This system is generally applied in microprocessors (J. L. Kaschmitter, et.al. 1991). The outputs of the processors are matched, if they don’t agree, it proves that an error has occurred. Then the system may reinitialize or go into safe mode.

Voting is a method one step further to the lockstep system i.e. three identical circuits are operated and their outputs are matched, the output which is agreed by at least two circuits is selected. R. Katz et al., 1994 proposed TMR (Triple Modular Redundancy) voting scheme for FPGAs (Field Programmable Gate Arrays) i.e. for each logical flip flops, there are three voting flip flops. Despite some advantages of this TMR approach, it adds large overheads.

III. Mitigation of Destructive Effects of SEEs

Destructive SEEs may be of two types, recoverable and non-recoverable; depending upon the response of the device under attack. System level SEE hardening techniques may not be effective in most of the cases (SEECA, NASA report, 1996). In the hard error events like SEGR or SEB, the prime device fails permanently; therefore these require the redundant device or systems. Single Event Latch (SEL) is very device specific and may or may not cause permanent device failure (SEECA, NASA report, 1996). In microlatch event, the device’s current consumption may be in specified limits, so it is not easy to detect the effect. The use of a multiple watchdog timeout scheme demonstrated by LaBel, et al. 1992 is used as a preferred approach for SEE mitigation.

4.3.2 Mitigation of Total Ionizing Dose, Displacement Damage and Surface Charging

Space radiation experts use different techniques to mitigate TID effects such as shielding, conservative circuit design and derating of the devices (Richard H. Maurer et al. 2008). The Radhard devices are also used for specific cases. The first step is to draw the Dose-Depth curve for the spacecraft and estimate the total dose vs. the thickness of different shielding materials.

Shielding is, in most of the cases, a good approach for system design. The High atomic number (high-Z) materials are used for shielding, such as Tantalum for the spot shields and Tungsten for PCBs are used for special cases (Richard H. Maurer et al. 2008). These high-Z materials have density six times the density of aluminum, so resulting in a relatively thinner shielding. But these high-Z materials attribute the secondary electrons produced in the shielding, proving to be disas- trous for the devices. Shielding is a good approach to adopt for mitigating the effects of electrons and low energy protons. But can not mitigate SEEs produced by high energy GCRs (Richard H. Maurer et al. 2008). These high energy particles produce secondary particles in the shielding ma- terial which cause the further SEEs.

Derating is the reduction of electrical, thermal and mechanical stress levels applied to a part in order to decrease the degradation rate and prolong the expected life of the parts in hazardous radia- tion environment (Mukund, 2004). It is a good technique to mitigate TID effects on the devices.

Silica cover glass is used to protect the solar arrays against the displacement damage effects. The thickness of the silica cover glass is calculated relative to the total dose endured by the solar arrays.

4.3.3 Spacecraft Deep Dielectric Charging Mitigation Techniques

Spacecraft deep dielectric charging is considered to be an important factor for space system de- signs. It is caused by trapped high energy electrons in the geostationary orbit (A.C. A.C Tribble, 1995). To mitigate the effect of charging, the electron buildup of significant potential is prevented (A.L. Vampola, 1996). For this purpose, the bleed-off paths and leaky dielectrics are used. The LET is calculated for space mission and thickness level is determined. For a spacecraft in GEO, generally the shielding level used is almost 3 mm (NASA-HDBK-4002, 1999). Shielding reduces the charging current.

For mitigating the charging effects, the spacecraft active elements such as conductors and insulators are avoided to expose to the outer radiation environment, the cables are properly terminated and the circuit boards are kept free of floating metallization (A.L. Vampola, 1996).

Another important consideration is the signal conditioning. High level signals are required instead of low level signals because bulk charging and discharging comprise low level signal (A.L. Vampola, 1996).

The accumulation of charge in dielectrics creates high electric fields (M. D. Violet, 1993). In order to mitigate the deep dielectric charging, the metalized dielectrics are useful (Shu T. Lai, 2003).

4.4 Spacecraft Radiation Environment Design Guidelines

Space mission designers, on the basis of their experience, have proposed some design guidelines which are important considerations for safe and secure design and operations of a spacecraft under hazardous space environment conditions. Some of these techniques are introduced in this document, although these are explained in the references given.

NASA Hand book (1999) suggests some guidelines which are stated as;

1) Shielding: The spacecraft electronic elements should be shielded with a specified aluminum equivalent thickness to keep the internal charging rate under normal conditions. For GEO spacecraft generally 3 mm Al equivalent shielding is applied.

2) Grounding: All the structural elements should be grounded. The isolated conducting components are identified and the relevant grounds are provided.

3) Conductive path: All the circuitry of the sensitive parts should have a conductive path. A direct ground path is preferred.

4) Material Selection: The Dielectric with best performance should be used. For in- ternal charging, the materials of great concern are dielectrics. Mostly the dielectrics Teflon, Kapton and FR4 Circuit boards are used in spacecrafts. These dielectrics have the tendency to store the charge. Therefore, they should not be used especially in large blocks.

5) Filter Circuits: The low pass filter, low noise immune logic should be used on in- terface circuits. The CMOS circuits having higher interface noise immunity should be used. But care must be taken with CMOS because they are more Latchup sensitive.

6) Isolate windings: The primary as well as secondary windings of all the transformers should be isolated. For reducing common mode noise coupling, the primary to secondary winding capacitance should be reduced.

7) Bleed paths: The conductors should be provided with a conductive bleed path. The sensitive elements such as;

- Capacitor cans
- Signal and power transformers cores
- Metallic IC and hybrid device cases
- Unused connector pins
- Relay cans

are protected by small charge storage areas or normal bleed paths and stray leakage through their conformal coating.

8) Conformal coatings: The energetic particles in the radiation environment charge the conformal coatings. This issue is of utmost importance at the time of evaluation of the interior charging probability. The conductive coatings should be properly grounded.

9) Antenna feeds and parabolas: The antenna feeds covering should be considered with a great care. On the antenna systems, the isolated dielectric materials may store excess charge or energy.

10) Cable harness: The cable harness should be routed away from apertures.

11) External wiring: Additional protection should be provided for the cabling external to the spacecraft structure.

12) Thermal blankets: Very little protection is provided by the thermal blankets against the electron penetration. All the conductive layers of thermal blankets should be grounded.

13) ESD sensitive parts: The ESD sensitive parts should be paid special attention be- fore use.

The above stated design guidelines and mitigation techniques certainly improve the immunity of space mission design against the hazardous space radiation environment. But still, it needs further research. The space design standards developed by NASA and other space agencies are based on the thirty years old data, so it needs updating.

The discussion of space radiation effects reflect that a large scale careful planning is necessary at the satellite planning stage. The radiation environment should be simulated prior to the design and the relevant effects should be categorized and then the necessary design mitigation techniques should be applied in the systems. The overall system should be designed according to the design guidelines.

5 MODELLING THE SPACE RADIATION ENVIRONMENT IN GEOSTATIONARY ORBIT

Introduction: The space radiation environment in geostationary orbit is very dynamic; it varies with the spatial (longitude) and temporal (11-year solar cycle and diurnal) variations of the magnetic field. These variations have certain implications on the space system design and operations. Therefore, it is very important and crucial to model the radiation environment of the spacecraft before the spacecraft design phase.

The radiation environment is generally represented by the McIlwain B-L coordinate system (McIlwain, 1961), where B represents the magnetic field strength and L is the distance in Earth Radii at which a magnetic field line crosses the magnetic equator. The L-shell at the rotational equator varies over all longitudes because of the tilt of earth’s magnetic field relative to the rota- tional axis (S. Bourdarie et al., 2008). At the magnetic equator, a particular L-shell constitute almost constant radiation environment. Magnetic field is represented by the internal and external magnetic fields (S. Bourdarie et al., 2008). The internal component is well modelled by the mod- els such as, Jensen and Cain 1962 (Jensen and Cain, 1962) and IGRF (Sawyer and Vette, 1976).The external field is due to the interaction of the solar wind with magnetosphere and there- fore is not symmetric; stretched on the night side and compressed on the day side (S. Bourdarie et al., 2008). It is modelled by different researchers such as Olson-Pfitzer (Olson and Pfitzer, 1974) and Tsyganenko (Tsyganenko, 1989).

In this chapter, the variations of GEO space radiation environment are simulated first and then the trapped electron models are compared with each other followed by the comparison of the solar protons models.

5.1 Variation of GEO Radiation Environment

The study and simulation of the GEO radiation environment is divided into three categories in this report as;

1) Variation of GEO Radiation Environment with longitude
2) Variation of GEO Radiation Environment with Solar Cycle
3) Diurnal variations of GEO Radiation Environment

These variations of GEO radiation environment are studied in detail on next pages.

5.1.1 Variation of GEO Radiation Environment with Longitude

Longitudinal variation of GEO radiation environment is analyzed by simulating the environ- ment of the satellites orbiting at different longitudes from 0-360 degrees with spacing of 20 degrees. Satellites operational years are 4 from 01-01-1999 to 01-01-2003.The analysis is done by using SPENVIS (www.spenvis.oma.be) online system.

The AE-8 model (Vette, 1961) analyses only internal magnetic field. The CRRESELE model (D. H. Brautigam et al. 1995) considers both internal and external magnetic fields. The analysis is done first by using AE-8 MAX model (Vette, 1961) and then by using CRRESELE model (D. H. Brautigam). As stated previously, the L-shells represent the state of the radiation environment, so L values for all of the satellites are simulated in SPENVIS and results are plotted in Matlab. Figure 5.1 shows the L-values for each satellite by using only internal magnetic field (with AE-8 MAX model) and Figure 5.2 shows the L-values by considering both the internal and external magnetic field components (with CRRESELE model). The AE-8 model (Vette, 1961) uses Jensen and Cain1962 internal magnetic field model (Jensen and Cain 1962).The CRRESELE model (D. H. Brautigam et al. 1995) uses internal magnetic field model IGRF 1990 (Brautigam and Bell, 1995) and external magnetic field model Olson-Pfitzer (1974) quite model (Olson and Pfitzer, 1974). Figure 5.3 shows the comparison plot of both the models.

illustration not visible in this excerpt

Figure 5.1: L-values are plotted with respect to corresponding Longitude position of GEO satellite using AE-8 Max model.

illustration not visible in this excerpt

Figure 5.2: Change in L-parameter with respect to the longitude of a GEO satellite using CRRE- SELE model.

illustration not visible in this excerpt

Figure 5.3: The comparison of the L-parameter calculated by AE-8 Max model and CRRESELE model.

The graphs show that the value of L is minimum at longitude of 2000 E, so this is the worst case for the flux because minimum L means the maximum flux. L is highest at 2800 E (800 W) and 1000 E so the flux is minimum at these longitudes. These two longitude positions i.e. 1000 E and 800 W are the most favorable locations for the satellites. The graph also shows that AE-8 Max model (Vette, 1961) underpredicts the environment as compared to CRRESELE model. The variation of the space radiation environment with the longitude of a satellite is a combination of the eccentricity and tilt of the inner magnetic field (S. Bourdarie et al., 2008).

For AE-8 MAX model (Vette, 1961), Lmin=6.590 and Lmax= 6.971, therefore ∆L= 0.381 where as ∆L for CRRESELE model (D. H. Brautigam) is 0.243, which means that the electron flux affected by only internal magnetic field at any longitude covers a wider L range as compared to the L range covered by the combination of the internal and external magnetic field components. The value of ∆L depends upon the energy.

It is also important to investigate the spectrum of electrons in GEO trapped electron belt as a function of longitude of the satellite. This analysis is carried out by operating the satellite at different longitudes from 0-360 degrees and calculated the flux vs. energy levels of electrons using AE-8 MAX model (Vette, 1961) and then a Matlab graph is plotted shown in Figure 5.4.

illustration not visible in this excerpt

Figure 5.4: Flux variations as a function of Longitude of geostationary satellite.

The Figure 5.4 shows that there is not a considerable difference in the flux levels with respect to the longitude.

5.1.2 Variation of GEO Radiation Environment with Solar Cycle

The variations of GEO radiation environment are studied by looking at the flux change with the solar activity cycle.

It is now a well established fact that solar activity does not remain constant but varies with time (W.W. Vaughan et al. 1996). It is believed that the frequency of occurrence of the solar activity reaches a peak approximately every 11 years. It seems that the variation in solar activity is pro- duced due to the interaction of magnetic field with the non uniform rotation of the sun; its equator rotates faster than its poles (W.W. Vaughan et al. 1996). As a result of this nonuniformity, the shearing effect on the gas contorts the field into the configurations that cause solar activity. This variation in solar activity has considerable effects on the overall space environment like ionosphere, magnetosphere, geomagnetic field and plasmasphere. As a result, a spacecraft launched in such a dynamic environment suffers sudden and regular problems. This variation is very important for space mission planners, designers and operators.

To analyze the variations due to solar activity, the satellite in geostationary orbit at 38 degree east is operated in different solar cycle years and the flux and energy levels of the particles are calculated as shown in Figure 5.5.The SPENVIS online system is used with ESP total fluence model (Xapsos, 1999, 2000) with 90 % confidence level.

illustration not visible in this excerpt

Figure 5.5: Variations of solar proton flux of energy 200 MeV as a function of satellite year of op- eration

The Fig. 5.5 shows that solar activity repeats after 11 years. The solar “active” years are nine in total, the five years of which are of peak solar activity and two years each before and after the peak level and the “quite” years are two, at the start and at the end of the solar cycle in which the flux drops down to zero. This is a little different behavior than stated by Feynman et al., 1990. Actually the problem is that ESP model is based on the data set of 7 active years of solar activity, so it pre- dicts higher flux than actual. In the quite solar years, ESP model assumes that there is a zero proton flux. This concept of zero flux has been proven as wrong by different researchers (ISO 15391, 2004). Therefore this model is not suitable to use for the lower solar activity (quite) years. The other problem with ESP model observed in the figure 5.5 is that it predicts the equal peak flux in the five peak years, which is against the actual solar proton fluences and generally there is a change in the peak solar years as well. So ESP model predicts a little higher flux than the actual. This ESP model behavior is now compared with the GOES data on the next page.

The solar proton flux from 1996-2006 is calculated by the ESP model in SPENVIS (Figure 5.6) and for comparison, the data of the GOES satellite for the same years is downloaded from NGDC website (http://www.ngdc.noaa.gov/stp/SOLAR/ftpsatenvir.html#SEP ) and is plotted in Figure 5.7.

illustration not visible in this excerpt

Figure 5.6: The solar proton flux data using ESP model.

illustration not visible in this excerpt

Figure 5.7: The solar proton data measured by the GOES satellite.

The data for comparison is not plotted on the same window because GOES measurements are almost two magnitudes lower than the ESP measurements. ESP model measures higher flux in the years from 1996-2006 than the GOES data. One reason of this difference is that GOES data incorporates an algorithm resulting in lower flux (http://www.ngdc.noaa.gov/stp/SOLAR/ftpsatenvir.html#SEP). The other reason of ESP model higher flux in the active years is that ESP model is based on the dataset of higher solar activity. GOES data plot is not representing an actual solar cycle. Its reason is that GOES data used here contains some problems. ESP model calculates zero fluence in 1996 and 2006 because ESP model assumes that no considerable event takes place in solar minimum phase and there is a minimum threshold level of fluence in ESP model, if the total protons are below this threshold level, the model results in zero fluence.

Moreover, ESP model assumes the similar behavior of solar flares in all the solar cycles. This assumption is negated by analyzing the GOES data of grouped solar flares from the previous three solar cycles. For this purpose, the GOES data of the grouped solar flares is downloaded from NOAA website http://sgd.ngdc.noaa.gov/sgdpdf/0309/Number_of_Solar_Flares_0309.pdf and plot- ted in Matlab as shown in Figure 5.8.

illustration not visible in this excerpt

Figure 5.8: The GOES data of grouped solar flares from 1976-2006

The Fig. 5.8 is showing that the number of grouped solar flares in one solar cycle is decreasing in the preceding solar cycle. This contradicts the assumption of ESP model. For observing the effects of solar activity on the Galactic Cosmic Rays (GCRs), The GOES data from 1986-1999 is downloaded from the NOAA website as shown in Figure 5.9.

illustration not visible in this excerpt

Figure 5.9: GOES satellite GCR Plot taken from NOAA/NGDC website http://www.ngdc.noaa.gov/stp/GOES/image/gd86_99.pdf

Fig. 5.9 is showing the GOES GCR data and it is obvious that heavy ions (GCRs) flux is increasing in solar minimum period and is decreasing in the solar maximum period.

5.1.3 Diurnal Variations of GEO Radiation Environment

Other than the 11-year solar cycle, the high energy electron flux at GEO orbit varies diurnally with the local time. The flux variation in the solar active days may be 10:1 from local noon to local midnight (NASA-HDBK-4002, 1999). The highest flux is observed at local noon and local mid- night. To verify this variation, GOES data predictions for 5th August, 2009 are downloaded from NOAA website, http://www.swpc.noaa.gov/ftpdir/latest/daypre.txt

And is plotted in Matlab as shown in Figure 5.10.

illustration not visible in this excerpt

Figure 5.10: GOES 3-hour prediction electron data plot

The plot shows that the higher flux is observed at local midnight and local noon, because of the external magnetic field variations.

5.2 Comparison between AE-8 Max and IGE-2006 Trapped Electron Models

The space designers have been using AE-8 model (Vette, 1961) developed by NASA since two decades, despite the fact that many research studies have pointed out some problems with this model. ONERA has developed a new model IGE-2006 (S. Bourdarie et al., 2008) for analyzing trapped electrons belt in geostationary orbit. The ECSS has recommended IGE-2006 model as the standard model for modelling GEO trapped electron belt (ECSS-E-ST-10-04C, 2008). Therefore, it is important to compare the two models to see the differences between the two models. The AE-8 model (Vette, 1961) uses Jensen and Cain 1962 model (Jensen and Cain 1962) for internal mag- netic field.

This comparison is carried out by operating a satellite at 38 degree longitude position from 1999-2003 and the graph obtained is shown in Figure. 5.11.

illustration not visible in this excerpt

Figure 5.11: Comparison between the electron flux deduced from AE-8Max model and electron flux from IGE-2006 model in the lower, mean and upper cases

At the longitude position of 38.0[illustration not visible in this excerpt]E, the AE-8 MAX model (Vette, 1961) starts the energy level reading from 0.04 MeV and goes upto 5MeV while IGE-2006 model (S. Bourdarie et al., 2008) starts from the energy level of 9.2 X 10-4 MeV and goes upto 3.97 MeV. At the longitude of 38.0[illustration not visible in this excerpt]E, the AE-8 MAX model (Vette, 1961) overestimates the energy level as well as flux of trapped electrons as compared to all the cases i.e. lower, mean and upper flux of IGE-2006 model (S. Bourdarie et al., 2008).

5.3 Comparison of Solar Flare Proton Models

Space researchers have developed different models for modelling solar proton events such as King’s model (J. H. King, 1974), JPL 91 model (Feynman et al. 1993) and now ESP model (Xap- sos, 1999, 2000). ECSS has recommended using ESP model for modelling solar proton events (ECSS-E-ST-10-04C, 2008). Therefore, it is necessary to compare the three models to evaluate the differences between them so that the space mission designers could choose the most suitable model for their analysis.

This analysis is done by operating a satellite at 38 degree East longitude from 1999-2005 and using SPENVIS software with JPL 91’s model (Feynman et al. 1993), King’s model (J. H. King, 1974), and ESP models (Xapsos, 1999, 2000) one by one with 90% confidence level and the results are plotted in Figure 5.12.

illustration not visible in this excerpt

Figure 5.12: Comparison of the SPE fluences estimated by King’s, JPL 91and ESP models .

It is obvious from the figure 5.12 that JPL 91 model (Feynman et al. 1993) and King’s model (J. H. King, 1974) only agree with each other for the energy levels of 70 to 80 MeV protons. In the energy range of 20-70 MeV, JPL 91 model (Feynman et al. 1993) calculates lower fluence level than King’s model. Then for higher energy levels from 90-200 MeV, the JPL 91 model (Feynman et al. 1993) estimates larger fluence than King’s model. This may be from the fact that JPL 91 model (Feynman et al. 1993) distributes SEP events according to log normal function which results in an increased fluence level. Another reason of this trend is that King’s model (J. H. King, 1974) is based upon the dataset of only one solar cycle 20 and there was no anomalously larger event noted in solar cycle 20.Therefore, King’s model (J. H. King, 1974) underpredicts the fluence levels at higher energies than JPL 91 model (Feynman et al. 1993).

It is obvious from the figure that ESP worst case model (Xapsos, 1999) estimates lower fluence levels at all the energies of protons than the rest of the three models. Therefore, it can be said that ESP worst case model underestimates the proton fluence as compared to other models.

The analysis in Figure 5.12 also shows that the ESP total fluence model (Xapsos, 2000) predicts higher fluence at lower energies as compare to the JPL 91 model (Feynman et al. 1993) and King’s model (J. H. King, 1974). It can be said that ESP total fluence model (Xapsos, 1999, 2000), as compared to the other models, overpredicts the fluence of lower energetic particles.

At the middle energy levels, the ESP total fluence model (Xapsos, 2000) predicts larger fluence than King’s model (J. H. King, 1974). It is due to the fact that King’s model (J. H. King, 1974) is based on the dataset of solar cycle 20 in which the average sunspot number was lower than the fol- lowing solar cycles. For very high energies above 100 MeV, The ESP total fluence model (Xapsos, 2000) predicts higher fluence than King’s model (J. H. King, 1974) but lower fluence than JPL 91 model (Feynman et al. 1993).

6 PREDICTION OF SPACE RADIATION ENVIRONMENT FOR PAKSAT-1R GEOSTATIONARY COMMUNICATION SATELLITE

Introduction

Paksat-1R is a Pakistani geostationary communication satellite which is going to be launched in 2011 at 380 E longitude. From the previous discussion of hazardous space radiation environment at GEO, it is very important to estimate the radiation environment for the satellite early in the mission planning and design process. This chapter deals with the predictions of expected radiation envi- ronment for Paksat1R. There are many software available for radiation environment predictions developed by NASA, ESA and others. In this report, SPENVIS (Space Environment Information System) online software (www.spenvis.oma.be) is used because almost all of the radiation envi- ronment models can be simulated in SPENVIS.

The ECSS (European Cooperation on Space Standardization) standard ECSS-E-ST-10-04C is- sued on 15 November 2008 is used for preparing the radiation environment specifications for Paksat-1R.

The radiation analysis is divided into seven steps as under;

Step 1: Creating a project in SPENVIS

Step 2: Simulation of the trapped radiation environment for Paksat-1R

Step 3: Simulation of the solar proton fluence for Paksat-1R

Step 4: Calculation of Dose-Depth curve

Step 5: Estimating Solar cell degradation and cover glass thickness requirement

Step 6: Simulation of the GCR fluence and LET spectra for Paksat-1R

Step 7: Calculation of Single Event Upset rate for Paksat-1R

These steps are performed one by one in SPENVIS online software and the results are obtained.

6.1 Step 1: Creating a Project in SPENVIS

A project in SPENVIS is created using orbit generator tab, according to the parameters of Pakistan’s Communication Satellite Paksat-1R; the major parameters are listed as;

illustration not visible in this excerpt

6.2 Step 2: Simulation of Trapped Radiation Environment

The trapped radiation environment in GEO is mostly due to the trapped electrons which have severe effects on the spacecraft. The trapped electrons and protons are estimated for Paksat-1R. Electron environment is simulated with IGE 2006 Average Model (S. BOURDARIE, et al. 2008) and proton environment is simulated with AP-8 MIN (Sawyer and Vette, 1976) and AP-8 MAX model (Sawyer and Vette, 1976). AP-8 model uses GSFC 12/66 120 Term updated to 1970.0 model (Cain et al. 1967) for internal magnetic field. AP-8MIN model (Sawyer and Vette, 1976) results are shown in Figure 6.1 and AP-8MAX model (Sawyer and Vette, 1976) results are shown in Fig. 6.2.

The electrons are responsible for causing severe effects on spacecraft; therefore these are estimated with great care.IGE-2006 (S. BOURDARIE, et al. 2008) average flux model developed by ONERA and recommended by the European Cooperation for Space Standardization (ECSS) (ECSS-E-ST-10-04C, 2008) is used in SPENVIS. ECSS also recommends the upper case model for conservative analysis. The electron model results are shown in Figure 6.3.

illustration not visible in this excerpt

Figure 6.1: Energy spectra of trapped protons for Paksat-1R using AP-8 MIN model.

illustration not visible in this excerpt

Figure 6.2: Energy spectra of trapped protons for Paksat-1R using AP-8 MAX model.

The Proton environment graphs show that at geostationary orbit altitude, there are no energetic protons present. The maximum energy of the protons is 1.5 MeV and total protons of this energy are 8 for the whole 15 years period. This ratio of the protons is not sufficient to cause a severe ef- fect on a satellite.

illustration not visible in this excerpt

Figure 6.3: The trapped electron spectra for Paksat-1R

The electron spectra graph in Figure 6.3 shows the energies of the electrons in trapped radiation belts along with their fluence level. This plot is obtained by using IGE-2006 (S.Bourdarie, et al. 2008) average flux model which has its maximum energy limit of 5.2 MeV. But the graph shows that the maximum energy level of the trapped electron at 380 E longitude is 3.97 MeV and there are 1117 electrons having this energy in 15 years of mission duration. It shows that the electron envi- ronment at this altitude is very severe and can cause spacecraft internal charging due to the higher penetrating power of the electrons. So the necessary internal charging mitigation techniques will have to be applied in the Paksat-1R design. On the basis of the electron spectra shown in Figure 6.3, a radiation budget is calculated which gives the estimations of the protective shielding to be applied for Paksat-1R.

The AE-8 Max (Vette, 1991) model, shown in Figure 6.4 predicts some electrons with the energies higher than shown by IGE-2006 (S.Bourdarie, et al. 2008) average flux model. The electrons with lower energy are important for solar cell degradations and to some extent surface accumulation leading to surface charging and discharging phenomenon.

illustration not visible in this excerpt

Figure 6.4: Trapped electron spectra by using AE-8 (Vette, 1991) MAX model calculated for Pak- sat-1R.

In Figure 6.4, the AE-8 Max (Vette, 1991) model plot shows that there are some electrons pre- sent at 380 E longitude with energies higher than calculated by IGE-2006 (S.Bourdarie, et al. 2008).

The European Cooperation for Space Standardisation (ECSS) (ECSS-E-ST-10-04C, 2008) has recommended IGE-2006 (S.Bourdarie, et al. 2008) model for estimating trapped electrons spectra for a geostationary satellite, so only IGE-2006 (S.Bourdarie, et al. 2008) model is considered for developing radiation budget for Paksat-1R.

Trapped Electron Peak Flux for Paksat-1R

As recommended by ECSS, The peak electron flux for Paksat-1R is calculated and shown in Table 6.1. The model used for this estimation is AE-8 Max (Vette, 1991) model.

Table 6.1 Trapped Electron Peak Flux predicted for Paksat-1R

illustration not visible in this excerpt

6.3 Step 3: Simulation of Solar Energetic Protons (SEPs) for Paksat-1R

Paksat-1R is a geostationary satellite; therefore, it is fully exposed to the solar energetic protons. The protons, with the energies lower than 40-50 MeV are normally attenuated geomagnetically. But during the events of solar flares and geomagnetic storms, this attenuation is broken down. Geosta- tionary orbit is about 6-7 earth radii, but during the higher solar activity phase, the field lines at GEO are compressed down to 4 earth radii. Therefore, the solar particles that were previously at- tenuated now have access to lower altitudes and latitudes. These energetic particles cause different types of degradations in the spacecraft components. Therefore, it is very necessary to compute the expected solar particles incident on spacecraft. The space mission analysts prepare an energetic solar protons budget so that the required protective measures could be taken during the spacecraft design process.

For Paksat-1R, the Solar Energetic Particle Environment is simulated according to the recommendations of ECSS (European Cooperation for Space Standardization) standard ECSS-E-ST-10- 04C , issued on 15 November 2008.

This standard recommends that

1) Proton fluence from Solar Particle Events integrated over mission durations (of 1 year or more) shall be derived using the ESP total fluence model (Xapsos, 1999, 2000).
2) For mission durations shorter than 1 year, the fluences for one year shall be used.

As Paksat-1R is a 15 years space mission, therefore point 2 does not apply for this mission and the solar proton events environment is simulated by using ESP total fluence model (Xapsos, 1999, 2000) as recommended in point 1.

The SPENVIS software predicts that Paksat-1R will spend 11 years in solar maximum and only 4 years in solar minimum. So, it will have to counter very severe and dynamic conditions of the solar flare protons. The solar flare proton environment is simulated by using ESP total fluence model (Xapsos, 1999, 2000) with 90 % confidence level and Størmer formula for quiet magnetosphere applied. The results are shown in Table 6.2 and the graph is shown in Figure 6.5.

Table 6.2: Solar Energetic Protons Predicted for Paksat-1R Mission.

illustration not visible in this excerpt

Figure 6.5: Solar proton spectra predicted for Paksat-1R mission

illustration not visible in this excerpt

The graph in Figure 6.5 is showing the expected solar protons incident on spacecraft; with loga- rithmic energy scale. For Paksat-1R, the minimum energy of protons calculated is 0.10 MeV with 2.0 x 10[illustration not visible in this excerpt](cm-[illustration not visible in this excerpt]) integral fluxes. These protons are important for solar cells degradation effect. But the actual matter of concern is the higher energy protons reaching upto 200 MeV. These protons can cause Single Event Effects (SEE) such as Single Event Upset (SEU), Single Event Gate Rup- ture (SEGR) and Single Event Burnout (SEB).Toa void these effects, the necessary SEU mitigation like Hamming code should be applied and Rad-Hard components should be used in sensitive sub- systems.

6.4 Step 4: Calculation of Dose-Thickness Curve

Dose-Thickness curve is important for the estimation of shielding level required to counter the particle dose in orbit. To a first approximation, the spacecraft is considered a spherical aluminium ball with the electronic part in the centre. Then Dose-Thickness curve is derived for the mission.

SPENVIS calculates the Dose-thickness curve for different materials. In this study, the dose- thickness graph is simulated with SHIELDOSE model (SPENVIS help) and shield configuration is taken at the centre of Al sphere. The target material taken is Si. The plot is shown in Figure 6.6.

illustration not visible in this excerpt

Figure 6.6: Dose vs. thickness curve for Aluminium material over 15 years of satellite mission life time.

Fig. 6.6 is showing the dose (in rad(Si)) deposited by electrons, protons and Bremsstrahlung (secondary radiation) to the Al absorber. The trapped protons are very low in number, so negligible. At 7 mm thickness level, the dose deposited by electrons meets the line of dose of Bremsstrahlung particles, but after that increasing the dose does not contribute against Bremsstrahlung particles. But 7 mm thickness can not be applied on the spacecraft due to mass/volume and cost concerns. For even higher thickness, the dose-depth curve continues to flatten for Bremsstrahlung particles. Therefore, it becomes less effective (in weight) to add shielding to reduce dose to electronics if much shielding is already present. The Rad. hard components may be used with lower shielding. It is obvious from the graph that for total particles;

illustration not visible in this excerpt

The graph states that 5 mm Al-thickness is enough to counter the effects of trapped electrons (82krad (Si)). But 5 mm thickness is too much to be applied due to mass/volume and cost concerns.

The results conclude that Bremsstrahlung radiation effect is very severe and these are increased largely when the thickness is increased from 4 mm. The Bremsstrahlung radiation are more severe than the electrons especially after 7 mm but at this level and after that, the shielding can not prevent the material from the Bremsstrahlung particles. The thickness greater than 3.5 mm is not generally preferred for the spacecraft design.

6.5 Step 5: Estimating solar cell degradation and cover glass thickness

It is very important to predict the displacement effects in solar cells due to the radiation environment. The total mission environment effects on the silicon and gallium arsenide solar cells by the effective 1 MeV electron flux are predicted and shown in the table 6.3.

Table 6.3 Solar Cell Displacement Damage Prediction for Paksat-1R

illustration not visible in this excerpt

The values stated in the table are used to evaluate how the two types of solar cells perform un- der a specified radiation environment and then these values are coupled with the radiation test data (typically supplied by the manufacturer) for predicting the electrical performance of solar cells at mission end.

6.6 Step 6: Simulation of Galactic Cosmic Rays (GCR) Environment for Pak- sat-1R

Galactic Cosmic Rays (GCRs) consist of the energetic particles originating outside of the solar system. The ions found in GCRs are from almost all of the periodic table from Z=1 to Z=92. They have higher energies ranging from 10s of MeV to 100s of GHz. Therefore, in spite of their lower fluence level, GCR ions cause intense ionizations while passing through matter. Moreover, Paksat- 1R, being a geostationary satellite is fully exposed to GCRs, therefore it has to face severe GCRs effects.

GCR flux varies with respect to the solar activity; it is higher in Solar Minimum and lower in Solar Maximum periods. Keeping it in mind, it is very necessary to model the GCR spectra for the spacecraft before design. The SEU rates are predicted from this analysis.

The space mission planners carryout the simulation of GCRs as an important part of the project planning phase.

For Paksat-1R, the GCR environment is simulated according to the recommendations of the ECSS; stated in a standard ECSS-E-ST-10-04C , issued on 15 November 2008. The ISO 15390 model (ISO 15390, 2002) is used with geomagnetic shielding applied for quite magnetosphere. The contributions from all ions from Z=1 to Z=92 is simulated with 3 mm of Al equivalent shielding. The Silicon is selected as a target material and trapped protons are taken into account.

SPENVIS produced Linear Energy Transfer (LET) spectra as an output of the simulation. The result of this simulation is shown in Figure 6.7.

illustration not visible in this excerpt

Figure 6.7: Paksat-1R shielded LET spectra

Figure 6.7 is used to estimate the SEU rates calculation at a specified shielding level, here it is 3 mm. To accurately predict the SEU rates, electronics testing of the devices is required. This test data is then combined with the environment data shown in Figure. 6.7. Figure 6.7 indicates the attenuation provided by 3 mm levels of shielding for the solar minimum galactic cosmic ray heavy ion integral and differential LET spectrum. Very limited attenuation of the GCR spectrum is provided by a significant range of shielding values presented in this figure.

The shielding level required to protect Paksat-1R from the hazardous particles is calculated from this LET curve. Moving of sensitive components and the use of additional shielding materials as a provision to reduce GCR single event upset rate in sensitive devices affords little benefit.

Taking this 3 mm of Al shielding, the LET vs. Fluence level are calculated from the graph as shown in table 6.4.

Table 6.4 LET vs. Fluence Level for Paksat-1R

illustration not visible in this excerpt

As stated above, the accurate SEU rates prediction depends upon the electronics test data. Thus, the design is an iterative process. First, the preliminary designs are completed and then it is evalu- ated for SEE. To the designer, this evaluation provides recommendations for circuit modification or device replacement. This iterative process continues till the satisfaction of both, the designer and radiation engineer.

6.7 Step 7: Paksat-1R Single Event Upset Rates Calculation

It is very important to estimate the single event upsets for the spacecraft. Electronic circuitry onboard the spacecraft has to counter a hazardous radiation environment comprising of heavy ions, protons and electrons which cause soft and hard errors to different components onboard the space- craft. Since Paksat-1R is a geostationary satellite; it is fully exposed to the radiation environment.

Therefore, a detailed analysis and prediction of the environment and Single event phenomenon is required before satellite design stage.

The SEU caused by two phenomena are calculated in SPENVIS, one is due to direct ionization and the other is due to proton nuclear interactions. On a single window, the parameters are set for both the effects and the simulation is run creating a table of SEUs estimated for the entirePaksat-1R mission.

For SEU predictions, the Paksat-1R environment is simulated with ISO 15390 model with 3 mm of Al shielding for Fairchild 93L422s device, the bendel parameters for this device are provided in SPENVIS (Tylka, 1996). The parameters of 93L422s are set as under and the results are shown in Table 6.5;

Direct Ionization effects:

Device Dimensions (micron): 38.70 by 38.70 by 2.00 Critical charge: 1.13 E -02 pC

Proton nuclear interaction effects: Bendel function parameters:

Proton Upset parameter/ threshold parameter, A= 4.88 MeV

Saturation cross section or Limiting Cross section, Sigmalim=1.87E-10 cm[illustration not visible in this excerpt]/bit

Table 6.5 SEU Rates Calculated for Paksat-1R

illustration not visible in this excerpt

The table 6.5 shows the SEU produced in the device caused by both the effects i.e. direct ionization and Proton nuclear interactions. The protons directly interacting with spacecraft systems cause direct ionization SEUs through LET (SPENVIS help). A very few devices onboard spacecraft are affected by direct ionization. The nuclear reactions occurring inside the shielding material produce recoil particles with LET high enough to cause SEUs. The SEUs produced due to the ionization effects are greater than the SEU produced due to Proton nuclear interactions (SPENVIS help). The location and number of the SEU is found out with this method and then relative shielding level or SEU mitigation technique is applied to counter these effects.

To have a more understanding about the SEU rates in devices, the Paksat-1R radiation environment is simulated and SEU rates are calculated for more devices on the basis of bendel function parameter. The thickness of Al is kept as 3 mm. The reference data of these devices is taken from W.J. Stapor et al. 1990. The table 6.6 gives the summary of the calculated SEU rates produced due to proton nuclear interactions of every device under consideration.

Table 6.6: SEU rates of various devices

illustration not visible in this excerpt

In this analysis,

Limiting cross section= (24/A)14 x 10-12 And the Al thickness is 3 mm.

The table 6.6 presents the SEU rates for different devices. But it can be concluded through this table that COTS devices are not usable for a geostationary satellite with 15 years of life time. Even, the shielding applied for these devices can not do much for protecting the devices from SEUs.

After calculating the SEE rates, the space designers apply a specific approach to counter SEE problems (J.W. Howard et al. 1999). All the spacecraft system is divided into three categories, i) vital, ii) mission important and iii) non critical.

The vital systems are the life support areas, attitude control etc. These system do not tolerate any effects (i.e. one upset may jeopardise mission safety cause complete mission failure) and must use radiation hardened electronics.

Mission important systems would substantially impact probability of mission success if they get affected by SEE. The effects on these systems are recoverable but have a significant impact on mission operations. A mixture of rad-hard and nonrad-hard technologies is appropriate for these systems.

The non critical systems can tolerate the errors without significant degradation or the effect can be removed through post processing.

7 CONCLUSION

7.1 Summary

This report presents an overview of the parameters of space environment in geostationary orbit. The radiation environment, being the theme of report, is discussed in details with its different ele- ments such as Trapped radiation environment, Galactic Cosmic Rays (GCRs) and Solar Proton Events (SPEs). The related references are provided in the text and in the end of report. The space radiation analyzing codes and models are provided with necessary background information and references, the widely used models for Trapped radiation belt particles, GCRs and SPEs are ex- plained in detail. The limitations and risks of using these models are outlined in the light of respective references.

The effects of the space radiation environment on the spacecraft are discussed in detail such as Total Ionizing Dose (TID) effects, Displacement damage, Single Event Effects (SEE), Internal charging and surface charging of spacecraft systems. The important aspect of this report is space- craft design mitigation techniques. The techniques developed and proposed by different space research institutes and agencies for mitigating the charging effects and SEEs in the spacecraft sys- tems are summarily stated with the relevant references for explanations. The guidelines for designing spacecraft in any hazardous radiation environment are briefly explained. To understand the temporal and spatial variations of the space radiation environment in geostationary orbit, the results of different studies undertaken in this project are presented in the form of graphs and neces- sary observations and discussions.

As a case study, the space radiation environment is estimated for a Pakistani geostationary communication satellite Paksat-1R with fifteen years of life time. The calculation of expected radiation belt particles, solar protons and GCR ions is carried out according to the recommendations of European Cooperation for Space Standardisation (ECSS) standard issued in 2008.The expected SEU rate, dose level required for shielding the spacecraft and the suitability of different COTS devices for Paksat-1R are also outlined in this report.

7.2 Conclusion

The space radiation environment in geostationary orbit is very dynamic; it varies with spatial and temporal variations. The magnetic storms and substorms largely affect the GEO radiation envi- ronment. The accurate modelling of the parameters of radiation environment such as Trapped radiation, GCRs and SPEs is very important prior to designing a spacecraft. The models are devel- oped in different times based upon the datasets collected by different space missions. But all of these models involve some risks and limitations which need to be addressed properly before using these models for radiation budget estimations. The detailed studies of the effects of radiation envi- ronment in this report reveal that a spacecraft has to suffer a catastrophic environment in the orbit. The discussion of TID and DD (Displacement Damage) show that the electronics onboard the spacecraft have a cumulative damage mode which tends to increase with the spacecraft lifetime in orbit.

The detailed discussion of SEEs show that the radiation environment causes recoverable and non recoverable problems which may lead to the complete failure of the systems. The studies per- formed in this project for GEO radiation environment analysis show that the external magnetic field has important considerations in spacecraft radiation environment and it should be taken into account while evaluating the spacecraft radiation budget calculations. The study of the solar cycle effect shows that the GCRs have inverse relationship while SPEs have direct relationship with solar activity. This is further verified by the GOES satellite data. The study of the longitudinal change of a spacecraft shows that 2000 E is the worst case for a GEO spacecraft. The comparison of different models shows that these models are old and need necessary updates and improvements for predict- ing the radiation environment.

The radiation environment predicted for Paksat-1R reveals that 15-years space mission in GEO is a crucial challenge to build for the spacecraft designers. Increasing the spacecraft shielding level does not contribute much against the GCR heavy ions. The Bremsstrahlung radiation are produced by the GEO energetic electrons which offer a severe challenges and the shielding is not a good option to counter Bremsstrahlung radiation. The SEU rates calculated for different COTS devices reveal that these are not a good option to be used for Paksat-1R.

7.3 Evaluation

This report is a very useful document for the beginners in the GEO Spacecraft mission planning and designing. All the objectives set for this project (stated in section 1.2) are met completely with a good spirit. The skills, techniques and the knowledge required for the analysis of a GEO radiation environment has been learnt and put in this report in a comprehensive way. There are two prob- lems in the analysis of Paksat-1R, i) ECSS latest standard ECSS-E-ST-10-04C, 2008 recommends CREME-96 model for the solar proton peak flux predictions, but this model has some restrictions and neither it is available in SPENVIS software package nor at SSC, so could not be used for Pak- sat-1R analysis and ii) ECSS standard ECSS-E-ST-10-04C, 2008 recommends the FLUMIC model for the prediction of worst case for trapped electrons for spacecraft internal charging analyses. This model is also unavailable in SPENVIS software package, that’s why it could not be used for Pak- sat-1R analysis. If some how these models get accessible, the spacecraft radiation analysis can really be improved. This document lacks in explaining the shielding and radiation hardening proce- dures for the spacecraft components and subsystems. The mitigation techniques in this document are stated very briefly, hence require more explanations and knowledge to apply for a real space- craft mission. But generally, this document is very beneficial for a quick start of the GEO radiation environment analysis.

7.4 Future Work

This project focused on the space radiation environment in GEO and as a case study, the expected radiation environment for Paksat-1R geostationary communication satellite is predicted. But to develop a more accurate and comprehensive space environment analysis, some areas need further investigations listed as under;

1) The SEU rates of some of the COTS devices have been calculated in this report, but for a real spacecraft mission design, more devices can be considered for SEU rate calculation.
2) Radiation testing of the systems and devices is very important for determining their dose levels, SEU rates and displacement damage effects. Therefore, these required tests can be carried out for a valuable space radiation design of the satellite.
3) The comparison and discussion of the radiation environment models reveal that these models need some corrections and improvements, but in this project these models are used for prediction of the radiation environment for Paksat-1R which may prove overestimation or underestimation of the radiation budget. To avoid this, the techniques to improve these models can be studied and applied for calculations.
4) The radiation calculations obtained by using the space environment models are compared with GOES data only but to have a broad analysis, more satellites data can be included such as the data from IMP satellites and GORID etc.
5) The radiation shielding and radiation hardening approaches can be studied and discussed for a detailed analysis.
6) For the prediction of solar proton peak fluxes, the CREME-96 model can be used for a better spacecraft radiation analysis.
7) The FLUMIC model can be used for trapped electron peak fluxes for a better internal charging analysis.

REFERENCES

Hasan Murtaza, Study of the Radiation Environment in GEO for applications in the satellite radiation environment.”

Hasan Murtaza, Study of the Radiation Environment in GEO three-spacecraft observations from geosynchronous orbit”, J. Geophys. Res., 98, p13, 453- 13,466, 1993.

Hasan Murtaza, Study of the Radiation Environment in GEO of space exploration, 2008”.

Hasan Murtaza, Study of the Radiation Environment in GEO 2008”

Hasan Murtaza, Study of the Radiation Environment in GEO 1999.

APPENDIX I - WORK PLAN

illustration not visible in this excerpt

Tasks:

1) Studying the elements of the Space Environment in GEO including space plasma, space radiations and space debris and studying the elements of GEO Space Ra- diation Environment such as Trapped Radiations, Solar Flare Protons and Galactic Cosmic Rays (GCRs).

2) Preparing and submitting Interim Project report in March 2009

3) Studying the effects of Space Environment on GEO spacecrafts such as Surface Charging, Internal Charging, Total Ionizing Dose (TID) and Single Event Ef- fects etc.

4) Studying the GEO Space Radiation Environment Simulation models alongwith their critical analysis.

5) Studying, learning and exploring the analysis features of SPENVIS system

6) Analysis of the variations of Space Radiation Environment in GEO along- with their comparison with the GOES data and comparing the models of Trapped Radiations and Solar Flares.

7) Simulation of the Space Radiation Environment for Paksat-1R, a Pakistani Geostationary Communication Satellite and Studying spacecraft design mitigation techniques and preparing spacecraft radiation design guidelines.

8) Preparing and Submitting Final Project Report

APPENDIX II - TRAPPED RADIATION BELT MODELS-I

(Adapted by Rami Vainio, et al. 2009)

illustration not visible in this excerpt

APPENDIX III - TRAPPED RADIATION BELT MODELS-II

(Adapted by Rami Vainio, et al. 2009)

illustration not visible in this excerpt

APPENDIX IV - SOLAR PROTON EVENT MODELS

(Adapted by Rami Vainio, et al. 2009)

illustration not visible in this excerpt

APPENDIX V - SPACE ENVIRONMENT INFORMATION SYSTEM (SPENVIS)

The spacecraft space environment analysis is very crucial for space mission planning. It in- cludes the analysis of radiation environment, space plasma, magnetic field etc. Different space environment analysing systems have been developed by different organizations such as EnviroNET system by GSFC and CREME-96 by Naval Research Laboratory (NRL). But these systems are not integrated and do not include all the models an tools developed for space environment analysis.

SPENVIS is an online free, easily accessible and user friendly system developed for ESA, for modelling space environment and its effects on spacecraft in any orbit. Different Space Radiation analysis can be carried out with this system, including;

- Orbit generation
- Magnetic field analysis
- Atmosphere and ionosphere analysis and LET spectra
- Spacecraft charging calculations
- Meteoroids and debris estimations
- Solar activity predictions
- Plasma interactions and effects
- End of life Solar Cell Calculations
- Single event upset rates
- Radiation doses calculation

In this project, specifically the space radiation environment is studied using SPENVIS. The space radiation environment analyzing models, developed by different organizations and researchers are accessible through SPENVIS. These models include;

- For trapped protons analysis

I. AP-8
II. CRRESPRO
III. SAMPEX/PET PSB97

- For trapped electrons analysis

I. AE-8
II. CRRESELE
III. AE-8MIN Update ESA-SEE1
IV. IGE2006/POLE

- For Solar Flare Protons analysis

I. King
II. JPL
III. ESP total fluence
IV. ESP worst case fluence
V. Rosenqvist et al. (2005, 2007)

- For GCR analysis

I. ISO-15390
II. CREME-86
III. Xapsos et al. Models

SPENVIS provides extensive defaulting and customizable input parameters and generates output in the form of tables, texts and graphs. The user specified input parameters are validated and necessary guidelines are offered for parameters values. The tables generated can be exported to Microsoft excel. The graphs can be generated in different formats such as PNG, GIF, JPEG, TIFF etc. and can be saved or copied to the clipboard.

Two satellite projects can be created at a time in SPENVIS.SPENVIS provides online help pages giving the background information of environment, models and the guidelines for using these models. The related references and websites are also provided in help.

SPENVIS is used for all the environment analysis in this report such as trapped electron analysis, solar proton event analysis, GCR analysis and SEU rates calculation etc. During its use, some problems are faced with SPENVIS including;

1) Only two projects can be created at a time, but environment analysis requires sometimes more than three or four projects at a time. But after two projects, one project needs to be deleted for any new project, all the data of the deleted project is lost.
2) The graphs generated as an output are not editable or customizable.
3) The Flumic model recommended by ECSS for worst case trapped electrons analysis is not in- cluded.
4) CREME-96 model recommended by ECSS for solar particle peak flux is not included.

[...]


1 A.B. Carlson, Communication Systems, New York: McGraw Hill Book Company, 1975.

2 A.C Tribble , “The space environment, implications for spacecraft design”

3 Adv. Mater., Ring-opening polymerization of a silaferrocenophane within the channels of mesoporous silica: Poly (ferrocenylsilane)-mcm-41 presursors to magnetic iron manostruc- tures,” Adv. Mater., vol. 10, pp. 144-149, 1998.

4 Anon. , “ANIK-E2 disabled,” Aviation Week Space Technol., vol. 140, p. 28, Jan. 1994.

5 Aran, B. Sanahuja, D. Lario, Adv. Space Res. 37, 1240 (2006)

6 A. Konradi, G. D. Badhwar, A. C. Hardy, B. L. Cash, M. J. Golightly, W. Atwell, and K. A. Hardy, “Space shuttle radiation measurements and the radiation models,” 23rd COSPAW World Space Congress, Washington D.C., 1992, paper F2.1-M.1.05.

7 A. Sicard-Piet A., S. A. Bourdarie, D. M. Boscher, R. H. W. Friedel, “A model for the geosta- tionary electron environment: POLE, from 30 keV to 5.2 MeV”, IEEE Trans. Nuc. Sci. 53 (4), 1844-1850, August 2006.

8 A. Sicard-Piet, S. Bourdarie, D. Boscher, R. H. W. Friedel, M. Thomsen, T. Goka, H. Matsu- moto, H. Koshiishi, “A new International Geostationary Electron model: IGE-2006, from 1 keV to 5.2 MeV”, submit to Space Weather Journal, 2007.

9 A. Sims, C. Dyer, C. Peerless, K. Johansson, H. Pettersson, and J. Farren, “The single event upset environment for avionics at high latitude,” IEEE Trans. Nuc. Sci., vol. 41, pp. 2361- 2367, Dec. 1994.

10 Bame, S. J., D. J. McComas, M. F. Thomsen, B. L. Barraclough, R. C. Elphic, J. P. Glore,

J. T. Gosling, J. C. Chavez, E. P. Evans, and F. J. Wymer, “Magnetospheric Plasma Analyzer for Spacecraft with Constrained Resources”, Rev. Sci. Instr., 64, 1026, 1993.

11 Barth, C. S. Dyer, and E. G. Stassinopoulos, Space, Atmospheric, and Terrestrial Radiation Environments, 2003.

12 Bernard Sklar, “Digital Communications; Fundamentals and Applications, Second edition”, 2000.

13 Boscher D. M., S. A. Bourdarie, R. H. W. Friedel, R. D. Belian,”A model for the geosta- tionary electron environment: POLE ” , IEEE Trans. Nuc. Sci. 50 (6), 2278-2283, December 2003.

14 B. Abel, R. M. Thome, and A. L. Vampola, “Solar cyclic behavior of trapped energetic electrons in Earth’s Inner Radiation Belt,” J. Geophys. Res., vol. 99, no. A10, p. 19427, 1994.

15 Cain, J. C., Hendricks, S. J., Langel, R. A., and W. V. Hudson, A Proposed Model for the International Geomagnetic Reference Field-1965, J. Geomag. Geoelectr. 19, 335-355, 1967.

16 Cane, H.V., G. Wibberenz, I.G. Richardson, and T.T. von Rosenvinge. 1999. Cosmic ray modulation and the solar magnetic field. Geophysical Research Letters 26(5):565-568.

17 Cladis, John B , The Trapped Radiation Handbook.. ; Davidson, Gerald T. ; Newkirk, Les- ter L.1971.

18 Cliver, E.W., and H.V. Cane. 2002. Gradual and impulsive solar energetic particle events. EOS Transactions 83:61-68.

19 Cliver, E.W., and A.G. Ling. 2001. Coronal mass ejections, open magnetic flux, and cos- mic-ray modulation. Astrophysical Journal 556(1):432-437.

20 C. S. Powell, “The blustery void,” Sci. Amer., vol. 275, pp. 24-25, Aug. 1996.

21 David P. Love, Donald S. Toomb, , Daniel C. Wilkinson, and J. B. Parkinson, Penetrating Electron Fluctuations Associated with GEO Spacecraft Anomalies, IEEE TRANSACTIONS ON PLASMA SCIENCE, VOL. 28, NO. 6, DECEMBER 2000.

22 Dalec Ferguson, “interactions between spacecraft and their environment”.

23 Daly E.J, J. Lemaire, D. Heynderickx, and D. J. Rodgers, Problems with Models of the Radiation Belts, 1996.

24 Dieter Bilitza ,Solar-Terrestrial models and Application software, , 1990

25 D. C. Wilkinson to D. P. Love, private communication, Oct. 29, 1998.

26 D. C. Wilkinson, “National Oceanic and Atmospheric Administration’s anomaly data base and examples of solar activity affecting spacecraft,” J. Spacecraft Rockets, vol. 31, pp. 160- 165, Mar.-Apr. 1994.

27 D. Dooling, “Stormy weather in space,” IEEE Spectrum, vol. 32, pp.64-72, June 1995.

28 D. H. Brautigam, M. S. Gussenhoven, and E. G. Mullen, “Quasistatic model of outer zone electrons,” IEEE Trans. Nucl. Sci., vol. 39, pp. 1797-1803, Dec. 1992.

29 D. Heynderickx, B. Quaghebeur, B. Fontaine, A. Glover, W.C. Carey, E.J. Daly, “New fea- tures of ESA’s Space Environment Information System (SPENVIS) Garrett, H. B., and C. P. Pike, eds., 1980, Space Systems and Their Interactions with Earth's Space Environment. New York: American Institute of Aeronautics and Astronautics”

30 D. Heynderickx, Belgisch Instituut voor Ruimte-Aëronmie/Institut d’Aéronomie Spatiale de Belgique (BIRA/IASB), Apr. 17, 1997.

31 D. Heynderickx and J. Lemaire, “Coordinate systems for mapping low-altitude trapped particle flux,” in Proc. Taos Workshop Earth’s trapped Particle Environments, Aug. 14-19, 1994.

32 D. Heynderickx, M. Kruglanski, V. Pierrard, J. Lemaire, M. D. Looper,and J. B. Blake, “A low altitude trapped proton model for solar minimum conditions based on SAMPEX/PET data,” IEEE Trans. Nucl. Sci., vol. 46, pp. 1475-1480, Dec. 1999.

33 D. HEYNDERICKX, J. LEMAIRE, E. J. DALY and H. D. R. EVANS, Calculating low altitude trapped particle flux with the NASA models AP-8 and AE-8, 1996.

34 D. Heynderickx, Lemaire J. and Daly E. J. (1996), Historical review of the different procedures used to compute the L-parameter, Radiat. Meas. 26, 325-331.

35 D. Heynderickx , Kruglanski M., Lemaire J. F. and Daly E.J. (1996c) A New Tool for Calculating Drift Shell Averaged Atmospheric Density. Proc. of Workshop on Radiation Belts: Models and Standards, Brussels, October 17-20, 1995, AGU Monograph, in press.

36 D.L Chenette and W.F. Dietrich” The solar flare Heavy Ion environment for Single Event Upsets:A summary of the observations over the last solar cycle, 1973-1983, 1984.

37 D. V. Reames, H. V. Cane, and T. T. von Rosenvinge, “Energetic particle abundances on solar electron events,” The Astrophys. J., vol. 357, pp. 259-270, July 1, 1990.

38 ECSS, ECSS-E-ST-10-04C, ESA, 2008

39 E. G. ‘ Stassinopoulos, G. J. Brucker, D. W. Nakamura, C. A. Stauffer, G. B. Gee, and J. L. Barth, Solar flare proton evaluation at Geostationary orbits for engineering applications, 1996.

40 E.L. Petersen, J.B. Langworthy, and S.E. Diehl, "Suggested Single Event Upset Figure of Merit," IEEE Trans. Nucl. Sci., NS-30,4533 (1983).

41 E. G. Stassinopoulos and J. H. King, Empirical Solar Proton Model for Orbiting Spacecraft Applications, IEEE Trans. Aerospace Electronic Systems 10, 4, 1974. (B17234)

42 E.G.Stassinopoulos and J.P.Raymond, The space radiation environment for electronics, 1983.

43 Feynman, J., T. P. Armstrong, L. Dao-Gibner, and S. Silverman, New Interplanetary Proton Fluence Model, J. Spacecraft and Rockets, 27, 403, 1990.

44 Feynman, J., G. Spitale, J. Wang, and S. Gabriel, Interplanetary Proton Fluence Model: JPL 1991, J. Geophys. Res., 98, 13,281-13,294, 1993.

45 Feynman, J., A. Ruzmaikin, and V. Berdichevs.ky, The JPL proton fluence model: an up- date, JASTP, 64, 1679-1686, 2002.

46 F. B. McLean and T. R. Oldham, "Charge Funneling in n- and p-Type Substrates," IEEE, Trans. Nucl. Sci., NS-29, 2018 (1982).

47 F.W. Sexton, Measurement of Single event Phenomenon in devices and ICs. IEEE NSRE Conference short course, chapter 3, 1992.

48 Goka, H. Matsumoto, H. Koshiishi, “A new International Geostationary Electron model: IGE-2006, from 1 keV to 5.2 MeV”, submit to Space Weather Journal, 2007.

49 G.P Ginet, “Air force research lab, AE9/AP-9 trapped radiation and plasma models re- quirement specifications”

50 Hastings, Daniel Spacecraft environment interactions / Daniel Hastings, Henry Garrett . - Cambridge : Cambridge University Press, 1996

51 Hovestadt D., Achtermann E., Ebel B., H~iusler B. and Paschmann G. (1972) New observations of the proton population of the radiation belt between 1.5 and 104 MeV. 1972.

52 Insoo Jun and Henry B. Garrett, Comparison of high-energy trapped particle environments at the earth and jupiter, 2005.

53 ISO 15390, Space environment (natural and artificial)-model of radiation impact by galac- tic cosmic rays, 2002.

54 ISO 15391, Probabilistic model for fluences and peak flux of solar energetic particles: Part I - protons, 2004.

55 I. A. Daglis, Space Storms and Space Weather Hazards. New York: Kluwer, 2001, vol. 38, ch. 3.

56 I. Manners, “Ring-opening polymerization of metallocenphanes: A new route to transition metal-based polymers,” Adv. Organometal. Chem., vol. 37, pp. 131-168, 1995.

57 Jacob I. Kleiman, Zelina Iskanderova, “Protection of materials and structures from space environment”

58 Janet L.Berth, “space and atmospheric effects from LEO to deep space. NASA Goddard space flight centre 301-286-8046”

59 Jensen, D. C., and J. C. Cain, An Interim Geomagnetic Field, J. Geophys. Res., 67, 3568, 1962.

60 Jerry John Sellers: “Understanding space, an introduction to Astronautics”

61 John B.Clasis, ET AL, Trapped radiation Handbook, 1971

62 Jun, R.T. Swimm, A. Ruzmaikin et al., Adv. Space Res. 40, 304 (2007a)

63 J. C. Anselmo, “Solar storm eyed as satellite killer,” AviationWeek Space Technol., vol. 146, pp. 61-62, Jan. 1997.

64 J. C. Cain, S. J. Hendricks, R. A. Langel, and W. V. Hudson, A Proposed Model for the International Geomagnetic Reference Field, 1965, J. Geomag. Geoelectr. 19, 335, 1967.

65 J.C. Pickel, B. Lawton, A.C. Freedman, and PJ. McNulty, "Proton Induced SEU in CMOS/SOS," Journal of Radiation Effects Research and Engineering, Vol. 7, #1, pg 69, 1989.

66 JE Mazur: “Space radiation environment”

67 J.-G. Wu’. L. Eliasson2, H. Lundstedt”, A. Hilgers4. L. Andersson2. 0. Norberg, “Space environment effects on geostationary spacecraft analysis and predictions”

68 J. H. Adams, Jr., R. Silberberg, and C. H. Tsao, “Cosmic ray effects of microelectronics, Part I: The near-earth particle environment,”, NRL Memo. Rep. 4506, Aug. 25, 1981.

69 J.L. Kaschmitter, D.L. Shaeffer, N.J. Colella, C.L. McKnett, P.G. Coakley, "Operation of Commercial R3000 Processors in the Low Earth Orbit (LEO) Space Environment", IEEE Trans. Nucl. Sci., vol 38, pp 1415-1428, Dec 1991.

70 J.W.Howard, Jr., D.M.Hardage, Spacecraft-Environment Interactions: Space Radiation and its Effects on Electronic Systems. NASA/TP-1999-209373, July 1999.

71 Kamide, Chian, Handbook of the Solar Terrestrial Environment, 2007.

72 Kenneth A. Label, Michael M getes, Paul M Martial, Janet barth, EG Stassinoponlos, Christina M seidleck, Cheryl J. dale, Army K moran, “Commercial microelectrics technologies

73 Kivelson, M. G. & Russell, C. T., eds. (1995). Introduction to Space Physics. Cambridge, UK: Cambridge University Press, 1995.

74 Knecht, D.J. & B.M. Shuman, in Handbook of Geophysics and the Space Environment, A.S. Jursa, ed., Air Force Geophysics Laboratory, 1985)

75 Konradi A., Hardy A. C. and Atwell W. (1987) Radiation Environment Models and the Atmospheric Cutoff. J. Spacecraft 24, 284--285.

76 Koons, H.C, Journal of spacecraft and rockets, 1983,0022-4650 vol.20 no.5 (425-431), “Summary of environmentally induced electrical discharges on the P78-2 (SCATHA) satellite “

77 K.A. LaBel, E.G. Stassinopoulos, G.J. Brucker, C.A. Stauffer, "SEU Tests of a 80386 Based Flight- Computer/Data-Handling System and Discrete PROM and EEPROM Devices, and SEL Tests of Discrete 80386, 80387, PROM, EEPROM and ASICS", Workshop Record for the 1992 IEEE Radiation Effects Data Workshop, pp 1-11, 1992.

78 K. A. Pfitzer, “Radiation dose to man and hardware as a function of atmospheric density in the 28.5 degree space station o rbit,” McDonnell Douglas Space Systems Co., Huntington Beach, CA, MDSSC Rep. H5387, 1990.

79 K.L. Bedingfield, R.D. Leach, and M.B. Alexander, “Spacecraft System Failures and Anomalies Attributed to the Natural Space Environment, August 1996, NASA Reference Publication 1390”

80 K. L Short, Microprocessors and Programmed Logic, Second Edition, New Jersey: Prentice Hall, Inc., 1987.

81 Lemaire J., Daly E. J., Vette J. I., Mcllwain C. E. and McKenna-Lawlor S. (1990) Secular Variations in the Geomagnetic Field and Calculations of Future Low Altitude Radiation Environments. In Proc. ESA, Workshop on Space Environment Analysis, 9-12 October 1990, ESTEC, Noordwijk, The Netherlands, pp. 5.17-5.30, ESA WPP-23.

82 Lieberman, B.S., and A.L. Melott. 2007. Considering the case for biodiversity cycles: Re- examining the evidence for periodicity in the fossil record. PLoS Computational Biology 2(8):e759, 2007.

83 Lionel S. Johns, Alan Shaw, Ray A. Williamson “Orbiting debris, a space environmental problem, Oct 1990, OTA-BP-ISC-72, NTIS order no. PB 91-114272”,workshop

84 MARTIN V. ZOMBECK, Handbook of Space Astronomy and Astrophysics, 3rd edition, Cambridge, USA,2007.

85 Mason, G.M., J.E. Mazur, J.R. Dwyer, J.R. Jokipii, R.E. Gold, and S.M. Krimigis. 2004. Abundances of heavy and ultraheavy ions in 3He-rich solar flares. Astrophysical Journal 606:555-

86 McComas D. J., S.J. Bame, B. L. Barraclouch, J. R. Donart, R.C. Elphic, J . T. Gosling, M. B. Moldwin, K. R. Moore and M. F. Thomsen, “Magnetospherics Plasma Analyzer: Initial

87 McCormack P. D. (1988) Radiation Dose and Shielding for the Space Station. Acta Astronautica 17, 231-241.

88 McIlwain, C. E., Coordinates for Mapping the Distribution of Magnetically Trapped Parti- cles, J. Geophys. Res., 66, pp. 3681-3691, 1961.

89 Messenger., Milton S.Ash, Single Event Phenomenon, 1997.

90 Moldwin, Mark, An introduction to Space Weather, 2008.

91 Mukund R. Patel, Spacecraft Power Systems, 2004.

92 M. D. Violet and A. R. Frederickson, “Spacecraft anomalies on the CRRES satellite corre- lated with the environment and insulator samples,” IEEE Trans. Nucl. Sci., vol. 40, pp. 1512- 1520, Dec. 1993.

93 M.N. Liu, S. Whitaker, "Low Power SEU Immune CMOS Memory Circuits", IEEE Trans. Nucl. Sci., vol 39, pp 1679-1684, Dec 1992.

94 M. S. Gussenhoven, E. G. Mullen, M. D. Violet, C. Hein, J. Bass, and D. Madden, “CRRES high energy proton flux maps,” IEEE Trans. Nucl. Sci., vol. 40, pp. 1450-1457, Dec. 1993.

95 M. S. Gussenhoven, E. G. Mullen, M. Sperry, and K. J. Kerns, “The effect of the March 1991 storm on accumulated dose for selected orbits: CRRES dose models,” IEEE Trans. Nucl. Sci., vol. 39, pp. 1765-1772, Dec. 1992.

96 M. S. Gussenhoven, E. G. Mullen, J. T. Bell, D. Madden, and E. Holeman, “APEXRAD: Low altitude orbit dose as a function of inclination, magnetic activity and solar cycle,” IEEE Trans. Nucl. Sci., vol. 44, pp. 2161-2168, Dec. 1997.

97 NASA-HDBK-4002, “Avoiding Problems caused by Spacecraft onboard internal charging effects” February 17, 1999.

98 NASDA Report, “DRTS getting ready for launch”, http://www.jaxa.jp/jda/library/nasdanews/ 2002/nr121.pdf, 2002.

99 Nymmik R.A., 1993.Averaged Energy Spectra of Peak Flux and Fluence Values in Solar Cosmic Ray Events, Proc. 23rd ICRC, Calgary, v.3, p.29,

100 Nymmik R.A., Models describing solar cosmic ray events, Radiation Measurements, 26(3), 417- 420, 1996

101 Nymmik R.A., Relationships among Solar Activity, SEP Occurrence Frequency, and Solar Energetic Particle Event Distribution Function, Proceedings of the 25th ICRC V.6, pp. 280-283, 1999a.

102 Nymmik R.A., 1999b., SEP event distribution function as inferred from Spaceborne meas- urements and Lunar rock isotopic data, Proceedings of the 26th ICRC, Salt Lake City, 1999.

103 NRC (National Research Council) report on “Managing space radiation risk in the new era

104 Owens, M.J., and N.U. Crooker. 2006. Coronal mass ejections and magnetic flux buildup in the heliosphere. Journal of Geophysical Research 111(A10).

105 Pickel, J. C., and Blandford, J. T., Jr., Cosmic-Ray-Induced Errors in MOS Devices, IEEE Trans. Nucl. Sci., NS-27, 1006-1015, 1980.

106 Purvis, C.K., Garrett, H.B., Whittlesey, A.C., and Stevens, N.J., 1984, “Design Guidelines for Assessing and Controlling Spacecraft Charging Effects,” NASA TP-2361, September, 1984”

107 Rami Vainio · Laurent Desorgher · Daniel Heynderickx · Marisa Storini, Erwin Flückiger Richard B. Horne · Gennady A. Kovaltsov · Karel Kudela, Monica Laurenza · Susan McKenna-Lawlor · Hanna Rothkaehl · Ilya G. Usoskin. “Dynamics of the earth’s Particle Ra- diation Environment” 2009.

108 Reames, H. V. Cane, and T. T. von Rosenvinge, “Energetic particle abundances on solar electron events,” The Astrophys. J., vol. 357, pp. 259-270, July 1, 1990.

109 Reames, D.V., and C.K. Ng. 2004. Heavy-element abundances in solar energetic particle events. Astrophysical Journal 610:510-522.

110 Richard H. Maurer, Martin E. Fraeman, Merk N. Martin, and David R. Roth, Harsh Envi- ronments:Space radiation environment, Effects, and Mitigation, 2008.

111 Robin Gubby, John Evans, “Space environment effects and satellite design”, Journal of Atmospheric and Solar-Terrestrial Physics 64 (2002) 1723- 1733, 2002.

112 R.J. Whitley, NASA Goddard Space Flight Center, Private communication, 1995.

113 R. Katz, R. Barto, P. McKerracher, R. Koga, "SEU hardening of Field Programmable Gate Arrays (FPGAs) for Space Applications and Device Characterization", IEEE Trans. Nucl. Sci., vol 41, pp 2179-2186, Dec 1994.

114 R-S Encoder Data Sheet, NASA VLSI Design Centre, 1994.

115 Sawyer, D.M., and J.I. Vette. 1976. AP-8 Trapped Proton Environment for Solar Maximum and Solar Minimum. NSSDC/ WDCA-R&S, 76-06. NASA Goddard Space Flight Center, Greenbelt, Md., December.

116 Significant achievements in particles and fields,1958-1964,NASA report No.SP-97, Wash- ington DC.1966.

117 SEECA, Single Event Effect Criticality Analysis (SEECA), NASA report, February 16, 1996.

118 Shu T. Lai, Critical Overview on Spacecraft Charging Mitigation Methods , IEEE TRANSACTIONS ON PLASMA SCIENCE, VOL. 31, NO. 6, DECEMBER 2003.

119 Space, Atmospheric, and Terrestrial Radiation Environments J. L. Barth, C. S. Dyer, and E. G. Stassinopoulos, 2003.

120 S. Bourdarie, D.Boscher, A.Sicard-Piet “GEO-handbook V1.1, , TN 1/13377 DESP - April,

121 S. L. Huston and K. A. Pfitzer, “A newmodel for the lowaltitude trapped proton environ- ment,” IEEE Trans. Nucl. Sci., vol. 45, pp. 2972-2978, Dec. 1998.

122 The solar flare heavy ion environment for single event upsets:A summary of the observa- tions over the last solar cycle,1973-1983,”IEEE transaction on nuclear science,1984.

123 Tsyganenko, A Magnetospheric Magnetic Field Model with a Warped Tail Current Sheet, Planet. Space Sci. 37, 5-20, 1989

124 Tylka, A.J, F. Dietrich, and P. R. Boberg, “Probability distributions of high-energy solar- heavy-ion flux from IMP-8: 1973-1996,” IEEE Trans. Nucl. Sci., vol. 44, pp. 2140-2149, Dec. 1997.

125 Tylka, A.J, F. Dietrich, and P. R. Boberg, “CREME96: A revision of the cosmic ray effects on microelectronics code,” IEEE Trans. Nucl. Sci., vol. 44, pp. 2150-2160, Dec. 1997.

126 Tylka, A.J., Dietrich, W.F., Boberg, P.R., Smith, E.C.and Adams, J.H., Jr., Single event up- sets caused by solar energetic heavy ionsNuclear Science, IEEE Transactions on Volume 43, Issue 6, Dec 1996 Page(s):2758 - 2766

127 Tylka, A.J., and M.A. Lee. 2006. A model for spectral and compositional variability at high energies in large gradual solar particle events. Astrophysical Journal 646:1319-1334.

128 Vampola, “Analysis of environmentally induced spacecraft anomalies. Journal of Space- craft and Rockets 31 2 (1994)”

129 Vampola, Alfred L. The hazardous Space Particle Environment” IEEE Transactions on Plasma Science, Vol. 28, NO. 6, December, 2000.

130 Vampola, A.L “The nature of bulk charging and its mitigation in spacecraft design”, 1996.

131 Vette, James, I. The AE-8 Trapped Electron Model Environment, 1991

132 V. I. Degtyarev, G. V. Popov and S. S. Sheshukov , Some Results of Empirical modelling of Quasi Trapped Electron Flux for Geostationary altitudes, Institute of Solar-Terrestrial Phys- ics, RAS, P.O. Box 4026, Irkutsk 664033, Russia.

133 W.J. Stapor, J.P. Meyers2, J.B. Langworthy, and E.L. Petersen “Two parameter bendel model calculations for predicting proton induced upset” IEEE Transactions on Nuclear Sci- ence, Vol. 37, NO. 6, December, 1990.

134 W.L. Pritchard, H.G. Suyderhoud, R.A. Nelson, Satellite Communication Systems Engi- neering, New Jersey: Prentice Hall, Inc., 1993.

135 W. P. Olson and K. A. Pfitzer, “A Quantitative Model of the Magnetospheric Magnetic Field”, J. Geophys. Res. 79, 3739, 1974.

136 W.W. Vaughan, K.O. Niehuss, and M.B. Alexander “Spacecraft Environments Interac- tions: Solar Activity and Effects on Spacecraft” 1996.

137 Xapsos, M. A., G. P. Summers, J. L. Barth, E. G. Stassinopoulos, and E. A. Burke, Probabil- ity Model for Worst Case Solar Proton Event Fluences, IEEE Trans. Nucl. Sci., 46, 1481-1485,

138 Xapsos, C. Stauffer, G. B. Gee, J. L. Barth, E. G. Stassinopoulos, and R. E. McGuire, “Model for solar proton risk assessment,” IEEE Trans. Nucl. Sci., vol. 51, no. 6, pp. 3394- 3398, Dec. 2004.

139 Xapsos, M. A., G. P. Summers, J. L. Barth, E. G. Stassinopoulos, and E. A. Burke, Probabil- ity Model for Cumulative Solar Proton Event Fluences, IEEE Trans. Nucl. Sci., 47, 486-490, 2000.

140 www.ieee.org/stamp/stamp.jsp

141 http://www.solarstorms.org/index.html

142 http://space-env.esa.int/R_and_D/SPENVIS.html

143 http://space.au.af.mil/primer/space_environment.pdf

144 http://standards.jpl.nasa.gov/jpl-nasa

145 http://parts.jpl.nasa.gov/asic/Sect.3.4.html#A0

146 http://radhome.gsfc.nasa.gov/radhome/papers/aspen.htm#ref

147 http://www.esa.int/TEC/Space_Environment/SEMEF3T4LZE_0.html

148 http://www.ngdc.noaa.gov/stp/SOLAR/ftpsatenvir.html#SEP

Details

Pages
98
Year
2009
ISBN (Book)
9783668072893
File size
3.2 MB
Language
English
Catalog Number
v308189
Institution / College
The University of Surrey
Grade
Tags
study space radiation environment

Author

Share

Previous

Title: Study of the Space Radiation Environment in GEO