Protection of Materials and Structures from Space Environment: ICPMSE-6 (Space Technology Proceedings) [1 ed.] 1402016905, 9781402016905, 9781402025952

Citation preview

PROTECTION OF MATERIALS AND STRUCTURES FROM SPACE ENVIRONMENT

Space Technology Proceedings VOLUME 5

PROTECTION OF MATERIALS AND STRUCTURES FROM SPACE ENVIRONMENT ICPMSE-6

Edited by JACOB I. KLEIMAN Integrity Testing Laboratory Inc., Toronto, Canada

and ZELINA ISKANDEROVA Integrity Testing Laboratory Inc., Toronto, Canada

KLUWER ACADEMIC PUBLISHERS NEW YORK, BOSTON, DORDRECHT, LONDON, MOSCOW

eBook ISBN: Print ISBN:

1-4020-2595-5 1-4020-1690-5

©2004 Springer Science + Business Media, Inc.

Print ©2003 Kluwer Academic Publishers Dordrecht All rights reserved

No part of this eBook may be reproduced or transmitted in any form or by any means, electronic, mechanical, recording, or otherwise, without written consent from the Publisher

Created in the United States of America

Visit Springer's eBookstore at: and the Springer Global Website Online at:

http://www.ebooks.kluweronline.com http://www.springeronline.com

v

Table of Contents Introduction Acknowledgements Organization

xi xiii xv

Section A New Era for Canada in Space M. Garneau

1

Space and Atmospheric Environments: From Low Earth Orbits to Deep Space J. L. Barth

7

Materials Interactions with the Space Environment: International Space Station May 2000 to May 2002 S. L. Koontz, M. Pedley, R. R. Mikatarian, J. Golden, P. Boeder , J. Kern, H. Barsamian, J. I. Minow, R. L. Alstatt, Mary J. Lorenz, B. Mayeaux, J. Alred, C. Soares, E. Christiansen, T. Schneider, D. Edwards

31

Photoconductivity in Transparent Arc-proof Coatings T. Cashman, J. Kaur, L. K. Muhieddine, M. Shanbhag, S. H. Ubaid, B. Welch, J. Vemulapalli, P. D. Hambourger

73

Effects of Space Environment Factors on Optical Materials for Space Application H. Liu,H. Geng, S. He, S. Yang, D. Yang, V.V. Abraimov, H. Wang

81

A Study of Synergistic Radiation Effects of Protons and Electrons on Teflon FEP/Al Degradation D. Yang, C. Li, H. Geng, S. He, S. Yang

91

Dose Rate Effects in Polymer Materials Irradiated in Vacuum B. A. Briskman, E. R. Klinshpont, V. F. Stepanov

99

Towards a Database for Assessment of Near-Earth Space Radiation Effects on Optical Glasses A.Gusarov, D. Doyle, M. Fruit

113

The Role of Proton and Electron “Abnormal” Formations in Radiation Influence on Construction Elements of Spacecrafts 123 Y. A. Grachov, O. R. Grigoryan, L. S. Novikov, I. V. Tchourilo A Study of Methylsilicone Rubber Damage Behavior Induced by Proton Irradiation L. Zhang, S. Yang, H. Geng, S. He, Q. Wei

131

vi Section B

A Unified Space Environment Effects Database for Russian and North American Organic and Inorganic Materials S. H. C. P. McCall, A. A. Clark, A. J. Clark, J. I. Kleiman, Z. Iskanderova, B. Briskman, E. Klinshpont, Y. Shavarin

139

The Effect of Heating on the Degradation of Ground Laboratory and Space Irradiated Teflon FEP K. K. de Groh, M. Martin

155

On the Thermal Stability of Polyimides for Space Application C.O.A. Semprimoschnig, S. Heltzel, A. Polsak, M. van Eesbeek

171

Synergistic Degradation of CV-1144-0 Due to Ultraviolet Radiation and Heat J. E. Haffke, J. A. Woollam

183

Behaviour of Thermal Control Coatings Under Atomic Oxygen and Ultraviolet Radiation S. Remaury, J. C. Guillaumon, P. Nabarra

193

Ground Testing of SCK5 White Silicone Paint for LEO Applications I. Gouzman, E. Grossman, G. Lempert, Y. Noter, Y. Lifshitz, V. Viel-Inguimbert, M. Dinguirard

203

Study of Polymer Coatings Resistance After the Long-Term Exposure on Space Station “MIR” E. N. Kablov, V. T. Minakov, I. S. Deev, E. F. Nikishin

217

Issues and Consequences of Atomic Oxygen Undercutting of Protected Polymers in Low Earth Orbit B. A. Banks, A. Snyder, S. K. Miller, R. Demko

235

Effect of Space Gaseous Environment on the Thermophysical Properties of Materials and Structures E. Litovsky, J. I. Kleiman, N. Menn

245

Cleanliness Support of the Launch Vehicle for Putting Into Orbit the Spacecraft Meteor-3M with the "SAGE-III" Instrument V.G.Sitalo, V.G.Tykhyy, L.P. Potapovych

257

Section C

Irreversible Shrinkage Effects of Carbon Fibers in Polymer Matrix Composites Exposed to the "MIR" Space Environment O.V. Startsev, D.A. Khristoforov, V.V. Issoupov, E.F. Nikishin, A.F. Rumyantsev

263

vii Combined Effect of Thermal and Mechanical Stresses on the Viscoelastic Properties of a Composite Material for Space Structures V. Issoupov, O. V. Startsev C. Lacabanne, P. Demont, V. Viel-Inguimbert, M. Dinguirard, E. F. Nikishin

271

Hyperthermal Reactions of Oxygen Atoms with Saturated Hydrocarbons T. K. Minton, D. J. Garton, H. Kinoshita

283

A Review of Lubrication Issues on the Canadarm 2 291

J. Antoniazzi, D. Milligan 3

Degradation of Polymers by O( P) in Low Earth Orbit A. Gindulyte, L. Massa, B. A. Banks, S. K. R. Miller

299

Iridium Metal as Potential Substrate for Experiments in Space L. Yan, J. A. Woollam

307

The Influence of the Atomic Oxygen Plasma on the Surface and on the Photoelectric Properties of Solar Arrays B.G. Atabaev, L.F. Lifanova, F. Rakhimova, A.V. Markov, I.V. Tchourilo

319

Some Aspects of Simulation of Outgassing Processes in Thermal Vacuum Exposure of Coatings Applied to Space Vehicles R..H. Khassanchine, A.V. Grigoresvskiy, Y.P. Gordeev

327

Section D Vacuum Ultraviolet Radiation Characterization of RF Air Plasma and Effects on Polymer Films J. Dever, C. McCracken, E. Bruckner

335

Studies of the Surface Oxidation of Silver by Atomic Oxygen M. L. Zheludkevich, A. G. Gusakov, A. G. Voropaev, A. A. Vecher, E. N. Kozyrski, S. A. Raspopov

351

Determination of the Energy Level of the Atomic Oxygen Flux Generated by the Space Simulation Apparatus using a Thermal Modeling Method X-X. Jiang, L. Lucier, D. Nikanpour, S. Gendron

359

Design and Testing of a Mini-Spectrometer System for On-Orbit Degradation Studies of Optical Materials M. Dinguirard, M. van Eesbeek, A. P. Tighe

367

Kapton as a Standard for AO Flux Measurement in LEO Ground Simulation Facilities: How Good Is It? E. Grossman, I. Gouzman, G. Lempert, Y. Noter, Y. Lifshitz

379

viii Temperature and Impingement Angle Dependences of Atomic Oxygen-Induced Erosion of Polyimide and Polyethylene Films Measured by Quartz Crystal Microbalance M. Tagawa, K. Yokota, T. Kida, N. Ohmae

391

Integrating Sphere Unit for Precision Measurement of Absolute Reflectance and Transmittance of Spacecraft Materials in a Vacuum Chamber

401

V. V. Eremenko, V. M. Naumenko, V. N. Fenchenko, V. G. Tykhyy

Numerical Simulation of Thermal Stress Induced by Thermocycling in Hot Rolled 1420 Al-Li Alloy H. Geng, S. He, D. Yang

407

Changes in Microstructure and Tensile Properties of Hot Rolled 1420 Al-Li Alloy Subjected to Thermocycling 413 H. Geng, S. He, D. Yang Section E

Photosil¥ Surface Modification Treatment of Polymer-based Space Materials and External Space Components Y. Gudimenko, R. Ng, J. I. Kleiman, Z. Iskanderova, P.C. Hughes, R.C. Tennyson, D. Milligan

419

Atomic Oxygen Resistant, Low α, Anti-Static Polyimides for Potential Space Applications A. J. Gavrin, S. W. Au-Yeung, R. Mojazza, K. A. Watson Jr., J. G. Smith, J. W. Connell

435

Deposition and Characteristics of Atomic Oxygen Protective Coatings Using Plasma Polymerized HMDSO 443 J. Wang, Y. Wang, X. Zhou, Z. Jin, Z. Yu Development of Protective and Passive Thermal Control Coatings on Carbon-Based Composite Materials for Application in Space 451 M. Francke, B. Fritsche, A. Moc, R. B. Heimann, Z. Iskanderova , J. I. Kleiman Structure and Composition of Non-Metallic Solar Array Materials Retrieved after Long-Term Exposure Overboard the “MIR” Orbital Space Station V. A. Letin, L. S. Gatsenko, I. S. Deev, E. A. Bakina, A. V. Malenkov, E. F. Nikishin 461 Micro- and Macrotribological Properties of Solid Lubricants in 5 Electronvolts Atomic Oxygen Exposures M. Tagawa, M. Muromoto, H. Kinoshita, N. Ohmae, K. Matsumoto, M. Suzuki 475 ZnSe Coatings for Spacecraft Electrochromic Thermal Control Surfaces L. Yan, J. A. Woollam, E. Franke

483

ix Perfluorocyclobutane (PFCB) Polyaryl Ethers for Space-Based Applications A. Gavrin, J. Nebo, N. Rice, L. Kagumba, D. W. Smith Jr., J. Jin, C. M. Topping

491

Section F

Fast Three-Dimensional Method of Modeling Atomic Oxygen Undercutting of Protected Polymers A. Snyder, B. A. Banks

503

Development and Verification of a Predictive Model and Engineering Software Guide for Durability Evaluation of Polymer-based Materials in LEO J. I. Kleiman, Z. Iskanderova, D. Talas, M. van Eesbeek, R. C. Tennyson 515 Comparative Study of Low Energy C and O Atoms Impact in a Hydrocarbon Surface M. Medvedeva, B. J. Garrison

527

A Direct Trajectory Dynamics Investigation of Fast O + Alkane Reactions R. Z. Pascual, G. C. Schatz, D. J. Garton

537

Mathematical Simulation Methods to Predict Changes of Integral and Spectral Optical Surface Characteristics of External Spacecraft Materials and Coatings 543 V. N. Vasliew, A. V. Grigorievskiy, Y. P. Gordeev Subject Index

551

Author Index

559

This page intentionally left blank

Introduction This publication presents the proceedings of ICPMSE-6, the sixth international conference on Protection of Materials and Structures from Space Environment, held in Toronto May 1-3, 2002. The ICPMSE series of meetings became an important part of the LEO space community since it was started in 1991. Since then, the meeting has grown steadily, attracting a large number of engineers, researchers, managers, and scientists from industrial companies, scientific institutions and government agencies in Canada, U.S.A., Asia, and Europe, thus becoming a true international event. This year’s meeting is gaining even stronger importance with the resumption of the ISS and other space projects in LEO, GEO and Deep Space. To reflect on these activities, the topics in the program have been extended to include protection of materials in GEO and Deep Space. The combination of a broad selection of technical and scientific topics addressed by internationally known speakers with the charm of Toronto and the hospitality of the organizers brings participants back year after year. The conference was hosted and organized by Integrity Testing Laboratory Inc. (ITL), and held at the University of Toronto’s Institute for Aerospace Studies (UTIAS). The meeting was sponsored by the Materials and Manufacturing Ontario (MMO) and the CRESTech, two Ontario Centres of Excellence; Air Force Office of Scientific Research (AFOSR/NL); MD Robotics; EMS Technologies; The Integrity Testing Laboratory (ITL); and the UTIAS. Summarizing, over 75 people from 15 countries including Canada, USA, France, Holland, Russia, Ukraine, China, Israel, Belarus, Japan, Belgium, Spain, Germany, India and England registered for the conference representing all major space agencies (NASA, Russian Space Agency, Canadian Space Agency, European Space Agency and the French Space Agency) and the major companies, institutions and government organizations involved in space activities, indicating a further increase in international co-operation in this critical area of protection of materials in space. A Plenary Session was held with 3 invited papers. Six Oral Sessions were organized with 34 papers, including three papers by Russian scientists invited by MMO through the “Distinguished Lecturer Program”. In addition, six Poster Sessions with over 38 papers and technical exhibitions by two companies were organized, with all presentations covering a variety of topics. For the ISS and other future space exploration projects, the safety of the crew and the soundness of the structures will be the major concern. Questions about thermal stability, resistance to soft and hard radiation sources, and combined effects of vacuum ultraviolet, atomic oxygen and micrometeoroids will continue to accumulate with the development of new materials and the increased use of polymers, plastics and composite materials. The papers in the proceedings are organized into six major sections as follows: a) Space Environmental Effects: Radiation and Charging Effects b) Space Environmental Effects: Synergism of AO/VUV/TC c) Space Environmental Effects: Synergism of AO/VUV/TC

xi

xii d) Space Environmental Effects: Instrumentation & Calibration e) New Materials and Processes f) Modeling and Computer Simulations

Jacob Kleiman, Chairman/Organizing Committee/ICPMSE-6 Integrity Testing Laboratory Inc., 20 January, 2003

Acknowledgements We would like to acknowledge the following for their generous support of ICPMSE-6, the Sixth International Conference on Protection of Materials and Structures from Space Environment; Air Force Office of Scientific Research (AFOSR/NL) Materials and Manufacturing Ontario (MMO) MD Robotics EMS Technologies The Integrity Testing Laboratory (ITL), CRESTech ASM International, Ontario Chapter; The University of Toronto Institute for Aerospace Studies As well, we would like to acknowledge all the people from ITL and UTIAS that contributed their time and effort and especially Sergei Sivolobtchik and Elissa Schaman two bright University of Toronto co-op students for their help in preparation of the materials for publication.

Jacob Kleiman, Integrity Testing Laboratory Inc. Conference Chairman

xiii

This page intentionally left blank

Organization Chairperson: Prof. Jacob Kleiman, ITL Inc./UTIAS, Canada

The Organizing Committee: B. A. Banks, NASA, Cleveland, USA D. L. Edwards, NASA, Huntsville, USA D. Nikanpour, Canadian Space Agency, Canada J. Golden, Boeing, Houston, USA E.F. Nikishin, M.V. Khrunichev State Space Scientific Production Center, Russia V.Sitalo, Design Bureau “Yuzhnoe”, Ukraine P. C. Trulove, Air Force, USA M. Van Eesbeek, ESA, Noordwijk, The Netherlands D. Yang, Harbin Institute of Technology, China

The Program Committee: M. Dinguirard, ONERA/DESP, France M. Finkenor, NASA, Huntsville, USA S. Koontz, NASA, Houston, USA T. Minton, Montana State University, USA G. Pippin, Boeing, Seattle, USA M. Tagawa, Kobe University, Japan E. Werling, CNES, France The Local Organizing Committee: Z. Iskanderova, ITL Inc./UTIAS, Toronto, Canada R. C. Tennyson, UTIAS, Toronto, Canada R. Worsfold, CRESTech, Toronto, Canada H. Pellegrini, MMO, Toronto, Canada P.Patanik, NRC-IAR, Ottawa, Canada Session Chairs Opening Session: Moderator: J. Kleiman-ITL Inc./UTIAS, Canada Session A: Space Environmental Effects: Radiation and Charging Effects Moderator: M. Dinguirard - ONERA, Toulouse, France Session B: Space Environmental Effects: Synergism of AO/VUV/TC Moderator: T. Minton - Montana State University, Bozeman, U.S.A.

xv

xvi Session C: Space Environmental Effects: Synergism of AO/VUV/TC Moderator: E. Nikishin M.V. Khrunitchev State Space Scientific Production Center, Moscow, Russia Session D: Space Environmental Effects: Instrumentation & Calibration Moderator: B. Banks, NASA Lewis Research Center, Cleveland, U.S.A. Session E: New Materials and Processes Moderator: M. Tagawa, Kobe University, Kobe, Japan Session F: Modeling and Computer Sumulations Moderator: Z. Iskanderova, Integrity Testing Laboratory Inc./UTIAS, Toronto, Canada

A NEW ERA FOR CANADA IN SPACE Speaking Notes

DR. MARC GARNEAU Canadian Space Agency 6767 Route de L ’Aeroport, St. Hubert, Quebec J3Y 8Y9, Canada

1 .0 Introduction It has only been 44 years since a small Russian aluminum sphere called Sputnik orbited the Earth for three months and, in doing so, launched the Space Age. Canada's own Space Program dates back to the launch of Alouette-1, which made us the third nation in space in 1962. Incredibly, for all the accomplishments along the way, our journey as a species into space has not yet spanned a human lifetime. Canada first went into space for practical reasons. The rationale behind the launch of our debut satellite, Alouette-1, was simple: Canadian scientists wanted to understand why ionospheric activity, prevalent in the far North, adversely affected radio communications. The best way to find out was to place a spacecraft in orbit for in situ measurements. Ten years later, mission accomplished: Alouette was switched off and Canada emerged as a world expert on ionospheric phenomena. Similarly, Canadians needed to find a way to connect communities scattered over our vast expanse. In 1972, we came up with a national communications satellite system called Anik, the first of its kind in the world. As a result, Telesat Canada, the system’s operator, is now the most experienced satellite control organization anywhere. The Canadian Space Agency was created in 1989 to manage Canada's civil space program, and to ensure that space science and technology benefit all Canadians.

2 .0 Building on a Legacy Building on a legacy of addressing the needs of Canadian society, the Canadian Space Agency has charted a vision for Canada's future in space—to expand and apply knowledge of space for the benefit of Canadians and of humankind and in so doing, inspire through excellence. To accomplish this, we are currently focusing on six strategic areas of importance to Canada: –Earth and Environment –ISS Utilization –Mars Exploration –Small Satellite Program

1

2 –Communicating with the Public; and –Preserving our technical expertise. Please do not assume that we are dropping or putting other agency programs on the backburner. These six strategic areas represent our decision to focus on certain areas in the face of limited budgets. Let’s look at each one briefly. 2.1 EARTH & ENVIRONMENT Our atmosphere is changing. We see evidence of climate change, air pollution, global warming and ozone depletion. Our planet is changing—there is relentless population growth, depletion of vegetable cover, soil erosion, loss of arable land, ground pollution, desertification, depletion of fresh water resources. Our oceans are changing due to pollution, depletion of fish stocks, and changing sea levels. The changes occurring in Canada and world-wide will have important long-term repercussions for our country. Using space as a vantage point to observe and monitor the Earth and our natural heritage will be our most important mission in the foreseeable future. Canada's RADARSAT-1, the world's first commercial Synthetic Aperture Radar, satellite, now has a six-year legacy of monitoring the Earth from space. The heart of the Canadian remote-sensing industry, RADARSAT-1 provides images that guide international rescue teams to disaster sites and dispatch Arctic icebreakers where needed; images that detect oil spills and supports scientists in the fields of cartography, mineral and oil exploration, hydrology, forestry, oceanography and agriculture. RADARSAT-1 currently has a network of stations and a global client base that includes more than 600 commercial and government users from more than 57 countries. It has also responded to over 50 emergencies—both natural and man-made disasters, and is now part of the International Charter of Space and Major Disasters. We are preparing to launch RADARSAT-2 in late 2003, which will build on the success of its predecessor with the most advanced satellite radar technology on the commercial market. 2.2 THE UTILIZATION OF THE INTERNATIONAL SPACE STATION Canada is one of 15 partner nations in the midst of building the most ambitious science and engineering project in the history of humanity—the construction of the International Space Station. The first element of the Canada's contribution to the project, Canadarm2, is already operating on board the Station, and our next element, the Mobile Base System, is scheduled for launch on May 30 of this year. Canada is fulfilling its promised contribution to the International Space Station. However, we must also focus on the main purpose of the International Space Station—to allow us to perform scientific experiments in the unique conditions of microgravity. We would also like to ensure that Canadian astronauts continue to be an integral part of building and working on the Station to perform research on behalf of Canadian scientists. This will be an important ongoing activity for the next 10 to 15 years, both through government and, hopefully,

3 through commercial channels. The success of the Station will depend on it. Canada is on its way to Mars 2.3 CANADA IS ON ITS WAY TO MARS Canada must also look outward into the universe as well as back at Earth. As our closest planetary neighbor, Mars holds a special appeal for the scientific community. Researchers hope that by looking outward into the depths of the universe to planets like Mars, we may find answers to some of the most basic questions about life here on Earth; the formation of our planet and our solar system; and whether life—and the water that sustains it—exists elsewhere in the universe. As we speak, a Canadian scientific instrument, known as the Thermal Plasma Analyzer, is on its way to Mars on board the Japanese satellite Nozomi. When it reaches Mars in 2004, the Thermal Plasma Analyzer, or TPA for short, will provide the scientific community with valuable information on the origin and composition of Mars’s atmosphere. The Canadian public supports our participation in Mars exploration. Mars exploration "inspires" Canadians. In a national public opinion poll conducted in March 2002, 77% of Canadians expressed support for Canada's participation in future missions to Mars. Exploring Mars is the next major international space program after Space Station. Canada can contribute but will only do so if: –It has a visible role –Our involvement is science-driven and technology-enabled –And, of course, if funding permits our involvement. We are now in the planning stages to determine what form our potential involvement may take. We are consulting with the Canadian scientific community and Canadian industry, as well as with our international partners. And we are currently conducting feasibility studies to help us identify opportunities for 2007 and 2009 missions. We are also assessing Canada’s potential contribution beyond 2009. Ideally, we would like to see a “distinctly Canadian” mission in 2011, conducted with our partners but one that would feature Canadian ideas, technologies and expertise. 2.4 SMALL SATELLITE PROGRAM Another area for future development is a small satellite program. There are compelling reasons why the Canadian Space Agency should be providing a Small Satellite capability on a regular basis, e.g. every 2 years. It would allow Canadian industry to space-test new hardware, i.e. technology demonstration. It would provide a more costeffective platform for new science instruments for Canadian researchers, and it would allow remote-sensing pilot projects. It would also enable Canada to compete internationally in the small satellite market, providing it can do so on a cost-competitive basis. 2.5 COMMUNICATING WITH THE PUBLIC The public is very interested in space.

4 We know this from polls and from all the requests we receive on a continuous basis. Canadians believe space is important for Canada. However, they know very little about many of our programs and how they serve Canadians. We need to better inform them— a challenging task. Young Canadians in particular, are fascinated by space. We need to continue to reach out to them in a variety of ways to stimulate their interest in Science and Technology. 2.6 PRESERVING OUR EXPERTISE The CSA is a multi-faceted organization. It manages space programs, but it also: • performs focused in-house R&D in specific areas; • performs satellite operations from St-Hubert; • tests space hardware at the David Florida Laboratory; • trains astronauts and cosmonauts in the use of the Mobile Servicing System; and • has a strong and visible interface with the public. We need to continue to foster the growth of in-house skills to maintain Canada's international leadership in its strategic areas of expertise.

3 .0 The Protection of Spacecraft and Materials from the Space Environment With the ambitious agenda we have laid out for ourselves in the coming years, the protection of spacecraft and materials from the space environment has taken on increased importance. Debris, of course, is a major concern. It has been estimated that more than 4000 space launches have taken place since 1957, leading to more than 8500 trackable objects above 10 cm in size in near-Earth orbit. Of these, between 600 and 700 are operational spacecraft. The remainder is debris, which poses a hazard to human space flight, as astronauts are well aware, as well as the safe operation of unmanned spacecraft. Orbiters have suffered “dings” on their windows. Spacecraft are also exposed to contamination from ultraviolet radiation and atomic oxygen in low Earth orbit, which can degrade sensitive optical surfaces. We are developing increasingly sophisticated hardware, like Canadarm2, which is designed to be maintained and repaired in space for the duration of its lifetime. Protecting the Space Station itself, as well as visiting vehicles, such as the space shuttle, the Soyuz, Progress, the European Space Agency and NASDA ATV (automated transfer vehicle) is a growing concern. A proactive approach to defending space hardware from both human-induced contamination and ever-increasing space debris is essential. Space debris is becoming increasingly concentrated in useful orbits, where activity is at its greatest. This includes geostationary orbit, where telecommunications satellites are located, medium Earth orbit, where numerous Earth Observation spacecraft are in orbit, and in low Earth orbit, where not only the Space Station and shuttle and transfer vehicles are affected, but also low Earth orbit satellites, particularly the micro- and nanosatellites.

5 The recent news of the go-ahead for the Galileo network of satellites by the European Space Agency is indicative of this trend. The only natural debris removal mechanism so far is the decay in the orbit of space debris, and their eventual fall back to Earth. This phenomena is only relevant to very low orbits, as you know, since for the medium Earth orbit, the orbital decay time is of the order of a couple of centuries. For higher orbits, it may take many thousands of years. But even the natural decay of space debris can be problematic. This is an issue of special concern to Canada , the only country to have been impacted by a satellite carrying a nuclear reactor, when the Soviet Union’s Cosmos 954 disintegrated over the Northwest Territories in 1978, spreading contaminated debris over several hundred kilometres. With a projected increase in the of number of launches, the debris population will undoubtedly increase. The Canadian Space Agency has recognized this trend, and has initiated research activity in space debris damage mitigation design. For instance, Canada's RADARSAT-2 spacecraft has introduced a debris damage mitigation design to protect its synthetic aperture radar antenna and reduce debris generation. Through the David Florida Laboratory, we will continue to ensure the rigorous qualification satellites and other forms of space hardware for space flight through its assembly, integration and testing facilities. And we will continue to support the fruitful exchange of ideas in for a such as this Conference as we collectively search for new approaches to address some of these pressing issues.

4 .0 Conclusions And we will continue to support the fruitful exchange of ideas in for a such as this conference as we collectively search for new approaches to address some of these pressing issues. In the coming years, the Canadian Space Agency will continue to build upon a 40-year tradition of Canadian excellence in space. We have charted our course for the future— Canada's future in space. We are committed to ensuring that Canada has a prominent role as we take our first steps off our home planet and venture out into the universe. And as our past achievements in space were guided by our desire to address the needs of Canadians, so, too, will our future be governed by issues Canadians take to heart as we strive to advance scientific knowledge for the benefit of Canadians, and indeed, for the benefit of all of humanity.

This page intentionally left blank

SPACE AND ATMOSPHERIC ENVIRONMENTS: FROM LOW EARTH ORBITS TO DEEP SPACE JANET L. BARTH NASA/Goddard Space Flight Center 301-286-8046 E-mail: [email protected]

Abstract Natural space and atmospheric environments pose a difficult challenge for designers of technological systems in space. The deleterious effects of environment interactions with the systems include degradation of materials, thermal changes, contamination, excitation, spacecraft glow, charging, radiation damage, and induced background interference. Design accommodations must be realistic with minimum impact on performance while maintaining a balance between cost and risk. The goal of applied research in space environments and effects is to limit environmental impacts at low cost relative to spacecraft cost, to infuse enabling and commercial off-the-shelf technologies into space programs. The need to perform applied research to understand the space environment in a practical sense and to develop methods to mitigate these environment effects is frequently underestimated by space agencies and industry. Applied science research in this area is critical because the complexity of spacecraft systems is increasing, and they are exposed simultaneously to a multitude of space environments.

1. 0 Introduction Spacecraft are exposed to a multitude of environments that are not present at the surface of the Earth, including plasmas, high-energy charged particles, neutral gases, x-rays, ultraviolet (UV) irradiation, meteoroids, and orbital debris. The interaction of these environments with spacecraft systems cause degradation of materials, thermal changes, contamination, excitation, spacecraft glow, charging, radiation damage, and induced background interference. The damaging effects of natural space and atmospheric environments pose a difficult challenge for spacecraft designers. Unfortunately, the need to perform applied research to understand and model the space environments, to understand the physics of the interaction of the environment with spacecraft systems, and to develop methods to mitigate environmental effects is frequently underestimated by space agencies and industry. At the same time that the complexity and performance requirements of spacecraft systems are increasing, other system drivers reduce our ability to meet

7

8 requirements. For example, the demand for commercial microelectronics reduces the availability of components suitable for space environments. Also, the need to design lighter and more complex spacecraft structures pushes the development of exotic materials for space use. Uncertainties in space environment and effects models used to predict the performance of new technologies translate into large design margins. Large design margins can preclude the use of new technologies that will meet mission requirements. The goal of applied research in space environments and effects is to reduce design margins used to account for the uncertainty of performance predictions, thereby enabling technology infusion into space programs. Design accommodations must have minimum impact on performance and budgets. The challenge is to achieve a realistic balance between cost of environment accommodations and mission risk. Several organizations have developed concurrent engineering approaches to accommodating environment effects and assessing risk [1,2]. Regardless of the environmental effect or the technology, the approaches follow these steps: Define the environment external to the spacecraft; Evaluate the environment interaction with the spacecraft; Define the requirements and define criticality factors; Evaluate design and performance characteristics of components; “Engineer” with designers and program managers including risk analysis and definition of design margins; Iterate the process with updated knowledge. Note that the first step for every mission is to define the level of the environments in metrics that are applicable the effect on the spacecraft technologies. This is accomplished by using environment and interaction models, and when models are not available, by using in-flight data. The purpose of this paper is to describe the natural space environments that must be taken into account when designing spacecraft. Specific effects on materials and structures can be found in other contributions to these proceedings. Further information on the radiation environment can be found in iii[3] and information on radiation effects can be found in [4].iv i

ii

[, ]

2. 0 Description of the Environments The space and atmospheric environments relevant to spacecraft effects can be roughly categorized into meteoroid and debris, ultraviolet irradiation, neutral thermosphere, cold and hot plasma, and particle radiation. The differences between the atmospheres and magnetospheres of the planets and interplanetary space are dramatic. Even within the atmospheric and magnetospheric systems of the Earth, there are large spatial and temporal variations in the constituency and density of the environments. For the purpose of discussion of environmental definitions, missions can be roughly categorized into low earth orbits (LEOs), middle earth orbits (MEOs), geosynchronous (GEO), geosynchronous transfer orbits (GTOs), interplanetary, and other planets. Table 1 gives a summary of the differences between the space environments around Earth and other planets

9 TABLE 1. Variation of planetary environments from Earth ENVIRONMENT Solar Wind Meteoroids Orbital Debris Galactic Cosmic Rays Solar Particles Solar Radiance Atmospheres Trapped Radiation

COMPARISON TO EARTH ~ Same ~ Same None at this time Small variation Large variation with radial distance from Sun Large variation with radial distance from Sun Large variations Large variations

2.1 THE SOLAR INFLUENCE ON SPACE ENVIRONMENTS The complex environment of Sun-Earth space consists of time varying ultraviolet, xray, plasma, and high-energy particle environments. Variations depend on location in space and on the year in the solar cycle, both somewhat predictable. However, large variations that depend on events on the Sun are not predictable with reasonable certainty and are known only statistically based on past history. Because the Sun is a gas, its solar magnetic field is convoluted and highly variable. Both the long-term variation in the magnetic field that occurs in a 22-year cycle and the short term variations in the form of intense, short lived storms are responsible for observable changes in the interplanetary and near-Earth environments.

Figure 1. The corona extends several solar diameters.

The sun’s outer atmosphere, the corona (see Figure 1), extends several solar diameters into interplanetary space. The corona continuously emits a stream of protons, electrons, doubly charged helium ions, and small amounts of other heavy ions, collectively called the solar wind. It was once thought that the region where the solar wind could no longer be detected, i.e., the boundary of the heliosphere, was not far beyond Jupiter (800 million km). However, the Pioneer 10 spacecraft, presently at > 12

10 billion kilometers from Earth, is still measuring solar wind. Scientists now believe that the boundary could lie as far as 17 billion kilometers from the Earth [5]. [v] The high temperature of the corona inputs sufficient energy to allow electrons to escape the gravitational pull of the sun. The result of the electron ejections is a charge imbalance resulting in the ejection of protons and heavier ions from the corona. The ejected gas is so hot that the particles are homogenized into a dilute plasma. The energy density of the plasma exceeds that of its magnetic field so the solar magnetic field is “frozen” into the plasma. The electrically neutral plasma streams radially outward from the sun at a velocity of approximately 300 to 900 kilometers per second with a temperature on the order of 104 to 106 K. While the solar wind is millions of metric tons of matter moving at a million kilometers per hour, its density is so low that the physics is that of a vacuum. The energies of the particles range from approximately 0.5 to 2.0 keV/n. The average density of the solar wind is 1 to 30 particles/cm3. Figure 2 shows that the solar wind velocity and density can vary greatly over a short time period. Table 2 gives the approximate particle composition of the solar wind.

Figure 2. The solar wind velocity and density are highly variable and are a function of the activity on the sun. SOHO/University of Maryland

TABLE 2. Solar wind particle composition

PARTICLE Proton He++ Other Heavy Ions Electrons

ABUNDANCE 95% of the positively charged particles ~4% of the positively charged particles < 1% of the positively charged particles Number needed to make solar wind neutral

11 It is well known that the level of activity of the sun varies with time defining “solar cycles”. The solar cycle as a recurrent pattern of solar magnetic activity was first identified in 1843 by the German observer, Schwabe, who found an approximately 11year cycle in the number of sunspots1 (see Figure 3). The 11-year cycle of sunspots corresponds to similar 11-year cycles of other features in the sun’s active regions, including the number of faculae, the rate of incidence of solar flares and coronal mass ejections (CMEs), and the intensity of coronal x-ray and radio-frequency emissions.

Figure 3. Sunspots are regions of highly dense magnetic field. after Lund Observatory

From Figure 4, it can be seen that the length of the solar cycle can be highly variable. From 1645 to 1715, the sunspot activity seemed to disappear. Because temperatures on Earth dropped during that time, those 70 years are known as the little ice age. From 1100-1387, there was an increase in the number of sunspots. Studies of recent solar cycles [5, 6], Cycles 19 through 22, have determined that the length of the solar cycle over the past 40 years has ranged from 9 to 13 years, with 11.5 being the average. For modeling purposes and for defining the environment for spacecraft missions, the solar cycle can be divided into a 7-year maximum phase of high levels of activity and a relatively “quiet” 4-year minimum phase. The space environment is dominated by the activity of the Sun, which acts as both a source and a modulator. It is a source of protons and heavier ions via the periodic highenergy solar events that accelerate large numbers of particles. The solar wind is also a source of the particles trapped in outer regions of the Earth’s radiation belts. The galactic cosmic ray heavy ion (GCR) levels follow a cyclic pattern reflecting the activity level of the sun because they originate outside of the solar system and must “fight” against the solar wind to reach interplanetary space. As a result, the GCR levels are highest during solar minimum and lowest during solar maximum. Atmospheric neutrons are secondary products of collisions between GCRs and oxygen or nitrogen atoms in the Earth’s atmosphere; therefore, their levels are also modulated by the solar cycle. Finally, the levels of particles trapped in planetary magnetospheres are modulated by both longterm variations in solar activity and solar storm events. The effect of the cyclic variation vi, vii]

1

cooler areas of the sun seen as dark “spots” through a telescope

12 of the sun’s activity will be discussed in more detail in later sections as it applies to specific environments.

Figure 4. Yearly sunspot numbers

Solar flares and coronal mass ejections (CMEs) are two storm phenomena occurring on the Sun that affect particle levels. Solar flares are seen as sudden brightenings in the photosphere near sunspots (see Figure 5). Flares are intense releases of energy involving tearing and reconnection of strong magnetic field lines. They are the solar systems largest explosive events. Large increases in the solar wind density are measured in interplanetary space after solar flare occurrence because the energy released from the flare accelerates particles in the solar plasma to high energies. CMEs occur in the chromosphere, the layer of the sun outside of the photosphere. The chromosphere can be seen only when filtering out the bright light of the photosphere. In Figure 6, the chromosphere is seen as a bright rim around the sun. CMEs are observed as large bubbles of gas and magnetic field (see Figure 7). A CME can release approximately 1017 grams of plasma into interplanetary space. The

Figure 5. Brightening seen with a solar flare.

Figure 6. Bright rim around the sun is the chromosphere.

13 mechanism for the plasma release is not completely understood. CMEs result in large increases in solar wind velocity. It is the shock wave of the plasma release that is associated with particle acceleration and magnetic storms at the Earth. CMEs are poorly associated with flares but, in very large event CMEs, both CMEs and flares occur together [8]. The particle composition of CMEs and solar flares is discussed in Section 2.9. [viii]

Figure 7. Bubble of gas associated with a coronal mass ejection. NASA/SMM 24 Oct. 1989

2.2 METEOROIDS AND ORBITAL DEBRIS Meteoroids are primarily remnants of comet orbits. Several times a year Earth encounters increased meteoroid exposure as it intersects a comet orbit. Also, sporadic particles are released on a daily basis from the asteroid belt. Orbital debris consists of operational payloads, spent rockets stages, fragments of rockets and satellites, and other hardware and ejecta. The United States Air Force Space Command’s North American Aerospace Defence Command (NORAD) tracks over 7,000 objects in LEO that are greater than 10 cm in size,

Figure 8. Location of objects tracked by NORAD

and there are tens of thousands smaller objects. Figure 8 shows the location of the objects tracked by NORAD. From the figure, it is possible to see the large number of objects in the LEO and GEO regions of space where most space agency, military, and commercial operations take place.

14 Meteoroids and orbital debris are a threat to spacecraft by causing structural damage and decompression, hypervelocity impacts from larger particles, surface erosion from collisions with smaller objects, and surface effects that cause changes in thermal, electrical, and optical properties. Mission risk factors include increased duration, increased vehicle size, vehicle design, solar cycle, orbit altitude, and inclination, and the threat is highly directional. Koontz et al. [9] give examples of micrometeoroid and orbital debris impacts on the International Space Station (ISS). ix]

2.3 ULTRAVIOLET IRRADIATION The sun is the natural source of ultraviolet irradiation, which has wavelengths of about 100 to 400 nanometers. UV irradiance can penetrate the atmosphere to reach the surface of the Earth. UV is an important component of the environment to evaluate due to its degradation effects on spacecraft surface materials. It is known to interact with atoms in the atmosphere, particularly oxygen, and ionizing particles to act synergistically on surface materials of spacecraft. UV radiation diffuses with distance from the sun at a rate of 1/R2 where R is the radial distance from the Sun. Table 3 lists solar UV irradiance at each planet. TABLE 3. Solar UV irradiance as a function distance from the Sun

PLANET Mercury Venus Earth Mars Jupiter Saturn Uranus Neptune Pluto

DISTANCE FROM SUN (AU) 0.39 0.72 1.00 1.52 5.20 9.54 19.19 30.06 39.53

IRRADIANCE (W/m2) 9,126.6 2,613.9 1,367.6 595.0 51.0 15.0 3.7 1.5 0.9

2.4 PLASMA ENVIRONMENTS Plasma is ionized gas in which electron and ion densities are approximately equal. Plasma is distinguished from the energetic particle population in that it does not cause radiation effects and has energies < 100 keV. Plasma sources are the ionosphere, geomagnetic sub-storm activity, and the solar wind. The solar wind plasma from the solar corona was discussed in Section 2.1. The ionosphere is the electrically charged portion of the atmosphere and is characterized by low energy (eV) and high density. Plasma from geomagnetic sub-storm activity, on the other hand, has high energy (keV) and low density. Figure 9 shows the plasma around the Earth as seen by the Extreme Ultraviolet (EUV) instrument on NASA’s IMAGE (Imager for Magnetopause-toAuroral Global Exploration) spacecraft. Plasma shows dramatic variation with altitude, latitude, magnetic field strength, and solar activity. The solar wind plasma was

15 discussed in Section 2.1. The other plasma environments will be discussed in Sections 2.5.2 and 0.

Figure 9. The Helium ion plasma around the Earth as seen by the EUV instrument on the IMAGE spacecraft. Note the auroral activity.

2.5 THE EARTH’S ATMOSPHERE The Earth’s atmosphere is composed of complex layers of matter that are loosely defined by their dominant constituents. Starting from the surface of the Earth, the layers are the troposphere, stratosphere, mesosphere, the neutral thermosphere, and the charged thermosphere (ionosphere). The layers overlap and form a connected system. Figure 10 shows the altitude domains of the regions of the atmosphere. Low altitude spacecraft (< 800 km) are exposed to the environments of thermosphere, so those environments will be discussed in more detail.

2.5.1 Neutral Thermosphere The neutral thermosphere is the neutral portion of the Earth’s atmosphere at 90 to 600 km altitude above the surface of the Earth, composed primarily of neutral gases. In the lower thermosphere, the neutral population is dominated by atomic oxygen and by hydrogen and helium in the higher thermosphere. The distribution of the thermosphere neutral gases varies with solar activity because of heating caused by absorption of solar extreme ultraviolet radiation (EUV). A proxy commonly used for EUV is the 10.7-cm radio flux (F10.7). The main effects of the neutrals on spacecraft are drag, degradation of surface materials, and spacecraft glow. Drag results in altitude decay and torques. Drag is a function of the density of the neutral gas, hence is strongly affected by solar activity. The impact of solar storms on the Earth’s atmospheric density often causes sudden changes in the location of tracked objects. Figure 11 is a plot of the number of objects that were lost after a large magnetic storm in March of 1989.

16 Figure 10. The altitude domains of the Earth’s atmosphere, after NASA/MSFC

Figure 11. A plot of the number of tracked objects lost after a large magnetic storm

The degradation of surface materials is also a serious problem in very low earth orbits due to the presence high levels of atomic oxygen at 200 to 400 km. As with other thermosphere constituents, the level of atomic oxygen varies with the solar cycle. The erosion of surface materials causes changes in thermal, mechanical, and optical properties. Micrometeoroid impacts, sputtering, UV exposure, contamination, and ionizing radiation can aggravate these effects. Optical emissions generated by excitation of meta-stable molecules can also cause spacecraft glow. The surface acts as catalyst, therefore, the effect is material dependent.

17 2.5.2 Charged Thermosphere (Ionosphere) The Earth’s ionosphere is the electrically charged portion of the upper atmosphere from 100 to 800 km altitude (see Figure 10). It is a low energy (eV) plasma with high density relative to the magnetospheric plasma and the solar wind. Supersonic spacecraft motion through background ions in the ionospheric plasma has detrimental effects on spacecraft in LEO orbits, including solar array coupling to the plasma causing current drain on solar arrays, generation and emission of plasma waves, and increased surface contamination. There is renewed interest in studying the ionosphere because military and civilian communications are severely degraded during storms in the ionosphere induced by solar activity. The disruptions of communications during storms are far reaching as they affect high frequency radio, backscatter radar, satellite communications, and global positioning system (GPS) location. 2.6 ATMOSPHERES OF OTHER PLANETS Other planets in our solar system have atmospheres which differ dramatically from the Earth’s atmosphere. Therefore, designers of missions to other planets must take into account the differences from the Earth’s atmosphere and those of other planets and adjust the evaluation and mitigation of the effects accordingly. The ability of a planet to have an atmosphere is dependent on the planetary surface pressure and gravity (Table 4). For the four “terrestrial planets” the table shows that Mercury and Mars have very thin atmospheres and that the atmosphere of Venus is 92 times that of the Earth. The Jovian planets have large variations in their atmospheric density as a function of radial distance from the center of the planet because they are primarily composed of gases and clouds.

TABLE 4. Planetary pressure and surface gravity

PLANET Mercury Venus Earth Mars Jupiter Saturn Uranus Neptune Pluto

SURFACE PRESSURE (BARS) 1.00x10-15 92 1 8.00x10-3 >>100 >>100 >>100 >>100 3.00x10-6

SURFACE GRAVITY (M/S) 3.70 8.87 9.78 3.69 23.12 8.96 8.69 11.00 0.66

In addition to the density of the atmosphere, it is critical to understand the composition of the atmospheres which is shown in Table 5. Notable is the similarity of the great hydrogen planets, Jupiter, Saturn, Uranus, and Neptune and the carbon dioxide atmospheres of Venus and Mars.

18 TABLE 5. Composition of the atmospheres of each planet

PLANET Mercury

Venus Earth

Mars

Jupiter Saturn Uranus

Neptune

Pluto

ATMOSPHERIC COMPOSITION 42% O2 29% Na 22% H2 6% He 96% CO2 3% N2 78% N2 21% O2 1% Ar 95% CO2 3% N2 2% Ar 90% H2 10% He 96% H2 3% He 83% H2 15% He 2% CH4 80% H2 19% He 1% CH4 ? CH4? N2 Ice?

2.7 THE EARTH’S MAGNETOSPHERE The Earth’s magnetosphere is a cavity formed by the interaction of the Earth’s magnetic field and the solar wind. In the absence of the solar wind, the Earth’s magnetic field would be shaped like the field of a bar magnet; non-varying, nearly symmetric about the magnetic axis, extending outward to long distances, and open at the poles. The bar magnet representation is accurate up to altitude of 4 to 5 Earth radii. The solar wind plasma, with its embedded solar magnetic field, compresses the geomagnetic field until there is balance between the magnetic pressure from the Earth and the momentum pressure from the solar wind forming a “bow shock”. On the dayside, during moderate solar wind conditions, the magnetosphere terminates at the magnetopause at ~10 Earth radii altitude. At the location of this “collision-less” shock, the solar wind plasma cannot penetrate deeply into the geomagnetic field because of its charged particle composition. In fact, 99.9% of the solar wind particles pass around the Earth’s magnetosphere. The flow of the solar wind around the flanks of the magnetopause stretches the geomagnetic field in the anti-solar direction into a long tail of up to ~300 Earth radii altitude. Some tail field lines are not closed and are connected to the solar magnetic field embedded in the solar wind. Figure 12 shows a depiction of the

19 magnetosphere.The magnetosphere is filled with plasma that originates from the ionosphere and the solar wind. The plasmasphere is at low and mid latitudes in the inner magnetosphere. The plasma sheet resides in the magnetotail. Overlapping the plasmasphere and the plasma sheet are the high-energy radiation belts or Van Allen belts (named for their discoverer, James Van Allen). Charged particles become trapped because the Earth’s magnetic field constrains their motion. They spiral around the field lines in a helicoidal path while bouncing back and forth between the magnetic poles. Superimposed on these spiral and bounce motions is a longitudinal drift of the particles because of the gradient of the magnetic field. Figure 13 illustrates the three motions. When the particle makes a complete azimuthal rotation, it has traced a “drift shell” (see Figure 14). The Van Allen belts will be discussed in more detail in Section 0.

Figure 12. The Earth’s magnetosphere, adapted from T. W. Hill by P.H. Reiff

Figure 13. The three motions of the trapped particles form drift shells. after Hess

20 2.7.1 Plasma Storms Geomagnetc substorms in the magnetotail plasma sheet can create “hot plasmas” which are injected into near-Earth regions of the magnetosphere. The effects of the plasma injections include biasing of spacecraft instrument readings, acing which causes upsets to electronics, increased current collection, re-attraction of contaminants, and ion sputtering which in turn leads to acceleration of material erosion. Missions affected by these injections are those in GEOs, GTOs, and MEOs. Conditions for the charging effects are large differentials, large fraction of total flux, darkness, and large spacecraft. Satellites at GEO have also measured strong local time effects on the rates of spacecraft charging with most occurring as the satellite passes into the dawn sector.

Figure 14. Drift shell of a trapped particle. Lamarie et al.

2.7.2 Van Allen Radiation Belts The Van Allen belts consist of protons, electrons and heavier ions that are “trapped” on the Earth’s magnetic field lines. The trapped electrons have energies up to 10s of MeV, and the trapped protons and heavier ions have energies up to 100s of MeV. These particles have complex spatial distributions that vary by several orders of magnitude depending on orbit inclination and altitude. The sun is a driver for long and short-term variations in the locations and levels of trapped particles. A feature of the Van Allen belts is the South Atlantic Anomaly (SAA). The 11° angle between the magnetic and geographic axes and the offset of the geographic and geomagnetic centers of the Earth causes a depression in the magnetic field in the South Atlantic. This magnetic field sink causes charged particles to be trapped at low altitudes ( 300 mils of shielding are dominated by the highly energetic trapped protons.

Figure 15. Total ionizing dose-depth curves for various orbits around the Earth

Total non-ionizing dose (also known as displacement damage or bulk damage) is another cumulative effect that causes degradation of solar cells, optocouplers, and focal plane arrays. As particles slow down in material and come to rest they knock atoms out of their lattice location creating defects which increase the resistance of the device. Electrons, protons, and neutrons cause displacement damage, and the energy spectra of the particles are used to evaluate the level of the hazard. Modern microelectronic systems are plagued by the effects of single particle strikes, namely SEEs, on sensitive regions of devices. There are several types of SEEs, including single event upsets, single event latchups, and single event transients. The consequences of SEEs in systems range from loss of data to the loss of a mission. SEEs are caused by ions from GCRs, solar particle events, and trapped protons. Figure 16 shows the geographic location of single event upsets on the SEASTAR satellite (98°

22 inclination, 705 km altitude) flight data recorder on a world map of latitude versus longitude. Trapped protons in the SAA cause the concentration of upsets near South America. In fact, the location of the SAA protons is clearly mapped out by the upsets. The upsets that occur in the polar regions are due to galactic cosmic rays and solar particles which, at high latitudes, have access to low altitudes due to the open magnetic field at the poles.

Figure 16. Upsets on the SEASTAR flight data recorder at 705 km altitude clearly show the location of trapped protons in the South Atlantic Anomaly

2.8 MAGNETOSPHERES OF OTHER PLANETS The minimum requirement for the existence of a planetary radiation belt is that the planet’s dipole magnetic moment must be sufficiently great such that the flow of the solar wind is arrested before the particles reach the top of the atmosphere where the particles will lose their energy due to collisions. The magnetic fields of some of the other planets are similar the Earth’s, however, they vary in strength. Figure 17 shows a schematic of the relative size of the planetary magnetospheres. Table 6 gives the dipole moments in nanotesta for each of the planets.

23

Figure 17. Relative size of planetary magnetospheres TABLE 6. Dipole moment for the planets

PLANET Mercury Venus Earth Mars Jupiter Saturn Uranus Neptune Pluto

DIPOLE MOMENT (NT) 330 0 30,760 0 428,000 21,000 22,800 14,200 0?

Table 6 shows that Venus, Mars, and possibly Pluto do not have magnetospheres and, therefore, cannot support particle trapping. Mercury has a weak magnetic field so it is expected it has a trapped particle population proportionally lower than that of the Earth. The Probos probe showed that Mars has a radiation environment, however, it is due to the thin atmosphere of Mars, which allows interplanetary GCRs and solar particles to penetrate to the surface. Interaction of these particles with the atmosphere produces neutrons, which penetrate to the planetary surface and then reflect back. Saturn, Uranus, and Neptune have magnetic fields with similar strength to that of the Earth but measurements indicate that the intensities of the trapped radiation

24 environments of Saturn, and Uranus are much lower than the Earth’s and do not pose a threat to spacecraft systems. Jupiter’s enormous magnetic dipole (Table 6) can support an intense particle environment. In fact, its magnetosphere is the largest object in the solar system. Measurements have shown that the radiation environment is considerably more intense than the Earth’s and is more extensive, therefore, mission planning for spacecraft that will spend time in trapping regions of Jupiter must include careful definitions of the radiation environment. For example, the electrons at Jupiter have energies of > 100 MeV whereas those at the Earth are in the 10s of MeV. Accurate dose calculations require a model that can transport high energy electrons through shielding. Figure 15 shows that the expected dose for a Europa mission is at the megarads level for 100 mils of shielding which is higher than a one year obit around the Earth in the intense MEO regions. Single event effects are also a problem at Jupiter. In addition to the protons trapped in Jupiter’s magnetosphere, single event effects calculations must include oxygen and sulfur ions injected by volcanic activity on Io. 2.9 INTERPLANETARY PARTICLES The sun dominates interplanetary space. Its magnetized plasma, the solar wind, distorts the magnetic field of Earth (see Figure 12) and even the outer planets. In addition to the solar wind plasma, interplanetary space contains high-energy charged particles. This radiation environment consists of galactic cosmic ray particles that are present at all times and particles from solar events that occur sporadically (coronal mass ejections and flares). GCRs cause single event effects on microelectronics on interplanetary missions and on Earth orbiting satellites that spend time over the poles (see Figure 16). Solar protons are both a single event hazard and cause degradation of detectors, microelectronics, and solar cells. The heavy ions from solar particle events can increase the rate of single event effects many factors above the background caused by GCRs. 2.9.1 Galactic Cosmic Rays In the early 1900s, scientists found that instruments used for studying x-rays and radioactivity measured a background source of unidentified radiation. Victor Hess, an Austrian physicist, measured gamma rays by designing ionization chambers and flying them on balloons. With his balloon experiments, he discovered an extremely penetrating radiation that increased in density as altitude increased. From his experiments, he concluded that this radiation was from an extraterrestrial source. Later, Jacob Clay was able to show that cosmic rays were the source of the on-ground radiation and that measured by Hess higher in the atmosphere. In 1936, Hess received the Noble Prize for the discovery of galactic cosmic rays. Although we now know that these “rays’ are really particles, they are still referred to as cosmic rays. The GCRs originate outside of the solar system. Although there are plausible models of how they are produced, their origin is still a matter of debate[10]. [x]Scientists believe that they propagate through all space that is unoccupied by dense matter. They are essentially isotropic outside of regions of space that are dominated by particles and fields of the sun. Galactic radiation consists of ions of all elements of the periodic table

25 and is composed of about 83% protons, 13% alphas (4He ions), 3% electrons, and about 1% heavier nuclei. Unlike the charged particles that originate at the Sun, the GCRs do not have a characteristic energy limit. Their energies range from 10s of MeV/n to 100s of GeV/n. Because they must pass through about 7 g/cm2 of interstellar gas, the GCRs of even the heaviest ions are probably fully ionized [11]. [xi] A second source of galactic particles is the so-called “anomalous component”. It is composed of helium and heavier ions with energies greater than 50 MeV/nucleon. It is believed that the anomalous component originates in the neutral interstellar gas that diffuses into the heliosphere, becomes singly ionized by solar radiation or charge exchange, and is then connected by the solar wind to the outer heliosphere. The ions are then accelerated and propagate to Earth. The anomalous component is seen only during solar minimum and the details vary from solar minimum to solar minimum. There is growing evidence that the anomalous component is singly ionized, therefore, the ions have greater ability to penetrate the magnetosphere. Our knowledge of the abundances of galactic cosmic rays comes from spacecraft and balloon experiments that have been conducted over a forty-year period. Figure 18 from Medwaldt [12] gives the abundances of the heavy ions at an energy of 2 GeV/n as a function of particle nuclear charge z. The values are normalized to silicon = 106. Note that the relative flux intensities vary by several orders of magnitude. The relative abundances are roughly proportional to the distribution in solar system material. Significant differences are discussed in Medwaldt who also gives a table of relative abundances. xii

]

Figure 18. Relative abundances of galactic cosmic ray ions in interplanetary space. after Medwaldt

The galactic particles are always present, however, their intensities rise and fall with the solar cycle variations. The sun modulates a set of local interstellar spectra at the outer boundary of the heliosphere.[13][ The modulation can be defined by a single parameter which is a function of distance from the sun, the speed of the radial solar wind, and a radial transport particle diffusion coefficient. GCRs are at their peak level during solar minimum and at their lowest level during solar maximum and we now xiii

]

26 know that the length of the GCR modulation cycle is 22 years and not 11 years as previously thought. The difference between the extremes of the solar minimum and maximum fluence levels is approximately a factor of 2 to 10 depending on the ion energy. Figure 19 shows the slow, long-term

Figure 19. IMP-8 measurements of interplanetary ions from the C-N-O group. Note the solar particle event spikes superimposed on the lower level, slowly varying galactic cosmic rays. after Nakamura

cyclic variation of the cosmic ray (C, N, O) fluences for a 20-year period as measured by the IMP-8 spacecraft. The sharp spikes superimposed on the cosmic ray background are caused by solar events. Measurements from Pioneer and Voyager show that the composition of cosmic rays is weakly dependent on the distance from the Sun. The radial gradient from 0.3 to 40 AU is < 10% per AU. For the anomalous component, the gradient increases to 15% per AU. During solar maximum there is 0% gradient out to 30 AU. Latitude gradients have also been studied and found to be 0.5% per degree and 3-6% per degree for the anomalous component [14].[xiv] The Earth’s magnetic field provides some protection from the galactic particles by deflecting the particles as they impinge upon the magnetosphere. The penetration power of these particles is a function of the particle’s energy and ionization state. The exposure of a spacecraft primarily depends on the inclination and, secondarily, the altitude of the trajectory. Cosmic rays have free access over the polar regions where field lines are open to interplanetary space (see Figure 16). 2.9.2 Solar Particles The sun is always active but it has been observed that there is a definite periodicity to the level of activity. Thus, the solar cycle is divided into minimum and maximum phases (see Section 2.1). During the maximum phase of the solar cycle, the numbers and intensity of coronal mass ejections and solar flares increases. This causes periodic increases in the levels of interplanetary particles up to orders of magnitude over the GCR environment. These particles consist of ions of all elements but protons usually dominate the abundances. As with the GCRs, spacecraft receive some protection from solar particles by the Earth’s magnetosphere depending on their orbit. Analysis of the exposure of spacecraft orbiting the Earth as a function of the geomagnetic disturbances

27 that are often associated with solar events is especially critical. For example, CRRES data showed that solar protons reached L shell values as low as 2 [15].[ Also, unlike galactic heavy ions, which are, for the most part, fully ionized, solar heavy ions are more often singly ionized because they pass through less matter before reaching the Earth. This must be taken into account when calculating the degree of penetration of the solar particles into the magnetosphere. Solar particles diffuse as they move through the interplanetary medium. Therefore, the solar particles abundances are a function of radial distance from the Sun. There are few accurate measurements of solar particles throughout the Solar System so the models of the particle diffusion are not accurate. Table 7 gives estimates of the scaling factors that are commonly used to calculate solar particle levels at distances other than the Earth. xv]

TABLE 7. Scaling factors for solar particles levels at regions ≠ 1 AU

RADIAL DISTANCE FROM THE SUN (AU) < 1 AU > 1 AU

RADIAL SCALING FACTOR RANGE 1/R2 to 1/R3 1/R to 1/R2

Solar Protons. Figure 20 shows the particle counts for E >10 and >30 MeV for some of the larger solar proton events for solar cycles 20, 21, and 22 (measured by the GOES spacecraft). Superimposed on the solar event data is the number of sunspots. Note that although the number of proton events is greatly reduced during solar minimum, they still can and do occur. Also, the figure shows that the peak of proton event activity for each solar cycle usually does not correspond to the peak sunspot number. Solar cycles vary in severity in terms of solar proton events. For example, in cycle 21 there were no proton events as large as the August 1972 event of cycle 20 whereas there were at least six events in cycle 22 where the intensity exceeded 109 protons/cm2 for energies greater than 30 MeV. The energies of the solar event protons may reach a few hundred MeV. The duration of the events is from several hours to a few days.

28

Figure 20. Large solar proton events for solar cycles 20, 21, and 22. The number of sunspots is superimposed on the graph.

Solar Ions – Helium and Higher. Some solar particle events are heavy ion rich with energies ranging from 10s of MeV/n to 100s of GeV/n. For the 26 events observed on CRRES [13] , the peak fluxes for the helium ions with energies E > 40 MeV/n were three times higher than the galactic cosmic ray heavy ion levels. Above energies of a few hundred MeV/n to approximately 1000 MeV/n (depending on the element), the solar helium levels merge with those of the galactic cosmic ray background. Early attempts to characterize the solar heavy ions at higher energies were restricted by a limited dataset. Later more space data became available. Dietrich et al[16] used data from the University of Chicago’s Cosmic Ray Telescope on the IMP-8 and GOES satellites to study the heavy ion events. They analyzed high energy spectra for C, O, and Fe using direct measurements and determined fluences in one or two energy bins for N, Ne, Mg, Si, S, Ar, and Ca. Also, He fluences were studied using carbon indices. This dataset provides the most comprehensive picture of high-energy solar heavy ions to date. [xiii]

[xvi]

3. 0 Summary There are many unknowns in space environments and the interaction mechanisms. The low level of funding in the applied science discipline of spacecraft environments and effects has resulted in model development and validation lagging behind rapid technology changes. Access to space for validating effects models and ground test

29 protocols is also critically low. Ground tests cannot duplicate the space environment, particularly when environments have synergistic effects and effects are complicated by enhanced low dose rates in space. Often these unknowns require that large design margins be applied to performance predications, resulting in overheads that reduce capability or that can preclude use of newer technologies in spacecraft systems.

4. 0 References

1. A. LaBel, A. H. Johnston, J. L. Barth, R. A. Reed, and C. E. Barnes, “Emerging Radiation Hardness Assurance (RHA) issues: A NASA approach for space flight programs,” IEEE Trans. on Nucl. Science, Vol. 45, No. 2, December 1998 2. D. Knison, “Radiation effects in the New Millennium - Old Realities and New Issues, Section V: Achieving Reliable, Affordable Systems,” 1998 IEEE Nuclear and Space Radiation Effects Conference Short Course, July 20, 1998, Newport Beach, CA. 3. J. L. Barth, “Applying Computer Simulation Tools to Radiation Effects Problems, Section II: Modeling Space Radiation Environments,” 1997 IEEE Nuclear and Space Radiation Effects Conference Short Course, July 21, 1997, Snowmass Village, CO. 4. A. Holmes-Siedle and L. Adams, Handbook of Radiation Effects Second Edition, Oxford Press, 2002. 5. J. M. Nash, “Cosmic Storms Coming,”, TIME, pp. 54-55, September 9, 1996. 6. 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,” IEEE Trans. on Nucl. Science, Vol. 43, No. 2, pp. 369-382, April 1996. 7. J. Feynman, T. P. Armstrong. L. Dao-Gibner, and S. Silverman, “New Interplanetary Proton Fluence Model,” J. Spacecraft, Vol. 27, No. 24, pp 403-410, July-August 1990. 8. D. V. Reames, “Solar Energetic Particles: A Paradigm Shift,” Revs. Geophys. (Suppl.), 33, 585, 1995. 9. S. L. Koontz, M. Pedley, R. R. Mikatarian, J. Golden, P. Boeder, J. Kern, H. Barsamian, J. I. Minow, R. L. Altstatt, M. J. Lorenz, B. Mayeaux, J. Alred, C. Soares, E. Christiansen, T. Schneider, and D. Edwards,” Materials Interactions with the Space Environment: International Space Station - May 2000 to May 2002,” these proceedings. 10. J. W. Cronin, T. K. Gaisser, and S. P. Swordy, “Cosmic Rays at the Energy Frontier,” Scientific American, January 1997. 11. J. H. Adams, Jr., R. Silberberg, and C. H. Tsao, “Cosmic Ray Effects of Microelectronics, Part I: The Near-Earth Particle Environment,” NRL Memorandum Report 4506, August 25, 1981. 12. R. A. Medwadlt, “Elemental Composition and Energy Spectra of Galactic Cosmic Rays,” Proc. from Conference on Interplanetary Particle Environment, JPL Publication 88-28, pp. 121-132, JPL, Pasadena, CA, April 15, 1988. 13. D. L. Chenette, J. Chen, E. Clayton, T. G. Guzik, J. P. Wefel, M. Garcia-Muñoz, C. Lapote, K. R. Pyle, K. P. Ray, E. G. Mullen, and D. A. Hardy, “The CRRES/SPACERAD Heavy Ion Model of the Environment (CHIME) for Cosmic Ray and Solar Particle Effects on Electronic and Biological Systems in Space,” IEEE Trans. on Nucl. Science, Vol. 41, No. 6, pp. 2332-2339 December 1994. 14. R. B. McKibben, “Gradients of Galactic Cosmic Rays and Anomalous Components,” Proc. from Conference on Interplanetary Particle Environment, JPL Publication 88-28, pp. 135-148, April 15, 1988. 15. M. S. Gussenhoven, E. G. Mullen, and D. H. Brautigam, “Improved Understanding of the Earth’s Radiation Belts from the CRRES Satellite,” IEEE Trans. on Nucl. Science, Vol. 43, No. 2, April 1996. 16. W. F. Dietrich, A. J. Tylka, and P. R. Boberg, “Probability Distributions of High-Energy Solar-HeavyIon Fluxes from IMP-8: 1973-1996,” to be presented at IEEE/NSREC Conference, 1997.

This page intentionally left blank

MATERIALS INTERACTIONS WITH THE SPACE ENVIRONMENT: INTERNATIONAL SPACE STATION - MAY 2000 TO MAY 2002 STEVEN L. KOONTZ(1)*, MICHAEL PEDLEY(1), RONALD R. MIKATARIAN(2), JOHN GOLDEN(2), PAUL BOEDER(2) , JOHN KERN(3), HAGOP BARSAMIAN(2), JOSEPH I. MINOW(6), RICHARD L. ALTSTATT(6), MARY J. LORENZ(2), BRIAN MAYEAUX(1), JOHN ALRED(2), CARLOS SOARES(2), ERIC CHRISTIANSEN(4), TODD SCHNEIDER(5), DAVE EDWARDS(5)

Abstract The set of materials interactions with the space flight environment that have produced the largest impacts on the verification and acceptance of flight hardware and on flight operations of the International Space Station (ISS) Program during the May 2000 to May 2002 time frame are described in this paper. In-flight data, flight crew observations, and the results of ground-based test and analysis directly supporting programmatic and operational decision-making are reported.

1.0 Introduction Orbital inclination (51.6o) and altitude (nominally between 350 km and 460 km) determine the set of natural environment factors affecting the functional life of materials and systems on ISS. ISS operates in the F2 region of Earth’s ionosphere in well-defined fluxes of atomic oxygen, other ionospheric plasma species, solar UV, VUV, and x-ray radiation as well as galactic cosmic rays, trapped radiation, and solar cosmic rays [1,2]. The micrometeoroid and orbital debris environment is an important determinant of spacecraft design and operations in any orbital inclination. The magnitude of several environmental factors varies dramatically with latitude and longitude as ISS orbits the Earth [1,2]. The high latitude orbital environment also exposes ISS to higher fluences of trapped energetic electrons, auroral electrons, solar cosmic rays, and galactic cosmic rays [3-6] than would be the case in lower inclination orbits, largely as a result of the overall shape and magnitude of the geomagnetic field [1-6]. As a result, exposure of ISS to many environmental factors can vary dramatically along a particular orbital ground track, and from one ground track to the next, during any 24-hour period. The induced environment results from ISS interactions with the natural environment as well as environmental factors produced by ISS itself and visiting vehicles. Examples include ram-wake effects, hypergolic thruster plume impingement, materials outgassing, venting and dumping of fluids, and specific photovoltaic (PV) power system interactions with the ionospheric plasma [7-16]. An induced ionizing radiation

31

32 environment is produced by cosmic ray interaction with the relatively thick ISS structure and shielding materials [6]. Vehicle size (L) and velocity (v), combined with the magnitude and direction of the geomagnetic field (B) produce operationally significant magnetic induction voltages (vxB.L) in ISS conducting structure during flight through high latitudes (>+ 45o) during each orbit [15,17,19]. In addition, ISS is a large vehicle and produces a deep wake structure from which both ionospheric plasma and neutrals species are largely excluded [7, 19, 21]. Finally, ISS must fly in a very limited number of approved flight attitudes [22], so that exposure of a particular material or system to environmental factors depends upon: 1) location on ISS, 2) ISS flight configuration, 3) ISS flight attitude, 4) variation of solar exposure (ȕ angle) with time, and 5) levels of solar and geomagnetic activity (space weather). The specific spacecraft-environment interactions that have had the greatest impact on ISS Program activities during the first two years of flight are: 1) spacecraft charging, 2) micrometeoroids and orbital debris impacts, 3) ionizing radiation, 4) hypergolic engine plume impingement, 5) venting/dumping of liquids, and 6) external contamination, atomic oxygen, solar ultraviolet effects.

2.0 Spacecraft Charging Phenomena 2.1.HIGH-VOLTAGE PHOTOVOLTAIC ARRAY DRIVEN CHARGING The relatively high plasma density, low plasma temperature, and high electrical conductivity characteristic of the F2 region ionospheric plasma preclude many of the spacecraft charging processes that are observed in lower density plasma environments [2,7]. Surprisingly, the most important identified spacecraft charging process for ISS requires a high-density, low-temperature plasma environment. An electrical interaction between the F2 plasma and the 160-V US PV arrays can produce an electrical potential difference between the conducting structure of ISS and the ambient plasma (i.e. a floating potential or FP) much greater than that usually observed for spacecraft in lowEarth orbit (LEO), most of which have 28-V PV array power systems [7-12]. Sky Lab, which employed 90-V PV arrays, is an important exception to be discussed below. As is shown below, ISS conducting structure becomes negatively charged with respect to the ambient plasma because the PV arrays and electrical power system utilize a negative-polarity grounding scheme, and the common ground point is ISS conducting structure. The severity of possible charging hazards is largely determined by materials interactions with the F2 plasma environment [7-12]. Spacecraft charging interactions lead to the application of electrostatic fields across the dielectrics that separate conducting structure from the ambient F2 plasma. The magnitude of the field gradient can be large enough to cause dielectric breakdown and arcing [7-12]. Degradation of some thermal control coatings, electrical system noise, and shock hazards to extra-vehicular activity (EVA) crew may result [7-12] if the FP is sufficiently negative. The following simple calculation is aimed at explaining the PV array driven charging process, while highlighting the important role of materialsenvironment interactions in both the charging process and the subsequent analysis of possible hazards. The physical basis of PV array driven spacecraft charging lies in the fact that the ions and electrons in the F2 plasma have nearly the same gas kinetic temperature and,

33 therefore, nearly the same kinetic energy. Because the electrons are much less massive than the ions (mi >> me), the mean gas-kinetic speed for electrons, ve= ¥(8kTe/2ʌme), is much larger than the mean gas kinetic speed for the ions, vi =¥(8kTi/2ʌmi). Therefore, the flux of electrons (electron current, I-) to any surface is much greater than the flux of ions (ion current, I+) until a steady state negative FP is established such that I+ + I- = 0. Each of the 400 photovoltaic cells in one string of the US PV array produces about 0.4 V in sunlight, yielding a total linear ǻV of 160 V from one end of the string to the other. There are 82 strings per PV array wing. In the real PV array, the string is mounted on one side of an insulating flat plate of length L (same as the string length). The plate is flying through the ionosphere at orbital velocity with the PV array string facing forward (ram orientation). The FP can be calculated point by point along the string given ǻV, orbital velocity (vISS), the electron and ion densities (Ne and Ni) and the corresponding gas kinetic temperatures, Te and Ti, in the F2 plasma. Geomagnetic field effects on current collection from the F2 plasma are small enough to neglect for this analysis. Orbit limited current collection, electrostatic focusing effects [1,2], and detailed PV array lay-out are also neglected, for the sake of simplicity, even though the subject effects are large and lead to smaller measured values of ISS FP than were predicted by simple early treatments [9]. The thermal velocity of the plasma ions, vi, is much less than orbital velocity of the spacecraft, vISS, so that only ram ion collection is considered. In contrast, the thermal speed of the plasma electrons is much greater than orbital velocity so that electron collection is by gas-kinetic diffusion to the Debye sheath [1,2] and then to the collecting surface. The faster electrons cannot catch up with the spacecraft from behind because separation from the slower moving ions in the wake region creates an opposing electric field (ambipolar diffusion [1,2]. As a result, simply turning the PV array strings to wake and exposing only the insulating plate to ram completely suppresses PV array driven charging, a prediction confirmed by in-flight ISS floating potential and plasma contactor emission current measurements made during 2001 [15,16]. Finally, the magnitude of the charging depends on the PV array voltage. When ǻV is zero, at night or when the PV strings are shunted, there is no charging. At steady state, plasma ion current to the string must equal plasma electron current to the string. Electron thermal current = I- = I+ = Ion ram current The positively biased end (area Ae, length Le) of the string collects electrons and the negatively biased end (area Ai, length Li) collects ions. Ae + Ai = A, and Le + Li = L, where A is the total exposed conducting area on the PV cell string of length L. I- = 0.25veNeqAe = vISSNiqAi = I+. The ionosphere is a neutral plasma so, Ne = Ni. The mean gas-kinetic speed of the electrons, ve, is multiplied by 0.25 to obtain the correct expression for thermal particle flux to a wall, and q is the value of the elementary charge leading to, 0.25veAe = viAi or Ae/Ai = vISS/0.25ve

34 Assuming a typical daytime F2 region plasma temperature of 0.1 eV (1,160o K) we have vi = 1.3 km/sec, vISS = 7.69km/sec, and ve =163 km/sec so that, Ae/Ai = vISS/0.25ve = (7.69)/(0.25x163) = 0.19. Since PV string voltage is a linear function of distance from the positive end, and ǻV = 160 V, FP can now be calculated as a function of distance from the positive end of the string, where FP(0) corresponds to L=0. FP(0) = ǻV(Ae)/(Ai+Ae) = ǻV(Le)/(Li+Le) = 160 x 0.1597 = +26 volts FP(L) = FP(0) – ǻV = -134 V If the negative end of the string is grounded to a spherical conducting structure that is 10 meters in radius (a reasonable size compared to ISS pressurized elements), the free space capacitance (Cfs = 4ʌİİor = 1112 pF; İ = dielectric constant, İo= free space permittivity) of the structure is charged to –134 volts giving a stored energy of only E = 0.5CV2 = 10 micro Joules. The sphere with a thin dielectric surface coating changes the character of the charging hazard dramatically. On the ram facing side of the sphere, the FP of the external surface of the dielectric film will approach 0 V as a result of positive charge collection from the ionosphere, and –134 V is applied across the dielectric. Now the sphere is best described as a parallel plate capacitor (the conducting structure is one plate and the conducting ionosphere is the other) able to store energy E = 0.5 CV2 = 0.5 İİo(Aram/d)V2 = (0.5)(8.85 x 10-12)İ(2ʌr2/d) V2, where Aram is the area of the hemisphere able to collect positive charge from the ionosphere. If d is 1 micron (1.3 microns is the thickness of the anodic coating on the US Lab and Node 1 meteoroid and debris shields) and İ = 5 for aluminum oxide, the stored energy becomes E = 250 Joules. Now, dielectric breakdown of the thin surface coating can discharge the parallel plate capacitor, releasing enough energy to damage the dielectric coating itself and producing enough voltage and current to present a possibly lethal hazard to any EVA crew in the discharge circuit. The high-density, low-resistance dielectric-breakdown arc plasma provides the conductive path connecting the negatively charged conducting structure to the positively charged dielectric film surface [8-10]. Note that the stored energy is inversely proportional to the dielectric film thickness. Simply increasing the film thickness from 1.0 micron to 100 microns reduces the stored energy from 249.5 Joules to 2.49 Joules while greatly reducing the risk of dielectric breakdown arcing. The thick (>120 microns) dielectric coatings on Sky Lab minimized any charging hazards that might have been generated by the 90-V PV array on that spacecraft. Similarly, the Russian elements of ISS contribute little to the charging hazard because surface dielectric coatings are thick. Stored energy is also directly proportional to V2, and reducing the FP at the negative end of the PV array to –40V reduces the stored energy to 0.9 micro Joules for the uncoated conducting sphere, and

35 22 Joules for the dielectric coated sphere. As discussed below, flight measurement and analysis of US Lab and Node 1 FP, with all FP controls disabled and PV array driven charging enabled, have not exceeded –28 volts during 2001 [16]. Plasma chamber testing (7-12) has shown that the dielectric breakdown voltage for the 1.3-micron thick anodic film on the US Lab and Node 1 meteoroid and debris shields is greater than 60 volts. Therefore, the plasma contactor system has not been in continuous operation since May 2001. The ISS FP not-to-exceed-limit for EVA safety is –40 V, however and two PCUs are operated routinely during EVA. A negative FP of –134 V is remarkably close to the predictions made before the US PV arrays were flown for the first time on ISS [7-10]. Using early charging models, a worst-case FP of –140 volts was predicted. The measured FP from PV array driven charging on ISS have been less negative than –28 V in all cases observed to date. The simple charging calculation presented above as well as the more elaborate pre-flight theoretical models consider only current collection by the PV array string. Ion collection by exposed conducting structure attached to the negative end of the PV array string can offset the effects of electron collection by the string, driving FP(0) toward +160V and FP(L) toward 0 volts, but only if the number of milliamps of electron current collected by the PV array is small. The number of square meters of ramoriented ion-collecting surface needed to hold the PV array FP (0) near +150 V and FP(L) near -10 volts is shown as a function of total electron current collected by the PV arrays in Table 1. TABLE 1. Area of Exposed Conducting Materials Compensating PV Array Electron Collection * Ionospheric Electron Current Collected by 160 V PV Arrays (milliamps) 10 30 60 100 * FP (0) =+150 V and FP (L) = -10 V; Ni = Ne =106/cc

Area of Ram Oriented Ion Collection Surface (square meters) 8 24 48 80

When the electron current collected by the PV arrays on an LEO spacecraft is less than about 60 milliamps, exposed ion conducting area connected to the negative ground plane can offer practical FP control. As the collected electron current grows beyond 100 milliamps, the ion collecting area requirements become unrealistic. The plasma contactor system was selected for ISS FP control precisely because the magnitude of the electron current collected from the ionosphere by the 160-V US PV arrays was estimated to be far too large to allow FP control by passive ion collection surfaces [9, 14]. The Russian segment of ISS provides significant ram-oriented conducting surface area (estimated to be greater than 30 square meters) as a result of Russian Program electrical conductivity/grounding requirements for thermal blanket materials. The plasma contactor system on ISS controls the FP by providing a low impedance return path to the ionosphere for electrons collected by the PV arrays or by other collection mechanisms [17]. The ISS telemetry stream provides measurements of electron emission current from the ISS ground plane to the ionosphere whenever the plasma contactor system is operating. Much of the plasma contactor emission current observed over the past 2 years is attributable to low-voltage non-PV-array-driven

36 charging processes [17]. However, direct measurements of PV array driven electron collection can be made by recording the change in emission current when ISS enters sunlight (eclipse exit) with sun-pointing PV arrays or by shunting and un-shunting the sun-pointing PV arrays while in sunlight. Figure 1 shows measured eclipse-exit plasma contactor emission currents since January 2001. The eclipse exit emission currents show considerable variation both during a given 24-hour day and over the last year. The well-known dynamic structure of the F2 region of the ionosphere [1,2] can account much of the variability. Large variations of Ne and Te with time of day, altitude, ISS latitude and longitude, geomagnetic field, solar activity, and season explain much of the observed variability in the eclipse exit emission currents [1,2,18]. Clearly, the magnitude of PV-array-driven charging will vary in a similar way with variation in natural environment. 0.09

0.08

0.07

0.06

Current (A)

One Array Shunted 0.05

0.04

0.03

0.02

0.01

0 1/1/01 1/29/01 2/26/01 3/26/01 4/23/01 5/21/01 6/18/01 7/16/01 8/13/01 9/10/01 10/8/01 11/5/01 12/3/01 12/31/0 1/28/02 2/25/02 1 Date

Figure 1. ISS plasma contactor emission current increase at eclipse exit; Jan 2001 to Feb 2002, US 160 V PV arrays sun tracking and un-shunted

With the plasma contactors operating, most of the PV array wing area is positively biased (+FP) so that electron collection is maximized and any ion collection by conducting structure is already accounted for in the observed emission currents. If the plasma contactors are turned off, the FP at the negative end of the PV array (and ISS conducting structure) moves toward more negative values. As a result, less of the PV array wing area is positively biased leading to reduced electron collection, while ion collection remains constant or increases slightly. At some negative value of the conducting structure FP, PV array electron collection will equal conducting structure ion collection stopping further movement of FP toward more negative values.

37 In the simple flat-plate charging calculation shown above, the reduction in PV array electron collection is a simple linear function of FP at the negative end of the array. In fact, as FP becomes more negative, the decrease in electron collection by the PV arrays is nonlinear as a result of: 1) dielectric surface charging of PV array cell materials, 2) detailed electrostatic focusing effects in the Debye sheath near the gap between PV array cells, and 3) PV array cell structure and lay-out. Small negative changes in FP cause relatively large reductions in electron collection while ion collection remains nearly constant [9b, 9c, 18]. Steady state FP values with the plasma contactor system off are expected to be closer to –30 V, not –134 volts, as has been observed to date [15,16,18]. As development of more accurate and detailed models of the PV array driven charging process continues, it becomes clear that a materials interaction with the ionospheric environment, specifically surface charging of dielectric materials in the photovoltaic cell structure, limits electron collection by the 160 V US PV arrays on ISS and places natural limits on the FP values that can be achieved [18]. The results of in-flight floating potential probe (FPP) [15,16] measurements of ISS FP characterizing both the PV array driven charging process and the contribution of the vxB.L (v = spacecraft velocity, B = geomagnetic field, L = length of conducting structure) magnetic induction voltages, with the plasma contactor system off, are shown in Figure 2 and Tables 2 and 3. Table 2 compares the worst-case pre-flight predictions of PV array driven charging with the worst-case measurement made to data. Measured electron collection by the two 160-V US PV arrays active during the April 2001 time frame is so low that exposed conducting structure can contribute to limiting the negative FP to the small value observed, as suggested above. FPP measurements of ISS FP were made during several days in 2001, including intervals when the Space Shuttle was docked to ISS. On January 31, FPP data measurements of ISS FP were made with active side (the side with PV cell strings) of the active surface of the PV arrays in shallow wake flight attitude verifying that wake orientation of the arrays prevents PV array driven charging. With the plasma contactor system off and PV arrays sun tracking, FPP data was collected on April 10-12, April 15, and on April 21 (before and after Space Shuttle docking). A total of 46 FP measurements characterizing PV array driven charging were made in 2001, encompassing a wide range of ionospheric conditions. Langmuir probe measurements of electron temperature, Te, at eclipse exit ranged from 0.08 to 0.23 eV while electron density, Ne, ranged from 109 to 1012/m3. To date, the observed range of PV array-driven charging FP values range from –4 to –24 V. It should be noted that the FPP could not provide Te or Ne if Fp exceeded –10V negative as a result of the limited sweep range of the Langmuir probe voltage. The April 11 data is fairly typical, despite the geomagnetic storm starting about 13:30 universal time (UT). Figure 2 shows the ISS FP at the FPP measurement point as a function of universal time on April 11, 2001. In Table 3 the total FP for the April 11, 2001 eclipse-exit charging peaks, shown in Figure 2, are broken down into the

38 magnetic induction and PV array driven components for the locations on ISS defined in Figure 3. Magnetic induction voltage is a significant fraction of the total FP in all cases, and must be considered in any ISS charging assessment. As shown in Figure 2, the agreement between calculated magnetic induction voltage and measurement is excellent in all cases. Figure 3 shows a calculated magnetic induction voltage map of ISS when passing south of Australia on April 11, 2001. Flight south of Australia generates the more magnetic induction voltage on ISS than any other ISS flight path. TABLE 2. PV Array Driven Charging - Pre-Flight Estimates vs. Flight Data Charging Hazard Related Quantity Maximum Negative FP 160 V PV Array Electron Collection Exposed Conducting Surfaces on ISS Duration of Max. Neg. FP

Pre ISS Flight 4A: Worst-Case Estimate -140 200 to 500 milliamps 0 m2 20 to 30 min. of day pass

Post Flight 4A: Worst-Case at US Lab Module -26 V 10 to 80 milliamps 15 to 40m2: PV array mast wires & ISS structure 4mm)

0.999983

1 in 58,824

0.999941

1 in 16,807

3.5 X

For 2700 EVA-hours (10years of EVAs)

meteoroids/ORDEM96 PNP

Odds

meteoroids/ORDEM2000 PNP

Odds

Risk Change

Any size leak

0.9747

1 in 39

0.8796

1 in 8

4.8 X

Critical leak (>4mm)

0.9924

1 in 131

0.9736

1 in 38

3.5 X

43

Figure 5. 3-5 mm MOD strike crater on the Service Module nadir view port

Figure 6. MPLM MOD shield penetration observed after ISS flight.

44 4.0 Ionizing Radiation I: EEE Parts and Avionics The ISS electrical power system and avionics suite (including commercial-off-the-shelf or COTS components) are performing well ahead of expectation with respect to the single event effects (SEE) and total ionizing dose (TID) effects produced by the ionizing radiation environment in LEO. No correlations between any electronic anomaly on ISS and: 1) exposure to solar energetic particle events, 2) flight through the South Atlantic Anomaly (SAA), or 3) flight at high latitude have been identified during the first two years of flight. Similarly, no galactic cosmic ray (GCR) effects, as revealed by a geomagnetic latitude dependence of electronic failures, have been identified, however, events that may be produced by high energy/rigidity GCR particles, which would show no latitude/longitude dependence, are the subject of an ongoing investigation. 4.1 THE RADIATION DESIGN ENVIRONMENT The ISS natural radiation design environment is specified in SSP 30512, Space Station Ionizing Radiation Design Environment [36]. For ISS design, we have specified conservative, single design point, ionizing radiation design environments to address Total Ionizing Dose (TID). The ISS TID radiation environment is specified for 500 km at solar maximum and includes both trapped protons and trapped electrons. Other minor contributions to TID, such as x-rays, as well as uncertainties in the trapped radiation models, are addressed through the application of a recommended design margin of 2x applied to the 500 km design environment. The ISS radiation design environment represents a conservative, low cost solution for ISS hardware design and verification. The selection of 500 km as a design point altitude is in itself a worst-case assumption because ISS will normally fly below 500 km. Both TID and SEE rates increase dramatically (about 3X) between 300 and 500 km largely as a result of the altitude structure of the SAA. Two different natural environments for design have been specified to address SEE on ISS [36]. The nominal SEE environment is based on AP8 solar minimum model for trapped protons and a solar minimum model for cosmic rays. The extreme SEE environment includes solar protons and heavy ions emitted during an extreme or worst case SEP event [36]. The nominal SEE environment is specified for a 500 km altitude at solar minimum so as to define a worst-case environment for both trapped protons and GCR [36]. The extreme SEE environment is also specified for 500 km. Major SEP events are relatively rare and are associated with powerful solar x-ray flares and/or coronal mass ejections [37, 38]. The SSP-30512 extreme SEE environment is based on the October 1989 SEP event, which is generally recognized as a 99th percentile worstcase extreme SEE environment with respect to both energetic proton and energetic heavy ion fluxes [39]. SSP-30512 contains no statements about the radiation-shielding environment produced by ISS structure or avionics cases or enclosures. A simple worst-case shielding environment has generally been assumed for predicting worst-case TID and SEE effects. For purposes of design and analysis, SEE and TID susceptible equipment is assumed to be at the center of a 0.5 cm (1.27 cm) thick aluminum sphere if located

45 outside (inside) the pressurized elements corresponding to shielding surface densities 1.35 g/cm2 (3.43 g/cm2), values substantially lower than the actual shielding mass distribution function inside the pressurized elements of ISS, which ranges from 10g/cm2 to 100 g/cm2 as is shown in the following paragraphs. ISS is exposed to the TID and nominal SEE environments on a continual basis and must meet all performance requirements during this exposure. ISS is protected from SEP events by the Earth’s geomagnetic field over most of its orbital flight path, though some fraction of the SEP heavy ions, which are not fully ionized, may penetrate to latitudes where cosmic rays of comparable rigidity are excluded [39]. The most important ISS exposure to extreme SEE particles fluxes happens when the orbital phasing is such that ISS passes over Canada or south of Australia during the SEP event. Electronic parts are susceptible to SEE induced destructive failure only if powered. The ISS program has recently established an extreme SEE caution and warning procedure to enable power down of any nonessential equipment that has a high risk of destructive failure in the extreme SEE environment [40]. 4.2 ON-ORBIT OBSERVATIONS ISS is performing well within expectations with respect to TID degradation and SEE impacts on EEE parts and avionics performance. Until recently, ISS has been flying at altitudes between 350 and 400 km during solar maximum, well below the 500 km specified for the worst-case radiation design environment in SSP 30512. TID accumulated to date is well below the performance degradation threshold for EEE parts. Ionizing radiation dose measurements, made within the habitable volume with thermoluminescent dosimeters and crew personal dosimeters, range from 5 to 12 µ Gy (0.5 to 1.2 milli rads) per hour, depending on location in the habitable volume, corresponding to an annual dose range of 44 to 105 milli Gy (4.4 to 10.5 rads) [41]. The variation in TID with location in the habitable volume is largely a result of variations in effective shielding mass with location [42-44]. No destructive SEE events of any kind have been observed during the first two years of flight. Only one ISS vehicle equipment item fault that may be uniquely attributed to SEE with a reasonable level of confidence has been observed. An S-Band Antenna System anomaly occurred at 18:16 GMT on 26 December 2001 during a moderate SEP event. The anomaly is best explained as the result of a single event upset (SEU) in the command register of the video signal processor. ISS was near latitude 35o south, longitude 157o east, altitude 390 km at 18:16 GMT on 26 December 20001. ISS was well away from the SAA, and well shielded by Earth’s geomagnetic field from both the SEP event and GCRs with energies below 10 GeV/nucleon, when the anomaly occurred. Using SEE test data for the subject device, the mean-time-between-failure (MTBF) analysis, and 2 gm/cm2 Al shielding as input parameters, calculations of the as-flown average radiation environment using CREAM-96 [45], predict an average upset rate of 0.0001 SEU/day. The S-Band Antenna System has been on orbit for 500 days as of June 26, 2002 for an observed failure rate of 0.002 SEU/day, in reasonable agreement with the CREME-96 calculation given the generic uncertainties in the analysis.

46 A study of those IBM Thinkpad 760 XD (ISS Personal Computer System or PCS) anomalies requiring reboot or power cycling that may be attributed to radiation effects, after excluding all other reasonable causes, shows no correlation with exposure to SEP events, flight through the South Atlantic Anomaly (SAA), or flight through high latitude regions, as shown in Figure 7. The GMT time of PCS anomalies are recorded in the PCS boot file and were reported by ISS crew. No destructive PCS events or anomalies have been observed to date. The observed PCS reboot/power cycling rate (reboots that may be attributed to radiation in the absence of any other identified causes) is in reasonable agreement with rates predicted using the SSP 30512 nominal SEE design environment combined with 0.129 GeV proton beam screening tests [46] and a worst-case mean time between failure (MTBF) analysis with less than 1 g/cm2 shielding, as shown in Table 6. However, the latitude, longitude, and SEP exposure anomaly dependencies predicted are not observed. The IBM 760 XD Thinkpads are commercial-off-the-shelf (COTS) items. One IBM 760 XD from the procurement lot was tested for SEE susceptibility using a proton beam test [46] on the assembled item. Proton testing showed lockups (requiring reboot or power cycling) only for the CPU, video board, and CD-ROM, in order of decreasing susceptibility. 90 No SPE in Progress >100 MeV Proton SPE in Progress

60

latitude

30

0

-30

-60

-90 -180

-150

W

-120

-90

-60

-30

0

30

60

90

120

150

180

E

longitude

Figure 7. PCS anomalies as a function of ISS latitude and longitude

TABLE 6. Predicted Radiation Induced Anomaly* Rate vs. In-Flight Anomaly Rate for Three IBM Thinkpad Laptop 760 XD Computers (PCS) on ISS Laptop

Predicted Reboots/Day (radiation)

Observed Reboots/Day (radiation)

Service Module PCS 0.04 0.02 Lab Robotics Work 0.04 0.01 Station PCS Lab PCS 0.04 0.04 All 0.13 0.08 * Anomalies requiring reboot or power cycling for recovery and not attributable to causes other than SEE causes

47 Figure 7 shows the timing of the laptop anomalies compared to the timing of SEP events observed by the NOAA GOES satellite constellation. Only 4 of the 22 ISS laptop anomalies were observed during periods of enhanced solar particle flux at GOES. However, ISS was not positioned to receive solar proton flux during three of these events (April 10, August 18 and September 26). For the fourth event, which occurred on August 16, the location of ISS at the time of the anomaly cannot be determined (no anomaly time tag). So, no laptop anomalies requiring reboot or power cycling can be uniquely associated with SEP event exposure while ground based proton testing combined with the SSP 30512 extreme SEE design environment predicts several reboots or power cycles per event day with SEP event exposure.

Daily Flux of >100 MeV Solar Protons (p/cm^2-sr-day at Geosynchronous Orbit)

1.E+7 Observed PCS Anomalies

>100 MeV Daily Solar Proton Flux

1.E+6

1.E+5

1.E+4

1.E+3 3/1

4/1

5/2

6/2

7/3

8/3

9/3

10/4

11/4

12/5

Date

Figure 8. IBM Thinkpad (PCS) Anomaly Data vs. Solar Particle Events for 2001 Figure 1 data sources: >100 MeV Daily Solar Proton Flux - Space Physics Interactive Resources (SPIDR) webpage (URL= http://spidr.ngdc.noaa.gov/ ). Observed PCS anomalies – ISS Mission Action Requests (CHITS) database

A review of the US Lab and Node 1 flight equipment nominal SEE susceptibility data indicates that the observed frequency of equipment functional interrupts (i.e. those requiring outside intervention and possibly caused by radiation), is at least an order of magnitude less than the single event rates calculated for the SSP 30512 design environment. As of January 2002, ISS has been exposed to three major SEP events as shown in Figures 9-11. ISS was exposed to SEP event particles only while at the high latitudes during the exposure periods shown in Figures 9-11.

48

Figure 9-11 Source Data: 1) Geosynchronous particle flux - Space Physics Interactive Resources (SPIDR) webpage (URL= http://spidr.ngdc.noaa.gov/ ). 2) ISS SPE exposure calculated with SPERT ( M.J. Golightly, M. J., Weyland M. D., Lin, T.; “Solar Particle Event-Real-Time (SPERT): A Real-Time Solar Particle Event Exposure Analysis System.” ESA Workshop on Space Weather, ESTEC, Noordwijk, The Netherlands, 11-13 Nov 1998.)

49 Each of these events produced a total fluence of >100 MeV protons greater than 1 x 108 protons/cm2 at geosynchronous orbit. For comparison the ISS extreme SEE design environment proton fluence at >100 MeV is 1 x 109 protons/cm2. No ISS equipment anomalies have been tied to any of these major solar particle events. There were not very many SEE susceptible items on ISS during the 2000 events, but a fully outfitted US Lab Module was present during the November 2001 event. Figures 9-11 show >100 MeV particle flux levels during these events, as well as the times ISS was exposed to the enhanced particle environment, when the orbital flight path carries ISS through high latitude (greater than + 45o) regions where geomagnetic shielding is relatively poor (e.g. Canada and Australia/Tasmania). The performance of the static and dynamic random access memories (SRAM and DRAM) in the ISS data system multiplexer-demultiplexer (MDM) units is of special interest in light of the critical nature of MDM function in ISS command and control systems. Several memory chips in the ISS MDM contain bit flip counters for error rate measurements. During 2000 and 2001 only multiple bit error data were downloaded from ISS. A software upgrade is being developed to enable downloading the single bit flip data in the near future. The predicted performance of the MDM SRAM and DRAM devices in the SSP-30512 design environment is compared with the observed performance (multiple bit error data only) in Table 7. TABLE 7. MDM RAM Multiple Bit Error Rates

1Mx4 DRAM

Predicted Nominal SEE multiple bit error rate 4.4 x 10-8/day

Observed Nominal SEE multiple bit error rate 0/730 days

4Mx4 DRAM

7.5 x 10-7/day

0/730 days

32Kx8 SRAM

1.1 x 10-6/day

0730 days

RAM Device

Predicted Extreme SEE multiple bit error rate 0.00034 per SEP event 0.013 per SEP event 0.0021 per SEP event

Observed Extreme SEE multiple bit error rate 0/2 SEP events 0/2 SEP events 0/2 SEP events

Application of a worst-case design environment, combined with rigorous EEE parts screening and MTBF analysis at the parts, assembly, and system levels, has produced an ISS electrical power and avionics system that is largely immune to the effects of ionizing radiation. The PCS and S-Band antenna systems are performing well ahead of expectations based on pre-flight test and analysis, verifying the validity of the test and analysis methods used as well as the SSP-30512 design environment as well as the test and acceptance methods described in SSP-30513. Is it possible that observed anomalies are produced by galactic cosmic rays (GCRs) of such high energy (Energy >10 GeV/nucleon, geomagnetic rigidity > 10 GV) that no latitude-longitude dependence of the anomaly rate is expected? Primary cosmic rays having energies greater than 10 GeV/nucleon also have very low flux as well as linear energy transfer (LET) values well below typical SEE thresholds (1 to 10 MeV-cm2/mg) because LET is inversely proportional to the square of particle velocity in the relativistic Bethe-Bloch equation [47, 48]. As a result, no detectable SEE effect is expected from the E > 10 GeV/nucleon GCRs themselves. However, energetic GCRs

50 have high cross sections for inelastic interactions with the Al nuclei in the relatively thick structure and shielding of ISS. Previous studies on GCR reaction and transport in spacecraft have shown that even those GCRs less energetic than 10 GeV/nucleon can produce showers of hadronic secondary particles (pions, protons, neutrons and alpha particles, nuclear fragmentation/spallation products) in the spacecraft shielding mass. The subject hadronic showers ultimately produce a significant flux of 100 to 200 MeV secondary particles inside the functional volume of the spacecraft [48-50] leading to SEE and TID effects [6, 48-51]. It has been previously reported that the CREME 96 SEE prediction software seriously underestimates SEU rates for shielding thickness on the order of 50 g/cm2 precisely because hadronic showers caused by high energy GCRs and other nuclear reactions in spacecraft shielding are neglected [51]. The energy spectrum of the primary galactic cosmic rays can be approximated by the following well-known power law equation [52]. IN(E) = 1.8 E-2.7 nucleons/(cm2 sec. sr GeV) Integrating the expression for IN(E) between 10 GeV/nucleon and 104 GeV/nucleon and multiplying by 2ʌ steradians yields the total flux of 10 to 104 GeV/nucleon particles to a zenith oriented 1 square centimeter target in low Earth orbit (cosmic rays from the nadir hemisphere are blocked by Earth). About 79% of the total nucleon flux consists of protons [52]. With the proton fluence and the relative abundances, the daily fluence of all nuclei with 10 GeV/nucleon < E < 104 GeV/nucleon can be calculated for specific ranges of Z and is shown in Table 8, along with the corresponding inelastic nuclear collision length, Ȝ, for Al (52). TABLE 8. Daily Fluence and Interaction Parameters of GCRs 10 GeV/nucleon < E < 105 GeV/nucleon Z

Elements

1 2 3-5 6-8 9-10 11-12 13-14 15-16 17-18

H He Li-B C-O F-Ne Na-Mg Al-Si P-S Cl-Ar

19-20

K-Ca

Relative Abundance 1 0.054 0.000825 0.00454 0.000619 0.00454 0.000392 0.0000619 0.0000206 0.0000412

Daily Fluence (1)

LET (2) MeV-cm2/mg x 103

Inelastic collision length(3), Ȝ

9,060 486 7.5 41 5.6 41 3.5 0.6 0.2

0.0002 0.0008 0.003 0.009 0.02 0.02 0.03 0.04 0.06

106 39.8 22.5 16.3 12.9 11.3 10.1 9.2 8.1

0.4

0.07

7.8

21-25 Sc-Mn 0.000103 0.9 0.10 6.6 26-28 Fe-Ni 0.000247 2.2 0.14 5.8 (1) Nuclei/cm2 per 24-hour day; (2) MeVcm2/g for 10 GeV/nucleus projectile in Si target (3) grams/cm2 Al:

51 The percentage of GCR particles (10 GeV/nucleon 10 GeV/nucleon primary GCR flux. TABLE 9. Percent of primary cosmic ray flux with E in the interval 10GeV/nucleon < E < 104 GeV/nucleon lost to inelastic nuclear interaction in Al shielding of thickness 1.0, 5.0, 10, and 20 g/cm2 Z 1 2 3-5 6-8 9-10 11-12 13-14 15-16 17-18 19-20 21-25 26-28

Elements H He Li-B C-O F-Ne Na-Mg Al-Si P-S Cl-Ar K-Ca Sc-Mn Fe-Ni

1.0 g/cm2 Al 0.94 2.5 4.3 5.9 7.4 8.5 9.4 10.3 11.6 12.0 14.0 15.4

5.0 g/cm2 Al 4.6 11.8 19.9 26.3 32.1 35.8 39.0 42.0 46.1 47.3 53.1 56.7

10 g/cm2 Al 9.0 22.2 35.8 45.8 53.9 58.8 62.8 66.3 70.9 72.2 78.0 81.2

20 g/cm2 Al 17.1 39.5 58.8 70.6 78.8 83.0 86.2 88.7 91.5 92.3 95.2 96.5

As is the analysis shown above suggests, inelastic nuclear interactions of high energy GCRs with thick Al structure and shielding can plausibly produce an induced SEE environment with a flux and LET spectrum capable of affecting ISS electronic systems. The GCR energy interval 10 GeV/nucleon < E < 104 GeV/nucleon and the secondary induced environment caused by the subject GCR population has not been explicitly or

52 completely included in previous studies [48-50] and typically isn’t included in SEE assessments at all [6, 51]. The absence of any dependence on SAA, latitude or SEP event exposure for the observed anomalies does not instantly confirm a non-radiation prime cause. Even in the previously reported examples of a strong SAA, latitude, and SEP exposure dependence of SEU rates in various LEO spacecraft [54-56], a substantial number of SEU events are also observed that are randomly distributed with respect to latitude and longitude and are unrelated to SEP events. A detailed study of the SEE environment induced in ISS by the GCR flux in the energy interval 10 GeV/nucleon < E < 104 GeV/nucleon using the Geant4 code (57) is planned and will be combined with continued tracking of the SEE performance of critical ISS avionic systems such as the MDMs. The solid-state memory devices in the MDMs are fully characterized with respect to the LET dependence of SEE processes. 5.0 Ionizing Radiation II: Crew Dose Environment Attempting to reduce ionizing radiation dose-equivalent (dose equivalent = dose x Q, where Q is the quality factor for the radiation field) to ISS expedition crews to As Low As Reasonably Achievable (ALARA) levels has raised a number interesting questions in the area of nuclear and radiation chemistry of materials. The crew radiation doseequivalent environment inside the ISS pressurized elements has a significant induced environment component, produced by inelastic nuclear interactions and nuclear reactions between the primary cosmic ray particle flux (trapped, solar, and GCR) and the relatively thick ISS structure and shielding [41-44, 58]. Design of augmented radiation shielding for ISS and verification of shielding effectiveness is complicated by the limited accuracy of existing theoretical models of cosmic ray reaction and transport in spacecraft materials [59, 60]. In addition, accurate measurement of the doseequivalent environment inside ISS, or any other spacecraft, is controversial and difficult [61]. Of more immediate importance is the fact that most ISS components are either already flying or in the final stages of preparation for flight, so that minimizing the radiation dose-equivalent environment by careful selection of materials and control of topology isn’t always possible. For example, carbon, polyethylene and water have well known advantages over aluminum as radiation shielding in energetic charged particle environments [62-66]. However, most external ISS structure is Al, and practical considerations preclude placing augmented radiation shielding outside the Al structure. As a result, GCR, trapped, and SEP event energetic particles interact with a thick Al target first, producing an induced radiation dose environment that is not effectively mitigated with reasonable amounts of hydrocarbon polymer or water [42, 44, 49, 58], though some flight results collected under relatively thin Al shielding do show that polyethylene can produce a more substantial reduction in dose equivalent than a comparable thickness of Al [67]. Results of the recently completed Japanese Space Agency (NASDA) Bonner Ball Neuron Detector (BBND) experiment on ISS demonstrate this point [68]. The Bonner sphere neutron spectrometer (69) consists of a set of identical thermal neutron sensors each situated at the center of polyethylene moderating spheres of different thickness.

53 Polyethylene is often chosen as a moderating material because it can slow down fast neutrons without significant absorption losses [69]. By design, the NASDA BBND detects neutrons less energetic than 10 MeV and BBND ISS data (68) clearly show significant crew dose-equivalent rates during SAA passage and at high latitude during the SEP events of April 2001. Measurements of quiet time (no SEP event) dose equivalent made on ISS using the German Aerospace Center charged particle detector telescope (DOSTEL) during 2001 yielded an average value of 20.8 micro-Sieverts/hour (2.08 millirem/hour) with a quality factor, Q, ranging from 2.5 to 2.6 [70], corresponding to an average 12-month dose-equivalent of 182 milli-Sieverts (18.2 rem). The total annual dose measured by the DOSTEL team during 2001 (solar maximum) was 0.055 Gray (5.5 rads), significantly less than the 0.091 Gray (9.1 rads) measured during the 1997 Shuttle-Mir missions (solar minimum), as expected. The DOSTEL measurements are comparable to ISS dosimeter measurements made during the same time frame [41], which ranged from an annual dose of 44 -105 milli-Gray (4.4 to 10.5 rads) depending on location in ISS. ISS expedition crews reside on ISS for between 4 and 6 months in most cases, so the practical crew dose environment is one-third to one-half the annual environment. 6.0 Ionizing Radiation III: Surface Dose to Exposed External Materials and Systems Energetic trapped electron dose is the principal threat to the long-term durability of Teflon®, silicone, and other radiation labile materials exposed on the exterior of ISS. Large uncertainties in electron dose predictions made with the AE-8 model [71], combined with uncertainties in the synergistic contributions of mechanical stress, thermal cycling, and atomic oxygen to the degradation of Teflon® by energetic electron radiation [72 a, 73] lead to a corresponding uncertainty in estimated degradation rates. The energy spectrum of trapped energetic electron environment leads to a rapid decrease in radiation dose with depth. Radiation dose and damage is always greatest near the surface of exposed materials. As an example, the dose vs. depth curve for an aluminum sphere (density = 2.7 g/cm3) is shown in Figure 12. Multiplying the thickness axis in Figure 12 by the ratio [density of aluminum]/[density of material] leads to an approximate dose vs. depth curve for any other material [72 b]. When electron range is expressed as g/cm2, the curve for aluminum is approximately correct for any material because density varies only slightly with the atomic number Z of the stopping medium [72c]. Teflon® is an excellent insulator which complicates accelerated ground based testing of energetic electron degradation effects as a result of target electrostatic charging in high dose rate electron beams. On-orbit, dose rates are much lower permitting continuous discharge of exposed Teflon® materials by dielectric relaxation processes, producing a dose-depth profile that is difficult to reproduce in ground based accelerated testing because the lower energy electrons responsible for most near surface dose are most easily decelerated and deflected by target charging. For example, in 10 years the SSP 30512 x 2 dose to surface exposed Teflon® materials is 2 Mrads, corresponding to 1.3 x 10-5 coulombs of charge, distributed over a depth of

54 2 mm if delivered by a 0.1 to 1.0 MeV electron beam (designed to approximate the trapped electron energy spectrum). A typical volume resistivity for Teflon® is on the order of 1016 ohm-m, and the dielectric constant is 2.1, allowing us to calculate the dielectric relaxation time, tr, given the permitivity of free space, İo = 8.85 x 10-12 Farad/m. tr = R İ İo = 51 hours Any charge distribution in Teflon® will dissipate by charge leakage to surrounding conductors with an exponential decay constant tr. Q(t) = Q(0)e(-t/tr) With a relaxation time of 51 hours, it is clear that delivery of 1.3 x 10-5 coulombs during10 years creates no target charging, but attempting to deliver the same charge during 10 hours can result in enough charge build-up in the target to deflect or decelerate the e- beam, producing a much lower dose than anticipated, especially in the near surface region. Consider the ISS Trailing Umbilical Assembly (TUS) cable, which provides power and data services to the Mobile Transporter (MT) platform that moves along rails on the ISS truss. The TUS cable is normally stowed in an uptake reel and is fed out to follow the MT as it moves along the truss. The mechanical design of the TUS cable and uptake reel assembly requires that the insulation and jacketing materials must be capable of surface elongation of not less than 3.5%, without cracking or tearing, for reliable system performance. The TUS cable is jacketed in 1.2 mm (6 layers, 0.2 mm per layer) of expanded polytetrafluoroethylene (PTFE) with a density of 1g/cm3. TUS cable temperature is expected to range between –100o C and +130o C. Polytetrafluoroethylene (PTFE) is especially susceptible to embrittlement by ionizing radiation with an onset of mechanical property loss on exposure to 50 and 150 krads in PTFE [72,73]. The SSP-30512 x 2 dose to each of the 6 layers of TUS cable jacket is shown in Table 10. The radiation dose in Table 10 includes the effects of ISS structural shielding and TUS cable geometry. Continuous exposure without shielding by the uptake reel housing is also assumed. Ionizing radiation breaks chemical bonds and embrittles PTFE causing an increase in stiffness and a decrease in elongation [72,73], suggesting the need for verification testing and analysis to assure TUS cable functional life on ISS. * TABLE 10. SSP-30512 x 2 radiation dose (kilorads PTFE) to the center of each layer of the expanded PTFE TUS cable jacket material as a function of time-on-orbit

Layer 1 2 3 4 5 6 1 year 250 50 22 13 9 6 2 years 600 110 50 29 19 12 4 years 300 120 64 40 29 8 years 1100 320 150 90 60 10 years 520 220 120 85 * Surface erosion rate of 0.03 mm/year from atomic oxygen attack is assumed, leading to the complete removal of the outermost two outermost layers in 10 years.

55 Estimating the useful life of exposed Teflon® materials and components on ISS starts with a worst-case estimate of the expected trapped radiation dose the SSP-30512 design environment (AE8 Max, 500 km) multiplied by 2. Materials are first subjected to the worst-case 10-year dose using energetic electrons (E = 1 MeV to minimize target charging effects) or Co60 gamma rays in vacuum. The test is designed to produce nearly uniform dose through the cross section of the sample because target charging could not be controlled for the reasons given above. Next, the irradiated materials are subjected to a worst-case thermo-mechanical environment to verify that mechanical properties are adequate to perform the required function on orbit. Mechanical testing results (%elongation at maximum load prior to failure) at –100o C, +25o C, and +130o C irradiation are shown in Table 11. TABLE 11. Mechanical test results (% elongation prior to failure) at three different temperatures for TUS cable wrap as a function radiation dose Radiation dose 2.7 Mrad Co60 Ȗ rays 2.7 Mrad 1 MeV electrons 900 krad 1 MeV electrons 300 krad 1 MeV electrons

-100o C

+25o C

+130o C

13.0

13.7

9.3

8.3

22.1

17.3

11.8

28.4

29.1

10.43

36.7

39.3

For the 51.6o inclination ISS orbit, AE8 predicts much higher doses than have been observed on MIR [75, 76] and on the two Russian-Canadian BION satellites [77]. As a result the SSP-30512 design environment, 500 km altitude AE8 Max, overestimates dose significantly, even without the recommended 2 x multiplier. Figure 12 compares dose vs. depth in a solid Al sphere for the 500 km AE8 Max SSP-50512 environment as well as the 390 km environment for both AE8 Max and AE8 Min. Table 12 compares the calculated SSP-30512 x 2-electron dose with corresponding dose measurements from Mir [75, 76] and BION [76]. Clearly, AE8 is seriously overestimating the actual trapped electron dose for the 51.6o, 350-400 km ISS (or Mir) orbit, and the use of the SSP-30512 x 2 design environment to define verification testing dose adds even more conservative margin, which is important in light of the uncertainties in any laboratory test method.

56

Figure 12. A comparison of the AE-8 dose/depth curves for an Al sphere: 1) SSP 30512 Design Environment, 2) 390 km altitude TABLE 12. A comparison of measured annual trapped electron dose with calculated AE8 annual dose and the SSP-30512 x 2 design-environment annual dose. Satellite, Experiment, Measured annual AE8 dose Shielding dose; rads/year rads/year Mir, REM (74), 21 122 0.7 mm Al Mir, TLD (75), 7 x 103 1 x 104 0.0 mm Al Mir, TLD (75), 673 645 0.5 mm Al equiv. Bion 10 & 11 MOSFET/TLD(76), 2 x 105 8 x 104 0.0 mm Al * SSP-30512 x 2 dose = conservative design margin Measured dose

SSP-30512 x 2 dose rads/year

SSP-30512 x 2 dose* Measured dose

1470

21

5 x 105

71

9 x 103

13

5 x 105

2.5

The right-most column of Table 12 shows the ratio of the measured trapped electron dose for a particular flight experiment to the SSP-30512 x 2 design and verification dose and is a direct measurement of the conservative margin built into the SSP-30512 x 2 design environment. The average conservative design margin from the data in Table 12 is 27. Despite the near 30 fold margin in the radiation dose for mechanical testing (Table 11), the TUS cable wrap retained more than adequate mechanical properties over the full temperature range expected on orbit. In addition, removal of the radiation embrittled near-surface layer by ram atomic oxygen further reduced the risk of mechanical failure. The large conservative margin in the radiation dose used for material verification testing assures that the TUS cable and other exposed Teflon® materials will retain useful performance properties for the full life of ISS, despite the many unknown or un-quantifiable factors affecting radiation-induced property loss.

57 7.0 Small Particle Impact and Molecular Contamination from Hypergolic Engine Plume Impingement and Fluid Venting Bipropellant thrusters (monomethyl hydrazine (MMH)/unsymmetrical dimethyl hydrazine (UDMH) and N2O4) are widely used for the purposes of spacecraft reboost, attitude control, and orbit correction. Experimental studies [77, 78] revealed the presence of small particles consisting of unburnt propellant in the engine exhaust plume. The origin of the droplets is commonly attributed to incomplete combustion [77, 78]. Such droplets can range from 1-100 microns in diameter and have velocities up to the exhaust velocity of the thruster exhaust-plume gases (~ 3 km/s). The unburnt propellant may be in the form of super-cooled liquid or ice. Hypergolic engine plume impingement can have important effects on the functional life of exposed materials [79-82]. Past flight experiments, including the Shuttle Plume Impingement Flight Experiment (SPIFEX), have shown the impact of these droplets on witness coupons. SPIFEX (80) was flown on STS-64 in September 1994 and measured in plume parameters during 84 firings of the Primary Reaction Control System (PRCS) and 17 Vernier Reaction Control System (VRCS) engines, all for engine-nozzle to witness-coupon distances of 2.4 to 24 meters, corresponding to more than 10 years of nominal Shuttle–ISS proximity operations. Post-flight analysis of witness coupons on SPIFEX showed mechanical impact pitting on aluminum films and a combination of mechanical and chemical erosion pitting on the Kapton® film samples. Examples of these effects are shown in Figures 13 and 14. It should be noted that particle impact effects were not visible to the unaided eye on the SPIFEX payload. Visual inspection and photographic survey of the SPIFEX payload before and after flight revealed no substantial damage [80]. Control of particle impact degradation of critical materials is based on operational protection of sensitive surfaces during proximity operations (thruster firings) by visiting spacecraft or during ISS venting or purging operations. PV arrays are positioned edgeon to the direction of plume flow. View ports and windows are generally covered, and cameras pointed away from the particle source. Low to medium velocity particles only affect the optical performance of view ports, windows and PV array glass cover slips. Both the Orbiter and US Laboratory module vent liquid water into vacuum as part of their normal operations. A model was developed to describe the two-phase plume (large and small ice particles interspersed with gas) resulting from these operations [83]. A comparison of on-orbit and model predictions shows excellent agreement, as is shown in Figure 15 [83]. The size (0.42 to 1000 microns radius) and velocity (3 to 23 m/sec.) distribution of the ice particles produced in the venting of water from both the Shuttle and ISS are well below the damage threshold for both the PV array cover slips and all ISS optical view ports and instruments (84).

58

Figure 13. SPIFEX Aluminum witness coupon showing craters from thruster plume droplets

Figure 14. SPIFEX Kapton witness coupon showing craters from thruster plume droplets

59

Figure 15. Orbiter Water Dump Plume: Model and Flight Comparison

8.0 External Contamination, Atomic Oxygen and Solar Ultraviolet Radiation 8.1 ISS EXTERNAL CONTAMINATION CONTROL & RISK MITIGATION Vacuum exposed materials contribute to contamination of spacecraft external surfaces as a result of materials outgassing and deposition processes [85]. Materials internal to non-pressurized shells also contribute to contamination of spacecraft external surfaces because outgassing from such materials is vented to the exterior of the spacecraft. SSP30426 [86] defines the external contamination control requirements for ISS. A process of careful materials selection and testing combined with integrated system level contamination transport and deposition analysis assures compliance with contamination performance requirements. The vacuum outgassing and contamination characteristics of ISS materials are determined in long duration ASTM E 1559 (ASTM, 1993b) tests [87]. In testing ISS materials, the sample (outgassing source) is tested over the expected operational temperature range identified by the ISS system level thermal model. Thermally Controlled Quartz Crystal Microbalances (TQCMs), or receivers, are held at temperatures covering the typical range of operational temperatures of ISS contamination sensitive hardware. ISS materials are typically tested for 144 hours according to ASTM E 1559, for each relevant source and receiver temperature, and the test data is curve-fitted to a conservative model describing the decay of outgassing rates, R(t), over the life of ISS using the scaling relation R(t) = R(144hrs)/(t/144[rs), which predicts more outgassing at long times than is expected for most materials [88]. R(t) has the units or mass/(area x time).

60 The NASAN (NASAN is not an acronym) software package [89] is used to analyze external contamination transport and deposition processes on all ISS flight configurations. The principal inputs to NASAN contamination analysis are: 1) the quantity of each material used, 2) materials outgassing data, 3) location on ISS, 4) location and directionality of vents, 5) outgassing source temperature, 6) receiving surface temperature, and 7) the three dimensional configuration of ISS. Using NASAN, contamination processes resulting from visiting vehicle approach, docking, and separations can also be accurately modeled. Direct Simulation Monte Carlo (DSMC) methods [90] have also been employed to model return flux and specific problems encountered during ISS design and development. When a violation of system level requirements is identified in a NASAN analysis, corrective actions are implemented. Materials substitution or vacuum baking of materials and assemblies may be required to remove volatile components and low molecular weight species. For example, the Ku-band antenna dish (coated an S-13 GLO type organic silicone based thermal control paint), Superflex-R silicone insulated electrical cables, and silicone electrical cable clamps were vacuum-baked to ensure compliance with ISS system level requirements. Russian spacecraft design practices require less severe external contamination control processes than those implied by SSP-30426. By working closely with the Russian program during ISS design and development, Russian elements were successfully integrated without volition of the SSP-30426 requirements. Extensive testing of Russian materials, combined with NASAN analysis of materials outgassing, venting, and impingement of thruster plumes has enabled successful and cost effective integration of Functional Cargo Block (FGB), Service Module (SM), Progress, and Soyuz, while also addressing any effects of the NASA, ESA, and NASDA elements on the Russian segment. Taking advantage of FGB and SM time on-orbit as a materials vacuum bake-out prior to the deployment of more contamination sensitive ISS elements also mitigated the impact of FGB and Service Module materials outgassing contamination on sensitive ISS surfaces. The principal FGB contamination sources are non-metallic materials. These materials exhibit high outgassing rates and are exposed to vacuum under multi-layer insulation. The major contamination sources are the impregnated mesh (BF-4) used as support to the solar cells in the solar arrays, AK-573 (acrylic based paint) on the solar array frames, the PArML (Aramid-type fabric) cable wrap which encapsulates the power and data cables, and the KO-5191 (organic silicone based) and EP-140 (epoxy based paint) paints which are widely used on exterior surfaces of the FGB. The source temperature, collector temperature, and collector surface area are vital factors that affect the degree of on-orbit contamination and material degradation rate. Modifications were made to the SM materials selection to minimize contamination risks for any more sensitive ISS surfaces, as there would be less on-orbit time for vacuum baking of materials. Of particular importance were cable insulation and paint changes. However, while changes were made to SM materials selection (cable insulation, surface preparation, paint, and use of anodizing for the pressure shell), evidence of limited self-contamination in the form of surface darkening along seams,

61 around vents, and plume impingement zones has also been identified. The AK-512 paint that replaced the high outgassing KO-5191 and EP-140 paints also shows extensive degradation in the form of flaking and darkening. Controlling surface contamination by hypergolic engine plumes is also an important part of ISS contamination control and risk mitigation. The SPIFEX [80], PIC [79] and DVICON [81] flight experiment projects were all directed toward a quantitative understanding of both short-term and long-term hypergolic engine plume contamination effects needed for verifying compliance with the ISS external contamination requirements [86]. ISS surface contamination resulting from docking and separation of visiting vehicles as well as operation of ISS thrusters for re-boosts and attitude control is estimated using the NASAN software package, verified models of engine plume transport [79,82], and the SPIFEX, PIC, and DVICON deposition and evaporation/sublimation measurements. NASAN simulations of various docking and separation procedures have been used to select procedures that use minimally contaminating thruster firing sequences. In addition, NASAN simulations have also supported development of flight rules that require protecting optically sensitive surfaces such as video camera lenses, ISS view ports, and ISS windows from thruster plume impingement in most cases. It should also be noted that the thrusters on the Progress or Soyuz docked at the aft end of the SM are used for ISS pitch control whenever possible to minimize the contamination impacts of the SM pitch thruster operation over the functional life of ISS. The SM roll jets will be used less in future if roll control can be accomplished using the thrusters on a Progress vehicle docked on nadir. The SM roll control thrusters fire automatically on depressurization of the Docking Compartment Module (DCM) airlock because depressurization is propulsive. As a result, it is possible for EVA crew to contact SM surfaces near the roll control thrusters shortly after thruster firing. Chemical analysis of fresh hypergolic thruster residues on Mir, conducted as a part of the DVICON flight experiment [81], suggests that contamination of EMU suit (space suit) surfaces with toxic substances is possible if insufficient time elapses between roll control thruster firing and entry of EVA crew into the thruster plume impingement area. PIC, SPIFEX, and DVICON data all demonstrate that the fresh contamination deposit, formed when a thruster fires, evaporates or sublimes rapidly if the temperature of the contaminated surface is near 25o C, leaving a persistent residue that is nontoxic if detectable at all [79-81]. However the fresh hypergolic thruster contamination deposit is significant and contains some highly toxic, though highly volatile, compounds [81. The time that must elapse between roll control thruster firings and complete evaporation or sublimation of the toxic species is very sensitive to the surface temperature of the SM and is expected to range from a few minutes to a few hours as surface temperature varies between –40o C to +25oC. EVA flight rules designed to minimize the risk of introducing toxic substances to the habitable volume of ISS have been developed, while laboratory studies aimed at a better definition of the temperature dependence of the toxic residue evaporation time are completed. Cosmonaut and astronaut observations during EVA on the aft end of the SM suggest that there is very little risk of toxic contamination of the EMU suits.

62 8.2 ISS EXTERNAL CONTAMINATION OBSERVATIONS In-flight imaging surveys of the US Lab, Node 1, the Pressurized Mating Adapters (PMAs), and the Z1 and P6 Truss structures, as well as the 160V US PV arrays show little or no evidence of large-scale self-contamination from molecular outgassing and deposition processes, providing some confirmation of the validity of the ISS contamination risk mitigation process. More specifically, high contamination risk items like the SuperFlex-R silicone cable, the silicone cable clamps and the Ku-band antenna dish are not producing visible molecular outgassing and contamination deposits on neighboring hardware after two years of flight. The FGB and SM both show limited self-contamination from molecular outgassing and deposition processes as well as surface discoloration and damage from hypergolic thruster plume impingement. The results of external contamination imaging surveys are summarized in Tables 13-15[91]. TABLE 13. Summary of FGB Contamination Observations Location FGB aft-end and radiator surfaces

First Flight Observation 2A (pre-docking)

FGB TORU antenna opening

2A

FGB MM/OD shields

2A

Gaps/vent paths in the FGB blankets and structure

2A

Description Dark patterns on FGB aft-end and radiators from Proton stage separation thruster induced contamination Darkening of the blanket surrounding the TORU antenna opening Darkening of the area surrounding the MM/OD shield gap and flaking of the MM/OD shield paint Darkening of the area surrounding gaps and vent paths on the FGB blankets and structure

8.3 ATOMIC OXYGEN AND SOLAR ULTRAVIOLET Basic ISS materials performance requirements for the atomic oxygen (AO), solar ultraviolet (UV), and solar vacuum ultraviolet (VUV) environments are defined in SSP30233, Rev. F., “Space Station Requirements for Materials and Processes,” [92] and require that materials be selected to perform in the LEO environment for the intended life cycle exposure. The effects of atomic oxygen (AO) and ultraviolet radiation on materials exposed to LEO were well known and reasonably well understood prior to materials selection for ISS as a result of a important series of space flight experiments [93-95] such as the Long-Duration Exposure Facility (LDEF), the several Space Shuttle flight experiments on Evaluation of Oxygen Interaction with Materials (EOIM-1, EOIM-2, EOIM-3), and the Passive Optical Sample Assembly (POSA) experiment conducted on Mir during ISS Phase 1. As a result, numerous design solutions were implemented to prevent environmental degradation of external materials and AO degradation problems experienced to date on ISS have been minimal.

63

TABLE 14. Summary of Service Module Contamination Observations Location Service Module aft-end and radiator surfaces

First Flight Observation 2A.2b

Service Module handrails

2A.2b

Gaps/vent paths in the Service Module structure

2A.2b

Service Module Instrument Section

2A.2b

Service Module Zenith facing thrusters

5A

Kurs antenna at the end of Service Module solar arrays

5A.1

Description Dark patterns on Service Module (and FGB) aftend and radiators from Proton stage separation thruster induced contamination Darkening patterns and flaking of handrail coating Darkening of the area surrounding gaps and vent paths on the Service Module structure Darkening patterns and surface degradation of painted surfaces Darkening of the area near the Zenith thrusters on the Service Module Darkening of the base of the Kurs antenna at the end of the Service Module solar arrays

The use of oxygen reactive materials in critical ISS applications has been minimized, as has the use of materials that darken or are otherwise degraded by UV/VUV. Where reactive materials must be used, for example in the Kapton® film thermal blanket material upon which the 160-V PV array wings are assembled, multiple layers of oxygen resistant protective coatings are applied to assure full 15-year functional life, as is described elsewhere in these proceedings [96]. Another area in which reactive materials can be used results from the wake environment produced by ISS structure. Because ISS can fly only a limited number of approved attitudes, structural wake shielding from ram atomic oxygen enables the use of some oxygen reactive materials in wake shielded locations with no risk of significant long-term AO induced degradation of performance.

64

TABLE 15. Summary of PMA1, PMA2 and Node 1 Observations Location Node 1 MM/OD anodized aluminum shields

First Flight Observation 2A

Node 1 SVS targets

2A.1

Node 1 stowage bags

2A.1

Node 1 aft-CBM supports

2A.2a

Metalphoto labels on Node 1 and both PMAs ORU Transfer Device (OTD) and APFR

2A.2a

PMA2 washers

2A.2b

PMA1 anodized surfaces

2A.2b

PMA2 Worksite Interface (WIF)

2A.2b

2A.2a

PMA2 APAS glass fabric surfaces and able straps

3A

PMA2 planar and hemispherical retroreflectors

3A

PMA1 zenith sunshade

3A

Description Discoloration patterns observed on several MM/OD aluminum panels “Bubbling” of SVS targets on Node 1 Darkening of the Beta cloth used on the Node 1 stowage bags Darkening of the Node 1 aftCBM supports Darkening of the anodized aluminum labels Darkening of the anodized layer at the base of the OTD and surfaces of the APFR Darkening of metallic washers on PMA2 Darkening of anodized aluminum surfaces on PMA1 Darkening of the anodized base of the WIF Russian glass fabric used on the PMA2 APAS, darkening of cable straps Coating used on PMA 2 planar and hemispherical retroreflectors shows degradation and exhibits flaking PMA1 zenith sunshade exhibits dark spots

8.4 ANODIZED ALUMINUM Aluminum is the most important structural material on ISS and the surface anodization on the aluminum also serves as the external passive thermal control surface for the NASA truss structure as well as on other NASA, Japanese Space Agency (NASDA), and European Space Agency (ESA) elements. In order to avoid the use of paints, major ISS external surfaces are anodized using controlled processes to produce the required thermo-optical properties. The truss structure is essentially all sulfuric acid anodized and the meteoroid/orbital debris shields are usually chromic acid anodized. The anodized surfaces are impervious to atomic oxygen and were shown to be stable to ultraviolet radiation in ground testing. So far, these surfaces have performed well with no visible darkening. Specialized forms of anodized aluminum have been used for some applications. Black anodize has been used in a few applications where a thermally hot environment was required. EVA handrails are required to be yellow; in the absence of a yellow paint that would be durable in the LEO environment, a gold anodize was used. This gold anodize

65 has also proved stable to date. These anodize coatings have also been used for external warning labels where yellow/black stripes were required. Almost all ISS external labels are manufactured using commercially available photosensitive anodized aluminum foil (Metalphoto). Monochrome Metalphoto labels were selected after ground based testing showed them to be stable in the atomic oxygen-ultraviolet radiation environment (many colored labels faded rapidly). However, some UV discoloration has been observed on orbit; although the labels are still readable, the clear-coated (originally silver in appearance) surfaces have turned yellow. Similar yellowing/browning of the anodic coating has been observed on a few components provided by NASA. Current speculation is that the UV sensitivity is caused by a nickel-acetate solution used to water seal anodic coatings (the nickel acetate water sealing process is definitely used for the Metalphoto labels and may have been used for some other anodized aluminum hardware); however, we have been unable to replicate the discoloration with vacuum-UV sources on the ground. Essentially all the other anodized hardware was water sealed with deionized water. 8.5 BLANKETS AND SHROUDS Standard multi-layer insulation (MLI) materials erode rapidly in atomic oxygen. Extensive use is made of Teflon impregnated, tightly woven glass fabrics (Beta cloth) as covers for multi-layer insulation blankets [95,96] where the composite fabric is expected to perform for the full life of ISS. The Beta cloth covers have excellent durability despite AO/UV erosion of the Teflon® coating. The remaining glass fiber fabric is impervious to AO and remains flexible enough to withstand 30 years of thermal cycling. Nearly all ISS MLI blankets have an outer layer of Beta cloth. They are sewn together with AO-compatible ceramic thread or Teflon-coated fiberglass thread. Beta cloth has also been used extensively for wrapping and protecting external wire and cable and for thermal control shrouds. The Beta cloth selected for ISS contains no silicone sizing, because Beta cloth with such sizing tends to darken in the UV environment. In a few locations, thermal requirements have dictated the use of MLI blankets with an outer layer of silver-Teflon®. The Teflon® layer is verified to be thick enough to survive in the environment to which it will be exposed, even with significant erosion by AO/UV. 8.6 PAINTS AND COATINGS Nearly all ISS radiators are coated with Z-93, a white inorganic paint with excellent optical properties and a long history of successful use on spacecraft. Silver-Teflon is again used for a few radiator surfaces, but has adequate durability in the application environment.

66 Very few other external surfaces are painted and the coating system used is almost always a specialty inorganic coating system that has been demonstrated to have good durability in AO/UV. Such coatings have been used for targets and logos. The ubiquitous Aeroglaze A276 polyurethane paint has been used for a few short-term applications. As expected, the A276 coatings on returned hardware are darkened by UV and powdery as a result of AO erosion; however, these coatings are adequate for 12 years exposure. Although nearly all labels are anodized aluminum, as noted above, cable labels are frequently poly(vinylidene fluoride) (Kynar®) overcoated with a thin layer of lowoutgassing silicone as protection against AO damage. No problems with these labels have been noted to date, although shrinking and cracking of the silicone coating is a long-term concern. Silicones have also been used for AO protection on solar array wing batons and silver-plated fasteners. The largest area of nonmetallic material on ISS occurs on the PV array wings. The front sides of these wings are protected by glass cover slips and low-outgassing silicone sealants. The backs of the arrays are Kapton®, protected from AO by a silica overcoating. In a few locations where silica could not be used, the Kapton is protected by a low-blocking, low-outgassing silicone. 8.7 ISSUES Despite the high sensitivity of the ISS Program to AO/UV effects, schedule, cost, and materials availability problems have occasionally forced selection of materials having low AO/UV resistance for exposure to the LEO environment for less critical or shortterm applications. The path followed to control such hardware depends on cost and schedule. Whenever feasible, such items are premeditated before launch. Nylon cable ties have been replaced by durable Teflon-coated fiberglass lacing cord. Nomex lacing cord has been oriented away from atomic oxygen and/or overwrapped with copper wire ties. Silver-plated fasteners have been used extensively, despite efforts to prohibit them because of the high reactivity of silver with both the high kinetic energy ram AO (atomsurface collision energy = 5 eV) and the epithermal/thermalized AO (atom-surface collision energy 30 in polycarbonate track detectors,” Nucl. Instr. and Meth. B 146, pp 114-119, 1998 Groom, D. E., The European Physical Journal, C15 (2000) 1, Chapters 6 & 23 Dementyev, A. V., Nymmik, R. A., Sobolevsky, N. M.; Nucleon Spectra Behind 1-100 g/cm2 Aluminum Shielding under Galactic and Solar Cosmic Rays Irradiation, Preprint n 95-28/392. SINP MSU 1995 http://techreports.larc.nasa.gov/ltrs/1995-cit.html , Shinn, J. L., Cucinotta, F. A., Wilson, J. W., Badhwar, G. D., O’Neill, P. M., Badavi, F. F.; “Effects of Target Fragmentation on Evaluation of LET spectra form space Radiation in Low-Earth-Orbit (LEO) Environment: Impact on SEU

70

51. 52. 53.

54.

55.

56. 57.

58.

59. 60.

61.

62. 63.

64.

65.

66.

67.

68. 69.

Predictions,” 32nd Annual International Nuclear and Space Radiation Effects Conference, Madison Wisconsin, 17-21 July 1995 http://www.spenvis.oma.be/spenvis/help/models/letorb/crème.html Groom, D. E., The European Physical Journal, C15 (2000) 1, Chapter 20 Atwell, W.; “ISS Radiation Monitoring: PTE Shielding Distributions, Calculations & Measurements,” 6th Workshop on Radiation Monitoring of the International Space Station, 12-14 September, 2001, Jesus College, Oxford England Seidleck, C. M., LaBel, K. A., Moran, A. K., Gates, M. M., Barth, J. M., Stassinopoulos, Gruner, T. D.; “Singe Event Effect flight Data Analysis of Multilpe NASA Spacecraft and Experiments; Implications to Spacecraft Electrical Design,” http://radhome.gsfc.nasa.gov/radhome/papers/chris.html Personal communication, NASA/JSC/EA44/Truong, A. J.; “STS-95, 88, 96, and 93 Post Flight Analysis Report of Radiation Effectson General Purpose Computer (GPC) Memory Devices.,” EA44-99-022, Oct. 25, 1999 Klausman, A. L.; Effects of Space Flight on Small Personal Computers, Masters Thesis, University of Houston at Clear Lake, December 1995 Truscott, P., Lei, F., Ferguson, C., Gurriaran, R., Neiminen, P., Daly, E., Apostolakis, J., Giani, S., Pia, M. G., Urban, L., Maire, M.; “Development of a Spacecraft Radiation Shielding and Effects Toolkit based on Geant4,” British crown copyright 2000/DERA, and http://wwwinfo.cern.ch/asd/geant4/geant4.html Berger, T., Hajek, M., Schoner, W., Vana, N., Noll, M, Ebner, R., Akatov, Y.,Shurshakov, V., Arkhangelsky, V.; “Measurement of the Depth Distribution of Average LET and Absorber Dose Inside a Water-Filled Phantom on Board Space Station MIR,” Physica Medica, Vol. XVII, Supplement 1, 2000 pp 1-11 http://www.nationalacademies.org/ssb/besrch2.html “Radiation hazards to Crews of Interplanetary Missions; 2 Issues of Concern to NASA” Strategic Program Plan for Space Radiation Health Research, Approved by Arnauld E. Nicogossian, M.D., Associate Administrator, Office of Life and Microgravity Sciences and Applications, NASA Headquarters, October, 23, 1998 Singleterry, R. C., Badavi, F. F., Shinn,, J. L., Cucinotta, F. A., Badhwar, G. D., Clowdsley, M. S., Heinbockel, J. H., Wilson, J. W., Atwell, W., Beaugean, R., Kopp, J., Reitz, G.; “Estimation of neutron and other radiation exposure components in low earth orbit,’ Radiat. Meas. 2001 June; 33(3). Pp 355-360 Kim, M-H. Y., Wilson, J., Thibeault, S., Nealy, J. E., Badavi, F. F., Kiefer, R. L.; “Perforamcne Study of Galactic Cosmic Ray Shield Materials,” NASA Technical Paper 3473, November, 1994 Doll, D., Van Hagen, T., Redler, K., Tooker, J., Baxter, A., Fikaini, M., Schneider, D., Spinos, F., Funk, W.; “Low to High Energy Beam Stops for ATP,” http://accelconf.web.cern.ch/AccelConf/l98/PAPERS/TU4090.PDF , and Maurer, R. H., Roth, D. R., Kinnison, J. D., Jordan, T. M., Hielbronn, L. H., Miller, J., Zeitlin, C. J.; “Neutron Production from Polyethylene and Common Spacecraft Materials,” IEEE Trans. Nucl. Sci., Vol. 48, No. 6, pp 2029-2033, Dec. 22001 Andronenko, M. N., Andronenko, L. N., Neubert, W., Seliverstov, D. M.; “Equilibrium, Isoscaling and Nuclear Isotope Thermometry Related to 1 GeV Proton Induced Nuclear Reactions,” arXiv:nucl-ex/0112014v1 27 dec. 2001 http://www.triumf.ca/safety/rpt/rpt_7/node20.html; http://www.triumf.ca/safety/rpt/rpt_7/node19.html; http://www-esh.fnal.gov/FRCM/Ch08/Ch08.html; Chapter 8 Accelerator Shielding and Radioactivation, Fermi National Laboratory Barschall, H. H., Chadwick, M. B., Jones, D. T. L., Meulders, J. P., Schumacher, H., Young, P. G., Cox, L. J., Hale, G. M., Schrewe, U., Siebers, J. V., Caswell, R. S., DeLuca, P. M., Wambersie, A.; Nuclear Data for Neutron and Proton Radiotherapy and for Radiation Protection, ICRU Report 63, Dec. 1999 Badhwar, G. D., Cucinotta, F. A.; “A Comparison of Depth Dependence of Dose and Linear Energy Transfer Spectra in Aluminum and Polyethylene,” Radiation Research, Vol. 153, pp1-8, 2000 http://spaceresearch.nasa.gov/research_projects/ros/bbndop.html Thomas, D. J., Alvera, A. V.; “Bonner sphere spectrometers-a critical review,” Nucl. Inst. and Meth. .in Phys. Res. A, Vol. 476, pp 12-20, 2002

71 70. Beaugean, R., Burmeister, S., Petersen, F., Reitz, G.; “Dosimetric Measurements on ISS during Quiet and Disturbed Periods,” 6th Workshop on Radiation Monitoring of the International Space Station, 12-14 September, 2001, Jesus College, Oxford England 71. Armstrong, T. W., Colborn, B. L.; “Evaluation of Trapped Radiation Model Uncertainties for Spacecraft Design,” NASA/CR – 2000-210072, March 2000. 72. a) Townsend, J. A., Hansen, P. A.,Dever,, J. A., de Groh, K. K., Banks, B. A., Wang, L., He C.; “ Hubble Space Telescope Metallized TeflonR FEP Thermal Blanket Control Materials: on-orbit degradation and post-retrieval analysis,” High Performance Polymers, Vol. 11, No.1 March 1999, b) Lilley, J.; Nuclear Physics, Principles and Applications, John Wiley and Sons, New York, 2001, pp. 139-137, figure 5.5, c) Katz, l., Penfold, A. S..; Reviews of Modern Physics, Vol. 24, page 28, 1952 73. Dever, J. A., de Groh, K. K., Banks, B. A., Townsend, J. A.; “Effects of Radiation and Thermal Cycling on Teflon FEP,” High Performance Polymers, Vol. 11, pp123-140, 1999 74. Buehler, P., Zehnder, A., Desorger, L., Hajdas, W., Daly, E., Adams, L.; “Measurements of the Radiation Belts from Mir and STRV,” http://www.estec.esa.nl/wmwww/wma/R_and_D/rem/cherbs97www/rem97.pdf 75. Schoner, W., Hajek, M., Noll, M., Ebner, R., Vana, N., Fugger, M., Akatov, Y., Shurshakov, V., Arkhangelsky, V.; “ Measurement of the Dose Depth and LET Distribution at the Surface and Inside of Space Station Mir,” http://www.magnet.oma.be/srew/08.pdf 76. Mackay, G. F., Thomson, I., Ng, A., Sultan, N.; “Applications of MOSFET Dosimeters on MIR and BION Satellites,” IEEE Trans. Nucl. Sci., Vol. 47, pp 2048 - 2051, December 1997. 77. Trinks, H.: Experimental Investigation of the Exhaust Plume Flow Fields of Various Small Bipropellant and Monopropellant Thrusters. AIAA Paper 87-1607, June 1987. 78. Rebrov, S., et al: “Monitoring of Contamination Actions of Flames and Exhausts of Propulsion Systems and Power Plants on Elements of Long-Term Space Stations to Minimize These Actions – Final Report,” Contract No. 251-5208/95, Russian Space Agency – Keldysh Research Center, Moscow, 30 January 1998 79. Soares, C.E.; International Space Station Provisional Plume Contamination Model (Phase I Report), Boeing, SSCN 2208 Phase I Report Deliverable. Houston, Texas. – September 30, 1999 80. Koontz, S.; Melendez, O.; Zolensky, M.; and Soares, C.: SPIFEX Contamination Studies. NASA Johnson Space Center Report JSC-27399, May 1996 81. Rebrov, S., Gerasimov, Y.; “Investigation of the Contamination Properties of Bipropellant Thrusters,” AIAA 2001-2818, 35th AIAA Thermophysics Conference, June 11-14, 2001 82. Soares, C., Mikatarian, R., Barsamian, H., Rauer, S.; “International Space Station Bipropellant Plume Contamination Model,” AIAA 2002-3016, 8th AIAA/ASME Joint Thermophysics and Heat Transfer Conference, June 24-17, 2002, St. Louis Mo. 83. Alred, J.W., Smith, L.N., Wang, K.C., Lumpkin F.E., Fitzgerald, S.M.; “Modeling of Space Shuttle Orbiter Waste Water Dumps,” AIAA Paper 98-2588, 7th AIAA/ASME Joint Thermophysics and Heat Transfer Conference, Albuquerque, NM, June 15-18, 1998. 84. SPHINX Impact Simulation Results, ref: Stellingwerf, Wingate; “Impact Modeling with Smooth Particle Hydrodynamics,” Int. J. Impact Engineering, Vol. 14, pp 707-718, 1993 85. http: history.nasa.gov/SP-404/ch7.htm 86. Thomas, D. and Peterson, C.E., “Space Station External Contamination Control Requirements,” SSP 30426, Revision D, NASA Johnson Space Center, Houston, Texas, 1994 87. Standard Test Method for Contamination Outgassing Characteristics of Spacecraft Materials, E1559, American Society for Testing and Materials, West Conshohocken, Pennsylvania, 2000 88. Alred,J.;” Outgassing Modeling and Preliminary Results,” ISS External Contamination Technical Interchange Meeting, Houston Texas, May 3-7, 1999 89. Hakes, Charles, The NASAN Program for Spacecraft Contamination Analysis – A User’s Guide, NASA Johnson Space Center, Houston, Texas, 1997 90. LeBeau, G. J., Lumpkin, F. E.; "Application Highlights of the DSMC Analysis Code (DAC) Software for Simulating Rarefied Flows", Comput. Methods Appl. Mech Engrg., 191 (2001), 595609 91. Soares, C.E, Mikatarian, R., Scharf, R., Miles, E.; “International Space Station Flights 1A/R – 6A External Contamination Observations and Surface Assessment, International symposium on Optical Science and Technology, SPIE’s 47th Annual Meeting, July 7-11, 2002, Seattle Washington 92. SSP-30233, Rev. F, “Space Station Requirements for Materials and Processes,” NASA Johnson Space Center, Houston, Texas, 1998

72 93. Koontz, S. L., Leger, L. J., Visentine, J. T., Hunton, D. E., Cross, J. B., Hakes, C. L.; “EOIM-III Mass Spectrometry and Polymer Chemistry: STS-46, July-August 1992,” J. Spacecraft and Rockets, Vol. 32, No. 3, pp 483-495, May, June 1995. 94. Linton, R. C., Whitaker, A. F., Finckenor, M. M.; “Space Environment Durability of Beta cloth in LDEF Thermal Blankets,” LDEF Materials Results for Spacecraft Applications, NASA CP-3257, Oct. 1992 95. Koontz, S. L., Jacobs, S., Le, J.; “Beta Cloth Durability Assessment for Space Station Freedom Multi Layer Insulation Blanket Covers,” NASA-TM-104748, March, 1993 96. Banks, B. A., Snyder, A., Miller, S. K., Demko, Rikako; “Issues and Consequences of Atomic Oxygen Undercutting of Protected Polymers in Low Earth Orbit,” ICPMSE-6, 6th International Space Conference; Toronto, Canada, May 1-3, 2002, B-8 97. Koontz, S., King, G., Dunnett, A., Kirkendahl, T., Linton. R., Vaughn, J.; “Intelsat Solar Array Coupon Atomic Oxygen flight Experiment,” J. Spacecraft and Rockets, Vol. 31, No. 3, pp 475481, May, June 1994

* Author to whom correspondence should be addressed. ES4/Dr. Steve Koontz, ISS Environments Manager NASA Johnson Space Center Houston, Texas, 77058 Phone:(281)-483-8860 & Fax: (281)-244-1301 e-mail: [email protected] (1)

Structural Engineering Division Mail code ES4 NASA Johnson Space Center Houston, Texas 77058

(2)

The Boeing Company Mail code HA1-20 13100 Space Center Blvd. Houston, Texas 77059-3599

(3)

Dynacs Engineering Co. Inc. 1110 NASA Rd. 1 Houston, Texas 77058

(4)

Solar System Exploration Division Mail Code SX2 NASA Johnson Space Center Houston, Texas 77058

(5)

NASA Marshall Space Flight Center Engineering Physics Mail Stop ED31 Huntsville, Alabama 35812

(6)

Jacobs Sverdrup, MSFC Group MSFC/ED44, Huntsville, AL 35812

PHOTOCONDUCTIVITY IN TRANSPARENT ARCPROOF COATINGS

T. CASHMAN, J. KAUR, L. K. MUHIEDDINE, M. SHANBHAG, S. H. UBAID, BRYAN WELCH, JYOTHI VEMULAPALLI, AND P. D. HAMBOURGER Cleveland State University Cleveland, OH 44115 USA

Abstract:

Highly transparent thin films with moderately high sheet resistivity ~108 ohms/square (Ω/ ) are needed for protection of photovoltaic arrays and other surfaces from static charging in space. Previous work at NASA Glenn Research Center has shown that codeposited mixtures of indium tin oxide (ITO) with MgF2 or SiO2 are promising for this application. We find that exposure of these films to low-intensity blue light substantially reduces the sheet resistivity, which then takes at least several hours to recover.

1. 0 Introduction Nonconductive spacecraft surfaces need protection from electrostatic charging due to particles in the solar wind. The coating must have a maximum sheet resistivity less than ~108-109 Ω/ and, in the case of photovoltaic arrays, must be as transparent as possible [1]. Indium tin oxide (ITO) has been used [2] but is actually too highly conductive for this application. The large number of free carriers in ITO results in unnecessary absorption and reflection of light, reducing solar array efficiency. In addition, accidental contact between power wiring and a low-resistivity coating could lead to large current flow between the power system and the conductive plasma present in low earth orbit. To produce a better coating for these applications, previous workers investigated mixed films made by co-depositing ITO with MgF2 or SiO2, both highly transparent insulators [1,3]. These films had the desired sheet resistivity, appeared to have enhanced atomic oxygen durability, and were considerably more transparent than ITO films of similar thickness, as shown in Figure 1. However, the sheet resistivity increased considerably in the first year after deposition.

73

Total Solar Transmittance

74

0.9 (a)

0.88 0.86 0.84

Avg. film thickness 650 Å Transmittance of SiO2 substrate 0.93

0.82 0.8 0

10

20

30

Sheet Resistivity (Ω/ )

Calculated Vol. % MgF 2 1.E+ 1.E+ 1.E+ 1.E+ 1.E+ 1.E+ 1.E+ 1.E+ 1.E+

10 09 08 07 06 05 04 03 02

(b)

Avg. film thickness 650 Å 0

Figure 1.

10 20 Calculated Vol. % MgF 2

30

Solar transmittance (a) and sheet resistivity (b) of ITO-MgF2 films vs composition (from Ref. 3) .

Due to renewed interest in these coatings, we are conducting a further investigation of their electrical properties as a function of deposition parameters and ambient conditions. Most notably, we find that the sheet resistivity of ITO-MgF2 and ITO-SiO2 are strongly affected by exposure to low-intensity light near the blue end of the visible spectrum. This paper focuses mainly on ITO-MgF2, which was previously found to have superior solar transmittance.

2. 0 Experimental Techniques Films ~700 Å thick were deposited on 0.75 mm thick fused silica substrates by simultaneous operation of two RF magnetron sputter guns. One target was 90%/10% indium/tin oxide, the other was MgF2. The guns were driven by separate 13.56 MHz RF generators; film composition was adjusted by changing the relative power outputs of the generators. Sputtering was carried out in Ar gas at 3 mTorr pressure. We did not

75 feed oxygen or air into the chamber because this system produces highly transparent conductive ITO films without it. The background pressure with Ar gas turned off and pump throttle valve set as for deposition was 130 °C)[5] than that of the MLI FEP (maximum temperature of 50 °C)[6]. During SM2, severe cracking of the 5 mil Al-FEP MLI outer layer was observed on the light shield (LS), forward shell and equipment bays of the telescope. Astronaut observations combined with photographic documentation revealed extensive cracking of the MLI in many locations, with solar facing surfaces being heavily damaged [2]. Figure 1 shows two large cracked areas on the LS. A very large vertical crack can be seen near the center of the photograph, and a smaller cracked area, in which free standing Al-FEP had curled-up tightly (with the FEP surface in

157 compression), is located above the vertical crack. The worst of the MLI outer layer cracks were patched during SM2. Prior to patching the upper LS crack, the tightly curled Al-FEP outer layer was cut off and retrieved for post-mission analyses. Patches of 5 mil thick (127 µm) Al-FEP were placed over the two LS cracks, and patches of 2 mil thick (51 µm) Al-FEP were placed over large cracks in MLI on Equipment Bays 8 and 10. Figure 2 shows one of these cracked areas from Bay 10. As determined through a HST MLI Failure Review Board, embrittlement of FEP on HST is caused by radiation exposure (electron and proton radiation with contributions from solar flare xrays and UV radiation) combined with thermal cycling [6].

SM2 sample location

Figure 1. Two cracked areas in the MLI outer layer on the HST LS as witnessed during SM2. The astronauts have cut off the curled upper LS material and are preparing to place a patch over the area.

Figure 2. Cracks present in 5 mil thick Al-FEP Bay 10 MLI, photographed after retrieval during SM3A.

During SM3A, original MLI from Bay 10, which experienced 9.7 years of space exposure, as well as 2 mil thick Al-FEP patch material, which experienced 2.8 years of exposure, were retrieved and available for degradation analyses. Surprisingly, FEP retrieved during SM2 after 6.8 years exposure was found to be more embrittled than FEP retrieved 2.8 years later during SM3A, after 9.7 years exposure [7,8]. Because

158 the retrieved SM2 material curled with the FEP surface in compression, exposing the lower emittance Al surface to space, it experienced a higher temperature extreme during thermal cycling (≈200 °C) than the nominal solar facing MLI experiences (≈50 °C). As this was the only sample retrieved during SM2, it was important to determine the difference between the measured properties of this excessively heated sample and nominally heated MLI FEP. Another back surface metallized Teflon insulation, 2 mil thick Al-FEP thermal shields that covered the solar array bi-stems of the second pair of HST solar arrays (SA-2), thermal cycled to a maximum temperature of 130 °C on-orbit. Because of the various temperatures experienced by HST FEP materials, it is necessary to understand the effect of temperature on FEP degradation on HST. Understanding temperature effects is important for determining degradation mechanisms, and for facilitating the prediction of FEP degradation in LEO. Investigations have been conducted by de Groh et al on the effects of heating pristine FEP and FEP irradiated in ground facilities or in the LEO space environment, as reported in [7,8]. For this study, samples of pristine FEP, x-ray irradiated FEP and SM3A-retrieved FEP were heated from 50 °C to 200 °C in 25°C intervals in a high vacuum furnace and evaluated for changes in tensile properties and density in order to improve the understanding of the degradation of this insulation material in the LEO space environment. X-rays were used for the source of irradiation because x-rays from solar flares are believed to contribute to the embrittlement of FEP on HST [6], and because previous ground tests have shown that solar flare x-ray energies are energetic enough to cause bulk embrittlement in 127 µm FEP [10]. Also, the mechanism of embrittlement of polymers is believed to be the same for all forms of ionizing radiation, therefore x-ray exposure is a very useful technique for understanding radiation damage effects in Teflon.

2.0 Materials 2.1. PRISTINE FEP AND FEP FOR X-RAY IRRADIATION Teflon FEP is a perfluorinated copolymer of tetrafluoroethylene (TFE) and hexafluoropropylene (HFP). The FEP material used for x-ray irradiation followed by vacuum heat treatment and for non-irradiated heat treatment tests was non-aluminized 5-mil thick (127 µm), and was purchased from Sheldahl (lot #96-16). 2.2. HST SM3A Al-FEP As previously mentioned, MLI blankets originally installed on HST on Bay 10 (exposed to the space environment for 9.7 years) and 2 mil Al-FEP patches installed on Bay 10 during SM2 (exposed for 2.8 years) were retrieved by astronauts during servicing mission SM3A. Figure 3a shows a close-up of the MLI and patches on Bay 10 prior to being removed from HST. Four different materials were retrieved: original MLI from the top section of Bay 10 (Top MLI (TM)), original MLI from the bottom section of Bay 10 (Bottom MLI (BM)), and patches installed during SM2 over portions of TM and

159 BM, designated as Top Patch (TP) and Bottom Patch (BP), respectively. The outer most layer of the TM and BM materials was 5 mil thick Al-FEP. Retrieved patch material was 2 mil thick Al-FEP. The Al was ≈1000 Å. Samples were sectioned from various regions of the BM, TP, and BP surfaces for post-flight analyses and the results are reported in [7,8,11]. For the BM surface, regions designated as R1 and R2 refer to the areas without a patch and covered by a patch, respectively. A large section of the original MLI material, which was not covered by a patch and thus had been exposed to the space environment for 9.7 years (section BM-R1), was cut from the MLI blanket and provided for these heating studies. This sample section can be seen in Figure 3b. Because this large sample was sectioned from the blanket in 2001, while most samples for post-flight analyses testing were sectioned shortly after the December 1999 flight in 2000, this particular sample is referred to as the SM3A 2001 BM-R1 sample. The 2001 BM-R1 sample was carefully examined and the location of all impact sites and cracks were documented in order to avoid these areas when punching out tensile samples or cutting density samples.

TP

TM

BM BP

(a)

(b)

Figure 3. HST Equipment Bay 10 material: (a) Bay 10 MLI and patches during SM3A prior to removal from HST, and (b) A section of MLI blanket showing the location and size of the SM3A 2001 BM-R1 sample.

3.0 The HST Environment Table 1 provides the space environmental exposure conditions for the retrieved SM2 and SM3A BM-R1 materials. For the samples, the table lists the direction the surface faces with respect to the coordinate system for HST, indicated by V2 and V3 axes (described below). The sample retrieved during SM2 faced the +V3 direction, which is

160 the solar facing surface of Hubble. Because this area had cracked and curled with the FEP surface on the inside of the curl at some point during the mission, all environmental exposure conditions (except thermal cycling) are indicated as being some amount less than the values calculated for the entire mission duration. Bay 10 faces the –V2 direction (SADA direction). The environmental exposure conditions of solar exposure hours, solar event x-ray fluence, electron and proton fluence, and atomic oxygen fluence for FEP surfaces on HST vary depending on the direction the surface was facing and position of nearby obstructing surfaces, whereas the number of thermal cycles is independent of direction. TABLE 1. Exposure Conditions for Retrieved HST FEP Materials. Exposure Thermal Cycles/Temperature Range (cycles/°C) Equivalent Solar Hours (ESH) X-ray Fluence (J/m2)

SM2 FEP (LS, +V3)12

SM3A BM-R1 MLI (Bay 10, -V2)

37,100/-100 to +50° C -100 to +200°C when curled

52,550/ -100 to +50°C

< 33,638

13,598

1-8 Å*: < 209

0.5-4 Å*: < 13

1-8 Å : 62

0.5-4 Å : 3.9

Electron Fluence (#/cm2), >40 keV

< 1.95 x 1013

2.74 x 1013

Proton Fluence (#/cm2), >40 keV

< 1.95 x 1010

2.77 x 1010

* Values reported in Ref. 12 incorrectly assumed that +V3 surfaces always face direct sun.

Because the Bay 10 MLI surfaces, which approximately faced the –V2 direction, were at an oblique angle to the sun, Bay 10 MLI retrieved at SM3A actually received less equivalent solar exposure than the SM2 sample retrieved 2.8 years earlier. However, the SM3A BM-R1 MLI experienced many more thermal cycles, and higher electron and proton radiation, and atomic oxygen fluence than the SM2 sample. The exposure levels for various retrieved materials are affected by the solar activity, and it should be noted that there was a solar minimum between the SM2 to SM3A time period. More details of environmental exposures are provided in [8, 12].

4.0 Experimental Procedures Exposures: 4.1 X-RAY EXPOSURE (PRISTINE FEP) A modified X-ray photoelectron spectroscopy (XPS) facility was used to irradiate the pristine FEP tensile samples. A copper target was irradiated with a 15.3 kV, 30 mA electron beam producing Cu x-rays (Cu Kα at 8048 eV, Cu L at 930 eV). The tensile samples were located 30.5 mm from the target, and the Cu x-rays were filtered through a 2 µm Al window (part of the x-ray tube). A 25 mil (635 µm) thick beryllium filter was placed over the FEP samples to absorb the low energy Cu L components, which would contribute significantly to damaging only the surface [13]. The x-ray flux was 13.28 W/m2[14]. The choice of target material, electron beam energy, and filter was chosen to produce a high flux, uniform distribution of energy absorbed, versus depth in the film. The energy deposition rate, or dose rate, versus depth below the surface for 127 µm FEP film at the specified exposure conditions are provided by de Groh and

161 Gummow in [15]. The technique used to characterize the x-ray source and energy deposition within the FEP film is described by Pepper and Wheeler in [13]. Pepper et al provide quantitative characterization of the Cu x-ray source and the absorbed energy deposition rate within a 75 µm film in [14]. X-ray irradiated samples were stored under vacuum until they were tensile tested or vacuum heat-treated. 4.2 VACUUM HEAT TREATMENT Pristine FEP, x-ray exposed FEP and HST SM3A Al-FEP were vacuum heat treated from 50 °C to 200 °C in 25° C intervals in a high vacuum facility adapted with a tube furnace. A Teflon lined Cu pipe was placed inside the tube furnace to promote uniform heating. The exposure temperature was monitored with a thermocouple attached to a Teflon witness sample, held in contact with the test samples. The pressure was 10-6 to 10-7 torr during heating. Samples were heated at the desired temperature for a target of 72 hours. 4.3 TENSILE PROPERTIES Samples for tensile testing were ‘dog bone’ shaped and die-cut using a tensile specimen die manufactured according to ASTM D638-95, type V. The tensile samples were 3.18 mm wide in the narrow section (neck), with a 9.5 mm gauge length. Samples were tested using a bench-top tensile tester with a 4.54 kg load cell and a test speed of 1.26 cm/min. Ultimate tensile strength (UTS) and elongation at failure were determined from the load displacement data. 4.4 DENSITY MEASUREMENTS Density measurements were obtained using density gradient columns calibrated using glass float standards of known densities (± 0.0001 g/cm3). The density solvents used were carbon tetrachloride (CCl4, ρ = 1.594 g/cm3) and bromoform (CHBr3, ρ = 2.899 g/cm3). The presence or absence of the thin (1000 Å) aluminized coating (as removed by NaOH solution) was found to have no effect on the density of the Al-FEP samples.

5.0 Results and Discussion 5.1 ROOM TEMPERATURE TENSILE PROPERTIES 5.1.1 Pristine FEP The room temperature (23 °C) tensile data for pristine FEP are listed in Table 2. The UTS and percent elongation at failure for the pristine FEP, for an average of 13 samples, was 24.1 ± 1.5 MPa and 271.2 ± 16.9%, respectively. 5.1.2 HST SM3A Al-FEP The retrieved SM3A Teflon from HST, after 9.7 years in the space environment, is substantially degraded. The tensile results for the as-retrieved SM3A FEP are listed in

162 Table 2. If compared to the pristine FEP tested in this study, the UTS of the retrieved HST FEP has decreased from 24.1 to 13.9 MPa, and the elongation at failure has decreased from 271.2% to 55.3%. These correspond to decreases of 42.3% and 79.6%, respectively. These tensile properties provide insight into the damage mechanism of Teflon in space. Because the UTS decreased, with the decrease in elongation at failure of the space-exposed FEP, chain scission is identified as the primary degradation mechanism on HST. TABLE 2. Tensile Properties of Vacuum Heat-Treated Pristine, X-ray Irradiated and HST Retrieved Teflon FEP. Vacuum Heat Treatment Temperature Room Temperature 23 °C 50 °C (+/- 1 °C) 75 °C (+/- 2 °C) 100 °C (+/- 1 °C) 125 °C (+/- 1°C) 150 °C (+/- 3 °C) 175 °C (+/- 12 °C) 200 °C (+/- 4 °C)

Material

Number of Samples

UTS (Mpa)

% Elongation at Failure

Pristine FEP

13

24.1 +/- 1.5

271.2 +/- 16.9

X-ray FEP

10

17.1 +/- 1.5

212.7 +/- 31

HST Al-FEP

4

13.9 +/- 0.4

55.3 +/- 9.3

Pristine FEP

4

23.4 +/- 0.7

264.1 +/- 12.6 162.6 +/- 35.5

X-ray FEP

4

15.1 +/- 1.1

HST Al-FEP

4

13.9 +/- 0.3

46.5 +/- 4.1

Pristine FEP

6

22.5 +/- 1.5

259.6 +/- 14.2 134.9 +/- 46.3

X-ray FEP

8

15.3 +/- 0.2

HST Al-FEP

4

14.4 +/- 0.1

25.7 +/- 5.5

Pristine FEP

4

22.1 +/- 0.7

250.2 +/- 5.7 43.1 +/- 6.6

X-ray FEP

4

15.5 +/- 0.3

HST Al-FEP

4

14.5 +/- 0.2

10.4 +/- 1.1

Pristine FEP

4

22.5 +/- 0.8

254.7 +/- 8.3 23.8 +/- 4.9

X-ray FEP

4

15.8 +/- 0.2

HST Al-FEP

3

14.2 +/- 0.7

9.4 +/- 3.3

Pristine FEP

4

22.8 +/- 0.6

271.0 +/- 6.4 22.8 +/- 5.6

X-ray FEP

4

15.4 +/- 0.3

HST Al-FEP

4

14.2 +/- 0.3

8.7 +/- 4.2

Pristine FEP

3

22.1 +/- 0.8

287.4 +/- 11.1 15.2 +/- 3.5

X-ray FEP

4

16.4 +/- 0.8

HST Al-FEP

4

14.9 +/- 0.8

7.9 +/- 2.0

Pristine FEP

4

23.7 +/- 0.8

306.8 +/- 9.6

X-ray FEP

2

16.1 +/- 0.4

9.7 +/- 5.2

HST Al-FEP

3

14.5 +/- 0.5

4.5 +/- 0.9

5.1.3 X-Ray Irradiated FEP The x-ray exposure was not intended to simulate the full extent of damage occurring on Hubble, but to cause irradiation induced polymer damage, and still have enough elongation at failure remaining to see the effects due to heating. Based on a series of prior tests, it was determined that a 2-hour exposure would provide the desired reduction in tensile properties [15]. Prior tests also indicated that the maximum number of samples that could be uniformly exposed at a time was two. The samples were centered in a holder that provided a 2.0 x 2.0 cm exposure area (the tensile sample

163 gauge length is ≈1 cm). The total energy absorbed per unit area integrated through the full thickness (the areal dose, D) of the 127 µm film for the 2-hour exposure was 33.8 kJ/m2.[16]. Density tests were conducted on pieces sectioned from a 200 °C heated xray irradiated tensile sample, which indicated that the irradiation exposure was uniform across the length of the exposed area. After x-ray exposure, the UTS and percent elongation at failure, for an average of 10 samples, was 17.1 ± 1.5 MPa and 212.7 ± 31%, respectively. This is a 29.0% reduction in the UTS and a 21.6% reduction in percent elongation at failure due to irradiation embrittlement. Although it was not the goal of this study to try to simulate the extent of damage on HST with the x-ray exposure, it was decided to compare the areal dose for xray irradiated FEP with that experienced by the FEP on HST. The areal dose for the 5 mil thick HST SM3A FEP is provided in Table 3 [17]. The total areal dose for the SM3A BM-R1 FEP was 427.2 J/m2. It should be noted that the HST FEP has an elongation at failure of 55% with an areal dose of only 427 J/m2, while the x-ray exposed FEP was much less embrittled (213% elongation), after orders of magnitude higher areal dose (33,800 J/m2). The factors that could contribute to these differences include the extreme differences in the dose rates (i.e. time factor), the variation in ionizing species and energies, temperature differences during irradiation exposure and the contribution from thermal cycling on HST (52,550 cycles from –100 to +50 °C), and possibly, surface effects from atomic oxygen and UV exposure in space. This stresses the difficultly in conducting simulated space environment durability tests, and emphasizes the potential complication when conducting durability testing based strictly on expected mission fluence or dose values. TABLE 3. Areal Dose for HST SM3A FEP. SM3A BM-R1 MLI (Bay 10, -V2)

Areal Dose (J/m2)

X-rays, 1-8 Å

29.80

X-rays, 0.5-4 Å

0.72

Electrons, >40 keV

389.6

Protons, >40 keV

7.11

5.2 VACUUM HEAT TREATMENT 5.2.1 Tensile Properties The results of tensile tests for the pristine, ground-laboratory irradiated and HST FEP after vacuum heat treatment are listed in Table 2 along with the room temperature data. The data are graphed in Figure 4. There was no degradation in the tensile properties of vacuum heat-treated non-irradiated FEP, in fact, with this batch of FEP an increase in the percent elongation at failure was observed for the higher temperatures (175 & 200 °C). Although heat treatment did not cause much change in the UTS of x-ray irradiated FEP with vacuum heat treatment, there was a dramatic decrease in the percent elongation at failure, as can be seen in Figure 4. The elongation decreased from 212.7% at 23 °C to only 9.7% after 200 °C exposure. This corresponds to a 95% decrease. And as can be seen in the graph, there is a rapid decrease in the elongation from 23 °C to 100 °C, with near complete losses of elongation from 125 °C to 200 °C. Only two of the

164 original four x-ray samples heated to 200 °C could be tensile tested because two stuck together slightly together during heating and then broke during separation. Although the FEP retrieved from HST was significantly embrittled in its asretrieved condition, it became even more embrittled with vacuum heat treatment (even after vacuum heat treatment at 50 °C, the maximum on-orbit temperature). The spaceexposed HST FEP followed a similar trend as the ground-laboratory x-ray irradiated FE, showing little changes in UTS and decreases in elongation from 23 °C to 100 °C, with near complete loss of elongation with heating to 100 °C and higher. 350

% Elongation at Failure

300 250 200

Pristine 150

X-Ray

100

HST

50 0 0

25

50

75

100

125

150

175

200

Vacuum Heat Treatment Temperature (C) Figure 4. Percent elongation at failure of pristine, ground laboratory x-ray exposed and retrieved HST Teflon FEP as a function of vacuum heat treatment temperature.

5.2.2 Density The density data for the pristine, ground-laboratory irradiated and HST space irradiated FEP, at room temperature and after vacuum heat treatment, are listed in Table 4. The data are graphed in Figure 5. The standard deviation is given when more than one sample was measured and averaged. As can be seen in the graph, the density of the retrieved HST FEP is essentially the same as pristine FEP, and the room temperature xray irradiated FEP is just slightly more dense than pristine FEP, even though these irradiated samples are significantly embrittled. This indicates that although irradiation induces scission in the polymer chains, resulting in embrittlement, the actual packing of the chains is not affected by irradiation exposure. There were very gradual increases in the density with heating up to 75 °C for all samples. Significant increases started at 100 °C, with larger increases corresponding to higher temperatures. Although the density increased with temperature for all samples, larger increases occurred for the samples that had been irradiated either in

165 space or in the ground facility than for pristine FEP. These results are consistent with de Groh's previous studies that show pristine FEP increases in density with heating, but FEP from HST has greater increases in density for the same heat treatment (200 °C exposure, references [7] and [9]). This is attributed to irradiation-induced scission of bonds in space, which allows for greater mobility and crystallization upon heating than that which occurs with non-irradiated FEP. Previous x-ray diffraction studies verify that the increases in density correlate to increases in polymer crystallinity.7,9 The density results further support chain scission as the primary mechanism of degradation of FEP in the space environment. TABLE 4. Density Data of Vacuum Heat-Treated Pristine, X-Ray Irradiated and HST Retrieved Teflon FEP. HST FEP Pristine FEP X-Ray FEP (SM3A 2001 BM-R1) Temperature (°C) Std. Density Std. Density Std. Density (g/cm3) Dev. (g/cm3) Dev. (g/cm3) Dev. 23

2.1373

0.0011

2.1407

-

2.1376

0.0005

50

2.1379

-

2.1414

-

2.1376

0.0005

75

2.1379

-

2.1428

-

2.1389

0.0005

100

2.1393

-

2.1477

-

2.1407

-

125

2.1414

-

2.1585

-

2.1456

-

150

2.1473

-

2.174

-

2.1577

-

175

2.1507

-

2.1775

-

2.1647

0.0016

200

2.1631

-

2.1856

-

2.1696

0.0031

2.19 Pristine

2.18

X-Ray

3

Density (g/cm )

HST

2.17

2.16

2.15

2.14

2.13 0

25

50

75

100

125

150

175

200

Vacuum Heat Treatment Temperature (C) Figure 5. Density of pristine, ground laboratory x-ray exposed and retrieved HST FEP as a function of vacuum heat treatment temperature.

166 When comparing the curves for the elongation and density data, it was observed that in each set of data there appeared to be a noticeable change in the slope of the data around 100 °C. The data was therefore graphed with linear fits for two sections of the data. The lines chosen were based on the best fit for each individual section of data. The resulting curves for the elongation at failure and density data are shown in Figures 6 and 7, respectively. The “change-of-slope” temperature has been highlighted in these graphs at the intersection of the two linear fits. The change-of-slope temperature of the pristine FEP (115 °C for the elongation data and 126 °C for the density data) correlates well with the glass I transition temperature (α relaxation), which is listed from ≈83 °C to 150 °C in the literature, dependent on hexafluoropropylene (HFP) content [9]. Eby and Wilson report transition temperatures for FEP with densities (2.136-2.135 g/cm3) similar to the pristine FEP examined in this report at ≈150 °C and ≈127 °C for 10.7 and 17.7 mol % HFP, respectively [18]. Commercially available FEP is reported to be 20 mol % HFP[19], which would indicate that the transition temperature for pristine FEP would be close to 125 °C based on the Eby study, which is consistent with the change-of-slope temperatures for the pristine FEP. Another interesting observation is that the irradiated samples have lower change-of-slope temperatures than pristine FEP. For example, the temperature in which the density of the ground-laboratory irradiated FEP starts to increase quickly is 82 °C, while it is 100 °C for the HST retrieved FEP and 126 °C for pristine FEP. These results indicate that irradiation causes changes in the polymer structure allowing increases in crystallization to occur at a lower temperature than which it occurs in pristine FEP.

6.0 Summary and Conclusions The objective of this research was to determine the effects of heating on ground laboratory irradiated FEP and FEP retrieved from the Hubble Space Telescope, in order to better understand the effect of temperature on the rate of degradation, and on the mechanism of degradation, of this insulation material in the LEO environment. Samples of pristine FEP, x-ray irradiated FEP and HST SM3A-retrieved FEP were heated from 50 °C to 200 °C in 25°C intervals in a high vacuum furnace and evaluated for changes in tensile properties and density. Results indicate that although heating does not degrade the tensile properties of non-irradiated Teflon, there is a significant dependence on the degradation of the percent elongation at failure of irradiated Teflon as a function of heating temperature, with dramatic degradation occurring at 100 °C and higher exposures. The density of non-heated irradiated FEP (ground or space irradiated) was essentially the same as pristine FEP, although these samples are significantly embrittled. This indicates that irradiation induces scission in the polymer chains, resulting in embrittlement, but chain packing is not affected. Gradual increases in the density occurred with heating from 23 °C to 75 °C for all samples, with significant increases occurring at 100 °C and higher exposures. Larger increases occurred for the irradiated samples than for the pristine FEP. These results were consistent with previous studies that show pristine FEP increases in density with

167 heating, but irradiated FEP experiences greater increases for the same heat treatment. This is attributed to irradiation-induced scission of bonds, which allows for greater mobility and crystallization upon heating than that which occurs with non-irradiated FEP. Changes in the rate of degradation were present in both elongation and density data. The change-of-slope temperatures of the pristine FEP (115 °C and 126 °C, for elongation and density, respectively) correlate with the glass I transition temperature of FEP. The change-of-slope temperature of irradiated FEP was lower than for pristine FEP, further indicating that scission damage has occurred. The tensile results and heated density data support chain scission as the primary mechanism of degradation of FEP in the space environment. The results show the significance of the on-orbit service temperature of FEP with respect to its degradation in the LEO space environment.

Figure 6. Change in the slope of the percent elongation at failure data of pristine, ground laboratory x-ray exposed and retrieved HST FEP as function of vacuum heat treatment temperature.

168

Figure 7. Change in the slope of the density data of pristine, ground laboratory x-ray exposed and retrieved HST FEP as function of vacuum heat treatment temperature.

7.0 Acknowledgments We would like to thank Dr. Stephen Pepper and Dr. Donald Wheeler of GRC for the use of, and characterization of, their x-ray facility. We thank Ed Sechkar of QSS, Inc. for build-up of the vacuum furnace facility. We appreciate technical contributions from Bruce Banks of GRC, and we thank Joyce Dever of GRC for providing areal dose values for the HST FEP. We would like to acknowledge John Blackwood and Jackie Townsend of NASA GSFC, and Ben Reed of Swales Aerospace, and the HST Project Office for providing the retrieved HST material for this study.

8.0 References 1. 2. 3. 4. 5. 6. 7. 8.

P. A. Hansen, J. A. Townsend, Y. Yoshikawa, D. J. Castro, J. J. Triolo, and W. C. Peters, (1998), SAMPE International Symposium, 43, 570. J. H. Henninger, (1984), NASA RP 1121. K. K. de Groh and D. C. Smith, (1997), NASA TM 113153. T. M. Zuby, K. K. de Groh, and D. C. Smith, ESA WPP-77, 385 (1995); NASA TM 104627, Dec. 1995. M. Van Eesbeek, F. Levadou, and A. Milintchouk, (1995), ESA WPP-77, 403. J. A. Townsend, P. A. Hansen, J. A. Dever, K. K. de Groh, B. A. Banks, L. Wang and C. He, High Perform. Polym. 11, 81-99 (1999). K. K. de Groh, J. A. Dever, J. K. Sutter, J. R. Gaier, J. D. Gummow, D. A. Scheiman and C. He, High Perform. Polym. 13, S401-S420 (2001). J. A. Dever, K. K. de Groh, R. K. Messer, M. W. McClendon, M. Viens, L. L. Wang and J. D. Gummow, High Perform. Polym. 13, S373-S390 (2001).

169 9.

K. K. de Groh, J. R. Gaier, R. L. Hall, M. P. Espe, D. R. Cato, J. K. Sutter and D. A. Scheiman, High Perform. Polym. 12, 83-104 (2000). 10. B. A. Banks, K. K. de Groh, T. J. Stueber, E. A. Sechkar, and R. L. Hall, SAMPE International Symposium, 43, 1523 (1998); also NASA TM-1998-207914/REV1. 11. J. R. Blackwood, J. A. Townsend, P. A. Hansen, M. W. McClendon, J. A. Dever, K. K. de Groh, B. B. Reed, C. C. He and W. C. Peters, SAMPE 2001 Conference Proceedings, May 6-10, 2001, Long Beach, CA, pp. 1797-1810. 12. J. A. Dever, K. K. de Groh, B. A. Banks, J. A. Townsend, J. L. Barth, S. Thomson, T. Gregory and W. Savage, High Perform. Polym. 12, 125-139 (2000). 13. S. V. Pepper and D. R. Wheeler, Review of Scientific Instruments, Vol.71, No. 3, March 2000, 15091515. 14. S. V. Pepper, D. R. Wheeler and K. K. de Groh, Proceedings of the 8th ISMSE & 5th ICPMSE Conference, June 5-9, 2000, Arcachon, France. 15. K. K. de Groh and J. D. Gummow, High Perform. Polym. 13, S421-S431 (2001). 16. S. V. Pepper, NASA Glenn Research Center, personal communication (1999). 17. Joyce Dever, NASA Glenn Research Center, personal communication (2002). 18. R. K. Eby and F. C. Wilson, J. of Applied Physics, 33, 2951-55 (1962). 19. Don Farrelly, DuPont, personal communication (1999).

This page intentionally left blank

ON THE THERMAL APPLICATION

STABILITY

OF

POLYIMIDES

FOR

SPACE

C.O.A. SEMPRIMOSCHNIG, S. HELTZEL, A. POLSAK, M. V. EESBEEK Materials Physics and Chemistry Section European Space Research and Technology Centre (ESTEC), European Space Agency (ESA), Keplerlaan 1, PO Box 299, NL-2200 AG Noordwijk, The Netherlands Abstract Currently planned missions of ESA (European Space Agency) to the inner part of the solar system will require the use of materials at an extreme radiation/temperature environment. This paper deals with the investigation of the thermal stability of two types of polyimides at a temperature of 350°C. Both materials were assessed by TGA (Thermo Gravimetric Analysis). Further tests were conducted in a high vacuum facility at 350°C. Test data were gathered up to a duration of 2500 hrs. The thermal stability was assessed by mass loss measurements and by UV/VIS/NIR spectrophotometric transmission measurements. Furthermore the degradation behaviour of the thermooptical properties versus time at this temperature was characterised.

1. 0 Introduction ESA (European Space Agency) is currently planning missions to the inner solar system. The first of such missions named Bepi-Colombo was approved at the end of 2000 as the fifth cornerstone mission of the Horizon 2000 science programme. This mission aims at a comprehensive exploration of the innermost terrestrial planet and aims to find answers about the understanding of planetary formation as well as the evolution in the hottest parts of our solar system [1]. As the environment closer to the sun will be harsher in terms of the impinging solar radiation (particle, UV etc.) such missions set a major challenge to materials and processes (M&P) and will require design solutions that are somewhat different to solutions used for spacecraft (s/c) in Earth orbit applications. The solar irradiance is inverse proportional to the square of the distance from the sun. At the Mercury perihelion for instance at about 0.3 AU it is more than a factor of ten above the averaged earth solar constant of about 1.3 kW/m2. Such high solar irradiances will naturally increase the temperature of any s/c in that environment. For the materials engineer it is therefore important to predict and understand the behaviour of materials at such an environment/temperature. This requires understanding the limits of the available materials, what drives their degradation

171

172 mechanisms and whether a material could fulfil its application in a certain environment within the s/c design life. This triggered us to adopt a certain testing & analysis philosophy within the Materials & Processes division. This philosophy is sketched in fig. 1 and has been presented in a previous paper [2]. In this paper results about the thermal stability of two different types of polyimides are presented. Polyimides have been widely characterised in the past and are well known for their thermal stability [3,4,5,6] and their intriguing mechanical properties [7]. Such materials are commonly used on the outside of s/c for thermal control purposes and are good candidates for sunshields [8,9]. The temperatures that such an external layer will reach depend on its solar absorptance (α) versus thermal emittance (ε) ratio, i.e. the ratio between the amount of the absorbed solar energy and the amount of the thermally emitted radiation. This socalled α/ε ratio can be varied in a wide area with various coatings. For inner solar system missions we want to assess the behaviour of such polymers at extreme temperatures. Figure 2 shows equilibrium temperatures as a function of the α/ε ratio for three different distances. The first would be at 1 Astronomical Unit (AU), the second APPROACH OF MATERIAL TESTING AND ANALYSIS

Materials

Degradation Tests

Property Analysis

High Temperature Resistance

Decomposition resistance

Adhesives

Structural adhesion

Mechanical Properties

Optical coupling

Optical stability (transmission)

Thermal Control Materials

Paints

Optical Solar Reflectors

Foils

Irradiation test Thermo-Optical

in combination with high temp resistance

(α/ε)

Accelerated UV radiation Accelerated X-ray, γ and particle radiation

Mechanical Properties (tensile, stress/strain)

Figure. 1. Approach to materials testing & analysis for inner solar system missions

173

Surface Temperature (sun facing) [°C]

500 0.31 AU

400

300

0.47 AU

200 1 AU

100

0

-100 0,0

0,2

0,4 0,6 Alpha/Epsilon ratio

0,8

1,0

Figure 2. Surface equilibrium temperature as a function of α/ε ration for 1, 0.47 and 0.31 AU.

around 0.47 AU and the third around 0.31 AU. The latter two correspond to the aphelion and the perihelion of Mercury. It is worth noting that this graph does not include any contribution of infra-red radiation or planetary albedo that will increase surface temperatures in low planetary orbits. 2.0 Experimental Approach on Assessing the Thermal Stability Various experimental thermal analysis techniques exist that can be used for this purpose. The thermal stability can be most easily measured by TGA that measures mass loss as a function of time and/or temperature. The change in mass is an indication of the thermal stability of the material. Experimentally this can be assessed in two different ways. By measuring in an isothermal mode precise data about the decomposition can be gathered. Results from such tests can be used to derive mathematical models in order to describe decomposition kinetics. The disadvantage of this method is that it is time intensive and test duration is naturally limited. The second method relies on heating the sample with a (mostly) constant heating rate. Testing a material with various heating rates offers the possibility to determine activation energies and pre-exponential factors. This enables the development of life-time models such as the one proposed in ASTM E 1641 [10]. Such models show a good correlation when single decomposition reactions occur and when higher heating rates do not activate normally covered reactions.

174 We have screened both materials by tests applying a constant temperature ramp as well as by isothermal ageing tests. The latter technique has however two limitations for being applicable to space missions. First it is generally not done under vacuum conditions. This reduces the outgassing potential. Second long term testing is naturally limited on one instrument. To overcome this we have built a so-called high temperature exposure system (HITES) that allows exposures of up to 50 samples at elevated temperatures under high vacuum conditions. To study the thermal endurance we have aged the polyimides thermally for more than 2500 hrs. Materials investigations focussed on determination of the mass loss and the optical (UV/VIS/NIR) and thermo-optical properties versus time. To cover a wide range of inner solar system missions at an extreme temperature on the one hand but to stay within a reasonably long duration for a defined service temperature on the other hand we have selected to perform thermal endurance tests at 350°C.

3.0 Experimental Results 3.1. MATERIALS UNDER INVESTIGATION The materials under investigation were supplied by Du Pont and by UBE Industries. They are commercially available under the name Kapton (K) and Upilex  (U). We have used various film thicknesses (with and without VDA (Vacuum Deposited Aluminium) coating) between 7.5 µm and 50 µm for our investigations. An overview is given in the table below. TABLE 1. Overview about materials used and investigations performed.

Material K-film 7.5µm K-film 25 µm K-film 7.5 µm/VDA K-film 50 µm/VDA U-film 7.5 µm U-film 25 µm U-film 7.5 µm/VDA U-film 25 µm/VDA

Investigation TGA, HITES, UV/VIS/NIR Transmission, TGA, HITES, UV/VIS/NIR Transmission TGA, HITES, Thermo-optical measurements TGA, HITES, Thermo-optical measurements TGA, HITES UV/VIS/NIR Transmission TGA, HITES, UV/VIS/NIR Transmission TGA, HITES, Thermo-optical measurements TGA, HITES, Thermo-optical measurements

3.2 Assessment of Thermal Stability by Thermal Analysis

For the mass loss experiments we used uncoated materials. In fig 3 a comparison of two TGA test results for both materials are shown. Testing was done according to ISO 11358:1997 [11]. A heating rate of 10K/min was applied and 5.0 dry nitrogen was used

175 as a purge gas with a flow rate of 55 ml/min. As can be seen the two materials show no apparent mass loss up to 300°C. The 5 % mass loss figure, the T5, was determined to be around 597°C for the K film and 634°C for the U-film. Furthermore the first derivative of the mass loss is also shown in that graph. It uses the right axis of the graph. We found the maximum rate of decomposition for the K-film around 616°C and for the U-film at 655°C. Figure 4 shows an enlargement of the TGA recordings between 200°C and 600°C. As can be seen above 300°C both films can be discriminated. The higher the temperature the bigger the difference gets. 0,2 Mass %

100

0,1

Derv. Mass %/T

MASS [%]

-0,1 -0,2

80

-0,3 70 U-film

-0,4

K-film

-0,5

DERV. MASS [%/°C]

0 90

60 -0,6 50 0

200

400

[°C]

600

800

-0,7 1000

Figure 3. Comparison of thermal stability between K-film and U-film by TGA. 100

Mass [%]

99,5

99 U-film

K-film

98,5

98 200

300

400

T [°C]

500

Figure 4. Comparison of thermal stability between K-film and U-film by TGA.

600

176 Figure 5 shows a compilation of two TGA isotherm runs of 25µm thick films at 350°C. Tests were done again according to ISO 11358:1997 and 5.0 dry nitrogen was used as purge gas with a flow rate of 55 ml/min. Again the films can be discriminated. Even at the longest isotherms of the U-film that lasted nearly 140 hrs only a marginal mass loss was detected. We determined the mass loss to be around 0.1% for the U-film and on the other hand we found a noticeable mass loss of about 0.3 % for the K-film material. Thus we observed in the same test duration about three times higher mass losses for the K-film material compared to the U-film material. 100,0%

mass%

99,9%

U-film (10,1 mg) K-film (8,8 mg)

99,8%

99,7% 0

50

t [hours]

100

150

Figure 5. Isothermal TGA tests at 350 °C, comparing thermal stability between K-film and Ufilm.

3.3 ASSESSMENT OF THERMAL STABILITY IN THE HITES The HITES is a custom built facility that enables to expose up to 50 samples at elevated temperatures under high vacuum conditions. Samples are sliding into a metallic samples holder and temperature is measured at several locations within the facility. We have found that we were able to control the set temperature of 350°C within ± 3°C over the duration of the test program. Vacuum levels within the facility were maintained below 2x10-5 mbar and reached bottom levels of high 10-8 mbar values after long exposure durations. The experiments performed within the HITES used the same type of uncoated materials as before for the mass loss and UV/VIS/NIR transmission analysis. For the thermo-optical analysis one-sided VDA coated materials were used. The thermo-optical properties were determined on the polyimide side as the front layer and were performed according to ESA-PSS-01-709 [12]. To determine the α we used a UV/VIS/NIR spectrophotometer (Cary 5) and measured in the wavelength range of 250 nm and 2500 nm. To measure the thermal emittance we used a Gier Dunkle DB100. Measurements were performed under

177 atmospheric condition outside the test facility. This required cooling the samples down to below 25°C and flushing it with dry nitrogen before opening and recovering the test samples. The total test duration of this test was 2540 hrs. After the initial determination of the beginning of life (BOL) values and properties the test was stopped three times for intermediate inspections/measurements. The first was done after approximately 172 hrs, the second after about 872 hrs and the third after 2540 hrs. Mass loss was determined with an ultra-microbalance that is able to resolve 100 ng. This balance is placed within a continuously dry nitrogen purged cabinet. Temperature and humidity are recorded before and after measurements. As measuring mass loss at such a resolution is a very delicate operation we had special reference samples that were weighed together with our test samples at each individual break. This was done to see whether variations within the reference material could be responsible for arbitrary mass losses or gains. The latter is commonly attributed to the adsorption and absorption of water. 3.3.1. Results on mass loss measurements The results gathered with the dedicated mass loss samples were conclusive in so far that we could discriminate both film materials. We observed a noticeable mass loss on the K-film samples (above 1%) and no significant change on the U-film material after the first inspection point at 172 hrs. Comparing these results with the test results of the TGA isothermal tests at 350°C shows that for approximately same durations (140 hrs versus 172 hrs) we find an increased outgassing/decomposition contribution on the HITES samples. The amount of mass loss on the K-film did however only marginally increase with time. The U-film material seemed to be stable and the highest variation was sometimes observed on the reference samples itself. We even noted sometimes increases on the mass of the aged test samples. One explanation could be that the aged material shows increased water adsorption that cannot be avoided during transfer of the sample. This could be facilitated for instance by a change in the surface morphology. To clarify this point we intend to perform further AFM images of the surfaces of the aged materials. 3.3.2. Results on UV/VIS/NIR transmission stability during isothermal ageing The transmission of dedicated transmission samples was measured BOL and at each of the three inspection points. A compilation of recorded values for 7.5 µm U- and K-film samples are shown in Figure 6. It shall be noted that each individual graph represents averaged values of three different samples.

178 100

Transmittance [%]

80 1

2

60

3

1

K7 - BOL

U7 - BOL

2

K7 - 172 h

U7 - 172 h

3

K7 - 872 h

U7 - 872 h

4

K7 - 2540 h

U7 - 2540 h

4

40

20

0 250

750

1250

λ [nm]

1750

2250

Figure 6. Compilation of UV/VIS/NIR transmission data of 7.5 µm K and U films vs. isothermal ageing duration at 350 °C

As can be seen the K-film shows a higher transmission than the U-film at the BOL. This was visually verified by the colour difference between the two films. The Ufilm appears darker at the BOL. At the first inspection break after 172 hrs a difference in the visible range can be noted already. The K-film increases its absorptance and shows already lower transmission behaviour in the visible range. This trend continues with time as can be deduced by the measurements performed at the next two breaks. Additionally it is worth noting that the cut-off wavelength of the K-film material shifts with increasing ageing duration whereas the cut off wavelength of the U-film material is shifting less. 3.3.3. Results on thermo-optical stability during isothermal ageing The results of thermo-optical properties of various samples measured between the BOL, after 172 hrs, 872 hrs and 2540 hrs of isothermal ageing time is shown in Figures 7 and 8. The first one shows the stability of the α vs ageing time of four test samples. Again the values show averaged data of three different samples. If we look at the two directly comparable 7.5 µm thick films one can see that the initial α at the BOL were nearly identical. At the next two inspection points the α’s of both films are noticeably different. This trend seems to increases and after 2540 hrs a

179 clear distinction can be made. The other two graphs give values for a 50 µm K-film and a 25 µm U-film. Even though the other two curves are not directly comparable due to a different thickness they show a similar trend. It is worth noting that the indicated trend lines show similar slopes for the two different films and the U-films shows a lower slope. This confirms the previously made measurements. The results of the thermal emittance measurements versus time are shown in Table 2. As was expected only a marginal change in the thermal emittance can be noted. TABLE 2. Thermal emittance data versus isothermal ageing duration at 350 °C

Material

U-film 7.5 µm

U-film 25 µm

0 hrs (BOL) 172 hrs 872 hrs 2540 hrs

0.444 0.446 0.442 0.449

0.639 0.638 0.614 0.620

K-film µm 0.476 0.464 0.454 0.465

7.5

K-film 50 µm 0.778 0.743 0.719 0.732

0,8 K-film - 50 um

0,7

Solar absorptance

U-film - 25 um K-film - 7 um

0,6

U-film - 7 um

0,5

0,4

0,3 0

500

1000 1500 ageing time [hours]

2000

2500

Figure 7. Compilation of solar absorptance of 7.5 µm K- and U- films vs. isothermal ageing duration at 350 °C

A plot of the α/ε ratio versus time is shown in Figure 8. It basically shows a similar picture as Figure 7, the U-film material tends to be more stable and degrades at a slower rate.

180 1,6

K-film - 7 um

1,4

Alpha / Epsilon

U-film - 7 um 1,2

U-film - 25 um 1,0

K-film - 50 um

0,8

0,6

0,4 0

500

1000 1500 ageing time [hours]

2000

2500

Figure 8. Compilation of α/ε ratio of K- and U- films vs. isothermal ageing duration at 350 °C

4.0 Conclusions In this paper results about the thermal stability at 350 °C of two different polyimides were presented. The thermal mass loss was assessed by various TGA measurements and it was found that the U-film material shows a higher thermal stability than the K-film. These results were confirmed by thermal endurance tests by TGA measurements. Further tests were conducted in a high vacuum facility at 350°C. Test data were gathered up to a duration of 2500 hrs. The thermal stability (mass loss) was again found to be lower for the K-film material. We also noted a significant vacuum effect on the Kfilm. Further UV/VIS/NIR transmission and thermo-optical measurements revealed the degradation behaviour of both films versus time at 350 °C. The test results gathered indicate that the U-film material tends to be more stable and that the space relevant thermo-optical properties degrade at a slower rate.

5.0 Acknowledgement The authors would like to acknowledge J. Sorenson for providing environmental data for the Mercury mission and A. Santovincenzo and H. Ritter for discussing thermal control aspects of the Mercury mission.

181 6.0. References 1.

2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.

“BepiColombo, An Interdisciplinary Cornerstone Mission to the Planet Mercury”, System and Technology Study Report, ESA-SCI (2000)1, April 2000, European Space Agency, ESTEC, Noordwijk, The Netherlands Semprimoschnig C.O.A., S. Heltzel, A. Polsak, “Materials Behaviour at Mercury – Challenges and First Experimental Results”, 33rd SAMPE Conf., Seattle, USA, 2001 Yokota R., “Recent trends and space applications of polyimides”, J. of Photopolymer Sc & Techn., Vol 12, 2, 209, 1999 Traeger et al., “Thermal Aging of polyimide films”, Polym Preprints 12, 292 (1971) Ortelli et al, “Pyrolysis of Kapton in Air: An in situ DRIFT Study, Appl. Spectr. 4, Vol 55, 412, (2001) Matsumoto T., “Nonaromatic polyimides derived from cycloaliphatic monomers, Macromolecules, 32, 4933, 1999 Hasegawa et al., “Structure and properties of novel asymmetric biphenyl type polyimides. Homo and copolymers and blends, Macromolecules, 32, 387, 1999 Russell et al, ‘Simulated space environmental testing on thin films”, NASA/CR-2000-210101, April 2000 Wooldrige et al, “Effects of manufacturing and deployment on thin films for the NGTS sunshade, AIAA, 2001-1349 ASTM E 1641-99, American Society for Testing and Materials, USA, 1999 ISO 11358:1997, International Organisation for Standardisation, PO 56, CH-1211 Geneva 20, Switzerland ESA PSS-01-709, ESA/ESTEC, Noordwijk, The Netherlands, July 1984

This page intentionally left blank

SYNERGISTIC DEGRADATION OF CV-1144-O DUE TO ULTRAVIOLET RADIATION AND HEAT JOSEPH E. HAFFKE JOHN A. WOOLLAM Center for Microelectronic and Optical Materials Research, Department of Electrical Engineering, University of Nebraska-Lincoln, Lincoln, NE 68588-0511

Abstract Nu-Sil’s CV-1144-O is a silicone-based co-polymer used on spacecraft in LEO. CV1144 has been spin-coated onto platinum sputter coated silicon wafers, and simulated LEO experiments made to study degradation. First, samples were exposed to ultraviolet radiation from a deuterium lamp with no additional heating. Next, samples were heated to 100°C during which there was no radiation exposure. Finally, samples were exposed simultaneously to ultraviolet radiation and heating. Variable angle spectroscopic ellipsometry (VASE£) was used to determine optical constants from vacuum ultraviolet (VUV) to mid-infrared (IR). VUV SE data were analyzed to determine thickness and optical constant changes, while IR data provided bonding-chemistry information. Multi-sample analysis of optical data eliminated correlation when determining polymer optical constants and thickness. Heating alone did not alter optical properties. Small changes were seen when exposed to ultraviolet radiation alone. Significantly larger changes occurred when the sample was exposed to ultraviolet light while being heated. Results show that ultraviolet radiation breaks apart benzene rings in the co-polymer, a process enhanced by heating.

1.0 Introduction Low earth orbit (LEO) exposure of space systems to atomic species, charged particles, temperature fluctuations, electromagnetic radiation, and debris are considered when designing spacecraft materials capable of long-term LEO exposure [1-8]. Protective coatings can protect space system surfaces from atomic oxygen destructive effects, and have been studied for years [9,10]. Recent attention has focused on polymeric protective layers, due to ease of application. Inorganic siloxane polymers, which convert to protective silica-like coatings, yield erosion rates one to two orders of magnitude lower than carbon based polymers when exposed to atomic oxygen [2, 9, 11]. CV-1144-O is a silicone dispersion developed by NuSil Technology [12] proposed

183

184 for use as a protective barrier in LEO [10]. It is based on a dimethyl diphenyl silicone co-polymer with chemical structure shown in Figure 1.

CH 3

95.2%

Si

+

O

CH

3

4.8%

Si

O n

m Figure 1

Chemical structure of CV-1144-O silicone-based copolymer.

Qualification of materials for use in space is difficult. Although in-flight testing for candidate materials is possible, it is usually time-and cost-prohibitive therefore laboratory simulation is important to development and testing of coatings [2,13,14]. In previous work we studied the effects of atomic oxygen on CV-1144-O [2,15]. In this study, the affects of vacuum ultraviolet radiation and thermal cycling on CV1144-O co-polymer were investigated, both independently and simultaneously. The optical properties of the copolymer were examined using variable angle spectroscopic ellipsometry (VASE) from vacuum ultraviolet (V-UV) to mid-infrared (IR).

2.0 Experimental 2.1 MATERIAL PREPARATION First optically thick platinum films were prepared on polished silicon wafers by magnetron sputtering. Thin films of CV-1144-O co-polymer were then spin-coated onto the smooth platinum substrates. CV-1144-O was thinned with Naphtha in different dilutions to lower the polymer viscosity for spin-coating. By changing the revolutions per minute during spin-coating and the ratio of polymer to solvent, uniformly distributed samples of unique thickness were obtained. After spin coating, and during drying the Naphtha evaporated [10]. Five nominally identical samples were prepared, labeled A, B, C, D, and E. 2.2 EXPOSURE EXPERIMENTS Thin films of CV-1144-O were subjected to heating and/or ultraviolet (UV) radiation in a simulation chamber capable of a base pressure near 1x10-6 torr. Samples were heated by placing them on a copper plate which was on top of a Kapton heating element with thermocouple-based temperature control. UV exposure was from an RF powered deuterium lamp, mounted in the vacuum chamber directly above the substrate. Three experiments were performed. First, the polymer was heated to 100°C with no ultraviolet radiation. Such heating might come from spacecraft equipment, the sun,

185 or other sources. In the second experiment a sample was exposed to vacuum ultraviolet radiation with no additional heating, as UV light is a strong component of sun radiation in LEO. The third experiment used a combination of heat and UV, as this better represents the actual space system in LEO environment. These experiments allowed study of cooperative effects of the two sources, UV light and heating. 2.3 ELLIPSOMETRY Variable angle spectroscopic ellipsometry data were acquired prior to and after each exposure condition. In ellipsometry the change of the polarization state of light on reflection from the surface is measured, and results displayed as psi (ψ) and delta (∆), which are related to the Fresnel reflection coefficients [16]. Optical constants and film thicknesses were determined from sets of measured ψ and ∆ values using regressionanalysis. The mean square error of the fit between experimental Ψand ∆ values compared to those calculated from Fresnel equations is minimized using the LevenbergMarquardt algorithm and varying the model parameters [17]. The optical parameters were then checked for correlation to make certain solutions were indeed unique and representative of each sample. The optical constants are expressed as ε1 and ε2, the real and imaginary components of the complex optical dielectric function of the material [16]. Spectroscopic data were taken from 0.75 to 9 eV, and 300 to 6000 cm-1, using vacuum-ultraviolet (VUV) and infrared (IR) ellipsometers, respectively. Ellipsometric data were taken every twenty hours, for each experiment, to establish trends in the optical constants or thickness. Data collected from the VUV and visible ranges helped determine thickness and optical constant changes, while IR data provided information on chemistry, and chemistry-related modifications of exposed samples. Multi-sample analysis was performed to minimize parameter correlation. In this procedure, five samples of varying thickness had variable optical model parameters coupled together assuming a Cauchy dispersion model in the regression. The Cauchy parametric model was used in the non-absorbing (300—700 nm) region of the spectrum This method forced the regression-determined optical parameters of all samples to be the same, and permitted an independent fit for film thickness of each sample along with the Cauchy parametric model parameters for the optical constants and dispersion representing all samples. Parameter coupling for the unique samples during optical model fit successfully eliminated correlation. Results are found in Table 1. 3.0 Results and Discussion 3.1 CV-1144-O—PRISTINE REFERENCE OPTICAL PROPERTIES Multi-oscillator models were developed to determine optical constants of the pristine polymer in both the VUV and IR regions where the polymer is not totally transparent. Both models contained mixtures of Gaussian and Lorentzian oscillators, whose dispersion parameters were used in regression fits of experimental data. In the VUV, a total of 4 oscillators provided exceptionally good fits to data. In the IR, 12 oscillators were used, and yielded excellent data fits. Oscillator parameters are given in Table 1.

186

TABLE 1. IR oscillators: Pristine, and after UV exposure with simultaneous heating to 100°C CENTER ENERGY

AMPLITUDE

BROADENING

Pristine

After 60hrs

Pristine

After 60hrs

Pristine

699.74

0

0.35

0

33.09

799.05

801.44

3.2

1.2

19.13

817.08

0

0.05

0

15.31

844.49

852.9

0.11

0.23

9.85

867.17

0

0.22

0

26.45

1019.4

1025.5

1.52

0.93

24.18

1047.2

0

0.85

0

70.45

1093.6

1083.8

1.09

0.98

1125.2

1153.9

0.17

1260.2

1261.9

1415.8 2962.3

Chemical ID

After 60hrs

35.82

C6H5 Methyl rocking & Si-C stretching Unidentified

27.77

Unidentified Unidentified

36.22

Si-O Stretch

42.1

97.88

0.06

13.2

63

1.31

0.58

9.9

15.49

1423.7

0.05

0.02

55.52

112.95

2961.7

0.11

0.05

36.52

63.93

Si-O Stretch Si-C6H5 Planar ring vibration Si-CH3 Symetric deformation Asymmetric CH3 deformation CH3

Si-O-Si

IR spectra oscillators were analyzed to determine chemical bonding and functionality [18,19]. Results are shown in Figure 2. Comparing oscillator locations in photon energy and relative strengths for all samples conclusions are made about chemical changes resulting from experiments performed.

4.0 CH3 rocking & Si—C stretching

3.0 ε2

Si—O stretch

2.0 1.0

Si—CH3 symmetric deformation

C6H5

CH3 asymmetric deformation CH3

0.0 0

1000 2000 3000 -1 Wave Number (cm ) Figure 2 Chemical identity of IR dielectric function peaks.

4000

187 3.2 CV-1144-O—AFTER HEATING TO 100°C Polymer sample A was heated to 100°C for 60 hours to determine changes in optical properties under thermal influence. Allowing the oscillator parameters from the pristine sample to vary in a regression fit with the new experimental data taken after 40 hours of heating, resulted in a new set of optical constants, ε1 and ε2. The absence of a wavelength shift of experimental data in the transparent spectral region indicates thicknesses were unchanged, and only small changes in the optical constants. In the same manner, the IR optical constants of the pristine sample were compared to those found after 60 hours of heating. ε1 and ε2 IR spectra before and after heating to 100°C are nearly identical. This evidence suggests that sample A is not chemically altered as result of heating alone.

3.3 CV-1144-O—IRRADIATING THE POLYMER WITH ULTRAVIOLET LIGHT Polymer sample B was exposed to ultraviolet radiation for 60 hours to observe possible optical property changes. New VUV optical constants were found by regression fit to experimental data taken after 20 hours of exposure. Significant changes in the optical constants were found, especially at photon energies higher than 5 eV. IR optical constants from the pristine reference sample were compared to the IR optical constants after 60 hours exposure. Close examination of Table 2 reveals a general decrease in optical constant amplitudes after exposure. Thus CV-1144-O is significantly affected by 20 hours exposure to ultraviolet radiation. TABLE 2. IR oscillators after UV exposure only CENTER ENERGY

AMPLITUDE

BROADENING

Pristine

After 60hrs

Pristine

After 60hrs

Pristine

699.74

695.57

0.35

0.23

33.09

After 60hrs 64.82

799.05

799.62

3.2

2.51

19.13

23.46

817.08

817.78

0.05

0.27

15.31

12.73

844.49

845.7

0.11

0.13

9.85

13.39

867.17

864.69

0.22

0.21

26.45

32.61

1019.4

1018.8

1.52

1.43

24.18

27.95

1047.2

1042.2

0.85

0.75

70.45

57.64

1093.6

1091.2

1.09

0.78

42.1

43

1125.2

1097.7

0.17

0.39

13.2

76.34

1260.2

1260.5

1.31

1.12

9.9

10.94

1415.8

1476.5

0.05

0.03

55.52

272.15

2962.3

2961.8

0.11

0.13

36.52

28.48

188 3.4 CV-1144-O—HEATING THE POLYMER TO 100°C WHILE EXPOSING IT TO ULTRAVIOLET RADIATION Polymer sample C was heated to 100°C during simultaneous exposed to UV radiation for 60 hours, to best simulate LEO conditions. Visible region experimental VUV data shifts results from a significant thickness reduction. Thickness after 20 hours exposure was determined by fitting experimental data in the non-absorbing spectral region (300— 700 nm) using a Cauchy dispersion model, similar to the method used to find thicknesses of the pristine references. Once the new thickness was determined, it was

3.2

(a)

3.0

After 20 hours of UV radiation and heat Pristine reference

ε1

2.8 2.6 2.4 2.2 2.0

0

1.2 1.0

2 (b)

6 4 Photon Energy (eV)

8

10

After 20 hours of UV radiation and heat Pristine reference

ε2

08 0.6 0.4 0.2 0.0 0

2

4 6 Photon Energy (eV)

8

10

Figure 3 Comparison of optical constants of sample C in the VUV-UV-Visible-NIR range before and after 20hr of UV radiation and heat. Dielectric function (a) ε1. (b) ε2.

189 used in a regression fit of the data. The optical constants, ε1 and ε2, obtained from this fit are compared to data from the pristine reference, and the results shown in figure 3. It is interesting that two additional oscillators were needed between 1 and 5 eV, to fit the experimental data in the VUV region. The presence of oscillator-like spectra indicate photon-induced transitions between new energy level sets generated by combined heating and UV exposure. The origins f these new spectra is unknown to us. The IR data were analyzed using a regression fit to the data taken after 60 hours

4.0

(a)

ε1

3.0 2.0 1.0 After 60 hours of UV radiation and heat Pristine reference

0.0 0 4.0

ε2

3.0

1000 2000 3000 -1 Wave Number (cm )

4000

(b) After 60 hours of UV radiation and heat Pristine reference

2.0 1.0 0.0 0

1000 2000 3000 -1 Wave Number (cm )

4000

Figure 4 Comparison of optical constants of sample C in the middle IR range before and after 60hr of UV radiation and heat. Dielectric function (a) ε1. (b) ε2.

190 exposure. The resulting ε1 and ε2 optical constants compared to those from the pristine reference, and results are shown in figure 4. Examination of IR data in table 1 reveals significant changes in peak amplitude and breadth, demonstrating major chemical changes. Most notable is disappearance of the mono-substituted benzene ring resonant signature seen in the pristine IR data at 699cm-1. A general trend of decreasing amplitudes is also observed, with the methyl associated peaks and Si—O stretching resonance peaks being the most affected. 4.0 Conclusions Heating the polymer to 100°C for 40 hours produces only very small changes in the VUV optical constants. Exposure of sample B to ultraviolet light without external heating produced significant change in chemical and optical properties of CV-1144-O. Although the visible region data exhibited no optical constant changes, substantial changes in the “B” material optical properties in the high photon energy range were seen after exposure. Decreasing peak heights seen in the IR ε2 spectra confirm that CV1144-O was chemically changed as result to 60 hours of exposure to ultraviolet radiation. The specific chemical bonds that were changed are listed in table 2. Most interesting, was extensive alteration of optical and chemical properties resulting from combined heating and UV irradiation of Sample C, where both optical properties and thickness were changed. In the VUV, a new model was needed to describe the material after 20 hours exposure due to large differences in experimental data. When compared to pristine film data, the new model required two additional oscillators between 2 and 5 eV to allow a good fit to data, indicating that two new sets of energy levels were created. Oscillator features in the UV are much more difficult to associate with specific chemistries as compared with IR resonance, thus the chemical origins of the new UV peaks are presently unknown to us. From the IR data summarized in table 2, the disappearance of several oscillators in the data shows that benzene rings in the CV-1144-O co-polymer are broken by UV radiation. It’s clear that exposing the sample to UV light and heating results in synergistic effects on material thickness and optical constants, not seen with heating or UV exposure alone. Chemical bond identifications are given in table 2 and show that benzene ring resonant spectra go away completely and others are drastically reduced in strength. Our explanation then is that UV radiation breaks apart benzene rings of the co-polymer, and heating enhances this process.

5.0 Acknowledgements This work was supported by the NASA Glenn Research Center, Grant NAG3-2219.

191 6.0 References 1. 2.

Tennyson, R. C., Can. J. Phys. (1991). 69, 1190 Yan, L., Bungay, C. and Woollam, J. A. “Surface chemistry changes and erosion rates for CV-1144-0 silicone under oxygen plasma and ultraviolet light exposure”, (2000) 5th international conference on “Protection of Materials and Structures from the LEO Space Environment”, European Space Agency. 3. Srinivasan, V., Banks, B.A., (1990) M,aterials Degradation in Low Earth Orbit (LEO), The minerals, Metals & Materials Society, Pennsylvania. 4. Kleiman, J. I., Tennyson, R. C., (1999) Protection of Materials and Structures from the Low Earth Orbit Space Environment, Space Technology Proceedings, Kluwer Academic Publishers, the Netherlands. 5. Zimcik, D. G. and Maag, C. R., (1998) J. Spacecr. Rockets 25, 162-168. 6. Banks, B. A., de Groh, K. L., Baney-Barton, E., Sechkar, E.. A., Hunt, P.K., Willoughby, A., Bemer, M., Hope, S., Koo, J., Kaminski, C. and Youngstrom, E. (1999) NASA Technical Memorandum 209180. 7. Dever, J. A., (1991), NASA Technical Memorandum 103711. 8. Banks, B. A., de Groh, K. K., Rutledge, S. K., and Difilippo, F. J.(1996) NASA Techinical Memorandum 107209. 9. Zimcik, D. G., Wertheimer, M. R., Balmain, K. B., and Tennyson, R. C. (1991) J. Spacecr. Rockets 28, 652-657. 10. Bungay, C. L., Tiwald, T. E., Thompson, D. W., DeVries, M. J., Woollam, J. A., Elman, J. F. (1998), “IR ellipsometry studies of polymers and oxygen plasma-treated polymers”, Thin Solid Films 313-314, 714. 11. Gilman, J. W., Schlitzer, D. S., and Lichtenhan, J. D. (1996), .J. Appl. Polym. Sci. 60, 591-596 12. NuSil Technology, 1050 Cindy Lane, Carpinteria, CA 93013, 805/684-8780. 13. Bungay, C. L., Synowicki, R. Spady, B., Hale, J. S., Woollam, J. A. (1999) “Laboratory Simulation of low earth orbit”, Protection of Materials and Structures from the Low Earth Orbit Space Environment, edited by J. I. Kleiman and R. C. Tennyson, Kluwer Academic Publishers, Norwell, MA. 14. Rutledge, S. K. and Banks, B. A., (1996), “A technique for synergistic atomic oxygen and vacuum ultraviolet radiation durability evaluation of materials for use in LEO”, NASA Technical Memorandum 107230. 15. Yan, L., Gao, X., Bungay, C., and Woollam J. A. (2001) “Study of surface chemical changes and erosion rates for CV-1144-O silicone under electron cyclotron resonance oxygen plasma exposure”, J. Vac. Sci. Technology A 19(2), 447. 16. Azzam, R. M. A. and Bashara, N. M., (1977) Ellipsometry and Polarized Light (North Holland, New York. 17. Jellison, Jr., G. E.,(1991) Appl. Opt. 30, 3354. 18. Colthup, N. B., Daly, L. H., and Wiberley, S. E., (1990) Introduction to Infrared and Raman Spectroscopy, Academic Press, San Diego. 19. The Sadtler Handbook of Infrared Spectra, (1978), edited by W. W. Simons, (Sadtler Research Laboratories, Inc, Philadelphia.

This page intentionally left blank

BEHAVIOUR OF THERMAL CONTROL COATINGS UNDER ATOMIC OXYGEN AND ULTRAVIOLET RADIATION

S. REMAURY, J.C. GUILLAUMON and P. NABARRA Centre National d'Etudes Spatiales (CNES) 18 avenue Edouard Belin 31401 Toulouse Cedex 04 FRANCE

Abstract The Thermal Control Department of CNES (French Space Agency) Toulouse Space Centre has developed many materials and thermal control coatings for satellites and launchers. Some of them have been specially developed to be used in low Earth orbit (LEO). These coatings are resistant to the conditions encountered at this orbit: vacuum, atomic oxygen, thermal cycling, and ultraviolet radiation. These coatings are: electrically conductive or non - conductive black and white paints; protective coating (MAPATOX K) for organic materials which are sensitive to atomic oxygen; new high performance low cost cold coatings named flexible solar reflectors (FSR) which could replace the second surface mirrors (SSM) in LEO. In this paper we present the results of the behaviour of these thermal control coatings under simulated LEO environment. Firstly, we describe the behaviour of black and white silicone and polyurethane paints under atomic oxygen (AO), and secondly, the behaviour of the new white silicone paints, of the MAPATOX K and of the FSR under AO and ultraviolet (UV) combined exposure. In the first part, it has been shown that the polyurethane paints are not stable towards AO but the silicone paints are stable. The new white silicone paints, SG121 FD (electrically non-conductive) and PCBE (electrically conductive) developed to replace SG120 FD and PCBZ, have a good stability towards AO. The new polysiloxane coating named MAPATOX K is a good solution to protect Kapton against AO, The new polysiloxane coating named FSR could replace the SSM for LEO applications. In the second part, the AO and UV combined effect is studied on the new CNES developed thermal control coatings: SG121 FD, PCBE, FSR and MAPATOXK. It has been shown that SG121 FD, PCBE and FSR are more sensitive to AO than to UV, MAPATOX K is not sensitive to AO and UV. The maximal degradation of the solar absorption factor is associated with the AO bombardment prior to UV irradiation. 1. 0 Introduction The Thermal Control Department of the CNES Toulouse Space Centre has been working for several years on the development of thermal control coatings for satellites and launchers. 193

194

For the thermal control of the low Earth orbit (LEO) satellites, we have developed different coatings particularly resistant to the conditions encountered at this orbit (vacuum, atomic oxygen, temperature, ultraviolet, etc.). These coatings are: - electrically conductive or non-conductive black and white paints - protective coatings on materials which are sensitive to atomic oxygen (MAPATOX K) - new low cost cold coatings named flexible solar reflectors (FSR) which could replace the second surface mirrors (SSM) in LEO. In this paper, we present the behaviour of thermal control coatings under simulated LEO environment: - firstly, we describe the behaviour of black and white silicone and polyurethane paints under atomic oxygen (AO), - secondly, the behaviour of the new white silicone paints, of the MAPATOX K and of the FSR under AO and ultraviolet (UV) combined exposure is described. 2. 0 Atomic Oxygen Effects In LEO, the surface of satellites materials are exposed to the oxygen atoms impact with a velocity of 8 km/s. The materials can be more or less oxidised during the mission and for some of them the resulting erosion can lead to the changes of the thermo-optical properties and the performances. The stability of thermal control coatings is thus a significant problem, particularly the stability of paints. The tests have been performed in the CASOAR chamber (Figure1) belonging to the Department of Space Environment (DESP) of ONERA in France. The fluence of the AO has been 2.1020 atoms/cm2 simulating the AO dose received by a satellite at 800 kilometres during 4 years. The tested materials have been black and white silicon paints and black polyurethane paints. 2.1. ATOMIC OXYGEN SOURCE The AO source is a pulsed source that is using a laser detonation technology. This laser detonation source was originally developed by Physical Science Inc. (PSI), and sources based on this design are now used in several laboratories. The source is capable of producing beams containing AO with average velocity that could be adjusted between 5 and 13 km/s, in practice it is maintained at 8 km/s. The amount of fluence received by exposed samples is estimated by exposing samples of a material with known reactivity (Kapton) and measuring the resulting erosion. 2.2. MATERIALS The tested thermal control paints and coatings are described in Table 1. The SG121 FD and the PCBE are the new developed paints which respectively replace the SG120 FD and the PCBZ paints. This replacement has been necessary because the commercial silicone resins are no longer available and because the CNES laboratory perfected a pigment treatment allowing a best stabilisation from UV and particles irradiation.

195

The FSR is a new low cost cold coating named Flexible Solar Reflector to replace the Second Surface Mirrors (SSM). It is a thin coating of polysiloxane (50 µm) with a low density (around 1), which is inexpensive, easy to use and able to be applied on metal deposits (such as polished aluminium or silver). The MAPATOX K is a new atomic oxygen resistant coating. It is a very thin film of polysiloxane (< 10 µm) which is applied on Kapton to avoid the debris production. The initial thermo-optical properties of the samples are given in Table 2. Figure 1. The CASOAR chamber

2.3. TESTS The paints were applied on 19×19 mm aluminium plates. The FSR was applied on the aluminium size of an aluminised Kapton (50 µm) and the MAPATOX K on Kapton (25 µm) and both were stuck on aluminium plates with transfer adhesive. The plates were placed on an aluminium sample holder and put on primary vacuum at 110°C during 24 hours, then hold at room temperature and put at 55 cm (for S2, PGN 7991, PGN AS, PU1, PUC, SG11 FD, SG120 FD and PCBZ) or at 40 cm (for SG 121 FD, PCBE, FSR, SSM and MAPATOX K) from the source. The fluence received by all exposed samples was 2.1020 atoms/cm2. 2.4. MEASUREMENTS The following measurements were carried out : - the mass of each sample (measurement accuracy ± 20 µg) - the spectral reflectance between 250 and 2500 nm using an integrating sphere with central sample put on a Varian Cary 2300 spectrometer. From this spectral value the solar reflectance Rs is calculated (measurement accuracy ± 1%). From Rs it can be deduced that the hemispherical solar absorption factor is: αs = 1-Rs.

196

- the total normal infra-red emissivity factor with a portable reflectometer (Gier Dunkle DB100). TABLE 1. Definition of the tested materials

Coatings S2 PGN 7991 PGN AS PU1

PUC SG11 FD SG120 FD PCBZ

SG121 FD PCBE FSR

SSM MAPATOX K

Characteristics Black electrically conductive Black electrically nonconductive Black anti-static Black electrically conductive Black electrically conductive White electrically conductive White electrically nonconductive White electrically conductive White electrically conductive White electrically conductive Transparent electrically non-conductive Transparent electrically non-conductive Transparent electrically non-conductive

Pigment Carbon black Carbon black Carbon black Carbon black Carbon black Zn2TiO4

Binder Commercial silicone resin Commercial silicone resin Commercial silicone resin Commercial polyurethane resin

Commercial polyurethane resin Commercial silicone resin

ZnO

Commercial silicone resin

Zn2SnO4

Commercial silicone resin

Treated ZnO

CNES synthesised silicone resin

Treated ZnO

CNES synthesised silicone resin

None

CNES synthesised silicone resin

None

Sheldahl G401900 FEP 125 µm silver-inconel and adhesive CNES synthesised silicone resin

None

TABLE 2. Initial thermo-optical properties of the tested materials

Coatings

αs ± 0.01

S2 PGN 7991 PGN AS PU1 PUC SG11 FD SG120 FD PCBZ SG121 FD PCBE FSR SSM MAPATOX K

0.96 0.96 0.96 0.94 0.93 0.15 0.19 0.15 0.21 0.20 0.12 0.08 0.36

εIR ± 0.03 0.89 0.88 0.88 0.89 0.77 0.85 0.85 0.90 0.89 0.87 0.84 0.81 0.80

197

2.5. RESULTS The erosion (deduced from mass measurement and density), the AO reaction efficiency, the variation of the solar absorption factor and the variation of the infrared emissivity factor are given in Table 3. For S2, PGN 7991, PGN AS, PU1, PUC, SG11 FD, SG120 FD and PCBZ, the results are given in details in [1]. For SG121 FD, PCBE and MAPATOX K, the results are described in [2], and for FSR and SSM in [3]. TABLE 3. Results after the AO test

Paints

Erosion (µm)

S2 PGN 7991 PGN AS PU1 PUC SG11 FD SG120 FD PCBZ SG121 FD PCBE FSR SSM MAPATOX K

-0.09 +0.10 +0.08 -1.22 -1.97 -0.14 -0.24 +0.01 +0.07 +0.08 -0.14 -2.39 -0.20

AO reaction coefficient, 10-24 cm3/atom 0.05 -0.05 -0.04 0.58 0.94 0.07 0.11 -0.01 -0.04 -0.04 0.08 1.40 0.12

∆αs

∆ε

-0.01 -0.01 -0.01 0.00 +0.02 +0.04 +0.01 -0.01 +0.01 +0.01 +0.04 +0.02 0.00

-0.01 0.00 0.00 +0.02 +0.08 0.00 0.00 +0.01 0.00 -0.01 0.00 0.00 -0.01

2.6. DISCUSSION Many comments can be deduced from Table 3. First of all, the polyurethane binder (C-C bond energy = 306.7 kJ/mol) is degraded much more by AO than the silicone one (Si-O bond energy = 797.5 kJ/mol). The attack of polyurethane bonds leads to the scission of the polymer chains, and then to the erosion of paints. Therefore surface changes degrade the emissivity factor. The commercial silicone resin used in the SG11 FD and SG120 FD paints is the same, and it appears that it is more sensitive to AO erosion than the CNES synthesised silicone resin used in the SG121 FD. The ZnO pigment (SG120 FD) and the treated ZnO pigment (SG121 FD and PCBE) are much more stable than the Zn2TiO4 (SG11 FD). The AO reaction coefficients and the variations of the solar absorption factor of PCBZ and PCBE are practically the same, so these two paints will be considered as equivalent. The erosion is practically the same for the FSR and the MAPATOX K. The solar absorption factor depends on the coating thickness then the more the thickness is important, the more the degradation is visible. Moreover, vacuum UV (< 200 nm) are produced by the AO source. The coatings are degraded in depth by these VUV, therefore the degradation is much more important for the FSR than for the MAPATOX K. The thermo-optical properties of the FSR are more degraded than the SSM ones. But the SSM is much more sensitive to erosion than the FSR.

198

3.0 Atomic Oxygen and Ultraviolet Combined Effects In LEO, the thermal control materials on satellites surface are exposed not only to AO impact but also to the solar flux, except for heliosynchronous orbits where particles (electrons and protons) are also present. A satellite or a space station attitude change leads to: - Direct and non-direct UV exposure of materials - AO exposure of materials, most critical when their direction is perpendicular to the AO velocity vector - UV and AO exposure Then materials degradation can be caused by expose to both those environmental factors, and also to cause successive degradation. Therefore it is interesting to study the combined effects of AO and UV on thermal control coatings. 3.1. ATOMIC OXYGEN AND ULTRAVIOLET SOURCES The AO testing was carried out in the CASOAR chamber (see 2.1). Irradiation tests by UV were conducted in the SEMIRAMIS chamber belonging to the DESP of ONERA (Figure 2). The test consisted of irradiating the coating samples with a lamp reproducing as faithfully as possible the solar spectrum without atmosphere in the UV range (200 to 300 nm band and 300 to 400 nm band). The UV generation was used with an acceleration factor between 3.1 and 3.4. The samples were positioned on a sample holder, held at 40°C throughout the test and placed in a vacuum chamber with a vacuum level of around 5.10-7 Torr. A new development of a transport and transfer system under vacuum (VESTA) between SEMIRAMIS and CASOAR allows us to perform the combined tests. Figure 2. The SEMIRAMIS chamber

199

3.2. MATERIALS The tested thermal control coatings are described in Table 4 and their initial thermooptical properties are given in Table 5. TABLE 4. Definition of the tested materials Coatings SG121 FD PCBE FSR MAPATOX K

Characteristics White non electric conductive White electric conductive Transparent non electric conductive Transparent non electric conductive

Pigment Treated ZnO Treated ZnO None None

Binder CNES synthesised silicon resin CNES synthesised silicon resin CNES synthesised silicon resin CNES synthesised silicon resin

TABLE 5. Initial thermo-optical properties of the tested materials Coatings SG121 FD PCBE FSR MAPATOX K

αs ± 0.01 εIR ± 0.03 0.21 0.20 0.12 0.36

0.89 0.87 0.84 0.80

3.3. TESTS The paints were applied on 19×19 mm aluminium plates. The FSR was applied on aluminised Kapton (50 µm) and the MAPATOX K on Kapton (25 µm) and both were stuck on aluminium plates with transfer adhesive. Ten samples were divided into two sets of samples. Three steps were carried out: step one, when the first set of samples was irradiated by UV during 500 equivalent solar hours (ESH) in SEMIRAMIS, and the second set was screened from irradiation; step 2, when the two sets were transferred with the VESTA system under vacuum to CASOAR and exposed to 2.1020 at/cm2, and step 3, when the two sets were transferred with the VESTA system to SEMIRAMIS and exposed to UV irradiation during 500 ESH. 3.4. MEASUREMENTS Before and after the step 1 and after the step 3 the spectral reflectance was measured in SEMIRAMIS under vacuum between 250 and 2500 nm using a Perkin Elmer Lambda 19 spectrometer with a lateral sample integrating sphere. The value of the solar reflectance Rs was calculated from the reflectance values in 250-2500 nm area of each reflection spectrum. The solar absorption factor is obtained by the formula: αs = 1-Rs with a relative error of ∆αs⁄αs = ± 1%. Before the step 1 and after the step 3 (beginning and end of the test) the total normal infra-red emissivity factor was measured with a portable reflectometer (DB100 Gier Dunkle).

200

3.5. RESULTS The results are described in [4]. The variation of the solar absorption factor is given in Table 6 and the variation of the infra-red emissivity before and after the test in air is given in Table 7. 3.6. DISCUSSION The thermal control coatings used in this study are more sensitive to AO than to UV irradiation (except the MAPATOX K). The maximal degradation of the solar absorption factor seems to be associated with AO bombardment prior to UV irradiation. One hypothesis could be imagine: the surface of the silicone resin is previously transformed to a defective SiOx form by AO; this structure could lead to the formation of colour centres under UV irradiation and therefore to the degradation of the thermo-optical properties. TABLE 6. ∆αs results after the AO and UV combined test Coatings

SG121 FD PCBE FSR MAPATOX K

∆αs

∆αs

∆αs

∆αs

after 500 esh

after 500 esh + AO 2.1020 at/cm2

+0.01 +0.01 0.00 +0.01

+0.01 +0.01 +0.03 0.00

after 500 esh + AO 2.1020 at/cm2 + UV 500 esh +0.02 +0.02 +0.05 +0.02

after AO 2.1020 at/cm2 + UV 500 esh +0.03 +0.02 +0.05 +0.01

TABLE 7. ∆ε results after the AO and UV combined test Coatings

∆ε

∆ε

SG121 FD PCBE FSR MAPATOX K

after 500 esh + AO 2.1020 at/cm2 + UV 500 esh 0.00 0.00 -0.01 0.00

after AO 2.1020 at/cm2 + UV 500 esh 0.00 0.00 -0.01 0.00

4.0 Conclusions In this paper, the AO effect and the AO and UV combined effect have been studied on thermal control coatings. The AO effect is tested on black and white silicone paints, black polyurethane paints and silicone varnishes. It has been shown that: - The polyurethane paints are not stable towards AO, on the contrary the silicone coatings have a good behaviour. - The new white silicone paints SG121 FD (electrically nonconductive) and PCBE (electrically conductive), developed to replace SG120 FD and PCBZ, have a good resistance towards AO. - A new polysiloxane coating named MAPATOX K is a good solution to protect Kapton against AO.

201

-

A new polysiloxane coating named FSR could replace the SSM for LEO applications. The AO and UV combined effect is studied on the new CNES developed thermal control coatings, SG121 FD, PCBE, FSR and MAPATOX K. It has been shown, that : - SG121 FD, PCBE and FSR are more sensitive to AO than to UV irradiation; the MAPATOX K is not sensitive to AO and UV. - The maximal degradation of the solar absorption factor is associated to an AO bombardment prior to UV irradiation. This test was important to perform in finding out which environmental component is responsible for the degradation of the coatings. It is the AO bombardment to consider first, but the AO source produces VUV which could also lead to the degradation of the coatings. A modification of the source for separating the AO and the VUV radiation could allow to find and distinguish which element is responsible for the degradation.

5.0 Acknowledgements: I thank Virginie Viel, Joseph Marco, and C. Pons from ONERA/DESP for their participation in this work.

6.0 References: 1. Paillous A., Oscar H., Riboulet M., and Siffre J. (1992) Bombardement par l'oxygène atomique de matériaux de régulation thermique, CERT/ONERA 437812 Report. 2. Viel V. and Siffre J. (1999) Tenue à l'oxygène atomique de revêtements de contrôle thermique, ONERA/DESP RF/472100 Report. 3. Viel V. and Chardon J.P. (2001) Essai de bombardement par l'oxygène atomique, ONERA/DESP RTS 2/06281 Report. 4. Marco J., Viel V., and Pons C. (2000) Tenue sous oxygène atomique et UV de revêtements de contrôle thermique, ONERA/DESPRF/CS0418801 Report.

This page intentionally left blank

GROUND TESTING OF SCK5 WHITE SILICONE PAINT FOR LEO APPLICATIONS

I. GOUZMAN, E. GROSSMAN, G. LEMPERT, Y. NOTER AND Y. LIFSHITZ Space Environment Division, Soreq NRC Yavne 81800, Israel V. VIEL-INGUIMBERT AND M. DINGUIRARD DESP, CERT ONERA Toulouse, France F-31055

Abstract SCK5 is a white antistatic silicone paint developed by CNES and manufactured by MAP (France). The present work summarizes durability tests of this paint in a simulated low earth orbit (LEO) atomic oxygen (ATOX) environment. The paint was applied on various substrates including Kapton film, Duroid 5880 (a glass/Teflon PTFE composite) and TMM3 (a ceramic/thermoset polymer). Two types of ATOX simulation systems were used: an RF oxygen plasma, and a laser detonation source (manufactured by PSI) producing a 5 eV ATOX beam. Both types of simulation facilities generate VUV radiation in addition to oxygen species. Dedicated experiments were performed to distinguish between VUV and ATOX effects. The SCK5 coated samples were also exposed to RF argon plasma, in order to separate between chemical effects of atomic oxygen and physical effects introduced by the RF plasma. The effects of ATOX exposure were studied by scanning electron microscopy (SEM), energy dispersive X-ray spectroscopy (EDS) and X-ray photoelectron spectroscopy (XPS). A comparative study of the erosion yield, the surface morphology and the chemical composition resulting from exposure to equivalent ATOX fluences in both types of simulation systems was performed. The SCK5 exposed to RF plasma showed significant cracking, partial delamination and enhanced embrittlement even for low ATOX fluence, equivalent to 2x1019 atoms/cm2. Similar exposures to the 5 eV ATOX (PSI source) exhibited no cracking. In both cases the exposed samples showed a decrease of the carbon atomic concentration and an increase of the oxygen concentration in the upper surface layer, indicating the formation of a silicon oxide skin, which was more significant for the samples exposed to the RF plasma asher. It may be concluded that the erosion of SCK5 by the RF oxygen plasma is considerably more severe than by the 5 eV ATOX, at least for the specific case of porous coating of siliconic material, tested in the present work.

203

204 This is most probably associated with a combination of factors, including the nature of the reactive species in the plasma asher, their omnidirectional flux and the high porosity of SCK5 coating, leading to a strong compressive stresses and consequently cracking of the brittle silicon oxide skin.

1.0 Introduction The SCK5 white antistatic silicone paint was developed by CNES and is manufactured by MAP (France) as a thermal control coating for high frequency circuit materials [1]. SCK5 has a very short shelf life (24h), implying that the application should be done at the paint manufacturing site. Technically this may be achieved either by painting the surfaces at the plant or by purchasing a ready-made SCK5 painted Kapton film and adhering it to the surface of interest in the customer’s facility. The present work involved the ground testing of atomic oxygen (ATOX) durability of different SCK5 painted surfaces, including a Kapton film, Duroid 5880 (a glass/Teflon PTFE composite) and TMM3 (a ceramic thermoset polymer composite). The effects of ATOX, the most prominent hazard for polymeric materials in LEO, have been widely investigated in recent years by in-flight experiments as well as by laboratory simulation techniques [2-7]. The difficulty of the ground simulation experiments to forecast the long-term LEO durability of external spacecraft materials stems from basic differences between the actual environment and the simulated one. These include: (i) differences in energy distribution and directionality of reactive species; (ii) low ATOX flux in space as compared to accelerated tests, (iii) neutral ground state ATOX in space as compared to ionic and excited species in some of the experiments; (iv) synergistic effects of ATOX, ultraviolet (UV) and ionizing radiation in space. Two complementary types of ground simulation techniques were used for simulation of the ATOX LEO environment in the present work. The first type included two RF plasma systems located at Soreq NRC. The RF plasma asher is a widely accepted simulation facility for screening tests in many laboratories [2,3]. It contains a mixture of reactive species (excited, neutral and ionized oxygen atoms and molecules) that impinge upon the exposed surface omni-directionally, as well as vacuum UV (VUV) radiation. The source is capable of simulating high LEO equivalent ATOX fluences within reasonable experimental times. The second system was a laser detonation source (CASOAR) located in CERT ONERA, Toulouse. This source generates a highly directional, pulsed 5 eV ATOX beam accompanied by high doses of VUV radiation. The instantaneous ATOX flux during the pulse is 4 orders of magnitude larger than in the LEO environment, but the average value is similar to the LEO ATOX flux (~1x1015 atoms/cm2sec) [4,5]. A significant discrepancy between the results of the RF plasma and the laser detonation systems was observed in the present work, initiating detailed studies and reevaluation of both techniques.

205 2.0 Experimental Details The following samples were studied in the present work: (i) Kapton coated with SCK5, further referred to as Kapton/SCK5; (ii) Kapton coated with SCK5 and bonded by a pressure sensitive adhesive to a Duroid 5880 (a glass/Teflon PTFE composite manufactured by Rogers Corp.) substrate, further referred to as Duroid/Kapton/SCK5; (iii) Duroid 5880 coated directly with SCK5, further referred to as Duroid/SCK5, and (iv) TMM3 (a ceramic/thermoset polymer composite manufactured by Rogers Corp.) coated directly with SCK5, further referred to as TMM3/SCK5. A conventional RF plasma asher (Harrick, Model PDC-3XG) system (15W, 13.56 MHz), operating at 100 mTorr was used for most of RF plasma tests. The LEO equivalent ATOX flux in the central part of the RF plasma reactor was ~5x1015 atoms/cm2sec. The SCK5 coated samples were exposed to different ATOX fluences in the range from 2x1019 atoms/cm2 to 1.7x1021 atoms/cm2. Argon plasma was used to distinguish between reactive and non-reactive plasma interactions. A high power RF plasma reactor (1200 W, Litmas company) was used to evaluate the relative effects of different plasma components, such as in-glow exposure (all plasma components) and after-glow exposure, 100 mm away from the reactor (reduced amount of excited species, an increased amount of neutral atomic oxygen). A specially designed target holder assembly was used to eliminate the direct VUV irradiation allowing all other after-glow plasma components to react with the sample. The system was operated at 300 and 1200 W RF power at an oxygen pressure of 120 mTorr. The VUV flux was assessed by a Phototube sensor (Hamamatsu Model R1187) positioned at the location of the exposed sample. The measurements were performed using a 1% transmission filter, in order to reduce the intensity to the working range of the sensor. The VUV flux was 1.6x1016 photons/cm2sec at 300W, 100 mm away from the reactor. At CERT ONERA, samples were tested using a laser detonation source (CASOAR, manufactured by PSI). Under the operational conditions used in the present study a mean LEO equivalent ATOX flux of ∼1x1015 atoms/cm2sec was obtained. The VUV flux of 1.85x1016 photons/cm2sec was measured by a phototube detector. Modification of the CASOAR system allowed the separation between ATOX, VUV and ATOX/VUV synergistic effects [8]. Due to limitations of the system, only relatively low LEO equivalent ATOX fluences, amounting to 4x1019 atoms/cm2 for ATOX without VUV and 2x1020 atoms/cm2 for ATOX/VUV were achieved. The samples were also exposed to VUV alone at a fluence of 2x1021 photons/cm2. In all ATOX simulation experiments the LEO equivalent atomic oxygen fluence was evaluated by measuring the mass loss of a Kapton HN polyimide reference sample. The erosion rate of Kapton was assumed to be equal to 3x10-24 cm3/atom and independent of the ATOX fluence [9]. The effects of ATOX exposure were studied by several complementary techniques including scanning electron microscopy (SEM), energy dispersive X-ray spectroscopy (EDS) and X-ray photoelectron spectroscopy (XPS). SEM micrographs were obtained using a FEI Quanta 200 microscope operating in the low vacuum mode. Surface elemental composition was estimated by means of energy dispersive X-ray

206 spectroscopy (EDS) using JSM 5410LV JEOL instrument with a primary electron energy of 25 keV. To minimize surface charging effects, all samples were coated with a thin (~25 nm) gold layer. The elemental composition and chemical bonding states in the near surface region were assessed by XPS measurements. The measurements were carried out using a VG system with Mg KĮ (1253.5 eV) 200 W X-ray source and a triple channeltron 150 mm hemispherical analyzer, CLAM2. Mass loss measurements were carried out for all the samples exposed in the laser detonation system and for some samples exposed in the RF plasma asher system. The results were used to determine the erosion yields of the SCK5 samples.

3.0 Results 3.1 MORPHOLOGICAL CHANGES AND EROSION YIELDS 3.1.1. RF plasma simulation All types of SCK5 coated samples were exposed in a low power RF plasma asher to different LEO equivalent ATOX fluences, ranging from ~2x1019 to 1.7x1021 atoms/cm2. Duroid/Kapton/SCK5. Duroid 5880 substrates (3cm×5cm) were prepared, onto which SCK5 coated Kapton was adhered by a pressure sensitive adhesive layer. The SCK5 coated Kapton applied on both sides of the substrate in order to prevent its direct exposure to RF oxygen plasma. The edges were protected with CV1144-0 clear silicone coating, manufactured by NuSil, to eliminate the possibility of oxygen undercutting. It was found that RF oxygen plasma exposure to 2x1020 atoms/cm2 and up resulted in well-distinguished changes in surface morphology, detected by the naked eye. The changes included the formation of numerous cracks on the initially smooth paint surface, resembling cracks in dry earth (see Figure 1(a)). Cracks were observed also at lower fluences but they were hardly detected by the unaided eye. An average crack width of about 5 µm was estimated from SEM micrographs (see Figure 1(b)). The width, shape and areal density of cracks did not change significantly with the ATOX fluence above 2x1020 atoms/cm2. The distance between cracks differed from about 100 µm to 1mm. The morphology of the regions between the cracks was similar to that observed for unexposed samples. It should be noted that the unexposed SCK5 coating was highly porous (see Figure 2). Pores with dimensions similar to the width of a crack were found in the unexposed sample. The size and areal density of these pores did not change after ATOX exposure to different fluences. Kapton/SCK5. A free-standing Kapton film (~40 µm thick, 1.5cm×2.0cm) painted with SCK5 was exposed in a low power RF plasma asher to an ATOX fluence of about ~5x1019 atoms/cm2. The sample was supported in a glass sample holder in order to reduce Kapton etching from the backside. After exposure, the sample was found to be strongly bent. No cracks were found on the exposed surface by visual inspection. However, an attempt to flatten the bent film resulted in cracks, as shown in Figure 3.

207 Note that due to the omni-directional nature of the ATOX in the RF plasma system, it is difficult to expose only one side of the sample to ATOX and completely avoid the exposure of the backside. As Kapton itself undergoes erosion under ATOX exposure (~1.5 µm at 5x1019 atoms/cm2 of ATOX), the observed bending effect could be partially associated with a thinning of the peripheral part of the Kapton substrate, in addition to bending caused by ATOX-induced stresses in SCK5. To prevent Kapton erosion from the backside of the sample, in the separate experiment Kapton/SCK5 sample was exposed in a glass frame, which maintained fixed position of the film. In this case welldefined cracks were observed on the SCK5 surface after ATOX exposure, similar to those shown in Figure 1.

(a)

(b)

Figure 1 SEM micrographs of Duroid/Kapton/SCK5 samples after exposure to RF oxygen plasma (1.0x10 21 atoms/cm2 LEO equivalent ATOX fluence): a general view (a) and a magnified view of cracks (b).

Figure 2. SEM micrograph of an unexposed SCK5 coating.

208 The erosion yield of the SCK5 was calculated using an ATOX fluence estimated from the mass loss of a Kapton witness sample exposed at the same experiment. The erosion yield was ~5x10-25 cm3/atom and ~1x10-24 cm3/atom after exposure to a LEO equivalent ATOX fluence of 3.6x1019 and 8x1019 atoms/cm2, respectively. Duroid/SCK5. After exposure to the LEO equivalent ATOX fluence of about 2.5x1019 atoms/cm2, the initially smooth paint surface was cracked into well-distinguished domains. The size of an individual domain was approximately 1mm×1mm (see Figure 4). The average width of the cracks was 5-7 µm. However, due to the poor adhesion of the coating to the Teflon-based Duroid substrate, many delaminated regions were observed, leading to the formation of the paint flakes.

Figure 3 A free standing SCK5 coated Kapton film exposed to RF oxygen plasma (LEO equivalent ATOX fluence of 19 2 ~5x10 atoms/cm ). The apparent cracks while trying to flatten the bent exposed sample.

TMM3/SCK5. No cracks or other changes were observed by the naked eye after an ATOX exposure to a maximum fluence of 1.7x1021 atoms/cm2. However, SEM observations revealed cracks, similar to those shown in Figure 1(a). In this case neither delamination nor separation between substrate and coating were observed. Argon RF plasma exposure. In order to distinguish between chemical effects of ATOX and physical effects of RF plasma, three samples were subjected to argon RF plasma instead of to oxygen plasma. The exposed samples were: (i) a free-standing Kapton/SCK5 film, (ii) a Kapton/SCK5 film in a glass frame, and (iii) a Duroid/Kapton/SCK5. Cracks or bending of the Kapton/SCK5 film were not observed after Ar plasma exposure of all samples for 34 hr (a 34 hr exposure of RF oxygen

209 plasma is equivalent to ATOX fluence of 6x1020 atoms/cm2), and the only effect was a very light coloration of the samples. High power RF plasma reactor. Dedicated experiments were carried out to distinguish between VUV and ATOX effects. This was achieved by using a high power RF plasma apparatus and a specially designed target holder assembly located in the afterglow region. Kapton/SCK5 samples were exposed to 300W and 1200W RF oxygen plasma afterglow flow including and excluding direct VUV radiation. In all cases similar cracking of the exposed area was observed after exposure to a LEO equivalent ATOX fluence of about 2x1019 atoms/cm2.

Figure 4 SEM micrograph of the Duroid/SCK5 sample after exposure to 1.0x1021 atoms/cm2 of equivalent ATOX fluence (RF oxygen plasma simulation).

3.1.2 Laser detonation ATOX source Several samples of each type (Kapton/SCK5, Duroid/SCK5, TMM3/SCK5 and Duroid/Kapton/SCK5) were exposed simultaneously in the laser detonation 5 eV ATOX source (CASOAR). The samples were exposed to a LEO equivalent ATOX fluence of 4x1019 atoms/cm2, to synergistic ATOX/VUV fluences of 2x1020 atoms/cm2 and 2x1021 photons/cm2, respectively, and to a VUV fluence of 2x1021 photons/cm2. Surface morphology of the exposed samples was studied by SEM (data not shown). No cracks or other changes in surface morphology were found after the ATOX, ATOX/VUV or VUV exposures. The erosion yields for Kapton/SCK5 exposed to various environments are shown in Table 1. The erosion yield was calculated using the LEO equivalent ATOX fluence obtained from the mass loss of Kapton witness samples exposed at the same

210 time. The erosion yield of samples exposed to VUV was calculated using the VUV flux of 1.85x1016 photons/cm2sec, as measured by a VUV detector. TABLE 1. Erosion yields of Kapton/SCK5 samples exposed in the laser detonation source. Environment

Fluence

Erosion yield

ATOX/VUV

2x1020 atoms/cm2 and 2x1021 photos/cm2

9.1x10-26 cm3/atom

VUV

2x1021 photos/cm2

1.3x10-26 cm3/photon

19

ATOX

2

3.3x10-25 cm3/atom

4x10 atoms/cm

3.2. CHEMICAL COMPOSITION CHANGES 3.2.1. EDS results The elemental composition of SCK5 coated samples exposed in both simulation systems was determined by EDS and the results are summarized in Table 2. The elemental composition of the reference sample (unexposed SCK5) was measured at several points to assess the uniformity of the coating. For all samples the most prominent effect is a decrease of carbon atomic concentration from about 20 at.% for the pristine coating to about 11-15 at. % after exposure to different fluences of ATOX. This was accompanied by an increase in oxygen atomic concentration in the analyzed layer. The modification of elemental composition in the analyzed region does not show a fluence dependence. No changes in Si, Ti, Zn, and Sn atomic concentrations were observed. TABLE 2. Elemental composition of SCK 5 coated samples determined by EDS before and after exposure to the RF oxygen plasma or the laser detonation source. Element Treatment Unexposed, area I area II 2x1019 atoms/cm2 9x1019 atoms/cm2 1x1021 atoms/cm2 ATOX, 4x1019 atoms/cm2 ATOX/VUV, 2x1020 atoms/cm2 and 2x1021 photons/cm2

C 21.3 19.6

O

Si

46.1 6.5 48.1 6.7 RF oxygen plasma 13.0 52.1 6.8 10.8 57.4 6.4 13.8 54.5 6.5 Laser detonation ATOX source 15.0 50.7 7.4 15.7 50.5 7.2

Ti

Zn

Sn

9.8 9.7

7.1 6.9

9.2 9.0

10.6 9.7 9.6

7.7 6.7 6.6

9.8 9.0 9.0

10.3 10.2

7.0 7.0

9.6 9.4

3.2.2. XPS Results Chemical changes in the irradiated surface layer were detected by XPS. The following Duroid/Kapton/SCK5 samples were analyzed: an unexposed SCK5 coating (reference), two samples exposed using the CASOAR system to ATOX/VUV and ATOX alone,

211 respectively, as well as two samples exposed in a low power RF plasma asher to the LEO equivalent ATOX fluence of 2x1019 and 1x1020 O-atoms/cm2. A typical XPS survey spectrum obtained from SCK5 surface included the Si 2p, C 1s, O 1s, Zn 2p, Sn 3d and Ti 2p core level lines. High-resolution XPS spectra (not shown) obtained from the surfaces before and after exposure to various environments were used for chemical composition analysis. The near surface composition analysis was done by assuming that this region is homogeneous and using published atomic sensitivity factors [10]. The results are presented in Table 3. TABLE 3. Surface composition (at.%) determined from XPS data for Duroid/Kapton/SCK5 samples before and after exposure to ATOX alone and to ATOX/VUV using the CASOAR system, as well as after exposure to an RF plasma system.

Peak Origin

Silicone matrix, Contamination Silicone matrix, Silicon oxide Si-O-Si, metal oxides, adsorbed O, H2O Metal oxides

Bonding Unexposed CASOAR, CASOAR, state ATOX, 4x1019 ATOX/VUV, 2 atoms/cm 2x1020 atoms/cm2 C 1s 36.2 10.2 9.8

RF plasma, 2x1019 atoms/cm2 4.3

RF plasma, 1x1020 atoms/cm2 2.2

Si 2p

24.1

30.5

30.4

23.7

23.9

O 1s

37.0

58.8

59.1

67.3

69.2

Sn 3d, Zn 2p, Ti 2p

2.7

0.7

0.7

4.7

4.7

4.0 Discussion The white paint SCK5 is a thermal control coating, applied on external satellite surfaces that may suffer from electrostatic discharge (ESD) problems. The paint is composed of a purified silicone binder, “doped” metallic oxides that are responsible for its antistatic properties and white color appearance, aromatic solvents and probably some other additives. Its outgassing properties are well within the ASTM E-595 limits (TML=0.187%, CVCM=0.036%, according to Soreq NRC outgassing tests). One issue of primary importance for this coating is its atomic oxygen durability. The SCK5 coatings (applied on Kapton, Duroid 5880 and TMM3) were tested using two types of ATOX simulation systems, an RF plasma asher and a laser detonation source. RF oxygen plasma sources are commonly used for material screening with respect to ATOX degradation in LEO. The advantage of such a facility is its relatively low cost and simplicity. The source generates omni-directional flux of oxygen atoms at thermal energies (~0.04 eV) [3]. However, other species are also present in the RF plasma environment, including molecular oxygen, atomic and molecular oxygen ions

212 and electrons at energies of tens of eV, excited neutral and ionic species, as well as ~130 nm VUV radiation with an approximate flux of 1013 -1016photons/cm2.sec [11,12]. The absolute and relative concentrations of these species depend crucially on the plasma operating conditions, such as plasma power, gas flow, pressure, chamber geometry and the sample’s position inside the reactor. Therefore the plasma - surface interactions are extremely complex and may introduce various artifacts as compared to the real ATOX environment in LEO. The laser detonation atomic oxygen source provides a highly directional beam of approximately 5 eV oxygen atoms. The mean flux provided by this source is about 1x1015 atoms/cm2sec. However, the atomic oxygen is generated in a pulsed mode and the fluence in a single pulse may be as high as 1x1014 atoms/cm2 in a period of time of about 100µsec, which is equivalent to a flux of 1x1019 atoms/cm2sec. This flux, higher by 4 orders of magnitude than in the LEO environment, may react differently with various materials, although, as yet this is not supported by experimental evidence. The formation of atomic oxygen in a laser detonation source is also accompanied by a high flux of VUV (1.85x1016 photons/cm2sec). Therefore this source may provide a strong synergistic effect between ATOX and VUV. In the present study, a modification of the sample holder enabled separation of the ATOX beam from the VUV radiation. However, due to technical limitations, the maximum ATOX fluence that could be simulated in this system within reasonable exposure time was far below the expected values in LEO applications. The interactions of the isotropic, omni-directional flux of thermal RF oxygen plasma vs. the highly directional ATOX beam in LEO with uncoated and coated polymer surfaces were discussed by Banks et al. [3]. It was demonstrated that atomic oxygen undercutting at defect sites (e.g. cracks) progresses much more rapidly in thermal energy plasma systems, due to its omni-directional impingement as compared to this effect in space. Note that from this point of view, the ATOX interaction in the CASOAR simulation system resembles much more closely the LEO ram direction interaction. However, as was already noted, the very high instantaneous acceleration factor may introduce artifacts in the mechanism of chemical interaction and relaxation processes in ATOX - surface interactions. Let us now discuss observed results with particular attention to the basic differences between the two simulation systems. The main visual effect of SCK5 exposure in the RF plasma simulation systems was its cracking. The cracking was observed after low equivalent ATOX fluence of about 2x1019 atoms/cm2. Increasing the ATOX fluence beyond this value up to a maximum ATOX fluence of about 1.7x1021 atoms/cm2 did not affect significantly the surface morphology. However, at some regions and for some materials delaminations of SCK5 were observed in the vicinity of the cracks at high ATOX fluences. This was most prominent for SCK5 coated Duroid substrates and Duroid/Kapton/SCK5 samples, where after exposure to RF oxygen plasma the SCK5 coating became brittle and was affected by vibrations and handling, causing formation of flakes. Such separated/delaminated fragments and flakes possess a potential source of particulate contamination in space, which should be avoided. Micro-cracks in SCK5 could allow the penetration of ATOX and erosion of the underlying substrate by undercutting. It is noted that the cracks on the

213 Kapton/SCK5, Duroid/SCK5 and Duroid/Kapton/SCK5 samples were readily observable by the naked eye, though their average width was only 5-7 µm. Such narrow cracks are visible only in the case of a significant undercutting of the underlying substrate. In the case of TMM3 substrates coated with SCK5 and exposed to a maximum equivalent ATOX fluence of 1.7x1021 atoms/cm2 no cracks were detected by the unaided eye. This may be explained by the better adhesion of SCK5 to TMM3, reducing SCK5 undercutting. The results described in Chapter 3.1.1 indicate a strong surface contraction of the SCK5 during interaction with RF oxygen plasma. Surface contraction leads to an increase of compressive stresses that finally result in fracture of the coating. In the case of a freestanding Kapton/SCK5 film, surface contraction caused the bending of the flexible film and its cracking while attempting to flatten it (see Figure 3), whereas fixed position of the film resulted in a stress relaxation via cracking during the ATOX exposure. To separate between the chemical effects of atomic oxygen and possible artifacts of RF plasma caused by the electromagnetic field, UV radiation, electrons and energetic ions, the SCK5 coated samples were exposed to an RF argon plasma. No cracks or other changes in surface morphology were found. Only some coloration was observed, probably associated with VUV exposure characteristic of RF plasma systems. Thus, the fracture and cracking of the SCK5 coating is clearly associated with oxygen reactivity. The exposure of the various samples in the CASOAR system demonstrated milder changes, as compared to RF plasma systems. The only detected morphological effect was some slight bending of Kapton/SCK5 film. The same trend, namely a milder deterioration of the SCK5 coating in the CASOAR system, was also reflected in the erosion yield and chemical composition results. The observed erosion yields for the same type of tested samples differed for the RF plasma and the laser detonation sources. The erosion yield measured after exposure to various environments in a laser detonation source (CASOAR) was found to be dependent on the irradiation type (see Table 2). It was maximal in the case of ATOX exposure alone, less in the case of synergistic ATOX/VUV exposure and minimal during exposure to VUV irradiation alone. The inhibiting effect of VUV is clearly observed from these results. Based on our previous studies [13] it is suggested that VUV radiation induces cross-linking in the siliconic matrix, reducing the erosion rate due to ATOX impingement and leading to a negative synergistic effect. For Kapton/SCK5 samples the erosion yield was measured both after RF plasma and after laser detonation source (CASOAR) treatments. The erosion yield after the RF plasma exposure was higher (~5x10-25 and ~1x10-24 cm3/atom after exposure to a LEO equivalent ATOX fluence of 3.6x1019 and 8x1019 atoms/cm2, respectively), compared to an ATOX or an ATOX/VUV irradiation in the CASOAR system (~3.3x10-25 and ~9.1x10-26 cm3/atom after exposure to 4x1019 ATOX and 2x1020 atoms/cm2 ATOX/VUV, respectively). It should be noted that the precise mass measurements of the exposed Kapton/SCK5 were problematic due to the unstable weight of the sample, probably due to humidity absorption by the exposed surfaces and/or charging problems.

214 Chemical composition of the SCK5 coating after exposure to RF plasma and laser detonation sources was studied by EDS and XPS. The sampling depths of EDS are matrix dependent and lie within the range of 0.1 µm for light elements, to about 3-5 µm for some metals [14]. XPS is a surface sensitive technique and probes the material to a depth of about 10 nm [15]. In addition, the sensitivity and accuracy of XPS measurements are higher as compared to EDS. The EDS and XPS analyses detected the following elements: C, O, Si, Ti, Zn and Sn. Comparison of the EDS and XPS results indicated that the main changes induced by ATOX are localized in a thin layer at the surface. Besides, it was found that this layer is composed predominantly of organic matrix. This is due to the fact that XPS analysis showed only small amounts of metals on the surface (~2.5%), whereas EDS revealed about 25% of Ti, Zn and Sn. EDS studies did not reveal any significant changes in the metals and silicon concentrations after exposure to both RF plasma system and laser detonation source, indicating that only a thin near-surface region is involved in the ATOX-material interaction. Accurate quantitative analysis of chemical composition using XPS was problematic due to possible surface contamination, as well as high roughness and porosity of the SCK5. Particularly, the concentration of small metal oxide particles embedded in the organic matrix can be affected by the surface morphology. In spite of these limitations significant differences in the concentrations of different components were observed after exposure to 5 eV ATOX and RF oxygen plasma. The results presented in Table 3 clearly show that exposure to the RF oxygen environment caused a drastic decrease of carbon atomic concentration (erosion of organic matrix), while exposing the metal oxide particles and increasing the oxygen content, which partially can be explained by an adsorption of water vapor and oxygen from air. The data clearly indicate the higher deterioration of the SCK5 surface under RF exposure, supporting the surface morphology and erosion yields evidence. Based on the above observations, the following phenomenological model is suggested. Cracking of the silicone-based SCK5 coating as a result of RF oxygen plasma is attributed to strong compressive stresses generated in the exposed coating due to a very effective erosion of the organic component in the silicone binder and formation of a brittle silicon oxide layer. Due to the omni-directional character of atomic oxygen impingement and the high porosity of SCK5 coating, this erosion process includes inner defects and pores. Consequently, a fast degradation of the silicone matrix takes place, leading to the exposure of the metal oxide particles to the RF plasma. This model cannot be applied to the laser detonation source since it interacts in a highly directional manner, creating a surface layer of silicon oxide, which might be non-continuous and/or thinner in this case and consequently less brittle. It may be speculated that the interaction of the SCK5 coating with the RF plasma environment may also be affected by (i) electromagnetic field interaction with exposed metal oxide particles, (ii) accompanying VUV irradiation, (iii) electrons and (iv) energetic ions. To gain a deeper insight into the RF plasma - surface interactions, an experimental study has been initiated at Soreq NRC, in order to understand in more detail the contributions, both individually and synergistically, of the different RF plasma components. Techniques have been developed to separate the plasma components in the plasma afterglow, where the effect of RF electromagnetic field is eliminated. However

215 this location of the sample did not prevent cracking of the SCK5 coating. In addition, a specially designed target holder assembly enabled the sample to be irradiated in the RF plasma afterglow, with and without direct VUV irradiation. No visible effect of the direct VUV flux on the surface morphology was observed. In the future we will expose the sample in a Faraday cup assembly in order to control the amount of electrons and energetic ions arriving at the sample surface, achieving a full separation of the RF oxygen plasma components.

5.0 Summary and Conclusions Atomic oxygen durability of SCK5 white antistatic silicone paint was studied using two types of simulation systems: a conventional RF oxygen plasma system and a laser detonation oxygen source (CASOAR, PSI system). The effects of equivalent ATOX exposure on the surface morphology and surface composition of SCK5 coating applied on different substrates were studied by several complementary techniques, including SEM, EDS and XPS. The tested materials were exposed to different equivalent atomic 21 2 oxygen fluences, ranging from 2x1019 up to 1.7×10 atoms/cm . The SCK5 exposed to RF plasma showed significant cracking, partial delamination and enhanced embrittelment at a relatively low ATOX exposure. Similar samples exposed to the laser detonation source (5 eV ATOX) exhibited no cracking. Thus, the RF plasma simulation demonstrated a more severe degradation of SCK5 paint, evidenced by the morphological changes, as well as by erosion yields and chemical composition changes as compared to the laser detonation system. These results are most probably associated with a combination of omnidirectional flux of reactive species and high porosity of SCK5 coating, which result in strong compressive stresses and consequently cracking of a brittle silicon oxide layer. It is suggested that RF oxygen plasma overestimates the ATOX interactions in LEO, at least for the specific case of a porosive coating of siliconic material, tested in the present work.

6.0 References 1. Guérard, F. and Guillaumont, J.C. (1997), “Thermal control paints and various materials for space use”, Proceedings of the 7th International Symposium on “Materials in a Space Environment”, Toulose, France, 457-458. 2. Golub, M.A., Wydeven, T., and Cormia, R.D. (1988), “ESCA study of Kapton exposed to low Earth orbit or downstream from a radio-frequency oxygen plasma”, Polymer Commun. 29, 285-288. 3. Banks, B.A., Rutledge, S.K., de Groh, K.K., Stidham, C.R., Gebauer, L., and LaMoreaux, C.M. (1995), “Atomic oxygen durability evaluation of protected polymers using thermal energy plasma systems”, NASA Technical Memorandum 106855, 1-15. 4. Caledonia, G.E., Krech, R.H., and Green, B.D. (1987), “A high flux source of energetic oxygen atoms for material degradation studies”, AIAA J. 25, 59-63. 5. Caledonia, G.E., Krech, R.H., Oakes, D.B., Lipson, S.J., and Blumberg, W.A.M. (2000), “Products of the reaction of 8 km/s N(4S) and O2”, J. Geophys. Res. 105(A6), 12,833-12,837. 6. Minton, T.K., Garton, D.J. (2001), “Dynamics of atomic-oxygen-induced polymer degradation in low earth orbit”, in “Chemical Dynamics in Extreme Environments: Advanced Series in Physical Chemistry”, ed. Dressler, R.A., World Scientific, Singapore.

216 7. Koontz, S.L., Albyn, K., and Leger, L.J. (1991), “Atomic oxygen testing with thermal atom systems: a critical evaluation”, J. Spacecraft, 28, 315-323. 8. Grossman, E., Gouzman, I., Viel, V., and Dinguirard, M., “Modification of the Atomic Oxygen Laser Detonation Source: separation of atomic oxygen and UV radiation”, to be published. 9. Minton, T.K. (1995), “Protocol for atomic oxygen testing of materials in ground based facilities, ver. No. 2”, JPL publication 95-17. 10. Chastain, J. and King, R.C., Jr. (eds.) (1995) Handbook of XPS, Physical Electronics, Inc. USA. 11. Townsend, J.A. (1996) A comparison of atomic oxygen degradation in low earth orbit and in a plasma etcher, Proc. 19th Space Simulation Conf., Baltimore MD, 249-258. 12. Kearns, D.M., Gillen, D.R., Voulot, D., McCullough, R.W., Thompson, W.R., Cosimini, G.J., Nelson, E., Chow, P.P., and Klaassen, J. (2001) Study of the emission characteristics of RF plasma source of atomic oxygen: measurements of atom, ion, and electron fluxes, J.Vac. Sci. Technol. A , 19, 993-997. 13. Grossman, E., Noter, Y., and Lifshitz Y. (1997) Oxygen and VUV irradiation of polymers: atomic force microscopy (AFM) and complementary studies, Proceedings of the 7th International Symposium on Materials in a Space Environment, 16-20 June, Toulouse, France, ESA-SP-399, 217. 14. Goldstein, J.I., Newbury, D.E., Echlin, P., Joy, D.C., Roming, A.D., Jr., Lyman, C.E., Fiori, C., and Lifshin, E. (1992) Scanning Electron Microscopy and X-ray Microanalysis, Plenum Press, New York. 15. Walls, J.M. (1990) Methods of Surface Analysis, Cambridge University Press.

STUDY OF POLYMER COATINGS RESISTANCE AFTER THE LONGTERM EXPOSURE ON SPACE STATION “MIR” E. N. KABLOV, V. T. MINAKOV, I. S. DEEV All-Russian Institute of Aviation Materials 17, Radio Str., Moscow, 107005, Russia Phone: +7 095 261-86-77, Fax: +7 095 267-86-09, E. F. NIKISHIN M.V. Khrunitchev State Space Scientific Production Center 18, Novosavodskaya Str., Moscow, 121087, Russia Phone: +7 095 142-50-36, Fax: +7 095 956-24-41

Abstract The study of the resistance of various thermo-regulating (thermal control) protective coatings to space environment was conducted on carbon fiber-reinforced polymer (CFRP) epoxy composites and three-layered honeycomb specimens after the longterm exposure of “Komplast” removable cassettes outside of “Kvant-2” module of Space Station “MIR”. Scanning electron microscopy (SEM), X-ray spectral analysis and visual inspection were used as the major tools in the study. Surfaces of white thermal control coatings, based on acrylic and epoxy polymers, in pristine condition and after 839 and 1218 days of the exposure were comparatively evaluated. In addition, similar analysis was conducted for a number of coatings based on organosilicone polymers, for inorganic coatings based on liquid glass, and aluminum foil, after 1024 days of space exposure. During the space exposures, some specimens with coatings of similar types were exposed in holders allowing to keep them in a stack, one under another with positioning of the upper specimen with the coating on outside and of the lower specimen’s coating facing the wall of the module that allowed to expose coatings of the same structure to different space conditions. Different types of macro- and micro-defects that appeared in the coatings were revealed, in particular, long cracks with a partial delamination of the coating from the substrate, a network of thin web-like cracks, scaly swelling-up and ring-like delamination. The change of the coatings surface composition, including the substances deposited on the surface during the long-term exposure in space, was evaluated. The possible coatings degradation mechanisms under the complicated space environment factors were discussed, with the account for the coating’s chemical nature. 1. 0 Introduction It was shown earlier that the thin surface layers of the matrix and the reinforcing fibers of polymer composite materials without protective coatings after long-term exposures on “Salute-5”, “Salute-6” orbital stations and Space Station “MIR” were subjected to various degradation effects that lead to changes of their structural 217

218

properties [4-7]. In order to improve the resistance of polymer-based composites to the long-term space environment factors and to prevent the deterioration of the thermal optical properties of the structural components it is necessary to use the thermal control protective coatings that should possess both the high environmental resistance and the ability to reduce the material’s surface temperature that is of great importance for its long-term thermocycling. The long-term space exposure effects on optical characteristics of protective coatings have been discussed already [7-8]. The use of thermal control coatings for the protection of polymer composites during the long-term flight in the outer space requires the confirmation of their resistance to the combination of space factors. Therefore, the evaluation of the long-term resistance of various thermal control protective coatings on CFRP specimens and skins in three-layered honeycomb structures to outer space factors was included into the program of space experiments with the use of “Komplast” removable cassettes. Those experiments are important for both research contribution and practical application. Results of a comparative evaluation of different types of thermal control protective coatings deposited onto polymer composite specimens after a long-term exposure on Space Station “MIR” are presented in this paper. 2. 0 Materials and Methods Various white thermal control protective coatings such as polymeric (acrylic, epoxy, organosilicone) and non-organic (based on liquid glass), and also aluminum foil of 30 µm thick were the subjects of this study. The KMU-4* epoxy CFRP composites in the form of plates (60x60x2 mm) and KMU-3l* in the form of skins for three-layered honeycomb specimens (60x60x15 mm) served as substrates for the coatings. Prior to the coating deposition process KMU-3l CFRP three-layered specimens with skins underwent additional thermal stabilization to eliminate any volatile substances (air, sorbed water vapor, organic solvent residuals, etc.) that may evolve. The specimens of above-mentioned CFRP’s with the protective coatings were exposed in “Komplast” removable cassettes on the “Kvant-2” module outside of Space Station “MIR” (Figure 1) for 839, 1024 and 1218 days, and returned back to Earth. A comparative study of the surfaces of thermal control protective coatings prior to and after the natural exposure in space was conducted using optical and scanning electron microscopy, X-ray microanalysis and visual inspection. The micro-structural analysis was performed using the JSM-35C SEM (“JEOL”) microscope and MBS-10 optical microscope with magnifications from x4 to x5000. A thin layer of gold (15-20 nm) was deposited using a sputter deposition vacuum system, FC-1100 FINE COAT (“JEOL”), before the structural examination by SEM to prevent charging. The surface chemistry of the pristine coatings and of the specimens, subjected to full-scale test in the space was determined by X-ray microanalysis using the JXA-840 micro-analyzer (“JEOL”) and energy-dispersive detector (“LINK”). The results are presented below.

*

traditional Russian types

219

Figure 1. “Komplast” removable cassette with protective coating on composite specimens on “Kvant-2” module exterior surface of Space Station “MIR”

3.0 Results and Discussion Epoxy coating. The homogeneous surface structure, as seen in Figs. 2a and 2b, is typical for the initial white epoxy coating both on KMU-4 CFRP and KMU-3l CFRP skin.

Figure 2. Microstructure of the initial (a, b) and shielded by another specimen surfaces of the white epoxy coating on KMU-4 CFRP after 1218 days (c, d) of exposure on Space Station “MIR”: a, c – x2000; b, d – x5000.

220

Powder-like, rounded filler particles (bright-contrast spots) are distributed comparatively uniformly in the film-forming polymer, with their sizes not exceeding 1 µm. No defects like microcracks, de-lamination, etc. are present in the initial coating. According to X-ray elemental microanalysis (Fig. 3a), the epoxy coating surface layer contained titanium (most probably in the form of titanium dioxide pigments) and silicon, magnesium and aluminum in smaller concentrations.

Figure 3. X-ray energy-dispersive microanalysis spectra of the white epoxy coating surface on KMU-4 CFRP prior to (a) and after 1218 days (b) of exposure on “MIR” orbital complex.

221

Visual examination of KMU-4 CFRP and KMU-3l exposed specimens (as well as shielded specimens) for 1218 days on “MIR” orbital complex showed that the morphology of the coating on the thermo-stabilized specimen surfaces remained very close to the initial, without the formation of visible macro-defects (Figure 4a). However, KMU-4 CFRP specimens that were not thermo-stabilized exhibited local macro-defects in the coating in the form of extended cracks with partial delamination and lifting, scale-like formations and ring-like de-laminations from the CFRP surface (Figure 4b). It should be mentioned that the observed local swelling and de-lamination also occur on shielded KMU-4 CFRP specimens, but to a considerably lesser extent as compared to unshielded specimens (Figure 4b).

Figure 4. General view of the surface of epoxy (a, b, c) and acrylic (d) protective coatings of KMU-4 CFRP specimens prior to (a) and after 1218 days (b, c, d) of exposure on “MIR” ; x1.

It was established from electron microscopy analysis that the shielding of KMU-4 CFRP specimen with epoxy coating by another similar specimen for 1218 days completely protected the coating surface from the degradation and erosion, with the structure of the shielded coating only negligibly changed (Figure 2b, d). The surface of the open specimen coating shielded within the period of 1218 days of exposure by a cassette frame (Figure 5a) was covered by a network of randomly distributed microcracks, with the structure of the matrix with dispersed in it filler particles being retained (Figure 5b). The matrix film-forming epoxy polymer on the

222

open coating surface, opposite to the surface covered by the frame, on the other hand, was subjected to degradation and erosion. As a result, the filler particles became loose and were partially removed from the surface. (Figure 5c, d). Also, as a consequence of the degradation processes and the matrix polymer coating film erosion, the surface became enriched in filler particles, giving rise to stronger signals of titanium, silicon and magnesium in the EDS spectra collected from the surface (Figure 3b).

Figure 5. Microstructure of the white epoxy coating surface on KMU-4 CFRP after 1218 days exposure on “MIR” orbital complex: a – surface, covered by a cassette frame, x200; b – surface, covered by a cassette frame, x2000; c – open surface, x2000; d - open surface, x5000.

A comparative investigation of pristine (Figure 6a) and exposed (Figure 6b, c, d) surfaces of the epoxy coating on the CFRP (KMU-3l) skin in the three-layered honeycomb specimen showed, that their microstructure and elemental composition changed in the same way as the KMU-4 CFRP after 839 (Figure 6b, c) and 1218 (Figure 6d) days of exposure. Acrylic coating. The surface structure of acrylic coating on CFRP (KMU-3l) skin in the three-layered honeycomb specimen prior to the space exposure (Figure 7a) doesn’t contain macro- and micro-defects, the filler particles are distributed uniformly. X-ray analysis of this coating showed (Figure 8a), that it’s characteristic feature is the content of titanium and aluminum, which is related to the filler composition. The visual inspection of acrylic coating specimens on KMU-4 CFRP surface and KMU-3l CFRP skin in three-layered honeycomb specimens, exposed for 839 and 1218 days showed, that the coating was retained on the majority of specimens without visible macro-defects.

223

Figure 6. Microstructure of the white epoxy coating surface on (KMU-3l) CFRP skin in the three-layered honeycomb specimen prior to (a) and after 839 (b, c) and 1218 (d) days of exposure on “MIR” orbital station; x2000.

At the same time macro-defects in the form of scale-like swelling-up and cracks with local de-lamination (Figure 4d) were observed on KMU-4 CFRP specimens with acrylic coating, having the lesser thermo-stabilization degree as well as with epoxy coating though in this case it occurs more rarely. The electron microscopic study of acrylic coating specimens on KMU-3l CFRP after 839 days of space exposure (Figure 7b) has shown the fine coating structure was changed insignificantly (only a weak polymer film erosion and negligible filler particle stripping took place). Due to the acrylic coating structural stability, its elemental composition practically did not change after 839 days of exposure in space, as compared to the pristine. An increase in the exposure time to 1218 days (Figure 7c, d) leads to the appearance of separate microcracks on the coating surface The electron microscopic study of acrylic coating specimens on KMU-3l CFRP after 839 days of space exposure (Figure 7b) has shown the fine coating structure has changed insignificantly (only a weak polymer film erosion and negligible filler particle stripping took place). Due to the acrylic coating structural stability, its elemental composition practically did not change after 839 days of

224

exposure in space, as compared to the pristine. An increase in the exposure time to 1218 days (Figure 7c, d) leads to the appearance of separate microcracks on the coating surface and some indication of erosion of the polymer film, however, the elemental surface layer composition is slightly changed (only small amount of compounds with silicon was revealed, Figure 8b).

Figure 7. Microstructure of the white acrylic coating surface on (KMU-3l) CFRP skin in the threelayered honeycomb specimen prior to (a) and after 839 (b) and 1218 (c, d) days of exposure on “MIR” orbital station; x2000

225

Figure 8. X-ray energy-dispersion microanalysis spectra of the white acrylic coating surface on (KMU-3l) CFRP skin in the three-layered honeycomb specimen prior to (a) and after 1218 days (b) of exposure on “MIR” orbital complex.

Figure 9. Microstructure of the open (a) and shielded (b) surfaces of the white acrylic coating surface on KMU-4 CFRP after 839 days of exposure on “MIR” orbital complex; x2000.

226

Figure 10. X-ray energy-dispersion microanalysis spectra of open (a) and shielded by another specimen (b) surfaces of the white acrylic coating surface on KMU-4 CFRP after 1218 days of exposure on “MIR” orbital complex.

A similar behavior of structural transformations of acrylic coating at the macro- and micro-levels is observed after 839 days of exposure in space on KMU-4 CFRP specimens both on the open surface (Figure 9a) and the one shielded by another specimen (Figure 9b). The presence of small amount of zinc (Figure 10b) is observed in the elemental composition of protective acrylic coating, containing titanium, aluminum, and silicon compounds on the shielded surfaces and the content of which is somewhat larger on the open coating surface (Figure 10a). Organo-silicone coating. The micro-heterogeneous structure (Figure 11a), in which the continuous dispersion medium (silicon-organic polymer) contains multiple inclusions of dispersed phase (powder-like filler particles) is the characteristic feature for the thermal control organo-silicone coating surface prior to the space exposure.

227

The coating surface has the roughness due to salient particles and their aggregates of dispersed filler. The size of such particles may be up to ~ 1 µm and their aggregates are up to 10-15 µm. The micro-cracks network (Figure 11b) appeared on the surface of organo-silicone coatings after 1024 days of exposure of honeycomb specimens in the space, however, it’s fine structure also was noticeably changed and become more friable and heterogeneous (Figure 11b). It should be mentioned, that any new structural elements were not been revealed on the coating’s surface after the long-term exposure in space. The X-ray energydispersion microanalysis of thermo-regulating organo-silicone coating, subjected to the long-term exposure in space showed that it included silicon and zirconium compounds (silicon is the polymer matrix base and zirconium is included into the powder-like filler composition (Figure 12a)).

Figure 11. Microstructure of the white thermo-regulating organo-silicone coating surface on (KMU-3l) CFRP skin in the three-layered honeycomb specimen prior to (a) and after 1024 days (b, c) of exposure on “MIR” orbital complex: a, c – x2000; b - x30.

228

The inorganic coating structure was negligibly changed after 1024 days of space exposure (Figure 13b). The fine-dispersed filler particle shape and sizes are retained on the coating surface without changes: microcracks are absent and the micrometeoroid particle effect traces or the condensed deposit were not revealed. According to X-ray spectrum microanalysis (Figure 12b) the inorganic coating includes the large amount of compounds, consisting of silicon, zinc and potassium atoms as well as chlorine- and titanium-containing compounds but in lesser amount into its composition. It’s known, that silicon and potassium are the base of liquid glass, entering into coating, and the zinc and titanium availability points to the content of fine-dispersed fillers in it (zinc and titanium oxides), ensuring the white color to coating.

Inorganic coating. The thermo-regulating inorganic coating on the base of liquid glass on KMU-3l CFRP surface before the space exposure (Figure 13a) has the homogeneous structure consisting of friable packed fine-dispersed (less than 1 µm) particles of a filler with the asimmetric shape. The filler particles of larger size (to 10 µm) along with some separate microcracks also can be though not so often. Aluminum foil. The aluminum foil macrostructure on the back KMU-4 and KMU3l CFRP specimen surfaces was considerably changed after 1218 days of space exposure. Rounded de-lamination (Figure 14a) in the form of swelling-up and folds, repeating the honeycomb cell configuration were formed in the foil (Figure 14). The skin surface composition with the aluminum foil wasn’t practically changed after the space exposure and it was determined only by the aluminum content and small silicon impurity.

229

Figure 12. X-ray energy-dispersion microanalysis spectra of the thermo-regulating inorganic coating surface (a) and organo-silicone coating (b) on (KMU-3l) CFRP skin in the three-layered honeycomb specimen after 1024 days of exposure on “MIR” orbital station.

230

Figure 13. Microstructure of the thermo-regulating inorganic coating surface on (KMU-3l) CFRP skin in the three-layered honeycomb specimen prior to (a) and after 1024 days (b) of exposure on “MIR” orbital complex; x2000.

231

Figure 14. Microstructure of the aluminum foil surface on KMU-4 CFRP after 1218 days of exposure on “MIR” orbital complex: a – x4; b – x20.

The observed changes of coating surface macrostructure (epoxy, acrylic and aluminum foil) are obviously associated with the liberation of gaseous lowmolecular compounds out of CFRP under the space conditions (air, absorbed water vapors, organic solvent residuals and other volatile substances), which can’t diffuse

232

through the coating film or foil because of their high tightness (low gas permeability). The liberated gaseous substances are accumulated under the coating during the long-term space exposure, break its adhesion to CFRP surfaces due to swelling-ups, so that local de-laminations and cracks are formed. At the micro-level the coating degradation under the space factor effect begins with the micro-crack net formations in the film-forming polymer, its erosion, which causes the stripping of powder-like filler particles, being placed on the surfaces. Most likely, the coating cracking is associated with the high inner stresses, appearing in the material during the long-term space exposure with the large thermo-cycle quantity (18 thermocycles per one day of flight) because of the difference of thermal linear expansion coefficients (TLEC), the elasticity of coating and CFRP on the surfaces of which it was deposited. 4.0 Conclusions The study of resistance for different types (epoxy, acrylic, silicon-organic, nonorganic) of protective thermo-regulating (thermal control) coatings on epoxy CFRPs composites and three-layered honeycomb specimens were conducted by SEM, X-ray spectrum microanalysis and visual inspection methods after the longterm exposure in the “Komplast” removable cassette composition on the exterior “Kvant-2” module surface of Space Station “MIR”. The comparative evaluation of white thermo-regulating coating surfaces based on acrylic and epoxy polymers is given prior to and after 839 and 1218 days of exposure as well as on the base of organo-silicone polymer, inorganic coating based on liquid glass and aluminum foil after 1024 days of space exposure. Different types of macro- and micro-defects appeared in coatings were revealed, in particular, long-length cracks with the partial coating de-lamination from a substrate, thin microcracks, scaly swelling-up and ring-like de-lamination. The elemental composition changes of coating surfaces and also substances, deposited on them during the long-term exposure in the space were studied. The feasible coating degradation mechanism was considered with the account of the complex space factor effect, chemical nature of coatings and their substrates. 5.0 References 1. Deev I. S., Nikishin E.F.) Effect of outer space factors on the polymer composite structure under longterm staying conditions in the nearearth orbit. Space Forum, OPA. 1 (1996), pp. 297-302 2. Deev I.S., Nikishin E.F.) Effect of long-term exposure in the space environment on the microstructure of fibre-reinforced polymers. Composites Science and Technology. 57 (1997), pp. 1391-1401 3. Shalin R.E., Minakov V.T., Deev I. S., Nikishin E.F. Study of polymer composite specimens surface changes after the long-term exposure in space. Proceedings of the 7th International Symposium on "Materials in a Space Environment", Toulouse, France, 16-20 June 1997 (SP-399, August 1997), pp. 375-383 4. Barbashev E.A., Bogatov V.A., Deev I.S., Dorofeev Yu.I, Konkin N.I., Milinchuk A.V., Naumov S.F., Nikishin E.F., Perov B.V., Skurat V.E. The contribution of different factors of a space environment in the changes of properties polymer materials. In: "Proc. of the Sixth International Symposium on “Materials in a Space Environment", ESTEC, Noordwijk, The Netherlands, 19-23 September, 1994, pp. 235-238 5. Startsev O.V. & Nikishin E.F. Structure and properties of polymeric composite materials during 1501 days outer space exposure at «Salut-7» orbital station, in: Proc. of the Third LDEF Symposium, Williamsburg, Wirginia, November 1993, pp. 8-12.

233 6. Startsev O.V. & Nikishin E.F. Properties of adhesive compounds of polymeric composite materials and thermoplastic polymers during 1501 days of outer space exposure, in: Proc. of the Sixth International Symposium on “Materials in a Space Environment”, ESTEC, Noordwijk, The Netherlands, 19-23 September 1994, pp. 223-235. 7. Naumov S.F., Gorodetsky A.A., Sokolova S.P., Demidov S.A., Kurilenik A.O., Gerasimova T.L. Study on materials and outer surface coatings aboard space station “MIR”. Proceedings of the 8-th International Symposium on "Materials in a space environment”/ Proceedings of the 5th International Conference on "Protection of Materials and Structures from the LEO Space Environment". Arcachon, France, 5-9 June 2000. 8. Kleiman J. and Iskanderova Z., Technological aspects of protection of polymers and carbon-based materials in space. Proceedings of the 8th International Symposium on "Materials in a space environment"/Proceedings of the 5th International Conference on "Protection of Materials and Structures from the LEO Space Environment". Arcachon, France, 5-9 June 2000.

This page intentionally left blank

ISSUES AND CONSEQUENCES OF ATOMIC OXYGEN UNDERCUTTING OF PROTECTED POLYMERS IN LOW EARTH ORBIT BRUCE A. BANKS, AARON SNYDER, SHARON K. MILLER NASA Glenn Research Center Cleveland, OH 44135, USA RIKAKO DEMKO Cleveland State University Cleveland, OH 44115, USA

Abstract Hydrocarbon polymers that are exposed to atomic oxygen in low Earth orbit are slowly oxidized which results in recession of their surface. Atomic oxygen protective coatings have been developed which are both durable to atomic oxygen and effective in protecting underlying polymers. However, scratches, pin window defects, polymer surface roughness and protective coating layer configuration can result in erosion and potential failure of protected thin polymer films even though the coatings are themselves atomic oxygen durable. This paper will present issues that cause protective coatings to become ineffective in some cases yet effective in others due to the details of their specific application. Observed in-space examples of failed and successfully protected materials using identical protective thin films will be discussed and analyzed. Proposed approaches to prevent the failures that have been observed will also be presented.

1. 0 Introduction The use of atomic oxygen protective coatings applied over conventional polymers that have traditionally been used in space has been the primary approach to date to achieve atomic oxygen durability in space. Metal atoms or metal oxide molecules have been used extensively for the protective coating materials. Typically silicon dioxide, fluoropolymer-filled silicon dioxide, aluminum oxide or germanium have been sputter deposited on polymers to provide atomic oxygen protection. For example, the large solar array blankets on International Space Station have been coated with 1300 angstroms of SiO2 for atomic oxygen protection [1]. Although protective coatings can provide excellent atomic oxygen protection of hydrocarbon or halocarbon polymers, the details of how the coatings are used and/or applied can result in widely varying protection consequences.

235

236 2. In-Space Protective Coatings Experiences 2.1 EUROPEAN RETRIEVABLE CARRIER (EURECA) The EURECA spacecraft, which was deployed into low Earth orbit on August 2, 1992 and retrieved after 11 months on June 24, 1993, was exposed to an atomic oxygen fluence of approximately 2.3x1020 atoms/cm2 [2]. To assist in its retrieval, the spacecraft used two thin adhesively mounted acrylic optical retro-reflectors for laser range finding. Prevention of atomic oxygen attack of the retroreflector surfaces, which would have degraded the specularity of the reflectance, was accomplished by coating the retroreflector surface with a ~1000 Angstrom thick film of sputter deposited SiO2 filled with 8% fluoropolymer (by volume). The LEO exposed and retrieved retroreflector was inspected and optically characterized. The results indicated that the protective coating provided excellent protection and the retroreflector performed as planned except in a small 3 cm patch where the protective coating was accidentally abraded prior to flight as a result of handling during preflight ground integration [3]. Figure 1 shows a close up picture of the retro-reflectors as well as their appearance during illumination after retrieval.

Figure 1. EURECA retro-reflectors after retrieval close up and during illumination.

2.2

INTERNATIONAL SPACE STATION (ISS) RETROREFLECTORS

ISS retro-reflectors, which serve in a similar role as the EURCA retro-reflectors, have been used which employ a glass corner cube retroreflector that is housed in a 10 cm diameter Delrin 100 polyoxymethylene mount. Polyoxymethylene is an oxygen rich polymer that is readily attacked by atomic oxygen. To prevent atomic oxygen attack of the Delrin, the machined polymer surfaces were coated by the same processes, in the same facility and with the same ~1000 Angstrom thin film of sputter deposited 8% fluoropolymer-filled SiO2 that was used for the EURECA retroreflector. Several of these retro-reflectors have been mounted on the external surfaces of the ISS structures at various locations that are exposed to LEO atomic oxygen. Figure 2 shows a close up of one of the coated retro-reflectors prior to use on ISS in space as well as a photograph from space of a retroreflector after attack by atomic oxygen. It is clear from the in-

237 space photograph that the coating was only partially attached allowing direct atomic oxygen attack of the unprotected areas.

Figure 2. ISS retro-reflectors prior to launch and during use in space on ISS after atomic oxygen attack.

2.3 ISS PHOTOVOLTAIC ARRAY BLANKET BOX COVERS Prior to deployment, the ISS photovoltaic arrays were folded into a box that allows the array to be compressed in a controlled manner against a cushion of open pore polyimide foam that was covered with a 0.0254 mm thick aluminized Kapton blanket. The Kapton was coated on both surfaces with 1000 Angstroms of vacuum deposited aluminum. The array was exposed to the LEO atomic oxygen environment from December 2000 through December 2001. Photographs of the array, taken in orbit, indicated that the Kapton blanket had been almost completely oxidized leaving only the thin largely torn aluminization in place as shown in Figure 3.

a. Distant photo b. Close up photo

Figure 3. ISS photovoltaic array showing effects of atomic oxygen erosion of the double aluminized Kapton blanket cover for the ISS photovoltaic arrays box cushions.

238 3.0 Analyses and Discussion 3.1 SURFACE ROUGHNESS AND DEFECT DENSITY The drastic differences in atomic oxygen protection provided by the same SiO2 coating filled with 8% fluoropolymer on the EURECA retro-reflectors and the ISS retroreflectors is thought to be due to drastic differences in the protective coating defect densities. The acrylic EURECA retro-reflectors surfaces were extremely smooth as required to produce high fidelity specular reflections. Such smooth surfaces result in low-defect-density protective coatings that have also been demonstrated, in ground laboratory testing, to perform acceptably. For example smooth surface (air-cured side) Kapton when coated with 1300 Angstrom thick SiO2 resulted in ~ 400 pin window defects/cm2. However, the same coating on the rougher surface (drum-cured side) has been found to result in 3500 pin window defects/cm2 [1]. Similar experiences with graphite epoxy composite surfaces formed by casting against another smooth surface produce defect densities of ~262,300 defects/ cm2 [3]. Surface leveling polymers applied over such surfaces have been found to reduce the defect densities by an order of magnitude to ~22,000 defects/cm2 [3]. The machining of the Delrin 100 (polyoxymethylene) retroreflector mount surfaces produces machine marks or rills in the surface resulting in a highly defected atomic oxygen protective coating. Such rills allow atomic oxygen to oxidize and undercut the high erosion yield Delrin, causing the coating to gradually be left as an unattached gossamer film over the retroreflector mount which could be easily torn and removed by intrinsic stresses and thruster plume loads. The use of smoother surfaces, surface-leveling coatings over the machined Delrin or use of alternative atomic oxygen durable materials could potentially eliminate the observed problem. 3.2 TRAPPING OF ATOMIC OXYGEN BETWEEN DEFECTED PROTECTIVE SURFACES The lack of atomic oxygen protection provided by the aluminized Kapton blanket cover for the ISS photovoltaic arrays box cushion is thought to be due to the trapping of atomic oxygen between the two aluminized surfaces on the 0.0254 mm thick Kapton blanket. Defects in the space exposed aluminized surface allow atomic oxygen to erode undercut cavities. If the undercut cavity extends downward to the bottom aluminized surface, then the atomic oxygen becomes somewhat trapped and has multiple opportunities for reaction until it either recombines, reacts, or escapes out one of the defects in the aluminization. This eventually results in a complete loss of the Kapton with only the aluminized thin film remaining. The vacuum deposited aluminum has a slight tensile stress that causes stress wrinkling of the unsupported aluminum films. Figure 4 is a photograph of a vacuum deposited aluminized Kapton sample that was placed in a radio frequency plasma environment to completely oxidize the Kapton over a portion of the sample.

239

Figure 4. Photograph of a vacuum deposited aluminized Kapton sample bonded to a metal frame after ground laboratory oxidation of the Kapton.

As can be seen in Figure 4, where the ~1000 Angstrom aluminum film in the lower portion of the sample is free standing, stress wrinkles and tears develop similar to those seen in the ISS photograph of Figure 3. A two dimensional Monte Carlo computational model has been developed which is capable of simulating LEO atomic oxygen attack and undercutting at crack defects in protective coatings over hydrocarbon polymers [4]. Optimal values of the atomic oxygen interaction parameters were identified by forcing the Monte Carlo computational predictions to match results of protected samples retrieved from the Long Duration Exposure Facility [4]. These interaction parameters and values were used to predict the consequences of atomic oxygen entering a 2-dimensional crack or scratch defect in the top aluminized surface. This was accomplished using 100,000 Monte Carlo atoms entering a defect which was 20 Monte Carlo cells wide (representing a 13.4 micrometer wide defect) over a 38 cell thick (representing a 0.0254 mm thick) Kapton blanket. Figure 5 compares the Monte Carlo model computational erosion results for a 45-degree angle of attack (relative to the surface normal) of the atomic oxygen for both double surface-coated Kapton (which was the case for ISS) and single top surfacecoated Kapton.

240

a.

b.

Aluminized on both sides

Aluminized on exposed side only

Figure 5. Monte Carlo computational atomic oxygen erosion predictions for a 45 degree from perpendicular angle of attack of atomic oxygen at a crack or scratch defect in the aluminized Kapton surface.

As can be seen from Figure 5, even though the atomic oxygen gradually becomes less energetic with the number of interactions and has approximately a 13% chance of recombination, the trapped atoms undercut far more in the actual ISS case of a double aluminization as would have occurred if the Kapton was simply aluminized on one side. Thus, contrary to intuition, the use of two atomic oxygen protective coatings rather than a single coating appears to cause more rather than less undercutting attack. The extent of undercutting of trapped atomic oxygen is also dependent on the opportunity for the atoms to loose energy, recombine, or escape back out the defect opening. Figure 6 compares the results of 2-dimensional Monte Carlo modeling and 3dimensional pin-window computational predictions [5] for a 45-degree angle of attack atomic oxygen of a 13.4 micrometer wide crack or scratch for the 2-dimensional case and a 5.1 micrometer diameter circular aperture for the 3-dimensional case for both single side and double side aluminized Kapton.

241

Fraction of atoms reacted

0.25

Double-coated

0.20

0.15

0.10

0.05

Single-coated 0.00 0.0E+00

4.0E+20

8.0E+20

1.2E+21

1.6E+21

2.0E+21

Fluence, atoms/cm2

a.

2-Dimentional model of crack or scratch defect

0.20 Fraction of atoms reacted

0.18 0.16

Double-coated

0.14 0.12 0.10 0.08

Single-coated

0.06 0.04 0.02 0.00 0.E+00

2.E+21

4.E+21

6.E+21

8.E+21

2

Fluence, atoms/cm

b.

3-Dimentional model of circular pin window defect

Figure 6. Computational atomic oxygen erosion predictions for 45-degree incident atomic oxygen attack at defect sites protected Kapton.

As can be seen in Figure 6, for both 2-dimensional modeling of a crack or scratch defect and 3-dimensional modeling of a circular defect the growth characteristics of the under cut cavity have similar trends with fluence. Initially, as the undercutting starts the existence or absence of the back surface coating plays no role

242 and as the cavity grows the probability of atoms reacting increases due to trapping of the incoming atom. However, as the bottom surface is reached, atoms begin either to escape, or in the case of no back-surface coating, they recombine after collision with the SiO2 on the back surface. The double surface aluminized Kapton consistently reacts more atomic oxygen atoms than the single surface aluminized Kapton except at very low fluences where the erosion in either case does not reach the bottom of the polymer. For both cases, as the fluence increases, the atomic oxygen can escape out the bottom (only in the case of the single surface aluminized Kapton), recombine, or thermally accommodate and thus becomes less probable to react with the Kapton. Thus it appears that a single surface aluminized Kapton would have been much more durable because the unreacted atoms passing through the bottom of the polymer would simply enter into the open pore foam and gradually react with it, without causing much damage to the aluminized Kapton. The double-SiO2 coated ISS solar array blankets may show similar detachment of the outer surface SiO2 layer with time. However, the defect density appears to be much lower than for vacuum deposited aluminum coatings as shown in Figure 7 which compares the experimental results of RF plasma oxidation of double aluminized Kapton with double SiO2 coated Kapton.

0

Mass change / area, mg/cm2

Double SiO2 Coated Kapton

-0.05

-0.1

Double Aluminized Kapton

-0.15

-0.2

-0.25 0

1E+20

2E+20

3E+20 4E+20 Fluence, atoms/cm2

5E+20

6E+20

Figure 7. Comparison of RF plasma oxidation of aluminized and SiO2 coated Kapton.

4. 0 Conclusions Atomic oxygen protective coatings have been developed and used in space that perform acceptably. However, rough surface substrates cause defects in the protective coatings that allow atomic oxygen to react and gradually undercut the protective coating. In the case of machined Delrin ISS retroreflector mounts, such roughness has lead to detachment of portions of the protective film covering the retroreflector mount.

243 Atomic oxygen undercutting of the double aluminized Kapton blanket covers for the ISS photovoltaic array box cushions has occurred resulting in a torn and partially detached aluminum film. Based on computational modeling, atomic oxygen atoms that become trapped between the two aluminized films on each side of the Kapton blanket appear to cause accelerated undercutting damage in comparison to the use of a single top-surface coating.

5. 0 References 1.

Rutledge S., Olle R., “Space Station Freedom Solar Array Blanket Coverlay Atomic Oxygen Durability Testing,” 38th SAMPE Symposium, May 10-13, 1993.

2.

Banks B. A., Rutledge S., and Cales M., “Performance Characterization of EURECA Retroreflectors with Fluoropolymer-Filled SiOx Protected Coatings”, Long Duration Exposure Facility (LDEF) Conference, Williamsburg, Virginia, November 8-12, 1993.

3. De Groh K., Dever J., and Quinn W., “The Effect of Leveling Coatings on the Durability of Solar Array Concentrator Surfaces,” 8th International Conference on Thin Films and 17th International Conference on Metallurgical Coating,” San Diego, California, April 2-6, 1990. .

4. Banks B., Stueber T., and . Norris, "Monte Carlo Computational Modeling of the Energy Dependence of Atomic Oxygen Undercutting of Protected Polymers," NASA TM 1998207423, Fourth International Space Conference, ICPMSE-4, Toronto, Canada, April 23-24, 1998. 5. Snyder A. and Banks B., “Fast Three-Dimensional Method of Modeling Atomic Oxygen Undercutting of Protected Polymers,” Sixth International Conference on “Protection of Materials and Structures from Space Environment”, Toronto, Canada, May 1-3, 2002.

This page intentionally left blank

EFFECT OF SPACE GASEOUS ENVIRONMENT ON THE THERMOPHYSICAL PROPERTIES OF MATERIALS AND STRUCTURES E. LITOVSKY, J. I. KLEIMAN Integrity Testing Laboratory Inc., 80 Esna Park Drive, Units 7-9, Markham, Ontario, L3R 2R7, Canada; Tel. 905 415-2207, Fax. 905 415-3633, N. MENN LUMINOS Computerized Systems Ltd., Carmel Business Park, Tavor Bldg., Tnufa Str. 3, P.O. Box 104,Tirat Carmel, 39101, Israel

1. 0 Introduction Design and weight of aerospace and space apparatus, energy consumption, temperature regimes and reliability are connected with thermal physical properties of applied materials. Thermal physical properties such as thermal conductivity and thermal diffusivity depend strongly on the environment the materials are used in and on the microstructure of materials. The thermophysical properties of materials under the influence of environment may change 2-10 and even more times. The investigation of these properties in conditions of space where high vacuum, drastic temperature changes, outgassing problems, irradiation, etc, remains an acute problem to which not enough attention is paid. To shed more light onto the behavior of thermal conductivity and diffusivity of materials in conditions mentioned above, heat barrier resistance behavior of interfaces containing two major types of materials, i.e. densified and with microcracks, and highly porous insulation materials were investigated at normal pressures and in vacuum condition. The contact heat barrier resistances between mechanical contacting bodies we consider as partial case of the micro cracks analyses. Experimental and theoretical work was conducted and mathematical models were developed allowing estimating the influence of environment and the structure of the material on thermal physical properties of the above materials and structures [1-13]. This paper presents a short review of our current work devoted to solutions of these problems.

2. 0 Basic Theoretical Models In order to understand the conducting experiments and the results described below in section 4, a very brief introduction of the developed basic theoretical models that described the thermal behavior of materials and pores in various environments is given first.

245

246 2.1 APPARENT THERMAL CONDUCTIVITY OF MATERIALS AND PORES Equation (1) below provides the basis for estimations of relative influence of heat transfer mechanisms on apparent thermal conductivity λapp that is governed by all mechanisms of heat transfer [1, 4]:

λapp = λs (1 – Π)3/2 M + λΠΠ1/4 + λrad + λconv

(1)

λs is thermal conduction of the solid phase, λrad, λconv – are radiative and gas convection components respectively, M is a function taking into account the contact heat barrier resistance (HBR) of cracks between grains. λconv =0 in our case. The apparent thermal conductivity of pores in equation (1) can be written as

λΠ = λgas + λem + λs-d

(2)

λgas in Eq.(2) accounts for heat transfer through the gas filling the pores either in a continuum or a free molecular regime; λem denotes the contribution of gas emission due to chemical reactions, evaporation or sublimation, occurring within pores and λs-d denotes the contribution of segregation and surface diffusion of impurities. The quantity λΠ,̓defined in Eq. (1) affects also the HBR coefficient M appearing in Eq.(1). Each of the additive components in Eqs.(1) and (2) is considered below. At atmospheric and higher pressures, the pore size normally greatly exceeds the molecular mean free path and characteristic Knudsen number Kn