Protection of Materials and Structures from the Space Environment: ICPMSE-7 (Space Technology Proceedings, 6) 1580625665, 9781402042812, 1402042817

Citation preview

PROTECTION OF MATERIALS AND STRUCTURES FROM THE SPACE ENVIRONMENT

Space Technology Proceedings VOLUME 6

PROTECTION OF MATERIALS AND STRUCTURES FROM THE SPACE ENVIRONMENT ICPMSE-7

Edited by Jacob I. Kleiman Integrity Testing Laboratory Inc. Markham, Toronto, Canada

A C.I.P. Catalogue record for this book is available from the Library of Congress.

ISBN-10 ISBN-13 ISBN-10 ISBN-13

1-4020-4281-7 (HB) 978-1-4020-4281-2 (HB) 1-4020-4319-8 (e-book) 978-1-4020-4319-2 (e-book)

Published by Springer, P.O. Box 17, 3300 AA Dordrecht, The Netherlands. www.springer.com

Printed on acid-free paper

All Rights Reserved © 2006 Springer No part of this work may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, microfilming, recording or otherwise, without written permission from the Publisher, with the exception of any material supplied specifically for the purpose of being entered and executed on a computer system, for exclusive use by the purchaser of the work. Printed in the Netherlands

CONTENTS

Introduction

xi

Acknowledgements

xiii

Organization

xv

Radiation Effects of Protons and Electrons on Backfield Silicon Solar Cells Z. Hu, S. He, and D. Yang Solar Array Arcing in LEO: How Much Charge is Discharged? D. C. Ferguson, B. V. Vayner, and J. T. Galofaro

1

9

Self-Restoration as SEU Protection Mechanism for Reconfigurable On-Board Computing Platform L. Kirischian, V. Geurkov, I. Terterian, and J. Kleiman

21

Synergistic Effect of Protons and Electrons on Radiation Damage of Methyl Silicone Rubber L. Zhang, S. He, D. Yang, and Q. Wei

35

Influence of Electron Radiation on Outgassing of Spacecraft Materials R. H. Khassanchine, A. N. Timofeev, A. N. Galygin, V. I. Kostiuk, and V. M. Tsvelev Effect of Surface Charging on the Erosion Rate of Polyimide Under 5 eV Atomic Oxygen Beam Exposure M. Tagawa, S. Seikyu, K.-I. Maeda, K. Yokota, and N. Ohmae Influence of Space Environment on Spectral Optical Properties of Thermal Control Coatings V. M. Prosvirikov, A. V. Grigorevskiy, L. V. Kiseleva, A. P. Zelenkevich, and V. M. Tsvelev Mitigation of Thruster Plume-Induced Erosion of ISS Sensitive Hardware C. Pankop, J. Alred, and P. Boeder v

43

51

61

71

vi

CONTENTS

Degradation of Thermal Control Coatings Under Influence of Proton Irradiation L. S. Novikov, G. G. Solovyev, V. N. Vasil’ev, A. V. Grigorevskiy, and L. V. Kiseleva Mitigation of Damage to the International Space Station (ISS) from Water Dumps W. Schmidl, J. Visentine, and R. Mikatarian Investigation of Synergistic Effects of Proton and Electron Radiation on the Dyeing of Optical Quartz Glass H. Liu, S. He, H. Geng, D. Yang, and V. V. Abraimov The Role of “Abnormal” Electron Fluxes with Energy 115 nm) of various intensities on transmittance at 300 nm of 152 μm DC93-500 film Intensity (number of VUV suns)

Exposure level (Thousands of ESH)

Transmittance at 300 nm (%)

1.5

0.2 0.37 0.57 0.7 1.12 1.93 0

79.10 70.77 66.24 64.50 55.43 48.76 92.50

3 5.5 0

figure 5(a). It is also evident that degradation increases with increasing equivalent sun hours of exposure. Table 4 and figure 6 show transmittance at 300 nm (a wavelength at which significant transmittance degradation of DC93-500 is evident, based on figure 5) as a function of exposure, represented by thousands of ESH, for various intensities (number of suns). In the figure, symbols indicate the measured data points, and lines indicate exponential decay curve fits. In addition to examining data for individual intensities, data were considered all together, independent of intensity, as one of the curves shown in figure 6.

100

Transmittance, %

80

60 All Data 1.5 suns 3.0 suns 5.5 suns All Data 1.5 suns 3.0 suns 5.5 suns

40

20

0 0.0

0.5

1.0

1.5

2.0

Thousands of ESH

Figure 6. DC93-500 film transmittance at 300 nm wavelength as a function of VUV exposure duration expressed in thousands of equivalent sun hours

136

JOYCE A. DEVER ET AL.

When transmittance data for all intensities are considered together, the best curve fit equation is given in the format shown in eq. (3): τ = y0 + a exp(−kx)

(3)

where τ is transmittance (%), x is exposure duration (thousands of equivalent sun hours), and y0 , a, and k are constants. Eq. (3) represents an exponential decay reaching an asymptotic value, which is evident from examination of figure 5. We can further determine the expression for transmittance of unexposed materials, or τ0 , by setting x = 0: τ0 = y0 + a

(4)

Using the expression for “a” as a function of y0 and τ0 from eq. (4), and by expressing x (thousands of equivalent sun hours) as the product of equivalent suns, S, and thousands of exposure hours, h, eq. (3) can be rewritten as τ = y0 + (τ0 − y0 ) exp(−k Sh)

(5)

In order to determine whether the rate of transmittance decay shows a dependence upon S, the intensity (expressed as number of equivalent suns), eq. (5) is solved for k, the constant in the exponential expression, to give:   1 τ − y0 k=− (6) ln Sh τ0 − y0 The curve fit for all measured transmittance data at 300 nm (figure 6) produces a value of y0 = 46.7, which is the asymptote being approached by the decay in transmittance and which is assumed to be constant, independent of intensity. If we examine transmittance data at 300 nm for all intensities and plot k vs. S, it is possible to determine whether k is constant, which would mean it is independent of intensity, or whether it varies as a function of intensity, indicating intensity dependence. The values for S and h are obtained from the VUV exposure conditions (table 2), and τ and τ0 values are the measured transmittance data (shown in table 4). The plot of k vs. S is shown in figure 7. Based on the data shown in figure 7, it is evident that there is no statistically significant dependence upon exposure intensity between approximately 1.5 and 5.5 VUV suns. Table 5 and figures 8(a) and 8(b) show mechanical properties of 152 μm DC93500 silicone films as a function of exposure (equivalent sun hours). Both ultimate tensile strength (figure 8(a)) and elongation at failure (figure 8(b)) decrease with increasing exposure and indicate the approach of an asymptotic value near the 2000 ESH exposure level. An exponential decay curve fit is shown in each figure. Based on these data, especially because of similar degradation for similar ESH, regardless of intensity, there is no clear trend indicating an intensity dependence upon the rate of mechanical properties degradation.

137

VACUUM ULTRAVIOLET RADIATION EFFECTS 2.5

2

k

1.5

1

0.5

0 0

1

2

3

4

5

6

S (Number of Suns)

Figure 7. Coefficient, k, described by eq. (6), as a function of VUV exposure intensity for DC93-500 film samples

4. Conclusions DC93-500 silicone has been found to undergo degradation in optical and mechanical properties upon exposure to a laboratory deuterium lamp providing VUV radiation of wavelengths greater than 115 nm. In one experiment, samples of DC93-500 films were exposed to narrow bands of VUV radiation (∼20 nm) in wavelength ranges between approximately 140 and 200 nm by using a broad spectrum deuterium lamp as the VUV source and narrow bandpass filters over TABLE 5. Effect of VUV exposure of various intensities on mechanical properties degradation of DC93-500 Avg. number of VUV suns 5.5

3

1.5 0 (pristine)

Exposure duration (h)

Cumulative VUV equivalent sun hours

No. test samples

Avg. UTS (MPa)

118 356 238 118 379 262 117 379 262 0

700 1900 1200 370 1100 740 200 570 370 0

2 3 3 2 3 3 3 3 3 5

5.9 ± 0.1 3.8 ± 2.0 3.6 ± 0.4 6.2 ± 0.8 4.3 ± 0.6 5.3 ± 0.9 6.2 ± 0.4 5.0 ± 1.0 5.6 ± 1.6 9.2 ± 0.5

Avg. elongation (%) 79.0 ± 2.8 65.3 ± 2.0 69.3 ± 2.1 86.0 ± 2.8 69.0 ± 0 76.3 ± 2.1 91.3 ± 2.5 75.0 ± 4.4 82.3 ± 6.1 139 ± 6.9

138

JOYCE A. DEVER ET AL. 10.0 5.5 suns 3 suns 1.5 suns

9.0 8.0

UTS (MPa)

7.0 6.0 5.0 4.0 3.0 2.0 1.0 0.0 0

200

400

600

800

1000

1200

1400

1600

1800

2000

VUV Exposure (ESH)

(a) 160.0 5.5 suns 3 suns 1.5 suns S i 1

Elongation at Failure (%)

140.0 120.0 100.0 80.0 60.0 40.0 20.0 0.0 0

200

400

600

800

1000

1200

1400

1600

1800

2000

VUV Exposure (ESH)

(b) Figure 8. Mechanical properties of (a) tensile strength and (b) elongation at failure for 152 μm DC93-500 silicone as a function of VUV exposure equivalent sun hours

the DC93-500 samples. Results indicated that each wavelength range used for the exposures produced degradation in DC 93-500 optical properties. However, degradation per incident energy fluence indicated the highest rate of degradation for samples exposed from beneath filters which included wavelengths above 185 nm. Vacuum ultraviolet ellipsometric optical measurements were made on DC93-500 silicone to determine the depth of penetration of vacuum ultraviolet light as a function of wavelength. Data showed that vacuum ultraviolet of wavelengths below

VACUUM ULTRAVIOLET RADIATION EFFECTS

139

185 nm penetrate DC 93-500 to depths no greater than 1 μm, indicating that VUV degradation of these wavelengths can only occur in a shallow layer compared to typical spacecraft polymer film applications which use films on the order of tens to over a hundred micrometers in thickness. Compared to VUV exposures to wavelengths below 185 nm, a significantly more rapid rate of transmittance degradation was observed in DC93-500 for VUV exposures which included wavelengths between 185 and 200 nm, which correspond to depths of VUV penetration between 1 and 3 μm. In another experiment, the rates of optical and mechanical properties degradation for DC93-500 films were examined for exposures to broad spectrum VUV (above 115 nm) of various intensities. It was found that for both transmittance degradation and mechanical properties (ultimate tensile strength and elongation at failure) degradation, loss of these properties followed exponential decay functions approaching asymptotic values. Examination of the data indicated no clear dependence of degradation on the intensity of exposure within a range of intensities between 1.5 and 5.5 VUV suns. The lack of intensity dependence in these data indicates that DC93-500 can be tested using VUV intensities as high as about 5.5 suns without causing significantly different degradation rates compared to near real-time exposure rates. It remains to be determined whether these rates of degradation are similar to those caused by actual space exposure, especially considering the significantly different spectra between the Sun and the laboratory VUV source. It is hoped that space exposure data will eventually be available to make such comparisons:

Acknowledgments The authors gratefully acknowledge the technical support of Frank Lam and James Mazor (Akima), Michael DePauw (NASA), Scott Panko and Edward Sechkar (QSS Group, Inc.), Michael Piszczor (NASA) and Mark O’Neill, Don Spears, and A. J. McDanal (ENTECH, Inc.).

References 1. Dow Corning Corp. (2001) Product Information for Dow Corning® Space-Grade Silicone Sealants, 2001. 2. O’Neill, M. J., Piszczor, M. F., Eskenazi, M. I., McDanal, A. J., George, P. J., Botke, M. M., Brandhorst, H. W., Edwards, D. L., and Hoppe, D. T. (2003), In International Symposium on Optical Science and Technology, SPIE’s 48th Annual Meeting, SPIE Paper no. 5179–17, August 2003. 3. Edwards, D. L., Finckenor, M. M., O’Neil, M., and McDanal, A. J. (2000) In 8th International Symposium on Materials in a Space Environment, Arcachon, France, 2000.

140

JOYCE A. DEVER ET AL.

4. Dever, J., Messer, R., Powers, C., Townsend, J., and Wooldridge, E. (2001) High Performance Polymers 13(3), S391–S399. 5. Dever, J., Semmel, C., Edwards, D., Messer, R., Peters, W., Carter, A., and Puckett, D. (2002)Radiation Durability of Candidate Polymer Films for the Next Generation Space Telescope Sunshield, AIAA 2002-1564, April 2002. 6. Dever, J. and McCracken, C. (2004) High Performance Polymers 16(2), 289–302. 7. American Society for Testing and Materials (Reapproved 1992). Solar Constant and Air Mass Zero Solar Spectral Irradiance Tables, ASTM-E 490-73a. 8. Dever, J. A. (1994) Flight-vehicle materials, structures, and dynamics—Assessment and Future Directions, Vol. 2, Advanced Metallics, Metal-Matrix and Polymer-Matrix Composites, American Society of Mechanical Engineers, New York, 1994, pp. 422–433. 9. Adams, M. R. (1993) The Degradation of Polymeric Spacecraft Materials by Far-UV Radiation and Atomic Oxygen, UMI Dissertation Services, Ann Arbor, MI, 1993, p. 138. 10. American Society for Testing and Materials. (1995) Standard Test Method for Tensile Properties of Plastics, ASTM D 638–95. 11. Herzinger, C. M., Snyder, P. G., Johs, B., and Woollam, J. A. (1995) Journal of Applied Physics 77(4), 1715–1724. 12. Dever, J. A., Pietromica, A. J., Stueber, T. J., Sechkar, E. A., and Messer, R. K. (2001) Simulated Space Vacuum Ultraviolet (VUV) Exposure Testing for Polymer Films, AIAA Paper no. 2001– 1054, American Institute of Aeronautics and Astronautics, January 2001.

ENHANCEMENT OF ATOMIC OXYGEN-INDUCED EROSION OF SPACECRAFT POLYMERIC MATERIALS BY SIMULTANEOUS ULTRAVIOLET EXPOSURE KUMIKO YOKOTA,∗ NOBUO OHMAE, AND MASAHITO TAGAWA∗ Department of Mechanical Engineering, Faculty of Engineering, Kobe University, Rokko-dai 1-1, Nada, Kobe 657-8501, Japan

Abstract. Synergistic effect on atomic oxygen-induced erosion of polyethylene and polyimide with 172 nm vacuum ultraviolet was investigated using a quartz crystal microbalance. In order to adjust the relative intensity of atomic oxygen and vacuum ultraviolet, the sample was rotated with an axis perpendicular both to the axes of atomic oxygen and ultraviolet. The erosion rate of polymers by ultraviolet exposure alone is independent of the incident angle of ultraviolet, whereas that by atomic oxygen alone follows cosine function. It was observed that the erosion rate of polyethylene increased 30–100% by a simultaneous exposure of 172 nm vacuum ultraviolet and 5 eV atomic oxygen depending on the relative intensity. The erosion of oxygen covered-polyethylene was three times greater than that of nonoxidized polyethylene. These erosion properties suggested that two independent erosion pathways exist in a simultaneous atomic oxygen and ultraviolet exposure condition; the atomic oxygen-induced erosion and the ultraviolet-induced erosion associated with oxidation. Key words: atomic oxygen, ultraviolet, low Earth orbit, synergy, space environment effect

1. Introduction There are many environmental factors in low Earth orbit (LEO) such as microgravity, thermal cycling, plasma environment (ions and electrons), ultraviolet, radiation, neutral gas, high-energy charged particles, and space debris. However, it has been recognized that atomic oxygen is one of the most important hazards to

∗ Corresponding

authors

141 J.I. Kleiman (ed.), Protection of Materials and Structures from Space Environment, 141–152.  C 2006 Springer. Printed in the Netherlands.

142

KUMIKO YOKOTA ET AL.

the spacecraft polymeric materials in LEO. Moreover, synergistic effect of atomic oxygen and other space environmental factors on polymer erosion is unsolved problem in space engineering. Since atomic oxygen fluence in a material exposure testing is measured by the erosion depth or mass loss of a polymer witness sample, the enhancement of atomic oxygen-induced erosion of witness sample by simultaneous ultraviolet or other environmental factors spoils the accuracy of atomic oxygen fluence measurement. Polyimide has been widely used as a witness sample, but polyethylene is also a candidate of a reference material because of its simplest possible chemical structure in all polymers [1]. The reaction efficiencies of polyethylene and polyimide with atomic oxygen in LEO have been established to be 3.7 × 10−24 and 3.0 × 10−24 cm3 ·atom−1 , respectively [2]. The fluence of atomic oxygen in LEO is calculated from the polyimide erosion with the established reaction efficiency. This analysis is based on the hypothesis that polyimide is not influenced by the synergistic effect of atomic oxygen and ultraviolet [3, 4]. On the other hand, in a ground based experiment a 30% increase in erosion of a polymer by simultaneous exposure to ultraviolet radiation and atomic oxygen is reported [5]. The erosion properties of polyethylene and polyimide under various space environmental factors need to be well understood because they are used as a reference material. In this paper, we focused on the effect of the relative intensity between atomic oxygen and 172 nm ultraviolet on the erosion of polyethylene and polyimide in ground-based experiments. The relative intensity of atomic oxygen and ultraviolet was adjusted by changing the incident angle of atomic oxygen and ultraviolet with respect to the sample surface. The erosion rate of atomic oxygen-exposed polymers was measured from the change in resonant frequency of quartz crystal microbalance (QCM) during atomic oxygen and/or ultraviolet exposures. The surface properties of atomic oxygen-exposed polymers were analyzed by X-ray photoelectron spectroscopy (XPS).

2. Experimental Details The samples used in this experiment were low-density polyethylene (LDPE) and polyimide films. Both of the films were spin-coated on a QCM sensor crystal. The polyethylene solution containing 0.3 g of LDPE (average molecular weight: 6500) in 40 ml xylene was prepared for polyethylene film. The pyromelliticdianhydride (PMDA)-oxydianiline (ODA) polyimide, which was supplied by Toray Industries Inc. (Semicofine SP-510), was used as a polyimide sample. Precursor of PMDAODA polyimide was spin-coated on a QCM sensor crystal and then annealed at 150◦ C and at 300◦ C. Details of the sample preparation are reported in [6]. The polyimide film, thus prepared, showed an XPS spectrum similar to that of Kapton-H film, which is commercially available polyimide film.

ENHANCEMENT OF ATOMIC OXYGEN-INDUCED EROSION

143

XPS

Temperature-controlled Rotatable QCM

Nozzle AO

PSV Laser

QMS TOF System

Au Mirror UV Source (excimer 172 nm)

Sample

Figure 1. A schematic drawing of the space environmental simulation facility using a laser detonation atomic oxygen beam source

A space environment simulation facility at Kobe University was used in this study. The schematic drawing of the facility is shown in figure 1. This facility equipped with a laser detonation atomic oxygen beam source, which was originally designed by Physical Sciences Inc., as a hyperthermal atomic oxygen source. The atomic oxygen beam was produced using a carbon dioxide laser (wavelength: 10.6 μm, output power: 5–7 J·pulse−1 ). The translational energy of the atomic oxygen was approximately 5 eV. The flux of the atomic oxygen beam was calculated to be 2 × 1014 atoms·cm−2 ·s−1 at the sample position. These values are almost equivalent to those in LEO at the altitude of 200–300 km. An excimer light source with a wavelength of 172 nm was used as an ultraviolet UV source in this study. This ultraviolet source was attached to the atomic oxygen source chamber. The sample was exposed to the UV through an evacuated light guide in order to avoid absorption of vacuum ultraviolet by air (wavelength of 1 × 10−14 mJ·atom−1 ). However, this synergistic effect was not observed in the range of UV/AO ratio < =10−15 mJ·atom−1 , as demonstrated in figure 7(a); atomic oxygen-induced erosion data with ultraviolet (empty circle) and without ultraviolet (solid circle) are identical. In contrast, polyethylene shows clear effect of simultaneous ultraviolet irradiation even in the range of UV/AO ratio < =10−15 mJ·atom−1 as shown in figure 7(b); i.e., the erosion rates with ultraviolet exposure (empty circle) are 30–100% greater than those without ultraviolet exposure (solid circle). From the result of figure 7, the relationship between the relative intensity of UV/AO and flux normalized erosion rate of polyethylene was replotted in figure 8. The ordinate was normalized erosion rate, namely that means the erosion rate in the case of atomic oxygen and ultraviolet exposures was divided by those in the case of only atomic oxygen exposure. The erosion rate of atomic oxygeninduced polyethylene was enhanced 150–180% at the relative intensity of UV/AO 0.5–2.6 × 10−15 mJ·atom−1 by simultaneous ultraviolet exposure. It was observed that the erosion rate increased 300% when the ultraviolet intensity was high (not shown). In the case of polyimide, the similar effect by simultaneous ultraviolet exposure was observed at relative intensities one order greater than that of polyethylene [11]. This finding suggests that polyimide is a better material as a witness sample for measuring atomic oxygen fluence in LEO since synergistic effect was only in high UV/AO conditions.

ENHANCEMENT OF ATOMIC OXYGEN-INDUCED EROSION

149

0.025 Residual VUV only least square fits

Erosion Rate (Hz/s)

0.020

0.015

0.010

0.005

0.000 0

10

20

30

40

50

60

70

80

90

100

Incident Angle (deg.)

Figure 9. Erosion Rate of polyethylene-coated QCM under VUV exposure (◦), and the contribution of VUV in the synergistic effect (•). Note that the abscissa represents the incident angle of atomic oxygen

The erosion rates of polyethylene by atomic oxygen beam exposure (solid circle in figure 7(b)) were subtracted from those in simultaneous exposures of atomic oxygen and ultraviolet (empty circle). The results are shown in figure 9. As shown in figure 9, the result is independent of the incident angle at angles greater than 20◦ where the sample surface is fully exposed to UV. The empty circles in figure 9 indicate the erosion rates of polyethylene only by ultraviolet; no atomic oxygen exposure (same data in figure 5). Again, the erosion rates are constant with the incident angle of ultraviolet. Beside the independency of the incident angles, it is obvious that the absolute erosion rates in figure 9 are not consistent, i.e., the results are three times greater than the erosion by ultraviolet only. Even though the structure of low-density polyethylene does not absorb the ultraviolet at 172 nm, it is clear that the presence of atomic oxygen enhanced ultraviolet-induced erosion of polyethylene. Oxidized radicals such as carbonyl or carboxyl group formed by the atomic oxygen-induced oxidation of polyethylene may have a critical role on synergistic effect of atomic oxygen and ultraviolet at 172 nm. From the experimental results shown in figures 8 and 9, it is suggested that two independent erosion pathways exist in a simultaneous atomic oxygen and ultraviolet exposure condition; atomic oxygen-induced erosion which follows cos θ dependence with incident angle and ultraviolet-induced erosion which is independent of the incident angle of ultraviolet. The erosion rate of ultraviolet-induced erosion pathway is accelerated three times under the presence of atomic oxygen

150

KUMIKO YOKOTA ET AL.

in the present experimental condition. Since the absolute reaction rate is greater in atomic oxygen-induced erosion, it was considered that the overall erosion was rate-limited by atomic oxygen. 3.4. EFFECT OF ULTRAVIOLET EXPOSURE ON ADSORBED OXYGEN

In the previous section, the importance of the oxygen atoms adsorbed at the polyethylene surface on the ultraviolet-induced erosion was suggested. XPS was used for analyzing the oxygen adsorbed polyethylene surface. The experiment was carried out with atomic oxygen fluence of 3.0 × 1018 atoms·cm−2 . The sample was exposed to 172 nm ultraviolet radiation (0.45 mW·cm−2 ) for 40 min in vacuum. Figure 10 indicates the C1s core level XPS spectra of atomic oxygen-exposed polyethylene before and after ultraviolet exposure. As clearly observed in figure 10, atomic oxygen-exposed polyethylene showed a high-energy shoulder at 288.5 eV besides the main peak at 284.8 eV. This shoulder is contributed by carboxyl groups. The high-energy peak disappeared under the ultraviolet irradiation. Similar results were obtained from polyimide samples. Table 1 shows the composition of the polyethylene and polyimide surfaces analyzed by XPS. After atomic oxygen exposure, the oxygen composition increased to 27% for polyethylene and to 33% for polyimide. Increase in oxygen composition corresponds to surface oxidation forming carboxyl groups. However, the high-energy shoulder disappeared after ultraviolet exposure, and the surface composition recovered to normal values. This spectral change in XPS is explained by the photodissociation of carboxyl group created by exposure to atomic oxygen [12]. Similar conclusion was also obtained in polyimide [12]. It was, thus, confirmed by XPS that ultraviolet irradiation promotes the decomposition of carboxyl group formed by atomic oxygen exposure at polyethylene and polyimide surfaces. This photochemical reaction is the

2500

(a (a) 2000

(b)

c/s

1500

(c) 1000

500

0

294

292

290

288

286

284

282

280

Binding Energy (eV)

Figure 10. C1S XPS spectra of atomic oxygen-exposed polyethylene before and after ultraviolet exposure. (a): Pristine, (b): AO exposed, (c): AO·UV exposed

ENHANCEMENT OF ATOMIC OXYGEN-INDUCED EROSION

151

TABLE 1. The composition of polyethylene and polyimide before/after atomic oxygen and ultraviolet exposure

Samples Pristine AO exposed∗ AO->UV exposed∗∗ UV exposed∗∗

Polyethylene

Polyimide

Atomic percent (%)

Atomic percent (%)

C

O

C

O

N

92.8 72.8 89.7 92.5

7.2 27.2 10.3 7.5

75.5 61.2 66.9 75.6

17.8 33.3 26.6 16.6

6.7 5.5 6.5 7.8

Polyethylene: ∗ AO fluence: 3.0 × 1018 atoms/cm2 , ∗∗ UV intensity: 0.45 mW/cm2 for 40 min. Polyimide: ∗ AO fluence: 9.6 × 1017 atoms/cm2 , ∗∗ UV intensity: 0.6 mW/cm2 for 4 hr.

origin of the oxygen-enhancement of ultraviolet-induced mass loss of polyethylene observed in figure 9.

4. Conclusion Synergistic effect on atomic oxygen-induced erosion of polyimide and polyethylene with 172 nm vacuum ultraviolet was investigated using QCM crystals coated with polyethylene and polyimide. In order to change the relative intensity of atomic oxygen and vacuum ultraviolet, the sample on QCM was rotated around an axis perpendicular both to the axes of atomic oxygen and ultraviolet. It was found that the erosion rate of polymers by ultraviolet exposure is independent of the incident angle of ultraviolet, whereas that by atomic oxygen follows cosine function. It was observed that the erosion rate of polyethylene increased 30–100% by a simultaneous vacuum ultraviolet exposure depending on the relative intensity of ultraviolet and atomic oxygen. The increase of mass loss is due to the creation of new reaction pathway of ultraviolet-induced decomposition of carboxyl group, which was created by atomic oxygen exposure. Since the synergistic effect of atomic oxygen and ultraviolet is obvious in polyethylene at one order lower relative intensity of UV/AO, polyimide is a better material as a witness sample for measuring atomic oxygen fluence in LEO.

Acknowledgments This study was partially supported by the Grant-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science, and Technology, Japan

152

KUMIKO YOKOTA ET AL.

under contact no. 13750842, 14350511, and 15560686; and the Space Utilization Promotion from the Japan Space Forum. Financial support from Kawanishi Memorial Shinmaywa Education Foundation is also acknowledged. The authors express their acknowledgments to S. Seikyu and K. Maeda of Kobe University for their help with this experiment.

References 1. Standard Practices for Ground Laboratory Atomic Oxygen Interaction Evaluation of Materials for Space Applications, ASTM Designation, E2089-00. 2. Atomic Oxygen Effects Measurements for Shuttle Missions STS-8 and 41-G, NASA TM-100459, 1988. 3. Rutledge, S. K., Banks, B. A., and Kitral, M. (1998) A Comparison of Space and Ground Based Facility Environmental Effects for FEP Teflon, NASA TM 207918. 4. Brinza, D. E., Chung, S. Y., Minton, K. T., and Liang, R. H. (1994) Final Report on the NASA/JPL Evaluation of Oxygen Interactions with Materials-3 (EOIM-3), JPL Publication, pp. 94–31. 5. Koontz, S. L., Leger, L. J., Albyn, K., Cross, J. (1990) Journal of Spacecrafts and Rockets 27, 346–348. 6. Kinoshita, H., Tagawa, M., Umeno, M., and Ohmae, N. (1998) Transaction of the Japan Society for Aeronautical and Space Science 41(132), 94–99. 7. Tagawa, M., Yokota, K., Ohmae, N., and Kinoshita, H. (2002) Journal of Spacecraft and Rockets 39(3), 447–451. 8. Yokota, K., Tagawa, M., and Ohmae, N. (2002) Journal of Spacecraft and Rockets 39(1), 155– 156. 9. Tagawa, M., Yokota, K., Kida, T., and Ohmae, N. (2003) In J. Kleiman and Z. Iskanderova (eds.), Protection of Materials and Structures from Space Environment Space Technology Proceedings, Vol. 5, Kluwer Academic Publishers, Dordrecht, pp. 391–400. 10. http://www1.ushio.co.jp/tech/ 11. Yokota, K., Ohmae, N., and Tagawa, M. (2004) High Performance Polymers 16, 221–234. 12. Scnabel, W. (1992) Polymer Degradation-Principles Practical Applications, Carl Hanser Verlag, Munich.

GROUND SIMULATION OF HYPERVELOCITY SPACE DEBRIS IMPACTS ON POLYMERS R. VERKER,1,2 E. GROSSMAN,1 N. ELIAZ,2 I. GOUZMAN,1 S. ELIEZER,3 M. FRAENKEL,3 AND S. MAMAN3 1 Space Environment Division, Soreq NRC, Yavne 81800, Israel 2 Department of Solid Mechanics, Materials and Systems, Tel Aviv University, Ramat Aviv, Tel Aviv 69978, Israel 3 Plasma Physics Department, Soreq NRC, Yavne 81800, Israel

Abstract. Hypervelocity space debris impacts can lead to degradation of satellite performance and, in extreme cases, might cause a total loss of a spacecraft. The increase in space debris population provides the motivation for this study, which focuses mainly on the mechanical behavior of space-qualified polyimide Kapton films impacted by simulated hypervelocity debris. Kapton is used extensively on spacecrafts, especially in thermal control blankets. Kapton films 25, 50, and 125 μm-thick were studied at different impact velocities of up to 2900 m·s−1 generated by a laser driven flyer (LDF) system. The Kapton-impacted sites revealed ductile-type fractures for low-velocity debris, which changed gradually into mixed ductile–brittle fractures with crack formation when debris impact velocity was increased. Fractures created by impacts at velocities above 1700 m·s−1 showed central impact regions which experienced the highest strain rate and revealed a ductile-type fracture, while the outer regions which experienced a lower strain rate failed through brittle cracking. A model explaining this phenomenon, based on the temperature profile developed within the impacted region at the time of impact, is presented. Key words: Hypervelocity Impact Simulation, Polymer damage, Space Debris

1. Introduction Many of the satellites nowadays are being launched into low Earth orbit (LEO), ranging from 200 to 700 km. The LEO space environment possesses many obstacles to a successive spacecraft mission. These obstacles include ambient space conditions such as ultrahigh vacuum, as well as man-made obstacles such as spacecraft debris. In LEO, spacecrafts are subjected to various destructive environmental components, such as ionizing radiation (electrons, protons), 153 J.I. Kleiman (ed.), Protection of Materials and Structures from Space Environment, 153–165.  C 2006 Springer. Printed in the Netherlands.

154

R. VERKER ET AL.

vacuum ultraviolet (VUV) radiation, hyperthermal atomic oxygen (AO), and other factors such as extreme temperature variations, micrometeoroids, and orbital debris. Due to either singular or synergistic interactions with these space components, structural materials—in particular polymer-based materials—suffer a relatively rapid erosion (mass loss), structure modification, and surface roughening, leading to irreversible degradation of their physical characteristics (optical, thermal, electrical, and mechanical) [1, 2]. Therefore, a careful selection of surface materials, namely polymer films and paints, is required [3]. Micrometeoroids originate naturally from planetary or asteroidal collisions and cometary ejecta [4]. The large debris population at LEO altitudes comprises the waste products of spacecraft operations. Typical velocities of debris particles range from few kilometers per second up to 16 km·s−1 , making these particles a threat to spacecrafts. The debris issue must be quantified over the projected lifetime of a space system to determine the life expectancy of exposed systems and to quantify necessary shielding requirements [5]. Artificial space debris consists of large objects such as spent satellites and rockets, and mostly of small objects such as aluminum oxide fuel particles, paint chips, and fragmentation objects from collisions of these bodies in orbit [4]. The recovery of several spacecrafts in the last decade offers information concerning the directionality of the LEO meteoroids and space debris fluxes [6]. Such recovered spacecrafts and spacecraft’s parts include one of the Hubble space telescope solar arrays (retrieved in 1993), the European retrievable carrier—EURECA (retrieved in 1993), and the long duration exposure facility—LDEF (retrieved in 1990 after 69 months in LEO). Spacecraft debris impact damage can degrade the performance of exposed spacecraft materials and, in some cases, destroy a satellite’s ability to perform or complete its mission [3]. The Hubble space telescope solar array, for example, suffered impacts at ultrahigh velocities ranging from 2.9 to 11.5 km·s−1 from particles 7 to 98 μm in diameter [7]. Particles traveling at ultrahigh velocities generate temperatures in the range of 5000 K and pressures of several megabars when they collide with a surface [8]. Accumulation of impacts over the large surface area of solar panels leads, in some cases, to degradation in efficiency [9]. Impacts into metals form craters, which have diameters averaging about 5 times the impact diameter. These craters are of concern because they can prevent impacted components from operating. In the case of composites, if a complete penetration occurs, this can lead to further breakdown of the composite during subsequent exposure to AO or VUV. Debris impacts into polymer films occurs quite often, since they are used extensively onboard spacecrafts, mainly as thermal blankets. Mostly, these materials are thin laminated layers; thus, the impacts cause delamination of these layers into many times the diameter of the crater [3].

GROUND SIMULATION OF HYPERVELOCITY SPACE DEBRIS IMPACTS 155

Thermal control materials on the LDEF have demonstrated a significant synergism of orbital debris with other space environmental components. Such synergism further expanded the damaged areas caused by impacts. For example, the top surface of a metalized Mylar sample aboard the LDEF was completely eroded, exposing the interior surfaces to VUV radiation, AO, and thermal cycling. As the number of missions sent into LEO is increasing, the frequency of debris impacts is expected to increase as well [3]. Such an increase may lead to further complications in operation of satellites in LEO environment. These complications may be in the form of (i) accelerated development of molecular and particle contamination, and (ii) an increased change in optical and mechanical properties due to debris impact. Thermal blankets that cover large parts of a spacecraft, will particularly be subjected to these changes. The expected increase in impacts frequency and the amount of polymeric thermal blankets onboard spacecrafts provides the main motivation of this study. The study deals mainly with the mechanical behavior of space-qualified polymer films subjected to ultrahigh velocity impacts.

2. Experimental 2.1. THE LASER DRIVEN FLYER METHOD

The laser driven flyer (LDF) method was used for simulating space hypervelocity debris with dimensions ranging from 10 to 100’s μm and velocities of up to 2900 m·s−1 . LDF is attractive as an acceleration technique for debris simulation due to its relative simplicity, relatively low cost, ease of incorporation into a vacuum facility, and high shot rate capability [8, 10, 11]. Figure 1 shows a schematic diagram of the LDF process. In this method, a high-intensity laser beam is shot into a metal foil (3–25 μm-thick) tenaciously bonded to a glass substrate (hereafter referred as target). The beam passes through the glass and hits the aluminum–glass interface. At the interface, high-pressure

Plasma (Shock wave) Laser beam

Glass/Aluminum laminate

Figure 1. Schematic description of the laser driven flyer (LDF) process

156

R. VERKER ET AL.

plasma is formed at the range of giga-Pascal (GPa). The instantaneous high pressure generates a compressive shock wave that propagates into the aluminum faster than the speed of sound. As the shock wave reaches the aluminum surface and the laser pulse ends, two (tensile) rarefaction waves are generated propagating one toward the other. When the rarefaction waves, and their tension pressure is higher than the spall pressure, a spall is produced. A pressure gradient between the plasma pressure on one side and the vacuum zero pressure on the other side causes the spalled layer acceleration which results in an aluminum layer to fly away at ultrahigh velocity. The aluminum flyers in this work were accelerated toward polymer samples located at a selectable distance of 2–12 mm from the laminate structure. The whole system was placed inside a vacuum chamber operating at a base pressure of 20 mTorr. Soreq’s LDF system is using a Titanium:Sapphire laser operating under a single-pulse mode at 810 nm wavelength. The length of each pulse is 300 ps, and the energy of the pulse can be controlled within the range of 250–750 mJ. The experiments were carried out with single laser shots. After each shot, a new sample was positioned and a new unexposed area of aluminum–glass target was placed into the laser beam path. 2.1.1. Flyer velocity The size of the formed aluminum flyer is identical to the beam spot size, which is controlled by a focusing lens. Changing the laser spot size leads to a change in the laser’s surface intensity, thus affecting the flyer velocity and size. The flyer velocity is affected also by the laser pulse energy. By changing the latter parameter, velocities of up to 2900 m·s−1 were attained. Figure 2 shows the theoretical and measured flyer velocity as a function of the laser’s pulse energy. The theoretical flyer velocity is the maximum possible velocity obtained when assuming a system’s

Velocity (m/s)

5000

Theoretical Measured

4000

3000

2000

1000 200

300

400

500

600

700

800

Laser pulse energy (mJ) Figure 2. Flyer measured and theoretical velocities as a function of the laser pulse energy

GROUND SIMULATION OF HYPERVELOCITY SPACE DEBRIS IMPACTS 157

Intensity (arb. units)

0.008 0.007 0.006 0.005 0.004 0.003 0.002 0.001 0.000 0.0

−6

3.0x10

−6

6.0x10

−6

9.0x10

−5

1.2x10

−5

1.5x10

Time (sec)

(a)

(b)

Figure 3. (a) Schematic description of the LDF velocity measurements, and (b) a scope display during a velocity measurement experiment

hydrodynamic coefficient ηH = 1.0. This value means that the whole laser pulse energy is transferred into flyer kinetic energy. The hydrodynamic coefficient is defined as ηH =

E KF V2 = M2 E LP VT

(1)

where E KF is the flyer kinetic energy, E LP is the laser pulse energy, and VM and VT are the measured and theoretical velocities, respectively. According to the theoretical and measured velocities presented in figure 2 and eq. (1), the Soreq’s LDF system hydrodynamic coefficient was calculated to be ηH = 0.23. The flyer velocity was measured using the system shown schematically in figure 3(a). A continuous He:Ne laser beam was set orthogonal to the flyer’s trajectory, so that the beam crossed the flyer path twice in the presence of a prism. The two parallel beams were set at a known distance of 13 mm from each other. A photodiode attached to a scope received the continuous laser signal. As the flyers crossed the continuous laser’s path, two peaks were detected by the scope, allowing the velocity calculation. Figure 3(b) shows a typical scope display with peaks time difference of 6 μs obtained at laser pulse energy of 650 mJ, indicating a flyer velocity of 2000 m·s−1 . 2.1.2. Flyer size In order to estimate the flyer dimensions, series of experiments were conducted using a 1.6 mm-thick BK7 glass as the impacted sample. All experiments were done using similar parameters: pulse energy of 250 mJ, vacuum pressure of 100

158

R. VERKER ET AL.

100 μm Figure 4. A typical ESEM image of an impact-induced crater on a glass surface. Impact conditions were 250 mJ pulse energy, 300 ps pulse length, 100 mTorr system pressure, and flyer velocity of 1400 m·s−1

mTorr, and pulse length of 300 ps. The flyer velocity was 1400 m·s−1 . Sample morphology was characterized using an environmental scanning electron microscope (ESEM, model Quanta 200 from FEI), which allows characterization of nonconductive samples (e.g., Kapton and glass) without the need for a conductive coating. Figure 4 shows a typical SEM image of an impact-induced crater on a glass surface. Flyer velocities and crater diameters, as determined from the ESEM images, were applied into the “conchoidal cracking diameter equation [7]”: −0.5 Dco = 5 × 10−4 ρS ρP0.71 dP1.13 vp0.754

(2)

where Dco is the concentric cracking region diameter, ρs and ρp are the sample and flyer densities, respectively, dp is the plate diameter, and vp is the plate velocity. The flyer dimensions were calculated to be 23–29 μm in diameter. This calculation was conducted only for a velocity of 1400 m·s−1 . In practice, each laser shot at the target produced several major flyers being part of a cloud of such flyer fragments, all traveling at ultrahigh velocity. In order to evaluate the cloud’s dimensions, a 12 μm-thick aluminum foil was used as the impacted sample. The cloud of flyers punctured the foil, creating a hole of 1.5 mm in diameter. The diameter of the hole resembles the dimensions of the cloud. 2.2. POLYMER SAMPLE CHARACTERIZATION

The morphology of the fractures created by the debris impacts was analyzed using the same ESEM.

GROUND SIMULATION OF HYPERVELOCITY SPACE DEBRIS IMPACTS 159

3. Materials The materials studied in this work were commercial KaptonR HN Polyimide (Du-Pont Inc.) films, 25, 50, and 125 μm-thick. Kapton possesses a unique combination of properties that makes it suitable for a variety of applications onboard of spacecrafts. Its main use is as the outer layer of multilayer thermal control insulation blankets, and also as flexible substrates for high-power solar arrays. Among its main properties are inherent strength, temperature stability, excellent insulation properties, and stability under ionizing and UV radiation. Kapton is also known for its superior optical properties including low solar absorbance and high thermal emittance.

4. Results and Discussion 4.1. FLYER VELOCITY EFFECTS

Figure 5 demonstrates the effect of flyer velocity on the extent and nature of damage developed in impacted 25 μm-thick Kapton films. The fractures were created using final flyer velocities VF of 1400 (figure 5(a)), 1650 (figure 5(b)),

Figure 5. ESEM images of 25 μm-thick Kapton films impacted by debris at velocities of (a) 1400 m·s−1 , (b) 1650 m·s−1 , (c) 1730 m·s−1 , and (d) 2900 m·s−1 , respectively

160

R. VERKER ET AL.

1730 (figure 5(c)), and 2900 (figure 5(d)) m·s−1 . All ESEM images were taken from the impact exit side. The volcano-like puncture sites obtained at the lowest velocity (figure 5(a)) may indicate ductile rupture of the polymer. At this relatively low velocity, only few flyers could penetrate the film and create the punctures. At a higher flyer velocity of 1650 m·s−1 (figure 5(b)), ductile rupture is still dominant, but some cracks begin to form around these volcano-like punctures. A further increase in the flyer velocity resulted in radial cracking around the central impact zone (figure 5c). The results indicate also that all flyers in the cloud had sufficient energy to penetrate the Kapton film. At the highest tested velocity of 2900 m·s−1 , these radial cracks completely developed into a brittle fracture of the polymer (figure 5(d)). The transition from ductile to brittle fracture may be expected because such transitions are strongly dependent on the strain rate. As the flyer velocity increased, the strain rate also increased. Brittle fractures are associated with less energy absorbance compared to ductile fractures [12]. At relatively low strain rates (figures 5(a) and 5(b)) the kinetic energy lost by the flyers was transferred into pronounced deformation energy. At relatively high strain rates (figures 5(c) and 5(d)), on the other hand, the successive kinetic energy was transformed into crack propagation energy and the associated formation of new surfaces. 4.2. FILM THICKNESS EFFECT

The effect of film thickness on the extent and nature of damage introduced into the Kapton film is demonstrated in figure 6. Laser driven flyers with velocity of 1730

Figure 6. Impacts at velocity of 1730 m·s−1 into (a) 25 μm, (b) 50 μm, and (c) 125 μm-thick Kapton films. Note the different scale bar in (c)

GROUND SIMULATION OF HYPERVELOCITY SPACE DEBRIS IMPACTS 161

m·s−1 were shot against 25, 50, and 125 μm-thick Kapton films. All images in figure 6 were taken from the impacted sample exit side. It is evident that as the film thickness increases, a transition from brittle (figure 6(a)) to ductile (figure 6(c)) fracture occurs and the overall extent of damage caused by the impact is reduced too. The 25 μm-thick Kapton film (figure 6(a)) experienced significant damage with radial brittle-like cracks emanating from a central impact zone. The 50 μmthick Kapton film (figure 6(b)) experienced less damage, lacking any radial cracks; only few punctures were noticed. The least damage was introduced into the 125 μm-thick Kapton film (figure 6(c)); only a single penetration zone is observed, exhibiting a volcano-like puncture. The following argument may be given to explain the results aforementioned. As the Kapton film becomes thicker, its ability to absorb energy and slow the flyer increases. Consequently, the strain rate associated with the impact process is reduced. For the 25 μm-thick film, the strain rate is high enough to catalyze the formation of brittle radial cracks. For the 50 μm-thick film, intermediate strain rates probably existed, leading to a semiductile fracture of the polymer. In this case, no sufficient energy was left to allow radial cracking. Finally, in the case of the 125 μm-thick film, only a single puncture was formed—most likely by a single flyer. The strain rate under which this process took place was low enough to enable ductile fracture.

4.3. A THERMAL MODEL

Close examination of the sample impacted at the highest velocity of 2900 m·s−1 is shown in figure 7 at a higher magnification. This figure shows also the fracture surface morphology (fractography) around the circumference of the penetrating hole. It is clear that this morphology changes significantly, indicating the possible involvement of different mechanisms of fracture. The characteristics of fracture morphology seem surprising at first—while the central penetration zone that experienced the highest strain rate failed in a fairly ductile manner, the radial cracks that formed subsequently under lower strain rates (i.e., as a secondary process) exhibit a more brittle fracture morphology. We believe that this behavior may be explained in terms of a high temperature gradient that is established within the polymer sample as the flyer hits its surface and penetrates through the film. It is well known that ductile–brittle transitions depend strongly on the local temperature. Whereas a ductile fracture is expected above the glass transition temperature, Tg , below this temperature brittle fractures are most likely to occur. It should be noted that in Kapton, a second-order transition occurs within the temperature range of 360–410◦ C, which is assumed to be the glass transition temperature [13]. This temperature dependence of fracture mode also reminds the deformation map suggested by Spaepen for metallic glasses [14]. Ultrahigh velocity impacts generate temperatures in the range of 1727–6727◦ C

162

R. VERKER ET AL.

Figure 7. (a) A 25 μm-thick Kapton film impacted at hypervelocity of 2900 m·s−1 . Two distinct modes of fracture are evident: (b) a fairly brittle fracture in the radial cracking region, and (c) a more ductile fracture around the central penetration region

and shock pressures of 30–100 GPa when striking ceramic surfaces [9, 15]. Highdensity polyethylene projectiles shot at an average velocity of 5 km·s−1 were observed to generate temperatures of up to 7927◦ C in the case of head-on impacts on aluminum targets [16]. Hence, the following model can explain the phenomenon shown in figure 7. Due to the high temperature generated at the penetration zone and despite the ultrahigh strain rate involved in impact, the Kapton film exhibits a fairly ductile fracture in this zone. On the other hand, the significantly lower temperatures of T < Tg are not sufficient to compensate for the still high strain rates, thus the secondary cracks far from the impact point exhibit brittle fracture characteristics. 4.4. IMPACT TEMPERATURE EVALUATION

In order to support the above model, an attempt was made to estimate the mean temperature developed within a sample at the time of penetration. Figure 8 shows a 25 μm-thick Kapton film impacted at an entrance flyer velocity Vi of 2200 m·s−1 . The flyer velocity at the exit, Vo , was also measured and found to be 1540 m·s−1 . The sample was divided into three regions: (a) the penetration hole, at which a piece of Kapton was sheared-off, (b) a region where ductile fracture was observed, and (c) a region where brittle fracture was noticed.

GROUND SIMULATION OF HYPERVELOCITY SPACE DEBRIS IMPACTS 163

Figure 8. A 25 μm-thick Kapton film impacted at a hypervelocity of 2200 m·s−1 . The sample is divided schematically into three regions: (a) the penetration hole, at which a piece of Kapton was sheared-off, (b) a region where ductile fracture was observed, and (c) a region where brittle fracture was noticed

The system energy balance was expressed as follows: E K,F = E K,S + E Sh + E C + E H

(3)

where E K,F is the flyer kinetic energy change due to the impact, E K,S is the kinetic energy provided to the piece of Kapton that was sheared-off, E Sh is the energy required for the shearing process, E C is the energy required for crack formation, and E H is the energy transformed into heat which results in temperature increase. The following assumptions were made in the calculation: (i) the impact was taken as an equilibrium process. Although this assumption is inaccurate, it greatly simplifies the calculation and provides the ability to calculate a mean temperature within a certain region of the sample; (ii) the velocity of the sheared Kapton piece is equal to the flyer velocity at the exit side; (iii) the energy required for deformation of the Kapton surface is relatively small, and may be neglected. E K,F and E K,S were calculated according to 1 E K = mv 2 2

(4)

Using a flyer mass m = 5.7 × 10−5 g, entrance velocity Vi = 2200 m·s−1 , and exit velocity Vo = 1540 m·s−1 , one obtains E K,F = 70.2 mJ, E K,S = 40.0 mJ. E Sh was calculated according to PS = 0.85 · h · 1 · (UTS) E Sh = 0.5 · PS · h

(5) (6)

164

R. VERKER ET AL.

where PS is the shearing force, h is the film thickness, l is the length of cut, UTS is the ultimate tensile strength of the material being sheared, and E Sh is the shearing energy. For h = 25 μm, l = 3.5 × 10−3 m, UTS = 2.31 × 108 N·m−2 , and PS = 17 N, we get E sh = 0.2 mJ. E C was evaluated according to E C = 2 · δa · h · (γS + γP )

(7)

where δa is the total length of cracks which were formed by the impact, γs is the specific surface energy, and γp is the plastic deformation energy. Using the measured value δa = 0.37 cm, and the typical values γs = 5.8 × 10−4 N·cm −1 and γp = 0.5 N·cm−1 , we find E C = 9.3 × 10−3 mJ. Now, by substituting these values into eq. (3), we find the amount of energy transformed into heat, E H = 30 mJ. Using the common expression for adiabatic heat transfer equilibrium process E H = m · Cp · T

(8)

where m is the mass of Kapton confined within region (b) in figure 8 (m = 5.7 × 10−5 g), CP is the Kapton’s specific heat (CP = 1.09 J·(g·K)−1 ), and T is the temperature difference, a mean temperature increase of T = 920◦ C was calculated. This calculation thus shows that a flyer penetrating at a velocity of 2200 m·s−1 causes a significant temperature increase in the penetration zone (figure 8, region (b)), raising the local temperature to as high as approximately 947◦ C. This estimated temperature is much higher than the glass transition temperature of Kapton, supporting the suggested model and explaining the ductile-like fracture within the penetration zone. This large temperature gradient also predicts a temperature lower than Tg in region (c), where the high strain rate results in a brittle-like fracture.

5. Summary and Conclusions A laser driven flyer (LDF) system was developed at Soreq NRC for simulation of space debris impacts. The system produces a cloud of flyers with diameters of up to 30 μm and measured velocities of 1400–2900 m·s−1 . The effect of simulated hypervelocity debris impacts on space-qualified Kapton films was studied for different film thickness (25–125 μm) and impact velocities. As the Kapton film thickness was increased from 25 to 125 μm at a fixed impact velocity, a transition from brittle to ductile fracture was observed. At a constant thickness of 25 μm, low impact velocities of 1400–1600 m·s−1 resulted in a ductile-like fracture. Increasing the flyer velocity resulted in cracks formation and a fracture that was mostly brittle-like. At impact velocities higher than 1700 m·s−1 , the central impact region (which is exposed to the highest strain rate) was characterized by a ductile-like fracture, while remote radial crack regions were

GROUND SIMULATION OF HYPERVELOCITY SPACE DEBRIS IMPACTS 165

characterized by a brittle-like fracture despite the lower strain rates. A model explaining this phenomenon was suggested based on the high impact temperature (T > Tg ) developed at the central impact region and the low temperatures (T < Tg ) at remote regions.

Acknowledgment This work was partially supported by Israeli Space Agency.

References 1. Grossman, E., Gouzman, I., Viel-Inguimbert, V., and Diguirard, M. (2003) Journal of Spacecraft and Rockets 40, 110. 2. Houdayer, A., Cerny, G., Klemberg-Sapieha, J.E., Czeremuszkin, G., and Wertheimer, M.R. (1997) Nuclear Instruments and Methods in Physics Research B 131, 335. 3. Silverman, E. M. (1995) Space Enviromental Effects on Spacecraft, LEO Material Selection Guide, NASA Contractor Report 4661, Langley Research Center, pp. 4–1. 4. Tennyson, R. C. and Shortliffe, G. (1997) In A. Paillous (ed.) Proceedings of the 7th International Symposium on Materials in Space Environment, Toulouse, France, ESA Publication, ESTEC Noordwijk, The Netherlands, 16–20 June 1997, p. 485. 5. Hastings, D. and Garrett, H. (1996) Spacecraft-Environment Interactions, Cambridge University Press, Cambridge, U.K., pp. 45–99. 6. Miao, J. P. W. (2001) Stark Planetary and Space Science 49, 927. 7. Paul, K. G., Igenbergs, E. B., and Berthoud, L. (1997) International Journal of Impact Engineering 20, 627. 8. Stein, C., Roybal, R., and Tlomak, P. (2000) In E. Werling (ed.) Proceedings of the 8th International Symposium on Materials in Space Environment, Arcachon, France, CNES Publication, Toulouse, France, 5–9 June 2000. 9. Medina, D.F., Wright, L., and Campbell, M. (2001) Advances in Space Research 28, 1347. 10. Tighe, A., Gabriel, S., and Van Eesbeek, M. (2000) In E. Werling (ed.) Proceedings of the 8th International Symposium on Materials in Space Environment, Arcachon, France, CNES Publication, Toulouse, France, 5–9 June 2000. 11. Roybal, R., Tlomak, P., Stein, C., and Stokes, H. (1999) International Journal of Impact Engineering 23, 811. 12. Callister, W. D. (2000) Materials Science and Engineering, John Wiley and Sons Inc., New-York, pp. 124–134. 13. Du-Pont Inc. Technical bulletin, http://www.dupont.com/kapton/general/H-38492-2.pdf. 14. Spaepen, F. (1977) Acta Metallurgica 25, 407. 15. Akins, J.A. (2003) Dynamic Compression of Minerals in the MgOFeO-SiO2 System, Ph.D. thesis, California Institute of Technology, CA. 16. Ramjaun, D., Shinohara, M., Kato, I., and Takayama, K. (2001) In Proceedings of the 23rd International Symposium on Shock Waves (ISSW23), Fort Worth, TX, 22–27 July 2001.

TESTING OF SPACECRAFT MATERIALS FOR LONG FLIGHTS IN LOW EARTH ORBIT L. S. NOVIKOV,1 V. N. CHERNIK,1 S. F. NAUMOV,2 S. P. SOKOLOVA,2 T. I. GERASIMOVA,2 A. O. KURILYONOK,2 AND T. N. SMIRNOVA3 1 Skobeltsyn Institute of Nuclear Physics Moscow State University, 119992 Moscow, Russia 2 RSC “Energia,” Korolev, Russia 3 Khrunichev State Space Scientific Production Center, Moscow, Russia

Abstract. Results of simulation tests of protective and functional coatings influence on resistance of polymeric constructional spacecraft materials to impact of atomic oxygen with fluences up to 3.5 × 1022 cm−2 are presented. It was demonstrated that oxygen plasma beams can be used in accelerated tests of carbon-based and polymeric materials (with the exception of fluorinated hydrocarbons) to evaluate their resistance to the AO impact in LEO. For the unprotected materials sharp fall of mechanical properties and optical characteristics deterioration were observed. Application of protective coatings had shown to reduce their degradation. Key words: atomic oxygen, protective coating, spacecraft materials

1. Introduction One of the principal damaging factors of space in low Earth orbits (LEO) is the atomic oxygen (AO). In prolonged exposures of materials on the outside surfaces of spacecraft, their resistance to AO attack is of major importance. Since polymerbased materials are widely used on spacecrafts and most of them are susceptible to AO attack, their protection by various means, including protective coatings, is necessary. The long-term LEO flight simulation requires irradiation of materials with high-fluence AO up to 1022 –1023 cm−2 . Simulator beam intensities do not exceed 1017 cm−2 ·s−1 (usually 1015 –1016 cm−2 ·s−1 ) [1] that results in practically unacceptable test duration. The reduction of test duration is achieved by an increase of the particle beam energy within the limits of conservation of the interaction mechanism with the test material [2]. 167 J.I. Kleiman (ed.), Protection of Materials and Structures from Space Environment, 167–174.  C 2006 Springer. Printed in the Netherlands.

168

L. S. NOVIKOV ET AL.

In our work, the accelerated simulation material tests were carried out in oxygen plasma beams, formed by a plasma accelerator, with the atomic oxygen particle energy of 20 eV. The changes of weight, mechanical and optical properties of prospective materials were investigated at equivalent AO fluences up to 3.5 × 1022 cm−2 .

2. Test Materials Various purpose materials were investigated: structural coating (black reinforced plastics, fabric, threads, films), functional coatings (protective, thermo-control (TCC), colored). The black reinforced plastic is composed of carbon fibers and epoxy matrix. Samples in form of plates 1 mm thick without coating and with TCC white conductive enamel EKOM-1 were used. The enamel EKOM-1 consists of ZnO pigment with acrylic binder, 0.1 mm thick. One more test object was a fragment of the protection shield of external ISS equipment used to protect from influence of space shuttle jets. The shield consists of the Terlon fabric (Kevlar-type polyamide) separate parts overcasted by the polyimide fibers threads. Another synthetic fiber-based material was a sennit “PARML.” It is an electrically conductive whisker material applied to cable shielding. It is twisted with polyimide fibers threads that are the sennit load-bearing basis. The threads are winded by a copper tinsel ribbon for high electroconductivity. Tensile, stress–strain measurements were used to evaluate the changes in the breaking strength of Terlon fabric and sennit “PARML.” For evaluation of the efficiency of different coatings, polyimide films 20 and 40 μm thick were chosen that were coated with silica layers 0.2–0.4 μm thick, deposited by plasma deposition. Silicone coatings with a thickness of 10–20 μm were made by spraying of varnishes of two types. Color enamels with epoxy-based binders and silicone pitches were studied. The efficiency of the epoxy enamels protection by a silicone varnish layer was studied also. The red, dark blue, white, black, and yellow coatings were applied to fiber glass fabric samples 20 × 40 mm2 in size. 3. The Test Technique The experiments were carried out in the plasma accelerator [3] in oxygen plasma beams consisting of ions, atoms, and molecules of oxygen with mean velocity of 16 km·s−1 (mean atom energy of 20 eV) and a flux density of (2.5–3.5) × 1016 cm−2 ·s−1 . Due to the dissociation of fast molecules and neutralization of the ions

TESTING OF SPACECRAFT MATERIALS

169

at collision with the surface, only the atoms with mean velocity of 16 km·s−1 impact the material. The AO equivalent fluence was evaluated using the witness film mass loss in an assumption of the erosion yield Y = 4.4 × 10−24 g·atom−1 O. The fluence is equal to a fluence of a fictitious 5 eV-AO beam that initiates the same polyimide mass loss. We followed the standard technique of AO fluence determination in ground-based facilities [4]. When material properties were studied the changes of weight and thickness were measured, and the reflection spectra in the range 0.2–2.5 μm were registered. The stress–strain properties were determined on a tensile test machine.

4. Experimental Results 4.1. PROTECTION EFFICIENCY OF TCC WHITE ENAMEL COATING OF BLACK REINFORCED PLASTIC

80 70 60 50 40 30

1.2 no coating with coating

20 10 0

Thickness, mm

Specific mass loss, mg/cm2

The specific mass loss and thickness of the black reinforced plastic samples uncoated and with the coating as a function of the AO equivalent fluence are given in figure 1. As can be seen, the mass loss and thickness of the uncoated samples are linear functions of the AO fluence. At the same time, the mass loss of the coated sample is lower by a factor of 6.5, and thickness change is negligibly small. Reactivity of the unprotected black reinforced plastic consisting of carbon fibers in epoxy matrix, measured in the tests, was 1.2 × 10−24 cm3 per O atom. It corresponds to the flight data known for the composite components: 1 × 10−24 cm3 per O atom for carbon and 1.7 × 10−24 cm3 per O atom for epoxy [3] that confirms that the plasma accelerator conditions can imitate adequately the LEO conditions. The polymer binder on the coated samples was etched away after plasma impact

1 0.8 0.6 0.4 no coating

0.2

with coating

0 1 2 3 4 0 AO equivalent fluence, 1022 1/cm2

0

1 2 3 4 AO equivalent fluence, 1022 1/cm2

Figure 1. Specific mass loss and thickness of samples of the black reinforced plastic uncoated and with the coating as functions of the AO equivalent fluence

170

L. S. NOVIKOV ET AL. 1400

Load, N

1200 1000 800 600 400 200 0

0

10

20

30

40

Elongation, % before exposure

after exposure

Figure 2. Tensile stress–strain diagrams of the sennit “PARML” samples before and after the AO exposure

exposing the fibers that were also partially destroyed with an observed increase in the surface roughness [5]. The picture is similar to observations in flight experiments onboard the space shuttle spacecraft, Salute-6 and Mir space stations [6]. 4.2. DEGRADATION OF MECHANICAL PROPERTIES OF SYNTHETIC FIBER MATERIALS

The results of measurements of specific mass loss of the Terlon fabric samples and the sennit “PARML” samples show, that the Terlon fabric is close to polyimide in AO resistivity, as the materials mass losses are close to 5.4 and 5.6 mg·cm−2 accordingly. A complete etching of the polyimide fibers in a significant part of bundles in the sennit “PARML” was observed. The tensile stress–strain diagrams of the sennit “PARML” samples before and after the AO exposure to an equivalent fluence of 3.7 × 1021 cm−2 are given in figure 2. The breaking load reduction by a factor of 5, and a drop of relative elongation at rupture by a factor of 1.7 was observed. After AO exposure the destruction of seam strings and Terlon fabric was observed on the fragment of the Terlon screen with the seam. The tensile stress– strain diagrams of reference and AO exposed samples of the shield fragments are presented in figure 3. When loading up to approximately 20-fold working load identical elongations were observed in both samples. By further increasing the loading, the exposed sample fails that manifests itself as a sharp break in the curve and an increase in elongation is observed without an increase or even on reduction of loading. The elongation on rupture of the exposed sample also decreases by a factor of 2.

171

TESTING OF SPACECRAFT MATERIALS

Load, N

800 600 400 200

1 2

0 0

2

4

6

8

10

12

14

Elongation, mm Figure 3. Tensile stress–strain diagrams of reference (1) and AO exposed (2) samples of the shield fragments

4.3. EFFICIENCY OF PROTECTIVE SILICON BASED COATINGS

In figure 4, the dependences of the mass loss of the polyimide films protected by deposited silica layers and by silicone coatings of two varnish types of AO fluence are presented. We quantify the AO protection efficiency γ as a ratio of bare film mass loss dm 0 to the protected film one dm i under equal fluence increment dF, i.e. γ = dm 0 /dm i |d F = const.

Mass loss, mg/cm2

The protection efficiency is greatest for silica layer (γ = 830) and decreases up to γ = 250–430 for varnish coverings [6]. The exposure to AO causes formation of microcracks in the silica layers. On the contrary, no cracking occurs in the silicone coatings. Thermo-optical characteristics of the samples changed very little. For the silica coating an insignificant and uniform in wavelength range reflectance reduction is observed that can be associated with scattering from the grid of microcracks. Thus, the solar absorptance αs increases from 0.366 to 0.380. The reflectance increase is typical for varnish coatings, especially appreciable in long wavelength part of the spectrum. 60 50 40 30 20 10 0

silicon 3 0

oxide

5 10 AO equivalent fluence, 1020 1/cm2

silicon 5 15

Figure 4. Dependences of the mass loss of the polyimide films with protection by silica layers deposited and silicone coatings of two varnish types on AO fluence

172

L. S. NOVIKOV ET AL.

4.4. TESTING THE RESISTANCE OF COLOR PAINTS

The paint test results have shown different AO resistance for various types of enamels [7]. Color change and significant mass losses were observed for the epoxy enamels. The erosion yields of the color epoxy enamels (0.3–0.5 × 10−24 g·atom−1 O) are much lower than that of the binder and of the witness polymers

Reflection coefficient

KO-811 blue 0.15 0.1

after before

0.05 0 300

400

500

600

700

Wavelength, nm

Reflection coefficient

EP-140 blue 0.35 0.3 0.25 after

0.2 0.15

before

0.1 0.05 0 300

400

500 Wavelength, nm

600

700

Reflection coefficient

EP-140 + KO-008 blue 0.2 after before

0.15 0.1 0.05 0 300

400

500

600

700

Wavelength, nm

Figure 5. Visible spectrum reflectance for three coatings after oxygen plasma exposure with AO equivalent fluence 1.4 ×1021 cm−2

TESTING OF SPACECRAFT MATERIALS

173

(3–4.4 × 10−24 g·atom−1 O) that is explained by protective action of pigments. The impact of oxygen plasma on silicone enamels almost does not change their color and mass. Protection of the epoxy enamel by silicone varnish layer increases the coating resistance up to the level of the silicon enamels. The erosion yields of these coatings are less than this of the witness film by two orders of magnitude. Figure 5 shows the reflection spectra of the dark blue paints in the visible wavelength range for three coatings: epoxy EP-140, silicon KO-811 K, and epoxy EP-140 with silicone varnish coating KO-008. Comparing the spectra, it is evident that the dark blue color peak seen in all initial spectra, almost completely disappears for epoxy enamel that is accompanied by increase of reflectance in the whole range. The color and the spectra of other dark blue paints practically do not vary.

5. Conclusions It was demonstrated that oxygen plasma beams can be used in accelerated tests of carbon-based and polymeric materials (with the exception of fluorinated hydrocarbons) and inorganic coatings to evaluate their resistance to the AO impact during the simulation of the long flights in LEO. The resistance of prospective spacecraft materials: polyimide films, synthetic Terlon fabric, sennit “PARML,” black reinforced plastic, polymeric paints, and thermo-control coatings under oxygen plasma beams simulating the AO fluxes in LEO was investigated. For the unprotected materials sharp fall of mechanical properties (manifesting in lower failure loads and relative elongation at rupture) was observed. Optical characteristics deteriorated as well. Application of protective coatings had shown to reduce the degradation of mechanical and optical properties. The protection efficiency is the greatest for coatings containing silicon.

References 1. Kleiman, J., Iskanderova, Z., Gudimenko, Y., and Horodetsky, S. (2003) In Proceedings of the 9th International Symposium on Materials in Space Environment, ISMSE-9, Nordwijk, 2003, pp. 313–324. 2. Rutledge, S.K., Banks, B.A., Dever, J., and Savage, W. (2000) In Proceedings of the 5th International Conference on Protection of Materials and Structures from the LEO Space Environment, ICPMSE-5, Arcachon, France ESA Publishing, Noordwijk, The Netherlands, 2000. 3. Akishin, A. I., Novikov, L. S., and Chernik, V. N. (2000) In L.S. Novikov and M.I. Panasyuk (eds.), New High Technologies and Technics, Vol. 17, Moscow, ENTSITEH, p. 100 (in Russian). 4. Protocol for Atomic Oxygen Testing of Materials in Ground-Based Facilities, JPL Publication, pp. 95–17.

174

L. S. NOVIKOV ET AL.

5. Novikov, L. S., Chernik, V. N., Babaevskij, P. G., Kozlov, N. A., Chalyh, A. E., Balashova, E. V., and Smirnova, T. N. (2001) Perspektivny materialy 5, 20–26 (in Russian) (Journal of Advanced Materials, Cambridge Interscience Publication). 6. Chernik, V., Naumov, S., Demidov, S., Sokolova, S., and Svechkin, V. (2000) In 5th International Conference on Protection of Materials and Structures from the LEO Space Environment, ICPMSE-5, Arcachon, France, 2000. 7. Chernik, V. N., Naumov, S. F., Sokolova, S. P., Gerasimova, T. I., Kurilyonok, A. O., Poruchikova, Y. V., and Novikova, V. A. (2003) In Proceedings of the 9th International Symposium on Materials in Space Environment, ESTEC, Noordwijk, 2003, pp. 281–285.

M/OD IMPACTS ON THE MULTIPURPOSE LOGISTICS MODULE Post Flight Inspection Results JAMES L. HYDE,1 RONALD P. BERNHARD,1 AND ERIC L. CHRISTINSEN2 1 Lockheed Martin Space Operations, Johnson Space Center, Houston, TX 77258 2 NASA, Johnson Space Center, Houston, TX 77258

Abstract. The International Space Station (ISS) program has manifested a Multipurpose Logistics Module (MPLM) on five launch packages since 2001, with the next flight scheduled for the STS-114 mission. The MPLM has been deployed each time on the nadir docking port of node 1 by the space shuttle for a period of 5–6 days, and then returned to Earth for refurbishment prior to the next resupply mission. MPLM flight module 1 (Leonardo) and flight module 2 (Raffaello) have accumulated about 700 h of low Earth orbit (LEO) exposure time on the ISS. Through five missions, there have been two perforations of the aluminum outer wall of the MPLM shielding and 24 craters due to meteoroid/orbital debris impact. This paper will document the results of an ongoing postflight inspection campaign that identifies hypervelocity impact (HVI) damage to the meteoroid and debris protection system (MDPS) of the MPLM through observations, data collection and analysis. Residual projectile materials are collected at each impact site and subjected to scanning electron microscope/energy dispersive X-ray spectrometric analysis to identify the elemental composition of the impactor and presumed source (meteoroid or orbital debris) of the impact damage. The observed impact damage exhibits a marked directionality on the module, with the majority of damages occurring to the forward half of the MPLM. Postflight predictions of damage to MPLM from the BUMPER code, which is utilized by the aerospace community in meteoroid/orbital debris risk assessments, are compared to the observed damages. Results from BUMPER very clearly show the directionality of the expected impact damage as was observed in the postflight MPLM inspections. Observations and analytical data from MPLM demonstrate the meteoroid/orbital debris impact hazards that the ISS, shuttle, and other spacecraft face in low Earth orbit environment. Implications of this work on the design of adequate meteoroid/orbital debris protection for future vehicles will be provided. 175 J.I. Kleiman (ed.), Protection of Materials and Structures from Space Environment, 175–191.  C 2006 Springer. Printed in the Netherlands.

176

JAMES L. HYDE ET AL.

Key words: orbital debris, meteoroids, hypervelocity impact, international space station 1. Introduction 1.1. PURPOSE

One of the primary reasons for tracking on-orbit damage is the fact that NASA has used observation data from returned spacecraft surfaces to calibrate the ORDEM 2000 engineering model of the low Earth orbit (LEO) environment [1]. Postflight inspections of shuttle surfaces such as the Crew Module Windows, Payload Bay Door Radiators has produced a database with thousands of entries [2]. The long duration exposure facility (LDEF), retrieved by the shuttle in 1990 after nearly 6 years on orbit, contributed a rich set of data [3] that has been used in the formulation of the orbital debris environment. Postflight inspections also provide an opportunity to validate meteoroid/orbital debris (M/OD) threat assessment codes such as NASA’s BUMPER-II [4, 5] or ESABASE/Debris. The ability to compare predictions with observations is a valuable component in the development life cycle of M/OD analysis software. The performance and adequacy of meteoroid/debris protection systems can be demonstrated with postflight damage inspections. The flight history of the MPLM has established the protection capability of the shielding system. 1.2. OBJECTIVE

This paper will document the results of the five postflight inspections that have been performed on the MPLM, providing data on damage sizes and sources of the impact. The results will also include an estimate of the equivalent diameter of the impacting particle. Due to the nature of hypervelocity impacts, projectile shape and impact angle data are not readily discernable. 2. MPLM Overview 2.1. DESCRIPTION

Three flight units, built by Alenia Spazio for the Italian Space Agency (ASI) as part of an agreement with NASA, were named after famous Italian engineers: Leonardo da Vinci, Donato di Niccolo, di Betto Bardi, and Raffaello Sanzio. The MPLM functions as a cargo container for ISS outfitting and resupply missions and as a pressurized module while attached to the ISS. The modules can be configured as “active,” when a payload requires power, data or cooling resources.

M/OD IMPACTS ON THE MULTIPURPOSE LOGISTICS MODUL

177

Figure 1. MPLM installation sequence

The MPLM can also be manifested in a “passive” configuration when the contents require no power, data or cooling resources (dry cargo). Each module has a design life of 10 years or 25 flights. Each module has an overall length of about 6.2 m and a diameter of about 4.5 m [6], giving a surface area of approximately 100 m2 . 2.2. INSTALATION

MPLM modules are carried into orbit and placed on the ISS by the shuttle arm as shown in Figure 1. At the end of a mission, the module is restowed in the payload bay of the orbiter and returned to Earth. 2.3. M/OD PROTECTION SYSTEM

Figure 2 provides an illustration of the meteoroid/debris protection system (MDPS) that is included on each MPLM. The MDPS consists of a sacrificial outer layer of 0.8 mm aluminum alloy (known as a “bumper”) mounted at a distance of ∼128 mm from the aluminum alloy pressure shell. The bumper-pressure wall cavity also contains a multilayer insulation (MLI) blanket loosely fastened to the 3 mm pressure wall of the MPLM. The MDPS consists of 8 forward cone panels, 48 side panels and 12 aft cone panels. Some of the module’s external features are highlighted in figure 3. MDPS panels on the side and forward cone are also shown. Some of these components will be referenced in a later section.

178

JAMES L. HYDE ET AL.

Figure 2. Meteoroid debris protection system (MDPS)

2.4. MISSION HISTORY

There have been five MPLM missions to date, with a cumulative docked exposure time of 701 h. Three missions used flight module (FM) 1 and two missions used FM2. Details of the missions are shown in figure 4. The modules were always SUPPORT BRACKET FOR ELECTRICAL PDA CBM RING

STABILIZER TRUNNION

FRGF

MDPS PANEL

MAIN TRUNNION

SUPPORT BRACKET FOR FLUID PDA STABILIZER TRUNNION

FRGF

MAIN TRUNNION

MDPS PANEL

Figure 3. External features of MPLM, including MDPS panels

M/OD IMPACTS ON THE MULTIPURPOSE LOGISTICS MODUL MPLM

Exposure

Name

Flight

FM1 Leonardo FM2 Raffaello FM1 Leonardo FM2 Raffaello FM1 Leonardo

1 1 2 2 3

STS Mission Vehicle 102 100 105 108 111

5A.1 6A 7A.1 UF1 UF2

OV-103 OV-105 OV-103 OV-105 OV-105

Docking Port

Hours

145.52 97.84 166.07 145.22 146.02 701 hrs 29.2 days

Node1 - nadir Node1 - nadir Node1 - nadir Node1 - nadir Node1 - nadir

179

ISS Attitude RPY sequence

0°, 23.5°, 0° 0°, 23.0°, 0° 0°, 23.0°, 0° 0°, 21.0°, 0° 0°, 23.0°, 0°

Figure 4. MPLM flight data

mated with the ISS at the nadir docking port on node 1. This placed the port edge of the MPLM (relative to its stowed position in the shuttle) in the ram direction when docked to the ISS. The predominant flight attitude of the ISS when the shuttle and MPLM are attached is a positive pitch bias of about 23◦ , as shown in figure 5, the front of the station is angled upwards ∼23◦ .

3. Impact Survey The following section gives details of the impact features observed during postflight processing. Figure 6 illustrates the measurement terminology that was used to characterize the impacts.

Figure 5. Typical MPLM flight attitude

180

JAMES L. HYDE ET AL. Spall Lip Crater

Depth

Figure 6. Impact crater measurement terminology

3.1. PROCEDURE

As part of the standard postflight processing workflow, all MDPS panels are inspected by technicians at a processing facility at the Kennedy Space Center (KSC). They are tasked with the documenting defects on external and internal surfaces. Instances of scratches, scuffs, ding, and rub marks have all been recorded, along with hypervelocity impact craters. Specialists from the Johnson Space Center with hypervelocity impact (HVI) inspection experience travel to KSC and independently examine as much as possible of the exterior for HVI damage. When an impact site is identified, the location is recorded and a graduated handheld magnifier is used to measure dimensions of the impact features (crater, lip, and spall diameters). For impacts into monolithic aluminum MDPS surfaces, a soft wooden probe is used to collect potential impactor residue from the crater or hole. Dental mold impressions are another technique that is used when surface damage needs to be sampled. Finally, when “soft goods” such as beta cloth are impacted, an adhesive agent is sometimes used to aid in data collection. Crater depth is obtained back at the lab by measuring the effected region of the sampling device.

Figure 7. FM2 at the KSC space station processing facility (SSPF)

M/OD IMPACTS ON THE MULTIPURPOSE LOGISTICS MODUL

181

Figure 8. MPLM inspection at the KSC SSPF

The figure above illustrates the inspection access limitations at the space station processing facility (SSPF). The mounting fixture has to be rotated at least once to allow a complete inspection of all MDPS panels. 3.2. FM1 FLIGHT 1 (STS-102/ISS 5A.1)

There were three hypervelocity impact sites observed after the first MPLM mission. One of the impacts produced a perforation of an MDPS bumper. The impact site, designated 1.1.1, had a 1.44 mm diameter hole with a 2.45 mm diameter lip. There was no discernable damage to the MLI or to the pressure wall of the MPLM. In an attempt to determine the origin of each impactor, samples collected from the three impact sites and subjected to a scanning electron microscope (SEM) and energy dispersive X-ray (EDX) analysis. The results are shown in the following table. Analysis indicated that the impactor that produced the bumper perforation was a particle of spacecraft paint. There was no discernable impact residue in the sample from site 1.1.2. When the particle type is known, an equivalent diameter can be estimated using eq. (1) with an assumed impact velocity of 9 km·s−1 and an impact angle of 45◦ . D = [(P/5.24 × H 1/4 × (ρB /ρP )1/2 × (S/Vn )2/3 ]18/19 IMP A C T D E TA IL S Impact Number Map ID Type A6 MDPS bumper 1.1.1 B5 MDPS bumper 1.1.2 D5 MDPS bumper 1.1.3

Location Region port/lower port/lower port/lower

Estimated

Diam (mm) Material Spall 0.8mm Al 6061-T6 --0.8mm Al 6061-T6 --0.8mm Al 6061-T6 ---

Depth

Lip Crater Hole

2.45 1.25 0.88

(1)

--0.78 0.54

1.44 -----

Figure 9. Impacts on MPLM FM1 flight 1

SEM/EDXA

Particle

Results Dia. (mm) --- Orbital Debris: Paint 0.46 1 0.51 Unknown --0.3 Meteoroid 0.19

(mm)

182

JAMES L. HYDE ET AL.

Figure 10. Impact 1.1.3–1.44 mm diameter hole in MDPS

where D—estimated particle diameter (cm) P—crater depth (cm) H —Brinell hardness of bumper (MDPS = Al6061-T6, 95) ρB —density of bumper (MDPS = Al6061-T6, 2.713 g·cc−1 ) ρP —density of projectile, 2.5 g·cc−1 (for spacecraft paint) S—speed of sound in target (Al6061-T6 MDPS bumper, 5.1 km·s−1 ) Vn —normal component of impact velocity (assuming a velocity of 9 km·s−1 and an impact angle of 45◦ , 6.36 km·s−1 ) Note: in the cases where a bumper perforation occurs, the value of P is extrapolated from the crater diameter. 3.3. FM2 FLIGHT 1 (STS-100/ISS 6A)

No hypervelocity impact features were observed on this module during postflight inspection activities. This was the second MPLM mission.

Figure 11. Results of SEM/EDX analysis—Impact 1.1.3

M/OD IMPACTS ON THE MULTIPURPOSE LOGISTICS MODUL IMP A C T D E TA IL S Impact

Location

Number Map ID 1.2.1 1.2.2 1.2.3

D2 B12 B11

Estimated Depth

Diam (mm)

Type

Region

Material

Spall

MDPS bumper MDPS bumper MDPS bumper

port/upper stbd/upper stbd/upper

0.8mm Al 6061-T6 0.8mm Al 6061-T6 0.8mm Al 6061-T6

-------

183

SEM/EDXA

Dia. (mm)

2

— — 0.14

(mm)

2.80 1.80 0.85

0.9 Unknown 2 0.75 Unknown 0.55 Orbital Debris: SS

2.0 1.5 0.8

-------

Particle

Results

Lip Crater Hole

Figure 12. Impacts on MPLM FM1 flight 2

3.4. FM1 FLIGHT 2 (STS-105/ISS 7A.1)

There were three hypervelocity impact craters observed on the aluminum MDPS bumpers after the third mission (Fig. 12). No perforations were detected. SEM/ EDX analysis of site 1.2.3 indicated that the crater was caused by a particle of stainless steel. The other two sites were not available for sampling. 3.5. FM2 FLIGHT 2 (STS-108/ISS UF1)

Eight impacts were observed after this mission, four on the MDPS bumpers and one each on a grapple fixture, trunnion brace, and scuff plate (Fig. 13). Impact 2.2.8 was the site of the second bumper perforation. The impact produced a 1.2 mm diameter hole with a lip diameter of 1.8 mm. As with the first perforation, there was no visible damage to the underlying structure or components. SEM/EDX analysis indicated that the impactor was a particle of stainless steel. Samples were taken at all eight impact sites. Two impacts were shown to be from orbital debris and two were meteoroids. The remaining four impact samples revealed no discernable impact source. 3.6. FM1 FLIGHT 3 (STS-111/ISS UF2)

Postflight inspections for the third mission of FM1 (Leonardo) recorded 12 impact sites (Fig. 16). Ten occurred on the MDPS panels, another in the material between IMP A C T D E TA IL S Impact

Location

Number Map ID LA4 2.2.1

Estimated Depth

Diam (mm)

Type scuff plate

Region port/lower

Material Spall yellow Sheldahl tape 1.4

Lip Crater Hole

SEM/EDXA

Particle

Results 0.15 Meteoriod: Fe,Ni,S

Dia. (mm)

(mm)

2.2.2 2.2.3 2.2.4 2.2.5

A6 FC9 LA4 D10

MDPS bumper MDPS bumper trunnion brace MDPS bumper

port/lower fwd cone port stbd/upper

0.8mm Al 6061-T6 0.8mm Al 6061-T6 Betacloth 0.8mm Al 6061-T6

---------

--0.70 0.60 --0.45

0.45 0.55 0.4 0.31 0.35

-----------

2.2.6 2.2.7 2.2.8

A2 A12 D12

grapple fixture MDPS bumper MDPS bumper

port stbd/upper stbd/upper

Al 6061-T6 0.8mm Al 6061-T6 0.8mm Al 6061-T6

2.1 -----

1.05 --1.80

0.85 0.45 ---

--0.95 Unknown 1 --0.50 Unknown 1.20 --- Orbital Debris: SS

0.25 0.50 0.10 0.50

Figure 13. Impacts on MPLM FM2 flight 2

1

Unknown 1 Unknown Orbital Debris: Al Meteoroid 1

0.14 ----0.10 0.24 ----0.19

Figure 14. Impact 2.2.8–1.2 mm diameter hole in MDPS

Figure 15. Results of SEM/EDX analysis—Impact 2.2.8 IMP A C T D E TA IL S Impact

Location

Number Map ID Type D6 MDPS bumper 1.3.1 D6 MDPS bumper 1.3.2

Estimated Depth

Diam (mm)

Region port/lower

Material Spall 0.8mm Al 6061-T6 ---

0.8mm Al 6061-T6 0.8mm Al 6061-T6

-----

Lip Crater Hole

Results 0.60 Meteoroid

(mm)

0.60 0.70 0.68

0.45 0.55 0.5

-------

0.60 Unknown 1 0.45 Unknown

1.3.3

D6

MDPS bumper

port/lower port/lower

1.3.4 1.3.5 1.3.6 1.3.7 1.3.8 1.3.9 1.3.10

D5 C6 C4 A2 B2 B2 3-2

MDPS bumper MDPS bumper MDPS bumper grapple fixture MDPS bumper MDPS bumper intercostal

port/lower port/lower port/lower port port/upper port/upper port/upper

0.8mm Al 6061-T6 --0.8mm Al 6061-T6 --0.8mm Al 6061-T6 --Al 6061-T6 4.25 0.8mm Al 6061-T6 --0.8mm Al 6061-T6 --0.8mm Al 6061-T6 ---

0.43 0.40 0.63 1.00 0.70 0.48 0.80

----0.5 0.82 0.55 0.4 0.65

---------------

----0.35 0.65 0.5 0.25 0.45

1.3.11 1.3.12

B2 C12

MDPS bumper MDPS bumper

port/upper stbd/upper

0.8mm Al 6061-T6 0.8mm Al 6061-T6

0.25 0.45

-----

-----

-----

-----

SEM/EDXA

Figure 16. Impacts on MPLM FM1 flight 3

Particle Dia. (mm) 0.28

1

— —

Unknown 1 Unknown Orbital Debris: Paint Orbital Debris: SS 1 Unknown Orbital Debris: Paint Meteoroid

1

— — 0.23 0.20 --0.17 0.21

1

-----

Unknown 1 Unknown

M/OD IMPACTS ON THE MULTIPURPOSE LOGISTICS MODUL

185

Figure 17. Summary of MPLM impacts from known sources

panels and one impact was observed on the port grapple fixture. Samples were obtained at all 12 impact sites. SEM/EDX analysis did not reveal impact residue in seven of the samples. Of the five remaining, three were determined to be orbital debris.

4. Summary Through five MPLM missions, 26 hypervelocity impact sites have been observed during postflight inspections. Although more than 80% of the impacts occurred on the MDPS panels, other parts of the module also sustained impact damage. Samples were obtained at 24 impact sites. SEM/EDX analysis was performed on each sample and 12 yielded results that could be interpreted as either meteoroid or orbital debris in nature. Of the known impact sources, seven were determined to be orbital debris and five were meteoroid. The table in figure 17 provides a summary of the 12 impacts. The estimated diameter of the particle that caused the damage is shown in the last column of the table, sorted from least to greatest.

5. Bumper Code Predictions An analysis of the expected number of MDPS outer wall perforations was performed using the BUMPER code, the meteoroid/orbital debris risk assessment software tool used by NASA to determine risk for the International Space Station [7] and the space shuttle [8]. The intent was to determine if BUMPER code predictions for the number of MDPS perforations came close to matching observations. Another type of “prediction vs. observation” activity that can be performed with the BUMPER code is comparison of observed impact locations to the predicted distribution on the MDPS.

186

JAMES L. HYDE ET AL.

1

2

3

4

Figure 18. Overview of BUMPER code

5.1. MDPS OUTER WALL PERFORATION CALCULATIONS

Using attitude information from the as-flown timeline in addition to the orbital and environmental parameters shown in the table below (Fig. 19), an analysis was performed with the BUMPER code to determine if the number of observed MDPS outer wall perforations agrees with predictions. Mission Duration Altitude Orbit Inclination Flight Year LVLH Attitudes - RPY Orbiter PYR Euler Sequence

(Exposure Hrs | % Duration) Orbiter Finite Element Model Orbital Debris Density Orbital Debris Environment Meteoroid Density Meteoroid Enviroment Meteoroid Velocity Distribution Meteoroid Showers

10 days 18 hours 258 hours 400 km 51.6° 2001 bias -XLV +ZVV = 0°, 113.5°, 0° 130 hrs bias +ZLV +XVV = 0°, 327°, 0° 5 hrs bias -XLV +YVV = 90°, 90°, 23.5° 5 hrs -ZLV -XVV = 0°, 180°, 0° 118 hrs ISS 5A.1 mated with orbiter, MPLM on Node 1 constant, 2.8 g/cc ORDEM2000 variable, 0.5 - 2.0 g/cc SSP30425 Rev. B variable, SSP30425 none

Figure 19. BUMPER code inputs

M/OD IMPACTS ON THE MULTIPURPOSE LOGISTICS MODUL Pre & Post Dock

187

ISS TEA

Space

RPY = 0°, 180°, 0° Velocity

RPY = 0°, 113.5°, 0°

Space Velocity

Reboost

Waste Dump

RPY = 0°, 327, 0°

RPY = 90°, 90°, 23.5°

Space Space

Velocity Velocity

Figure 20. BUMPER finite element models showing typical MPLM mission attitudes

Risk is calculated from the BUMPER results using eq. (2) Risk—(1 − (e−N )),

(2)

where N is the expected number of perforations. It is usually expressed in percent notation. Perforation odds are the reciprocal of the risk, expressed in the familiar “1 in x” notation. Figure 21 presents the calculated expected number of MDPS outer wall perforations for a 258 h mission. MPLM REGION

Expected Number of Perforations M&D Deb Met

fwd cone Bay 1 Bay 2 Bay 3 Bay 4 Bay 5 Bay 6 Bay 7 Bay 8 Bay 9 Bay 10 Bay 11 Bay 12 aft cone

0.0057 0.0433 0.0358 0.0172 0.0166 0.0351 0.0428 0.0272 0.0090 0.0021 0.0022 0.0086 0.0271 0.0185

0.0028 0.0048 0.0070 0.0056 0.0057 0.0073 0.0053 0.0027 0.0013 0.0006 0.0006 0.0011 0.0023 0.0025 TOTALS

0.0085 0.0481 0.0428 0.0228 0.0223 0.0425 0.0481 0.0299 0.0103 0.0028 0.0028 0.0097 0.0294 0.0211 0.34

Risk 0.9% 4.7% 4.2% 2.3% 2.2% 4.2% 4.7% 2.9% 1.0% 0.3% 0.3% 1.0% 2.9% 2.1% 28.9%

Odds 1 in 118 1 in 21 1 in 24 1 in 44 1 in 45 1 in 24 1 in 21 1 in 34 1 in 97 1 in 361 1 in 352 1 in 104 1 in 35 1 in 48 1 in 3

% Total 3% 14% 13% 7% 7% 12% 14% 9% 3% 1% 1% 3% 9% 6% 100%

Figure 21. Expected number of MDPS outer wall perforations for a 258 h mission

188

JAMES L. HYDE ET AL.

Figure 22. MDPS panel locations—forward and aft cones

5.2. MDPS PANEL LOCATIONS

Figures 22 and 23 show the numbering scheme for the MDPS panels. The “top,” “bottom,” “forward,” and “aft” designations refer to the position of the module when it is stowed in the shuttle payload bay. Figure 22 above illustrates the 12 rows or bays of MDPS panels that are referenced in the analysis results table in figure 21.

Figure 23. MDPS panel locations—port and starboard cylinder

M/OD IMPACTS ON THE MULTIPURPOSE LOGISTICS MODUL

189

Figure 24. MDPS impact risk distribution—port side

5.3. IMPACT RISK DISTRIBUTION

Figures 24–26 show the relative distribution of MDPS outer wall perforation risk from meteoroid and orbital debris impacts. The highest risk of bumper perforation can be seen in the red bands 30◦ –70◦ off of the velocity vector on the port and starboard sides. It can be seen in figures 24 and 25 that the MDPS panels on the

Figure 25. MDPS impact risk distribution—tarboard side

190

JAMES L. HYDE ET AL.

Figure 26. MDPS impact risk distribution—forward view (ISS and Orbiter removed for clarity)

end cones have considerably lower risk than the cylinder region. In figure 26, the meteoroid and orbital debris flux “shadowing” effect from the docked Orbiter can be seen in the forward view of the risk distribution. The correlation between observed and predicted impacts is shown in figure 27, where the values from the impact risk plots in figures 24 through 26 are mapped to the MDPS cylinder panels and observed impact locations. Of the 25 observed impacts in the cylinder region, 16 (64%) occurred in the top 2 “high-risk” areas.

6. Discussion This paper details the status of an ongoing inspection and analysis campaign for the MPLM. More flights are planned in the future and postflight inspections will be needed. One item for investigation is the determination of MDPS hole size at onset of significant MLI degradation/rear wall damage. Information of this type could aid in postflight serviceability requirements. The aluminum composition of the MDPS outer wall makes it difficult for the SEM/EDX analysis to discern aluminum projectile residue in the midst of the common background material. Many of the impacts in the database that were assigned to the “unknown” category are probably aluminum in nature. Since aluminum is one of the more common sources of orbital debris impacts, this discrepancy needs to be addressed.

M/OD IMPACTS ON THE MULTIPURPOSE LOGISTICS MODUL

191

Figure 27. MDPS cylinder region risk distribution with locations of observed impacts (dotted line indicates leading edge of module when docked at ISS)

References 1. Liou, J. C., Matney, M. J., Anz-Meador, P. D., Kessler D., Jansen M., Theall, J. R. (2002) The New NASA Orbital Debris Engineering Model ORDEM2000, NASA/TP – 2002-210780, May 2002. 2. Hyde, J. L. Christiansen, E. L., Bernhard, R. P., Kerr J. H., Lear, D. M. (2001) A History of Meteoroid and Orbital Debris Impacts on the Space Shuttle, Proc. Third European Conference on Space Debris, ESA SP-473, October 2001, pp. 191–196 3. See, T. H., Allbrooks, M. A., Atkinson, D. R., Simon, C. G. and Zolensky, M. (1990) Meteoroid and Debris Impact Features Documented on the Long Duration Exposure Facility, NASA documents published by the Johnson Space Center JSC-24608, August 1990. 4. Hyde, J. L. (2000) As-Flown Shuttle Orbiter Meteoroid/Orbital Debris Shield Assessment: Phase 1—Shuttle/Mir Missions, NASA documents published by the Johnson Space Center JSC-28768, January 2000. 5. Hyde, J. L. (2000) As-Flown Shuttle Orbiter Meteoroid/Orbital Debris Assessment: Phase 2 Flights, NASA documents published by the Johnson Space Center JSC-29070, September 2000. 6. MPLM Interface Definition Document (ISS-MPLM-IDD-006), December 2000. 7. Prior, T. G. (2003) International Space Station Meteoroid & Orbital Debris Integrated Threat Assessment #10, NASA documents published by the Johnson Space Center Revision C (JSC29951), April 2003. 8. Hyde, J. L. and Christiansen, E. L. (2001) Space Shuttle Meteoroid & Orbital Debris Threat Assessment Handbook, NASA documents published by the Johnson Space Center JSC-29581, December 2001. 9. Christiansen, E. L. (1991) Shield Sizing Equations, NASA inter-department communication SN-90-131, October 1991.

FUEL OXIDIZER REACTION PRODUCTS (FORP) CONTAMINATION OF SERVICE MODULE AND RELEASE OF N-NITROSODIMETHYLAMINE IN A HUMID ENVIRONMENT FROM CREW EVA SUITS CONTAMINATED WITH FORP WILLIAM SCHMIDL,1 RON MIKATARIAN,1 CHIU-WING LAM,2 BILL WEST,3 VANESSA BUCHANAN,4 LOUIS DEE,4 DAVID BAKER,5 AND STEVE KOONTZ6 1 Boeing, Houston, TX 2 Wyle Laboratories, Houston, TX 3 Hamilton Sundstrand, Houston, TX 4 Honeywell, White Sands, NM 5 NASA White Sands Test Facility,White Sands, NM 6 NASA Johnson Space Center, Houston, TX

Abstract. The U.S. Control Moment Gyros (CMGs) maintain the International Space Station (ISS) vehicle attitude by compensating for disturbances. It is preferred to use CMGs, instead of attitude control thruster firings. However, prior to an extravehicular activity (EVA) on the Russian Segment (RS), the Docking Compartment (DC1) must be depressurized, as it is used as an airlock. When the DC1 is depressurized, the CMGs’ margin of momentum is insufficient to compensate for the disturbance and the service module (SM) attitude control thrusters need to fire to desaturate the CMGs. The SM roll control thruster firings induce fuel oxidizer reaction products (FORP) contamination of the adjacent SM surfaces. One of the components present in FORP is the potent carcinogen N-nitrosodimethylamine (NDMA). Since the EVA crewmembers often enter the area surrounding the thrusters for tasks on the aft end of the SM and when translating to other areas of the Russian Segment, the presence of FORP contamination is a concern. This paper will discuss FORP contamination of the SM surfaces, the potential release of NDMA in a humid environment from crew EVA suits, if they happen to be contaminated with FORP, and the toxicological risk associated with the NDMA release. Key words: fuel oxidizer reaction products, FORP, NDMA, EVA

193 J.I. Kleiman (ed.), Protection of Materials and Structures from Space Environment, 193–208.  C 2006 Springer. Printed in the Netherlands.

194

WILLIAM SCHMIDL ET AL.

1. Introduction The U.S. Control Moment Gyros (CMGs) maintain the International Space Station (ISS) vehicle attitude by compensating for disturbances. However, when the docking compartment (DC1), the location of the Russian airlock, of the International Space Station (ISS) is depressurized for extravehicular activities (EVAs), the service module (SM) attitude control thrusters have to fire because the CMGs have an insufficient margin of momentum to compensate for the disturbance and must be desaturated. Thruster firings produce fuel oxidizer reaction products (FORP) that can contaminate adjacent surfaces. For EVAs on the aft end of the service module (SM) of the Russian Segment (RS), there is a concern that when the EVA crewmember translates around the FORP contaminated area they could inadvertently brush against the FORP and transfer some of it to their suit. FORP is composed of both volatile and nonvolatile components. How fast the volatile components leave varies. One of the components present in FORP that represents a toxicological risk to the crew is the potent carcinogen N-nitrosodimethylamine (NDMA). Although NDMA is volatile, it does persist long enough to be a concern. In addition, when dried FORP is reintroduced into a humid environment, like the ISS cabin, NDMA can reform from the components remaining in the FORP. So the concern is that when FORP (on the suit) is brought back into the humid environment of the ISS cabin, it can release NDMA into the atmosphere.

2. Background Discoloration around the SM zenith roll thrusters was observed during the ISS flight 5A Orbiter fly around of the ISS, as shown in the image in figure 1. In the image, the pitch thrusters are closest to the aft end of the SM (right side of the image). The roll thrusters are to the left of the pitch thrusters and raised above the surface of the SM. The discoloration (brown color) can be clearly seen close to the roll thrusters in the inset zoom image. The EVA handrails can be seen in the inset image close to the roll thrusters and on either side of the pitch thrusters. Contamination has also been observed around the SM nadir roll thrusters. Figure 2 shows the relative position and direction of the SM attitude control thrusters. It can be seen that the roll thrusters’ plume centerline is directed 47◦ away from the surface normal (lower diagram), while the pitch and yaw thrusters’ plume centerline is directed 13◦ away from the surface normal (top diagram). In addition, the pitch and yaw thrusters are recessed below the SM surface, while the roll thrusters are elevated above the surface to provide the roll control component. So the plumes from the roll thrusters are more likely to contaminate the adjacent

FUEL OXIDIZER REACTION PRODUCTS

195

Figure 1. Fight 5A observation of darkening near service module (SM) zenith roll thrusters. The inset image shows an enlarged image of the zenith roll and pitch thrusters. The brownish discoloration is visible near the role thrusters

SM surfaces than the plumes from either the pitch or yaw thrusters. The ISS External Contamination team has concluded that the discoloration is plume contamination due to the thrusters firings. In such areas, it must be assumed that FORP is present. Roll Thrusters

13°

Pitch Thrusters Yaw Thrusters

Contamination expected from roll Thrusters

Pitch Thrusters 47°

Roll Thrusters

Figure 2. Service module (SM) attitude control thrusters’ position and direction. The pitch and yaw thrusters point away from the vehicle surface (13◦ from the normal). The roll thrusters point 47◦ from the normal. contamination from the roll thrusters is expected on the adjacent SM surfaces

196

WILLIAM SCHMIDL ET AL.

Figure 3. (a) Service module (SM) Gas Dynamic Protection Unit (GZU) for the SM roll thrusters prior to flight and installation; (b) SM roll thrusters prior to installation of the GZU; and (c) SM roll thrusters (right side of the image) and pitch thrusters (left side of the image) prior to installation of the GZU

The image in figure 1 was acquired before shields, gas dynamic protection devices (GZUs), were installed on the thrusters in January 2002. Rocket and Space Corporation Energia (RSC-E) has designed the GZUs to constrain the thruster plume and limit contamination of the surrounding surfaces. Figure 3(a) shows the GZU for one of the roll thrusters prior to flight. It fits over the top of the roll thrusters. The handle on the top of the GZU is used during installation and is not used for translating during EVAs, as it becomes contaminated. The brackets on the sides of the GZU are used to install the GZU on fittings that were preinstalled on the SM surface prior to flight. Figure 3(b) shows an image of roll thrusters prior to installation of the GZU. Figure 3(c) shows the other side of the roll thrusters (right side of the image) and the corresponding pitch thrusters (left side of the image). It should be noted that the GZUs used for the pitch and yaw thrusters are different than the ones used for the roll thrusters because they are recessed below the SM surface. Images have been taken at regular intervals from the DC1 window of the SM nadir attitude control thrusters and the Russian Kromka experiment. Figure 4 shows one of these images. Kromka is an experiment to measure how well the GZUs are performing and to test how well some material samples perform in space [3]. The Kromka experiment is the material tray visible in the middle of the image. It has some small material samples mounted on it. In front of the Kromka are the pitch thrusters with its GZU installed. In front of the pitch thrusters are

FUEL OXIDIZER REACTION PRODUCTS GZU for SM pitch thrusters

197

Kromka experiment

GZU for SM roll thrusters

Figure 4. Image taken from Docking Compartment 1 (DC1) window. Kromka is visible in the middle of the image

the roll thrusters with its GZU installed. It can be seen that the GZUs are heavily contaminated compared to the preflight GZU in figure 3(a). Figure 5 shows a diagram of the nadir side of the SM and the relative position of the Kromka experiment and the thrusters. The Kromka experiment is visible near the SM pitch thrusters. EVA handrails can be seen on either side of the pitch thrusters.

3. Eva Constraints Due to the presence of FORP on the SM surfaces adjacent to the roll thrusters, additional EVA constraints were required to be implemented in these areas. The constraints were initially established through a nonconformance report (NCR) that discussed the removal of the Kromka 1-0 experiment and installation of the Kromka 1-1 experiment and in subsequent ISS Program Safety Review Panel (SRP) discussions. The constraints were later confirmed in the protocol from a

Figure 5. Nadir side of service module (SM). The Kromka experiment is visible near the SM pitch thrusters

198

WILLIAM SCHMIDL ET AL.

joint U.S./Russian FORP technical interchange meeting held in Houston, TX, April 15–26, 2002. The EVA constraints were initially developed because the Kromka experiment is in close proximity to the SM thrusters, as can be seen in figure 5, and the EVA crewmembers would need to enter that area. The constraints included establishing a one meter keep-out zone (KOZ) around the thrusters for 2.5 h after the last SM thrusters fired before the EVA crew members could enter the area, procedures for inspecting the EVA suits prior to ingress back into the airlock, and procedures for wiping the gloves and suit with towels that are jettisoned to retrograde. Also, once inside the ISS, the EVA gloves are bagged to mitigate any potential risk from FORP. Since EVAs are generally very time constrained, the ISS Program approved a test program at the NASA White Sands Test Facility (WSTF) to obtain FORP test data that could be used to determine if those EVA constraints could be relaxed.

4. NASA White Sands Test Facility (WSTF) Laboratory Tests 4.1. INTRODUCTION

A test program was setup at the NASA White Sands Test Facility to obtain the data that would be needed to determine what EVA constraints would be required to mitigate the risk of an EVA crewmember inadvertently contacting a FORP contaminated surface and bringing the FORP back into the humid ISS cabin. The program included tests to determine the evaporation rate of FORP on the zenith (25◦ C, hot) and nadir (−40◦ C, cold) sides of the ISS, the evaporation rate of NDMA within the FORP on the zenith and nadir sides of the ISS, the quantity of NDMA that would be released in a humid pressurized environment from the dried FORP, and the rate at which NDMA reforms when dried FORP is introduced into a humid environment. Two groups of tests were performed. The first group of tests were performed during Calendar Year 2003. Results from this group of tests are designated “CY 2003.” Results from these tests included 100 h evaporation rate data for FORP for both the zenith and nadir cases, NDMA evaporation rate data for the corresponding zenith and nadir cases, and NDMA formation rate data. Based on these results, the time to remain outside the keep-out zone (KOZ) after the last SM thrusters fired was reduced from 2.5 to 2 h. Additional tests were requested to determine if the time to remain outside of the KOZ could be further reduced from 2 to 1 h. This was the series of tests performed during early Calendar Year 2004 and designated “CY 2004.” Results from these tests included 1 to 6 h evaporation rate data for FORP for the zenith

199

FUEL OXIDIZER REACTION PRODUCTS TABLE 1. FORP composition ion results Ionic species CY 2003 Residue (10%)b CY 2004 CY 2004 Residue (14%)b a b

Ammonium (%/wta )

Methylammonium (%/wta )

Dimethylammonium (%/wta )

Nitrate (%/wta )

Nitrite (%/wta )

0.05

1.1

7.9

36

0.08

0.6 0.3

2 1.7

20 13

9 48

20 0.6

%/wt—Weight percent Ionic Species were analyzed after UFORP was subjected to vacuum for 5 days

and nadir cases, NDMA evaporation rate data for the corresponding zenith and nadir cases, and NDMA formation rate data.

4.2. PREPARATION OF FORP

For each group of tests, a batch of FORP was generated using a permeation technique developed at the NASA White Sands Test Facility (WSTF) [2]. In this technique, separate Unsymmetrical Dimethylhydrazine (UDMH) and NO2 gas streams are concentrated in a small controlled area. The batch of FORP needed for the tests was prepared over a couple of weeks. For the formation test, a sample of FORP from each batch was evaporated for 5 days at 25◦ C in a vacuum to generate a sample of dried FORP. Since the FORP was prepared over a long duration, the composition of the two batches of FORP varied. The composition of the FORP batches used in the tests is shown in table 1. For the CY 2003 FORP, only the composition of the evaporated sample was measured. When comparing the two evaporated samples, it can be seen that the CY 2004 FORP has a higher concentration of dimethylammonium, nitrates, and nitrites, 7.9% vs. 13%, 36% vs. 48%, and 0.08% vs. 0.6% respectively, than the CY 2003 FORP. The higher concentration of dimethylammonium, nitrates, and nitrites in the CY 2004 FORP likely explains why it has more mass remaining after the 5 day evaporation, 14% vs. 10%, than the CY 2003 FORP. The nitrite levels of the CY 2004 FORP before evaporation and after evaporation, 20% of the initial mass and 0.6% of the 14% mass remaining, indicate that the nitrite concentration is decreasing. A lower nitrite level significantly decreases the NDMA formation rate. This was seen in the results that will be discussed later. The concentration of dimethylammonium is also higher in the CY 2004 dried FORP, 13% vs. 7.9%, than in the CY 2003 FORP. The higher concentration of dimethylammonium and nitrites indicates that a higher NDMA formation rate

200

WILLIAM SCHMIDL ET AL.

would be expected for the CY 2004 FORP than the CY 2003 FORP. This was seen in the results. 4.3. MEASUREMENT TECHNIQUE

The technique used to measure the evaporation rate for FORP and NDMA and the NDMA formation rate is discussed in the WSTF report “Evaporation rate study and NDMA formation from UDMH/NO2 Reaction Product,” WSTF-IR-0188001-03 (2003) [1]. 4.3.1. Evaporation test The FORP evaporation rate was determined by measuring the difference between the initial weight of FORP on a slide before it was put in a vacuum chamber and the final weight of FORP on the slide after it was removed from the vacuum chamber. To determine the initial weight of FORP placed on the slide, an empty syringe was weighed, then ∼0.200 ml of FORP was put in the capped syringe and the syringe was reweighed. This was necessary because of the presence of NDMA in the FORP. A blank slide was also weighed. The weight of the blank slide, and the difference in weight of the syringe with and without the FORP is the weight of the initial FORP and slide. The evaporation test was performed in a vacuum chamber at 10−3 Torr. Earlier tests performed by Dee [2] showed that more FORP and NDMA would be removed for tests performed at 10−6 Torr. So the results from these tests will be more conservative than what would be expected on-orbit. After the samples were removed from the vacuum chamber, they were allowed to warm up to room temperature in a dessicator before being reweighed. The NDMA evaporation rate was measured by using gas chromatographymass spectroscopy (GC-MS). The NDMA concentration in the initial sample was measured before the evaporation test. After the evaporation test, the FORP remaining was rinsed from the slide and the GC-MS test performed again. 4.3.2. NDMA formation test The NDMA formation test was performed using a solid phase microextraction (SPME) test. The NDMA formation rate was measured by sampling the headspace above the sample at selected time intervals. 4.4. TEST RESULTS

4.4.1. NDMA evaporation rate The test results in figure 6 show the concentration of NDMA relative to the initial mass of FORP decreases rapidly. The test results are for the nadir (−40◦ C, cold) case. The red squares are from the CY 2004 tests and the blue squares are from

201

FUEL OXIDIZER REACTION PRODUCTS μg NDMA/g FORP(t=0) vs Time (hrs) 100000 μg NDMA/g FORP(t=0)[2003(-40°C)] μg NDMA/g FORP(t=0)[2004(-40°C)]20

10000

μg NDMA/g FORP(t=0)

1000

100

10

1 0

1

2

3

4

5

6

7

Time (hrs)

Figure 6. Concentration of NDMA relative to the initial mass of FORP (μg NDMA/g FORP) for the −40◦ C case (nadir, cold case) decreases rapidly with time

the CY 2003 tests. It can be seen that the CY 2004 FORP starts (time = 0) with a higher concentration of NDMA than the CY 2003 FORP. However, by the time it reaches the 2 and 4 h data points, the concentration of NDMA relative to the initial amount of FORP is comparable and the CY 2004 and CY 2003 data points overlap. The results in figure 7 are for the zenith (25◦ C, hot) case. It can be seen that the NDMA concentration drops more rapidly for the 25◦ C case than for the −40◦ C case. After 1 h the concentration has dropped approximately 2 orders of magnitude, compared with the 1 order of magnitude for the −40◦ C case. It can also be seen that by 1 h the concentration drop has reached a plateau. 4.4.2. FORP evaporation rate Figure 8 shows that the FORP volatilizes rapidly to a stable mass that persists over a longer period of time. The results show that within 1 h the FORP has reached a stable mass and that the CY 2004 FORP settles at a higher mass than the CY 2003 FORP. For CY 2004, the FORP remaining after 1 h is ∼36%. The results for CY 2003 showed 12–22% of the initial mass remained. The higher mass remaining is likely due to the higher concentration of nitrites and nitrates present in the CY 2004 FORP. The results in figure 9 are for the zenith (25◦ C, hot) case. It can be seen that the FORP volatilizes rapidly to a stable mass. These results showed that the CY 2004 FORP remaining after 1 h was 22% and that for CY 2003 it was 10–11%. Again,

202

WILLIAM SCHMIDL ET AL. μg NDMA/g FORP(t=0) vs Time (hrs) 100000 μg NDMA/g FORP(t=0)[2003(25°C)] μg NDMA/g FORP(t=0)[2004(25°C)])

μg NDMA/g FORP(t=0)

10000

1000

100

10

1 0

2

1

3

4

5

6

7

Time (hrs)

Figure 7. Concentration of NDMA relative to the initial mass of FORP (μg NDMA/g FORP) for the 25◦ C case (zenith, hot case) decreases rapidly with time. Within 1 h the concentration level has reached a plateau

FORP Weight % vs Time (hrs) 100

FORP Weight %

(FORP Weight %)[2003(−40°C) (FORP Weight %)[2004(−40°C)

10

1 0

1

2

3

4

5

6

Time (hrs)

Figure 8. FORP weight (%) vs. time for the −40◦ C case (nadir, cold case)

7

203

FUEL OXIDIZER REACTION PRODUCTS FORP Weight % vs Time (hrs) 100 (FORP Weight %)[2003 (25°C)] (FORP Weight %)[2004 (25°C)]

FORP Weight %

3 pts

2 pts

10 2 pts

2 pts

1 0

1

2

3

4

5

6

7

Time (hrs)

Figure 9. FORP weight (%) vs. time for the 25◦ C case (zenith, hot case). FORP volatilizes rapidly to a stable mass that persists for a longer period of time

the higher mass remaining is likely due to the higher concentration of nitrites and nitrates. 4.4.3. FORP formation rate Table 1 shows the nitrite levels of the CY 2004 FORP before evaporation and after evaporation, 20% of the initial mass and 0.6% of the 14% mass remaining. The nitrite levels indicate that the Nitrite concentration in the FORP is decreasing. A lower nitrite level in the FORP will decrease the NDMA formation rate. This was observed in the results, as no NDMA formation was detected in the CY 2003 FORP when moisture was introduced into the sample of dried FORP. For the CY 2004 sample of dried FORP, the NDMA formation rate was negligible when moisture was introduced. It was determined that to form NDMA, nitrite has to be present in the sample. Both the CY 2003 and CY 2004 dried FORP samples were spiked with nitrite before the formation rate was measured. The nitrite introduced was 25% of the mass of nitrate present in the sample. Table 1 also shows that the concentration of dimethylammonium is higher for the CY 2004 FORP, 13% vs. 7.9%. The higher concentration of dimethylammonium indicates that a higher NDMA formation rate would be expected for the CY 2004 FORP. Figure 10 shows the NDMA formation rate that was measured. The plot shows the rate for both the CY 2003 and CY 2004 FORP. The solid symbols indicate FORP where nitrite is added, and the open symbols are FORP with no nitrite

204

WILLIAM SCHMIDL ET AL. μg NDMA / g FORP Formed vs Time 1.00E+05

μg NDMA / g FORP

1.00E+04

1.00E+03

1.00E+02

2004 μg NDMA/g FORP (Nitriteadded) 2004 μg NDMA/g FORP (Nonitriteadded) 2003 μg NDMA/g FORP (Nitriteadded)

1.00E+01

2003 μg NDMA/g FORP (Nonitriteadded)

1.00E+00

1.00E-01 0

1

2

3

4

5

6

7

Time (hrs)

Figure 10. NDMA concentration (μg NDMA/g FORP) vs. time measured during the NDMA formation test. It can be seen that the NDMA forms rapidly

added. It can be seen that when no nitrite is present the NDMA formation rate is very low. For the CY 2003 FORP, no NDMA formation was detected. For the CY 2004, a low rate of ∼100 μg NDMA/g FORP was measured. The results in figure 10 also show that for the CY 2003 FORP, the NDMA formation rate is 1800 μg NDMA/g FORP present after 2 h and that the formation rate for the CY 2004 FORP was higher, 18400 μg NDMA/g FORP present after 2 h. One cause for the higher formation rate is the higher Dimethylammonium concentration in the CY 2004 FORP. The mass of CY 2004 FORP remaining after the 5 days of evaporation was also higher. This might indicate that there were other components present in the FORP that might have resulted in a higher NDMA formation rate.

5. Methodology A methodology was developed to determine the FORP and NDMA remaining on the SM surface after the SM roll thrusters fire prior to an EVA and the subsequent release of NDMA in a humid environment due to inadvertent contact by an EVA crew member with the service module (SM) surface in the vicinity of the SM roll thrusters contaminated with FORP. The first step is to calculate how much FORP would be deposited on the adjacent SM surfaces by the thrusters firing prior to an EVA. To calculate the FORP remaining on the SM surface, the Russian plume model was used. The thruster

FUEL OXIDIZER REACTION PRODUCTS Roll Thrusters with GZUs

205

Roll Thrusters

Pitch Thrusters with GZUs

Figure 11. Zenith side of service module (SM). Gas Dynamic Protection Units (GZUs) constrain the thruster plume. The distance from the roll thruster to the closest SM surface outside the GZUs is ∼3.2 in (∼8 cm)

firing times were obtained from the ISS Program’s Guidance, Navigation and Control group (GN&C). The value of 45 s was used because it is the longest thrusters firing time that has been observed during the previous EVAs. The next step is to calculate how much FORP would remain 1 h after the SM thrusters have fired. This was determined using the WSTF laboratory evaporation test data for the Nadir case. The conservative value of 36% FORP remaining used. Figure 11 shows a map of the distribution of FORP from the thruster remaining on the SM thruster based on the plume model and laboratory data. It can be seen that the FORP concentration drops rapidly with distance from the thrusters. Next, to be conservative, it is assumed that all the FORP in a 100 cm2 area is transferred to the suit by the inadvertent swipe. Based on the amount of FORP transferred, the amount of NDMA that would be released from this FORP inside the cabin in the first 2 h was calculated using the rate of 400 μg NDMA/g FORP present. This value was obtained from the WSTF data based on the evaporation rate of NDMA for the zenith hot case. The amount of NDMA that would be formed from the dried FORP in the first 2 h was also calculated based on the more conservative CY 2004 rate, 18400 μg NDMA/g FORP present. Figure 12 shows that since the GZUs were installed, the closest point on the SM surface that can be touched is ∼3.2 in (∼8 cm) from the edge of the thruster. The diagram is of the zenith side of the SM. The figure on the left shows the roll thrusters with the GZUs installed. The figure on the right shows the distance from the edge of the roll thruster to the closest SM surface outside of the GZUs. The roll thruster diameter is 1.8 in. The distance from the roll thruster centerline to the closest SM surface is ∼5 in (12.7 cm). The predicted concentrations of NDMA with distance from the SM thrusters that would be released by an inadvertent swipe into the DC1 and ISS compartments

206

WILLIAM SCHMIDL ET AL.

Figure 12. Map of the FORP predicted to remain after 1 h on the surface around the SM roll thrusters after a 45 s roll thruster firing. It can be seen that the FORP concentration drops rapidly with distance from the thruster

are shown in tables 2 and 3. It can be seen that the concentration drops off rapidly. Also the closest point is at the GZU itself, which is defined as a “no touch” area by flight rule. This data was given to the Toxicology group for assessment.

6. Toxicological Assessment The quantity of FORP present and NDMA released into the ISS from an inadvertent swipe by an EVA crewmember of the contaminated area around the SM TABLE 2. NDMA concentration predicted to be released in the cabin for FORP transferred to the EVA suit one hour after the roll thruster firing at different distances from the roll thruster for the −40◦ C case (Nadir)

Distance from roll jet (m)

FORP present (g·cm−2 )

NDMA released in pressurized environment

0.08 0.15 0.23 0.30 0.37 0.44

1.96E-02 6.22E-03 2.96E-03 1.72E-03 1.11E-03 7.77E-04

3.69E-02 1.17E-02 5.57E-02 3.23E-02 2.09E-02 1.46E-02

NDMA concentration (ppb) DC1

ISS

965 306 146 84 55 38

34 11 5 3 2 1

at GZU

207

FUEL OXIDIZER REACTION PRODUCTS

TABLE 3. NDMA concentration predicted to be released in the cabin for FORP transferred to the EVA suit one hour after the roll thruster firing at different distances from the roll thruster for the 25◦ C case (Zenith)

Distance from roll jet (m)

FORP present (g·cm−2 )

NDMA released in pressurized environment (g)

0.08 0.15 0.23 0.30 0.37 0.44

1.20E-02 3.80E-03 1.81E-03 1.05E-03 6.81E-04 4.75E-04

2.25E-02 7.15E-03 3.40E-03 1.97E-03 1.28E-03 8.93E-04

NDMA concentration (ppb) DC1

ISS

589 187 89 52 33 23

21 7 3 2 1 1

at GZU

roll thrusters was calculated and provided to the NASA Toxicology group for an assessment. This data is shown in tables 2 and 3. Acute toxicity has not been reported at concentrations below 10 ppm. Therefore, it can be concluded that at concentrations below 1 ppm, it is very unlikely that NDMA will produce any acute toxic reactions. So, the NASA Toxicology Group concluded that for the concentration levels expected, it is unlikely that NDMA will produce acute toxicity. The NASA Toxicology Group also found that the highest calculated risk from the projected NDMA concentrations is less than 8.46 × 10−5 (−40◦ C, distance 0.08 m from the thrusters). The NASA Toxicology Group, with the concurrence of the National Research Council Spacecraft Maximum Allowable Concentrations (SMAC) Subcommittee, accepts a cancer risk of 1/10,000 (i.e., 10−4 ) in deriving SMACs on carcinogenic compounds, such as benzene.

7. Summary Prior to an extravehicular activity (EVA) on the Russian Segment (RS), the Docking Compartment (DC1) must be depressurized, as it is used as an airlock. When the DC1 is depressurized, the CMGs’ margin of momentum is insufficient to compensate for the disturbance and the Service Module (SM) attitude control thrusters need to fire to desaturate the CMGs. The thruster firings result in FORP contamination of the adjacent SM surfaces. The FORP contamination of the SM surfaces, the release of NDMA in a humid environment from crew EVA suits, if they happen to be contaminated with FORP, and the toxicological risk associated with the NDMA release were calculated. It was determined that the FORP and NDMA evaporate rapidly and that their concentration drops off rapidly with distance from the thrusters. The

208

WILLIAM SCHMIDL ET AL.

FORP remaining after 1 h for the Nadir case was found to be 36% of the initial mass. For the Zenith case the FORP remaining after 1 h was 22% of the initial mass. The NASA Toxicology Group found that the highest calculated risk from the projected NDMA concentrations is less than 8.46 × 10−5 (−40◦ C, distance 0.08 m from the thrusters). The NASA Toxicology Group, with the concurrence of the National Research Council Spacecraft Maximum Allowable Concentrations (SMAC) Subcommittee, accepts a cancer risk of 1/10,000 (i.e., 10−4 ) in deriving SMACs on carcinogenic compounds, such as benzene. Based on these results the time to remain outside the 1 m KOZ could be reduced from 2 to 1 h.

Acknowledgments The authors gratefully acknowledge the ISS Program office for supporting this study, and the NASA White Sands Test Facility (WSTF) and its personnel for conducting the study.

References 1. Buchanan, V. D. and Dee, L. A. (2003) White Sands Test Facility (WSTF) Investigative Report, WSTF-IR-0188-001-03. 2. Dee, L. A., Davidson, V. D., and Baker, D. L. (2000) Protocol External Contamination Technical Interchange Meeting Fuel/Oxidizer Reaction Products (FORP) Plume Induced Contamination, 15–26 April 2002. 3. Naumov, S. F., Gerasimov, Y. I., Sokolova, S. P., Rebrov, S. G., Gerasimova, T. I., Kalistratova, O. V., Prokofyev, M. A., Grigorevsky, A. V., Prosvirikov, V. M., Buryak, A. K., and Chernik, V. N. (2003) In 9th International Symposium on Materials in a Space Environment, ESTEC, Noordwijk, The Netherlands, 16–20 June 2003 .

EFFECT OF VACUUM THERMOCYCLING ON PROPERTIES OF UNIDIRECTIONAL M40J/AG-80 COMPOSITES YU GAO,1 DEZHUANG YANG,1 SHIYU HE,1 AND ZHIJUN LI2 1 Space Materials and Environment Engineering Laboratory, Harbin Institute of Technology, Harbin 150001, P. R. China 2 The 39th Institute China Electronic Science and Technology Groups Inc., Xi-an 710065, P. R. China

Abstract. The vacuum thermocycling in the temperature interval of 93–413 K under vacuum of 10−5 Pa was performed for unidirectional M40J/AG-80 composites, and changes in the bending strength and modulus, the mass loss ratios, the surface morphology and the internal structure transitions were examined. Experimental results show that with increasing the vacuum-thermal cycles, both the bending strength and modulus exhibit an increase trend followed by decreasing after 40 cycles and leveling off after 97 cycles. The mass loss ratio increases firstly and then tends to level off after 48 cycles. The vacuum thermocycling could cause the debonding in the interfacial layers between the AG-80 epoxy matrix and carbon fibers, as well as the post curing in the epoxy matrix. The changes in the bending strength and modulus, the mass loss ratio and the surface morphology of the M40J/AG-80 composites can be related to the debonding and the post curing due to the vacuum thermocycling. Key words: carbon/epoxy composites, vacuum thermocycling, bending properties, debonding effect, loss tangent (tan δ) 1. Introduction Carbon/epoxy composites are extensively used in structural components for spacecraft, such as in the truss structure, antennas, and solar cell panels, etc. [1]. When flying in orbit, the spacecraft goes repeatedly into and out the shadow region of the Earth, leading to a change in the surface temperature [2]. For example, the surface temperature of spacecraft can vary in the interval 113–393 K. Since the thermal expansion coefficients are quite different between carbon fibers and the epoxy matrix, the thermocycling would result in thermally cyclic stresses and strains, leading to a degradation in mechanical properties and chemical structure. 209 J.I. Kleiman (ed.), Protection of Materials and Structures from Space Environment, 209–215.  C 2006 Springer. Printed in the Netherlands.

210

YU GAO ET AL.

Also, under high vacuum, the mass loss of the epoxy matrix composites could take place. Therefore, it is important to study the behavior of the composites under vacuum thermocycling, which might influence the performance of spacecraft in orbit [3–5]. The AG-80 resin is a new type of thermosetting matrix for advanced polymetric composites, which exhibits a unique resistance to wet and heat [6], showing a good prospect for the usage in spacecraft. The aim of this study was to examine the effect of vacuum thermocycling on the unidirectional M40J/AG-80 composite, in order to provide the basic information for its further development and application in spacecraft. The AG-80 resin is a type of tetraglycidyl diaminodiphenyl methane (TGDDM) with diaminodiphenyl sulfone (DDS).

2. Experimental A commercial grade of AG-80 resin was used as the matrix material, which was supplied by Shanghai Institute of Synthetic Resins in China. The chemical structure is shown in figure 1. Its epoxy value is approximately 0.80, and the curing agent is diaminodiphenyl sulfone (DDS). The M40J carbon fibers were chosen as the reinforcement for the composites, which were supplied by TORAY Company in Japan. The prepreg tape for the M40J/AG-80 composite specimens was laid on a mould board and placed in an autoclave for curing. The curing cycles were: heating to 130◦ C from room temperature with the heating rate 1.5–2.0◦ C·min−1 , and holding for 40–60 min at 130◦ C under the pressure of 0.6–0.7 Mpa; and then heating to 180◦ C with the rate 1.5–2.0◦ C·min−1 and holding for 120 min; finally, cooling to room temperature in the autoclave. The volume fraction of M40J fibers in the manufactured composites was approximately 60%. The vacuum thermocycling regime is shown in figure 2. Before and after the vacuum thermocycling, the bending strength and modulus of the M40J/AG80 composites were measured using a MTS 810 type test machine. The mass loss ratios for the specimens after the thermocycling were characterized by a high precision microbalance with a sensitivity of 10–5 g. The change in surface morphology of specimens due to the vacuum thermocycling was observed using an atomic force microscope (AFM) of Nanoscope IIIa Dimension 3100 type. H C

H2 C

CH2

O H C

H2C

CH2 N

CH2

CH2

H C

CH2 O

N CH2

O

H C

CH2 O

Figure 1. The chemical structure of AG-80 resin

211

EFFECT OF VACUUM THERMOCYCLING

Temperature/°C

200

(140°C,10min)

10−5Pa

100 0 −100 −200

(−180°C,10min) 0

10

20

30

40

50

60

70

Time/min

Figure 2. Temperature vs. time curve for the vacuum thermocycling

The laid direction of fibers was parallel to the observed surface of specimens. In order to reveal the effect of thermocycling on the chemical structure of the AG-80 matrix, the dynamic mechanics thermal analysis (DMA) was performed using a Rheometric Scientific DMTA Vtype spectrometer by means of a standard three-point mode. The loading frequency was 1 Hz, the temperature ranged from −130◦ to +300◦ C and the heating rate was chosen as 5◦ C·min−1 . The size of DMA samples was 40 × 7 × 1 mm3 , and the longitudinal direction of samples was parallel to the laid direction of the carbon fibers. 3. Result and Discussion 3.1. BENDING STRENGTH AND MODULUS

Bend strength/MPa

1800

(a)

1700

10−5Pa 93 413K

1600 1500 1400 0

50 100 150 200 250 300 350

Thermal cycles

Bending modulus/GPa

Figures 3(a) and 3(b) show the changes in bending strength and modulus with vacuum-thermal cycles for the M40J/AG-80 composites, respectively. The variation trends for the bending strength and modulus are similar. With increasing the 215

(b)

10−5Pa 93 413K

210 205 200 195 190 185 0

50

100 150 200 250 300

Thermal cycles

Figure 3. The bending strength and (a) modulus vs. (b) vacuum-thermal cycles for M40J/AG-80 composites

212

YU GAO ET AL.

Weight loss/%

0.4 0.3 −5

10 Pa 0.2

93

413K

0.1 0.0 0

40

80

120

160

Thermal cycles

Figure 4. The mass loss ratio vs. vacuum-thermal cycles for M40J/AG-80 composites

vacuum-thermal cycles, both the bending strength and modulus increase firstly and begin to decrease after approximately 40 cycles. After 97 cycles, the decrease tends to level off. 3.2. MASS LOSS EFFECT

The outgassing characteristics of the M40J/AG-80 composites were examined at 125 ± 0.5◦ C for 24 h under the vacuum 10−6 Torr. It was shown that the percentage of total mass loss (TML%), the water vapor reverse amount (WVR%) and the collected vapor condensate matter (CVC%) were 0.46, 0.19, and 0%, respectively. These values indicate that the M40J/AG-80 composite has a good resistance to outgassing under high vacuum environment, meeting the requirement for the materials to be used in spacecraft. Figure 4 shows the mass loss ratio vs. vacuum-thermal cycles. The mass loss ratio increases with the thermal cycles, and tends to level off after 48 cycles. The highest mass loss ratio is approximately 0.38% during the vacuum-thermal cycling. It is believed that the vacuum-thermal cycling could induce the formation of smaller molecules. Also, during the storage, the surface of M40J/AG-80 composites would absorb water molecules, which are vaporized under vacuum. The mass loss of M40J/AG-80 composites under vacuum-thermal cycling could be attributed to the smaller molecules and the water absorption. 3.3. DEBONDING EFFECT

Figure 5 shows the change in the surface morphology after vacuum-thermal cycling for the M40J/AG-80 composites, in which the carbon fibers, the epoxy matrix, and the interfacial layers are indicated by A, B, and C respectively. Before the vacuum thermocycling, the bonding of the matrix with carbon fibers is in good

EFFECT OF VACUUM THERMOCYCLING

213

Figure 5. AFM micrographs showing the change in surface morphology after vacuum-thermal cycling for M40J/AG-80 composites: (a) 0 cycles; (b) 40 cycles; (c) 274 cycles

condition, as shown in figure 5(a). After 40 cycles, debonding occurred in some interfacial layers, see the C areas in figure 5(b). But, with further increasing the thermal cycles, the debonding extent did not change a lot, figure 5(c). 3.4. DMA ANALYSIS

Figure 6 shows the change of the loss tangent (tan δ) vs. temperature spectrum after vacuum thermocycling for the M40J/AG-80 composites. The spectrum can be divided into two portions, as shown in figures 6(a) and 6(b), respectively. Figure 6(a) shows the change of the peaks for the primary or α-transition. The temperature at which the α-transition peak appears is corresponding to the glass transition temperature (Tg ). It can be seen that with increasing the vacuum-thermal cycles, the Tg value decreases slightly. After 40 cycles, the peak height for the αtransition drops obviously (almost by 18%), implying that the cross-linking extent increases for the M40J/AG-80 composites. The increase in the cross-linking extent demonstrates that the vacuum thermocycling would lead to a post curing effect on the epoxy matrix. However, when the thermal cycles increase from 40 to 274, the

214

YU GAO ET AL. 0.30

0.045

(a)

0.040

original state vacuum thermo-cycling for 40 cycles vacuum thermo-cycling for 274 cycles

(b)

0.035

0.25

tanδ

tan Ä

0.030 0.20 original state vacuum thermo-cycling for 40 cycles vacuum thermo-cycling for 274 cycles

0.15

0.10

165 170 175 180 185 190 195 200 205

Temperature/°C

Y 0.025 0.020

β

0.015 0.010 −100

−50

0

50

100

150

Temperature/°C

Figure 6. The tan δ vs. temperature spectra for the M40J/AG-80 composites after vacuum thermocycling for various cycles

peak height change a little, indicating that the post curing does not take place any more. Figure 6(b) shows the changes of the peaks for the secondary transitions, including the β and γ ones. The β peak height decreases with increasing the thermal cycles, and disappears after 274 cycles. The β transition occurring in the temperature range of –10–100◦ C can be related to the movement of benzene rings, which would weaken with increasing the cross-linking extent. The change in the β peak height could also demonstrate the post curing effect during vacuum thermocycling. Also, as can be seen in figure 6(b), the γ -transition takes place in the temperature range of –120–50◦ C, which can be related to the movement of the chain segment –CH2 CHOH(R)CH2 –. Such a chain segment could be formed by breaking the epoxy rings. The γ peak signals are stronger for the vacuum thermocycling of 40 cycles than those before the thermocycling. This phenomenon indicates that the amount of the –CH2 CHOHCH2 – segments in the epoxy matrix is larger in the former state than in the latter one, and thus the post curing would occur due to the thermocycling.

4. Conclusions M40J/AG-80 composite is a new type of carbon/epoxy materials, which could be used in spacecraft. Since the vacuum and thermocycling are important environment factors in space, it is necessary to study the damage effect of vacuum thermocycling on the M40J/AG-80 composites. It was found that with increasing the vacuum-thermal cycles, both the bending strength and modulus exhibited an increase trend followed by decreasing after 40 cycles and leveling off after 97 cycles. The mass loss ratio increased firstly and then tended to level off after

EFFECT OF VACUUM THERMOCYCLING

215

approximately 48 cycles. The vacuum thermocycling could cause the debonding in the interfacial layers between the AG-80 epoxy matrix and carbon fibers, as well as the post curing in the epoxy matrix. The changes in the bending strength and modulus, the mass loss ratio, and the surface morphology of the M40J/AG-80 composites can be related to the debonding and the post curing, which are caused by the vacuum thermocycling. The debonding and post curing mainly occur during the thermocycling less than 40 cycles.

References 1. Xiao, S. and Liu, Z. (1993) Aerospace Materials and Technology 4, 1. 2. George, P. E. and Dursch, H. W. (1994) Journal of Advanced Materials 25(3), 10–19. 3. Shin, K.-B., Kim, C.-G., Hong, C.-S., and Lee, H.-H. (2000) Composites Part B: Engineering 31(3), 223–235. 4. Seehra, S., Benton, D., Rosen, J., and Gounder, R. (1985) SAMPE Journal 21(2), 18–23. 5. Tennyson, R. C. and Matthews, R. (1995) Journal of Spacecraft and Rockets 32(4), 703–709. 6. Wang, R.-M. and Lan, L.-W. (2001) Thermosetting Resin 16(1), 36–38.

DAMAGE CHARACTERISTICS OF Zr41 Ti14 Cu12.5 Ni10 Be22.5 BULK METALLIC GLASS IMPACTED BY HYPERVELOCITY PROJECTILES C. YANG,1,2 C. Z. FAN,3 Y. Z. JIA,3 X. Y. WANG,1,2 X. Y. ZHANG,1 H. Y. WANG,1 Q. JING,1 G. LI,1 R. P. LIU,1 L. L. SUN,2 J. ZHANG,2 AND W. K. WANG1,2 1 Key Lab of Metastable Materials Science and Technology, Yanshan University, Qinhuangdao 066004, P. R. China 2 Institute of Physics, Chinese Academy of Sciences, Beijing 100080, P. R. China 3 School of Material Science and Technology, Harbin Institute of Technology, Harbin 150001, P. R. China

Abstract. The damage characteristics of Zr41 Ti14 Cu12.5 Ni10 Be22.5 bulk metallic glass under planar shock wave have been investigated by firing aluminum projectiles using a two-stage light gas gun. The SEM results show that radial and symmetric cracks formed on the shocked plane of the sample at the impacted location by aluminum projectile at the velocity of 2.7 km·s−1 . Parallel shear crack/bands in the sublayer under the shocked plane were formed. For a better understanding of the response features under shock wave, hypervelocity impact tests with conventional sphere aluminum projectiles were carried out. Besides the same adiabatic shear crack/bands and crack propagations, craters were formed and lamination cracks occurred. Key words: bulk metallic glass, impact, fracture

1. Introduction Multicomponent Zr41 Ti14 Cu12.5 Ni10 Be22.5 bulk metallic glass (BMG) with extremely high glass forming ability, one of the most widely studied BMGs [1], has recently gained considerable attention due to its low density, high strength and fracture toughness, good corrosion and wear resistance, excellent ductibility, and uniquely dynamical deformation characteristics such as adiabatic shear banding. In recent years, this specific alloy has already made commercial application potential such as headway in golf clubs. BMG is also a likely candidate as a 217 J.I. Kleiman (ed.), Protection of Materials and Structures from Space Environment, 217–223.  C 2006 Springer. Printed in the Netherlands.

218

C. YANG ET AL.

space vehicle material. The BMG used for the spacecrafts might face a danger of collision by space debris and micrometeoroids. In order to protect the space vehicles against the impact of space debris and micrometeoroids, related knowledge of high strain rate response must be obtained to evaluate the dynamic properties of the BMG impacted by hypervelocity projectiles. Extensive research on damage behavior of Zr41 Ti14 Cu12.5 Ni10 Be22.5 BMG under the uniaxial compression [2–4], tensile tests [5, 6], and shocking loading by a powder gun loading system [7] has been carried out in the past. However, these studies paid particular emphasis to low and medium strain rate response of the BMG. In this paper, we report the damage characteristics of a Zr41 Ti14 Cu12.5 Ni10 Be22.5 BMG under the planar shock wave treatment and the impact of hypervelocity projectiles by accelerating aluminum projectiles using two-stage light gas gun.

2. Experimental Procedure The ingots of the alloy with a nominal composition of Zr41 Ti14 Cu12.5 Ni10 Be22.5 were firstly prepared from a mixture of pure elements in an arc-melting furnace under Ti-gettered Argon atmosphere. The purity of Zr, Ti, Cu, Ni, and Be were 99.999, 99.9, 99.5, 99, and 99.5% respectively. The ingots were crushed and remelted in a quartz tube, and then were quenched into water to obtain a BMG rod. The structure of the BMG was identified by X-ray diffraction (XRD) to be fully amorphous and no crystalline phase was detected. Small cylinders 10 mm long, cut out from the alloy rod, were utilized as impacted target. Targets were recovered using a “soft” recovery device, as shown in figure 1. The amorphous BMG cylinder was embedded into a copper cylinder with a diameter of 40 mm.

Figure 1. Schematic diagram of the recovery device for high speed impact [(1) steel cylinder; (2) steel cylinder tube; (3) aluminum cylinder; (4) polyethylene cylinder; (5) copper cylinder; (6). Zr41 Ti14 Cu12.5 Ni10 Be22.5 BMG target]

DAMAGE CHARACTERISTICS OF Zr41 Ti14 Cu12.5 Ni10 Be22.5 BMG

219

A steel cylinder was mounted tightly beneath the copper cylinder and aluminum cylinder was mounted tightly beneath the steel cylinder. The polyethylene cylinder was mounted tightly beneath the aluminum cylinder. Finally, the assembly of these cylinders was inserted into a steel cylinder with 40 mm inside diameter and 120 mm outer diameter. The end facing the direction of the cylinder assembly was machined as shown in figure 1. The other end was fixed by bolts. The BMG cylinder samples of diameter 22 mm were impacted by the hypervelocity spherical aluminum projectiles of 5 mm in diameter at speeds of about 2.6, 3.2, and 3.7 km·s−1 , respectively. Though these speeds are not very high, they can cause a catastrophic failure of protected layers for space vehicles [8]. As compared with the damage behavior for hypervelocity impact, planar shock experiment was carried out using the same recovery device except that the BMG sample with a diameter of 18 mm was covered by a copper plate with a thickness of 2 mm. In the planar shock experiment, the cylinder aluminum projectile was fired at a speed of 2.7 km·s−1 . The projectile was designed to be 22 mm in diameter and 2 mm thick in order to avoid the influence of side rarefaction wave and reflected rarefaction wave on the sample. The axis of projectile is coincident with the steel cylinder assembly. The projectiles were fired by two-stage light gas gun. The velocities of the projectiles were measured by the electromagnetic induction method. The specimens were cut from the impacted recovered targets with an electric spark-cutting machine, and were grinded and polished mechanically after cutting. The morphology of the craters and microstructure changes in the region near the crater and deformation damage characteristics of cross section such as adiabatic shear bands and microcracks were observed by XL30 S-FEG scanning electron microscope (SEM).

3. Results and Discussion 3.1. TYPICAL DAMAGE MORPHOLOGY OF CRATERS FORMED IN Zr41 Ti14 Cu12.5 Ni10 Be22.5 BMG MEDIUM THICKNESS TARGETS

Figure 2 shows the craters that were formed in the Zr41 Ti14 Cu12.5 Ni10 Be22.5 BMG target after the hypervelocity impact with velocities of 2.6, 3.2, 3.7 km·s−1 , respectively. The samples’ projectile 22 mm in diameter and a thickness of 10 mm belong to the medium thick target, in which the impact wave can interact with the side and back faces, resulting in significant effect on the impact process. It can be seen that shape of the craters is almost spherical coronary without an obvious outer projecting fringe in the “mouth” of the crater, which is extruded from the crater by high pressure and commonly observed in the target impacted by hypervelocity projectiles. This may be related to the high strength and low ductility of the BMG

220

C. YANG ET AL.

a

b

c Figure 2. Cross-section morphologies of damage craters formed in the Zr41 Ti14 Cu12.5 Ni10 Be22.5 BMGs medium thick target: (a) sample A: v = 2.6 km·s−1 ; (b) sample B: v = 3.2 km·s−1 ; (c) sample C: v = 3.7 km·s−1

materials. The shape of the crater becomes spherical from spherical coronary with the increase of projectile velocity.

3.2. FRACTURE DAMAGE CRACKS/BANDS UNDER PLANAR SHOCK LOADING

The resulting dimensions of the sample after the planar shock loading with an aluminum flyer at a velocity of 2.7 km·s−1 were 20.5 mm in diameter and 7.3 mm thickness compared to 17.5 mm in diameter and 10 mm thickness for the original sample. The upper surface facing the flyer exhibits an annular crack and many densely radial cracks initiating from a central point. The reason for this phenomenon is that the central part of aluminum flyer collided firstly with the central part of the BMG sample because the flyer was hindered by air in the target housing. The central part of the flyer deforms during the flight. The compressive shock wave therefore acts initially on the center point of the BMG target, thus initiating the failure in this part first. Then, radial cracks propagate from the point under the action of strong shock wave. Annular cracks formed naturally by interaction between the wave (tensile wave) reflected from the cylinder copper wall and compressive shock wave. The recovered aluminum flyer with a prominent

DAMAGE CHARACTERISTICS OF Zr41 Ti14 Cu12.5 Ni10 Be22.5 BMG

221

Figure 3. SEM micrographs showing cross-section microstructures of Zr41 Ti14 Cu12.5 Ni10 Be22.5 BMG under planar shock loading of aluminum flyer with a speed of 2.7 km·s−1

feature formed along the direction of flight demonstrates experimentally the above discussed phenomenon. Figure 3 shows a SEM microstructure micrograph of Zr41 Ti14 Cu12.5 Ni10 Be22.5 BMG target hit by aluminum flyer with a speed of 2.7 km/s. Several fracture planes (shear cracks/bands) 2 mm apart and approximately parallel to each other were seen in the upper part of the sample (that are not shown in figure 3 due to limitation of the view field of SEM). Another kind of shear cracks/bands with an angle of 40◦ between the fracture plane and horizontal direction result from the interaction between the shock wave reflected from the cylinder copper wall and compressive shock wave. This kind of cracks/bands cannot appear in usually uniaxial compressive and tensile tests because the exterior walls of the cylinder sample has not been restrained under these conditions. In our case, as discussed above, the sample was reduced in thickness and expanded in diameter after the impact. Once the shearing occurs along the direction shown by the left arrow, it results in the formation of primarily adiabatic shear band and induces a new secondary shear band along the direction shown by the left arrow. Thus a triangular region with an apex angle of 115◦ is extruded from the sample because of convergence between a reflected shock compressive wave along the radial direction of the sample from the interior wall of the recovery copper cylinder and a compressive shock wave along the projectile trajectory. The upper two shear cracks/bands has already intersected each other, while the subjacent two haven’t. The zigzag-shaped cracks (indicated by right arrow) on the micrograph of the BMG suggest that the spalling of the BMG is characterized by brittle damage resulting from the shear localization even though the sample is under tension during decompression process. Shear cracks/bands propagate and end at a point where several secondary radial microcracks (shear

222

C. YANG ET AL.

Figure 4. Typical SEM micrographs of shear bands/cracks observed on the cross section of the recovered specimen after shock loading with a velocity of v = 3.7 km·s−1

cracks/bands) initiate. It can be predicted that these secondary microcracks would continue to propagate with the increase of shock pressure. 3.3. SHEAR CRACKS/BANDS UNDER THE IMPACT OF HYPERVELOCITY PROJECTILES

Figure 4 shows a typical SEM micrograph of shear bands/cracks observed on the cross section of the recovered specimen after shock loading by a hypervelocity aluminum projectile 5 mm in diameter with a velocity of v = 3.7 km·s−1 . Because the target belongs to medium thick type and the projectile is spherical, the shock wave reflected from the back and side faces interacts with the subsequent compressive wave, leading to sophisticated microdamages. Zigzag-shaped edge side shear band/crack (as indicated by the thick arrow) along vertical direction can be observed clearly, which is a macro-damage feature caused by edge side shear stress induced by hypervelocity projectile. The observed spallation of the BMG is due to propagation of higher energy shear cracks/bands. It can be suggested, therefore, that a crater/hole would be left at the BMG target face by a projectile having a high enough velocity to posses a sufficiently high kinetic energy. Lamination crack phenomenon (as indicated by the thin arrow) with a distance of 2 mm away from back surface can also be observed at our experimental velocity. The reason for this phenomenon is that the reflected rarefaction wave, from recovery casket copper of sample, interacts with compressive waves and form tensile waves in the BMG target due to lower impedance than the BMG. Lamination crack appears spontaneously if tensile stress exceeds the yield strength of the BMG. In our cases shear bands/cracks are visible clearly in the SEM micrographs. Furthermore, the

DAMAGE CHARACTERISTICS OF Zr41 Ti14 Cu12.5 Ni10 Be22.5 BMG

223

sizes of these shear bands/cracks increase gradually with the increase of the impact velocity.

4. Conclusions Zr41 Ti14 Cu12.5 Ni10 Be22.5 BMG glasses under shock loading by two-stage light gas gun revealed shear fracture cracks/bands formation. Radial symmetric cracks were formed at the shocked plane. Shear crack/bands parallel to each other on the part close to the shocked plane were formed readily on the cross section of the recovered sample. The depth of crater in the target increases linearly with increase of the projectile velocity in the range of experimental velocity. Edge sides shear band/crack, together with lamination shear band/crack has been observed on the recovered samples impacted by hypervelocity projectiles. Based on these results, it can be concluded that BMG can be used as a candidate for a space vehicle material that resist the impact of small space debris.

Acknowledgment We wish to thank the National Natural Science Foundation for its support of this research through Grant 50171077, 50171059, and 50325103.

References 1. Tang, X. P., Geyer, U., Busch, R., Johnson, W. L., and Wu, Y. (1999) Nature 402, 160–162. 2. Subhash, G., Dowding, R. J., and Kecskes, L. J. (2002) Materials Science and Engineering A334, 33–40. 3. Gilbert, C. J., Ager III, J. W., Schroeder, V., and Ritchie, R. O. (1999) Applied Physics Letters 74, 3809. 4. Wright, W. J., Schwarz, R. B., and Nix, W. D. (2001) Materials Science and Engineering A 319–321, 229–232. 5. Li, J. X., Shan, G. B., Gao, K. W., Qiao, L. J., and Chu, W. Y. (2003) Materials Science and Engineering A 354, 337–343. 6. Flores, K. M. and Dauskardt, R. H. (2001) Materials Science and Engineering A 319–321, 511–515. 7. Zhuang, S. M., Lu, J., and Ravichandran, G. (2002) Applied Physics Letters 80, 4522. 8. Zukas, J. A. (1989) Impact Dynamics (translated into Chinese by Zhiyun Zhang et al.), Publishing Company of Military Industry, P. R. China, p. 182.

EFFECT OF VUV RADIATION ON PROPERTIES AND CHEMICAL STRUCTURE OF POLYETHYLENE TEREPHTHALATE FILM

GUIRONG PENG, DEZHUANG YANG, AND SHIYU HE Space Materials and Environment Engineering Lab, Harbin Institute of Technology, Harbin 150001, P. R. China

Abstract. The effect of VUV radiation on polyethylene terephthalate (PET) film was investigated. A gas-jet type of VUV source with the wavelengths of 5–200 nm was used. The experimental results show that under the VUV irradiation, both the tensile fracture strength and elongation decrease slightly. The spectral absorbance of the PET film increases noticeably with increasing VUV dose. The absorption band mainly forms in the near-ultraviolet region. The XPS, FTIR, and ESR analyses indicate that in the skin layer of PET film irradiated with VUV, the C–O bonds could be broken and decarbonylation occurs, leading to the formation of free radicals of benzene rings as well as a trend of carbonification. The scission of the macromolecule chains, the increase in radical concentration, and the carbonification would cause the degradation in optical properties for the PET film under VUV exposure. Key words: PTFE, Radiation Stability, VUV 1. Introduction Polymeric materials have been widely employed in spacecraft for parabolic antennas, solar wings, and thermal control systems [1–3]. Their performance in space will directly influence the reliability and lifetime of spacecraft. Vacuum ultraviolet (VUV) radiation is one of the major environment factors causing the degradation of polymers. Although the VUV energy percentage in the total spectrum of the Sun is very limited, its photon energy is high enough to break most chemical bonds in the polymers [4]. Therefore, it is important to study the effect of VUV radiation on the chemical structure and properties of polymeric materials, such as the polyethylene terephthalate (PET) film. Most published work about the effects of VUV on polymeric films was performed using various VUV lamps, with which it could be difficult to obtain a VUV spectrum similar to the Sun. A gas-jet type VUV source was used in this study, which could give a spectrum similar to the Sun in the wavelength range of 5–200 nm [4]. The aim of this work was to examine the changes in chemical structure and properties of the PET film under VUV exposure. 225 J.I. Kleiman (ed.), Protection of Materials and Structures from Space Environment, 225–232.  C 2006 Springer. Printed in the Netherlands.

226

GUIRONG PENG ET AL. O

O

C

C

O

CH2 CH2

O

Figure 1. Chemical structure of the PET film

2. Experimental The PET film was 60 μm thick. Figure 1 shows the chemical structure of the PET film. The film was annealed at 70◦ C for 3.5 h, rinsed with analytically pure acetone and ethanol, and dried in a desiccator for more than 48 h at ambient temperature before the VUV exposure. The VUV source used in this study operates with supersonic argon gas, which is excited by a high-energy electron beam to give the VUV wavelengths ranged from 5 to 200 nm. The VUV intensity of 0.24 w·m−2 was acquired at a distance of 70 cm from the source, corresponding to 10 times the VUV solar constant (VUV suns). In this study, the gas-jet pressure was 5 atm, the electron-beam energy was 1000 eV and the electric current was 15 mA. The vacuum before and after the injection of argon gas into the chamber was 10−5 and 10−3 Pa, respectively. The samples were located at 72, 34, and 22 cm away from the VUV source, corresponding to the 10, 40, and 100 VUV suns respectively. The tensile tests were carried out at room temperature after the VUV exposure. The engaged surface area of the film samples was 20 × 5 mm2 , and the crosshead speed was 2.4 mm·min−1 . The transmittance of the VUV irradiated film samples in the wavelength range of 200–3200 nm was measured with a UV-3101PC scanning spectrophotometer. The X-ray photoelectron spectroscopy (XPS) was performed using a VG-ESCALAB Mark-2 type spectrometer with Mg Kα source. The vacuum in the chamber was 10−6 Pa, and the pass energy 20 eV. IR spectra were acquired using a Perkin Elmer System 2000 Fourier Transform Infrared (FTIR) spectrophotometer. The resolution of the spectrophotometer was 1 cm−1 . The test wave numbers were in the range of 400–4000 cm−1 . The electron-spin resonance analysis (ESR) was performed at room temperature in a JES-FE3AX type spectrometer.

3. Results and Discussion 3.1. TENSILE AND OPTICAL PROPERTIES

Figure 2 shows the changes in tensile strength and elongation for the PET film irradiated with various doses of VUV. It can be seen that both the strength and elongation decrease slightly with increasing the VUV dose.

227

EFFECT OF VUV RADIATION 240

120

(a)

(b)

90 δ, %

σf , MPa

200

160

60

100 VUV suns

30

100 VUV suns

120 0

2000 4000 Irradiation dose, esh

0

6000

0

2000 4000 6000 Irradiation dose, esh

Figure 2. (a) The tensile strength σ f and (b) elongation δ vs. VUV dose for PET film

Figure 3(a) shows the change in spectral absorbance Aλ for the PET film after VUV irradiation. The absorbance limit is at the wavelength of 322 nm and changes little with VUV dose. The Aλ increases with the irradiation dose. Figure 3(b) illustrates the effect of VUV intensity on Aλ . The Aλ in the near-ultraviolet region changes remarkably with the VUV intensity; but less in the visible to near-infrared regions. Moreover, under the same VUV dose, the Aλ in the near-ultraviolet region increases less for the PET film irradiated with 100 VUV suns than that under 40 VUV suns (see the curves 3 and 4), but more than that under 10 VUV suns (see the curves 1 and 2). Figure 4 shows the change in Aλ at 322 and 600 nm with VUV dose. With increasing the dose, the A322 increases rapidly and then tends to level off, while A600 changes almost linearly. 3.2. XPS ANALYSIS

Figure 5 shows the C1s spectra of PET film irradiated with 100 VUV suns and various doses. Before the irradiation, the C1s spectrum would be composed of 0.5 0.6

ΔAλ

3

0.3 0.2

2 1

3, 2940 esh 2, 1600 esh 1, 680 esh 100 VUV suns

0.1

0.4

4

0.3 ΔAλ

0.5 0.4

(b)

(a)

3 0.2

2

4, 1600 esh @ 40 VUV suns 3, 1600 esh @ 100 VUV suns 2, 680 esh @ 100 VUV suns 1, 680 esh @ 10 VUV suns

0.1 0.0

1

0.0 400

600 800 1000 1200 1400 Wavelength, nm

400

600 800 1000 Wavelength, nm

1200

Figure 3. The change in absorbance Aλ vs. (a) VUV dose and (b) intensity for PET film

228

GUIRONG PENG ET AL. 0.6 100 VUV suns

0.5

ΔA 322

ΔAλ

0.4 0.3 0.2

Δ A 600

0.1 0.0 0

600 1200 1800 2400 3000 Irradiation dose, esh

Figure 4. The change in absorbance at 322 nm (A322 ) and 600 nm (A600 ) vs. VUV dose for PET film irradiated with 100 VUV suns

three peaks with binding energy at 284.6, 286.3, and 288.4 eV in turn, as shown in figure 5(a). The three peaks might be related to the carbon in the benzene rings, the carbon singly bound with oxygen in the ethylene, and that in the carbonyl groups, respectively [5–6]. After the VUV irradiation, the C1s spectra could be still characterized with the three peaks, as shown in figures 5(b) and 5(c). Table 1 shows the change in the corresponding area ratio for the three characteristic peaks

(c)

2940 esh

CPS

(b) 680 esh

(a) unirradiated

291

288 285 Binding energy, eV

282

Figure 5. C1s spectra of PET film irradiated with 100 VUV suns for (a) 0, (b) 680, and (c) 2940 esh

229

EFFECT OF VUV RADIATION TABLE 1. The change in area ratios of the fitting peaks for C1s spectra of PET film after 100 VUV suns irradiation Dose (esh) 0 680 1600 2940

Binding energy (eV) and Area ratio (%) 284.6 (69.6) 284.6 (84.7) 284.6 (86.3) 284.6 (89.4)

286.3 (17.1) 286.7 (11.3) 286.5 (10.3) 286.6 (8.4)

288.4 (13.3) 288.8 (4.0) 288.6 (3.3) 288.7 (2.2)

with VUV dose. With increasing the irradiation dose, the area ratio for the peak at 284.6 eV increases, but those at 286.3 and 288.4 eV decrease obviously. Figure 6 shows the change in O1s spectrum after VUV irradiation. Before the exposure, the O1s spectrum is fitted into two Gauss peaks with the binding energy at 531.8 and 533.3 eV, respectively. The former could be related to the oxygen in the carbonyl groups, and the latter to the oxygen singly bound with carbon [5–6]. After the irradiation, the intensities of the two peaks decrease. From the above XPS analysis, it is believed that the single bonds of carbon with oxygen (C–O) in the PET molecules can be broken under the VUV radiation, and further resulting in decarbonylation. 3.3. FTIR ANALYSIS

The FTIR spectra for the PET film before and after VUV irradiation are shown in figure 7. The absorption peaks at 3627 and 3545 cm−1 are related to the vibration of hydroxy1 groups [7–8]. After the VUV irradiation, these two peaks are

(c)

2940 esh

(b)

CPS

680 esh

(a) 0 esh

540

535

530

525

Binding energy, eV

Figure 6. O1s spectra of PET film irradiated with 100 VUV suns for (a) 0, (b) 680, and (c) 2940 esh

230

GUIRONG PENG ET AL. 50

80 40

(b) unirradiated 100 VUV suns, 2940 esh

60

20

T,%

T,%

30

40 (a) unirrradiated 100 VUV suns, 2940 esh

0 3600

3200

2800 −1

Wave number, cm

2400

20 10 0 1800

1700

1600

1500 −1

Wave number, cm

Figure 7. FTIR spectra for PET film before and after irradiation with 100 VUV suns for 2940 esh in the wave number ranges of (a) 3700–2300 and (b) 1800–1480 cm−1

enhanced, indicating that the content of the end hydroxy1 groups might be increased to some extent. In addition, several new absorption peaks appeared in the region from 1680 to 1530 cm−1 . These peaks originate from the change in the substitute groups in the benzene rings and the formation of ethylene groups such as –CH=CH– [7]. 3.4. ESR ANALYSIS

Figure 8 shows the ESR spectra of PET film after VUV irradiation. The g value is 2.004 and the line width 9G. On the basis of the ESR analysis, the free radical concentration in the PET film samples could be estimated using the formula in reference [9], as shown in figure 9. It is shown that the free radical concentration increases rapidly and then tends to level off with the increase of the VUV dose. This trend is similar to that of A322 vs. dose in figure 4, demonstrating that the change of optical properties has a relation with the free radicals.

Figure 8. ESR spectra for PET film irradiated with 100 VUV suns for (a) 680 and (b) 1600 esh

EFFECT OF VUV RADIATION

Spin number, a.u

25

231

100 VUV suns

20 15 10 5 0 0

600 1200 1800 2400 3000 Irradiation dose, esh

Figure 9. The relative radical amount vs. VUV dose for the PET film irradiated with 100 VUV suns

4. Conclusion VUV radiation is a major factor of space environment. Under VUV radiation, the tensile properties of PET film did not change significantly, while the optical ones varied noticeably. With increasing the irradiation dose, the tensile fracture strength and elongation decreased slightly, and the spectral absorbance increased remarkably in the ultraviolet to visible regions. The XPS, FTIR, and ESR analyses indicated that in the skin layer of PET film irradiated with VUV, the C–O bonds could be broken and decarbonylation occurred, forming free radicals of benzene rings. As a result, the condensation extent of the benzene rings was increased, and then a trend of carbonification or enrichment in carbon appeared. The increase in the free radical concentration and the carbonification would cause the degradation in optical properties for the PET film under VUV radiation.

References 1. Nakayama, Y., Imagawa, K., Tagshira, M., Nakai, M., Kudoh, H., Sugimoto, M., Kasai, N., and Seguchi, T. (2001) Journal of High Performance Polymers 13(3), 433–451. 2. Dever, J. A., Pietromica, A. J., Stueber, T. J., Sechkar, E. A., Messer, R. K. (2002) “Simulated Space Vacuum Ultraviolet (VUV) Exposure Testing for Polymer Films”, NASA Report TM2002-211337. 3. Iwata, M., Tohyama, F., Ohnishi, A., and Hirosawa, H. (2001) Journal of Spacecraft and Rockets 38(4), 504–509. 4. Verkhovtseva, E. T., Yaremenko, V. I., and Telepnev, V. D. (1997) In The Proceedings of 7th International Symposium on Materials in Space Environment, Toulouse, France, 16–20 June 1997, p. 119. 5. Beamson, G. and Briggs, D. (1992) High Resolution XPS of Organic Polymers: The Scienta ESCA300 Database, John Wiley and Sons, Chichester, UK. 6. Ektessabi, U. and Yamaguchi, K. (2000) Thin Slid Films 377–378, 793–797.

232

GUIRONG PENG ET AL.

7. Grossetete, T., Rivaton, A., Gardette, J. L., Hoyle, C. E., Ziemer, M., Fagerburg, D. R., and Clauberg, H. (2000) Polymer 41(10), 3541–3554. 8. Miyake, A. (1959) Journal of Polymer Science 38, 497–512. 9. Yin, J. and Mo, Z. (2001) Modern Polymer Physics, Science Publisher, Beijing, 743–829, 679– 680.

STATUS OF SOLAR SAIL MATERIAL CHARACTERIZATION AT NASA’S MARSHALL SPACE FLIGHT CENTER DAVID L. EDWARDS,1 CHARLES SEMMEL,2 MARY HOVATER,1 MARY NEHLS,1 PERRY GRAY,3 WHITNEY HUBBS,1 AND GEORGE WERTZ1 1 NASA/MSFC, ED31, MSFC, AL 35812 2 Qualis Corporation, ED31, MSFC, AL 35812 3 ICRC, ED31, MSFC, AL 35812

Abstract. Near term solar sail propelled science missions are targeting the Lagrange point 1 (L1) as well as locations sunward of L1 as destinations. These near term missions include the Solar Polar Imager [1] and the L1 Diamond [2]. The Environmental Effects Group at NASA’s Marshall Space Flight Center (MSFC) continues to actively characterize solar sail materials in preparation for these near term solar sail missions. Previous investigations indicated that space environmental effects on sail material thermo-optical properties were minimal and would not significantly affect the propulsion efficiency of the sail [3–5]. These investigations also indicated that the sail material mechanical stability degrades with increasing radiation exposure. This paper will further quantify the effects of space environmental exposure on the mechanical and thermo-optical properties of candidate sail materials. Candidate sail materials for these missions include Aluminum coated MylarTM , TeonexTM , and CP1 (Colorless Polyimide). Experimental data will be presented on the response of sail material to charged particle radiation that indicates change in the mechanical and thermo-optical properties. Thermo-optical property data will also be presented indicating the effects of long-term Near Ultraviolet (NUV) exposure. Charged particle radiation with subsequent Hypervelocity Impact (HVI) results will be presented, indicating that sail material damage is primarily limited to the diameter of the incident projectile. Key words: solar sail, space environment, mylar, polyimide

1. Introduction The non-Keplerian mission opportunities offered by solar sail propulsion are driving many science missions to require the capabilities of solar sail propulsion. This 233 J.I. Kleiman (ed.), Protection of Materials and Structures from Space Environment, 233–246.  C 2006 Springer. Printed in the Netherlands.

234

DAVID L. EDWARDS ET AL.

renewed interest in solar sails has resulted in escalating development of solar sail materials, structures, and deployment mechanisms. Many challenges confront sail material technologists and among these is the development of low areal density sails. Present state-of-the-art sails have an areal density ranging between 5 and 15 g·m−2 . Sail materials investigated in this effort are on the order of 3 g·m−2 . The farterm goal of sail material manufacturers is a sail material on the orderof 0.01 g·m−2 [6]. As reported by McInnes [7], a solar sail is a thin membrane that uses the momentum carried by photons to propel spacecraft. These photons originate from the sun or can be beamed onto the sail with a laser. If the sail is a good reflector, the momentum transferred to the sail can be almost doubled. Since the momentum carried by a single photon is extremely small, the surface area of a sail must be large to produce a reasonable acceleration, as indicated by a = (2Ap t )/m.

(1)

where, a—nonrelativistic sail acceleration (m·s−2 ), A—surface area (m2 ), pt —incident radiation pressure (N·m−2 ), m—mass (kg). The radiation pressure ( pt ) varies as the inverse square of the distance from the Sun as shown by: 2 pt = 4.56 × 10−6 [(1 + R)/rAU ]

(2)

where, R—surface reflectivity (0 < R < 1), rAU —distance to the Sun in astronomical units (AU). A perfectly reflective sail (R = 1) at a distance of 1 AU from the Sun experiences a light pressure of 9.1 μN·m−2 . Substituting eq. (2) into (1) and solving for the acceleration gives: a = 9.12 × 10−6 [(1 + R)A/mr 2AU ]

(3)

For example, a 1000 kg spacecraft with a perfectly reflective (R = 1) sail area of 106 m2 at a distance of 1 AU from the Sun experiences an acceleration of approximately 1.8 cm·s−2 . As shown in eq. (3), the reflectivity, R, of a sail material is a primary factor in sail performance. A factor not specifically referenced in eq. (3) is the mechanical stability of the film. This property is implied in the variable defining the area, A. Area loss can occur through numerous mechanisms including rips, delaminating

STATUS OF SOLAR SAIL MATERIAL CHARACTERIZATION 6

Acceleration (cm/s^2)

1 10 2

5

8 10

Sail Area (m^2) 5 5 6 10 4 10

235

5

2 10

Sail Reflectivity

1.5

1

0.5

0

Sail Area

1

0.9

0.8

0.7

0.6

0.5

Ref lectivity

Figure 1. Sail characteristic acceleration dependence on sail material reflectivity and sail area Calculations were made assuming an initial sail area of 1 × 106 m2

of reflective coating, and meteoroid impact. Figure 1 shows the dependence of sail acceleration on the sail material reflectivity and sail area. The Environmental Effects Group at NASA’s Marshall Space Flight Center (MSFC) is tasked with characterizing the material properties of newly developed sail materials and further characterizing these materials in emulated space environments. The ultimate goal of this work is to determine the effect of space environment exposure on sail performance. This paper serves as a status of a work in progress and will report data obtained to-date, and describes the future work planned through September, 2004.

2. Description of the Facilities The process of characterizing materials in a space environment requires the utilization of specialized test facilities. The Environmental Effects Group operates a number of highly specialized facilities dedicated to understanding the effects of the space environment on materials. The facilities described below include: the Pelletron test system for combined exposure effects, Long-Term Ultraviolet radiation exposure facilities, and the MSFC Impact Facility (MIF) for characterizing the effects of micrometeoroid and orbital debris impacts. Material characterization equipment and analysis techniques will be introduced in chapter 3.

236

DAVID L. EDWARDS ET AL.

2.1. PELLETRON TEST SYSTEM

The Pelletron test system provides the unique capability to irradiate material to a simultaneous or sequential exposure to an emulated space environment, and perform in-vacuum reflectance measurements. This facility possesses the capability to irradiate materials to: high energy electrons (100–2.5 MeV), low energy electrons (1–100 keV), protons (40–700 keV) Vacuum Ultraviolet (VUV) (photon radiation over the wavelength range from 121 to 200 nm) and Near Ultraviolet (NUV) (photon radiation over the wavelength range from 200 to 400 nm). The exposure coverage is nominally a 7.6 cm × 7.6 cm (3 inch × 3 inch) area. In-vacuum reflectance measurements can be obtained on two 1-inch diameter coupons or on one 1-inch × 2-inch rectangular sample. The reflectance measurements are taken over the wavelength range from 250 to 2500 nm. The solar absorptance (α s ) is calculated from this spectral reflectance data. The Pelletron system is shown in figure 2. 2.2. LONG-TERM ULTRAVIOLET RADIATION

Ultraviolet radiation is a critical component of the space environment to which sails will be continuously exposed. To characterize the sail material response to long-term Ultraviolet (UV) radiation, sail materials are irradiated, under vacuum, to a solar intensity of 2 UV suns. Solar intensity is measured using a spectral

Protons

Test Chamber VUV

High Energy Electrons Low Energy Electrons

Reflectance Measurement

Figure 2. Photograph of the Pelletron combined space environmental effects exposure system

STATUS OF SOLAR SAIL MATERIAL CHARACTERIZATION

237

Figure 3. Long-term UV exposure test facility

radiometer with integrating sphere. Calibration on the Near UV (NUV) source is accomplished by measuring the source with a spectroradiometer over the 200–400 nm region. This bandwidth is then integrated to obtain the total energy in W·cm−2 . This value is divided by the sun’s output in that spectral region, as given in ASTM E-490 [8] to determine the number of UV suns. A thermopile that has a flat spectral response is placed at the same plane as the spectroradiometer’s integrating sphere and a millivolt reading is recorded that is compared to the total energy measured by the spectroradiometer. This provides a UV suns per millivolt reading that can be used to determine the source intensity inside the chamber at the sample plane using the thermopile voltage reading. This particular test facility has the capability to irradiate up to 20 1-inch diameter coupons. The long-term UV test system is shown in figure 3. 2.3. HYPERVELOCITY IMPACT

Hypervelocity Impact (HVI) testing was performed on sail material to determine the effect of micrometeoroid impact on sail material. The concern is that the sail material will become brittle by exposure to radiation over the mission lifetime and increase the probability of rip propagation. The HVI testing was performed using the MSFC Impact Facility’s (MSFC) Micro Light Gas Gun (MLGG). The MSFC MLGG has the capability to impact targets with a single projectile. Projectile sizes range from a diameter of 0.4–1.0 mm. Projectile velocity is selectable and ranges from 2 to 7 km·s−1 . Each projectile is encased in a sabot to keep the projectile centered in the barrel to assure shot accuracy. Prior to each HVI test series, calibration shots were performed using Nylon slugs. These calibration shots ensure system operation and validate experimental parameters for desired projectile velocities. Projectile velocity is measured with each test using

238

DAVID L. EDWARDS ET AL.

Figure 4. MSFC MLGG test facility

photodiodes located at each end of the flight tube. Sail materials, investigated for this work, were approximately 15.2 cm × 15.2 cm (6 inch × 6 inch) and mounted in a frame. Test coupons can be as large as 20.3 cm × 25.4 cm (8 inch × 10 inch). The MSFC MLGG is shown in figure 4.

3. Experimental Procedure Sail materials investigated in this on-going activity include: aluminized Mylar, aluminized Teonex, and aluminized CP1. Certain characteristics of the material designs are held as proprietary, so further descriptions will not be discussed within this paper. Test matrices were developed to guide the material characterization for each specific material exposure condition. These matrices will be discussed in the following sections. 3.1. CHARGED PARTICLE EXPOSURE

The charged particle exposure activity focuses on the determination of the mechanical and thermo-optical property response of sail material to various types of exposure conditions (e.g., electron, proton, combined electron and proton). The objective of this work is to determine the radiation tolerance of the candidate sail materials for each type of charged particle radiation. Each candidate

STATUS OF SOLAR SAIL MATERIAL CHARACTERIZATION 10

239

10

Rads

9E15 e-/sq.cm @ 50 keV 2E14 p+/sq.cm @ 700 keV

10

9

10

8

0

0.5

1

1.5

2 2.5 Microns

3

3.5

4

Figure 5. Flat dose profile of 50 keV electrons and 700 keV protons in aluminized Mylar

material type will be exposed to specific doses of charged particle radiation, ranging from 100 Mrad to over 5 Grad. The dose for each type of radiation is determined by modeling the radiation transport and energy loss during propagation through the sail material. Electron propagation and energy loss is determined using the Integrated Tiger Series 3 (ITS 3) [9]. Proton propagation and energy loss is determined using the Transport of Ions in Matter (TRIM) code [10]. To accurately characterize the sail material response to a dose from charged particles, the dose must be uniform throughout the entire thickness of the sail material. This dose is term to be a “uniform” or “flat” dose profile. Figure 5 shows a typical “flat” dose depth profile for 50 keV electrons and 700 keV protons in aluminized Mylar. 3.2. LONG-TERM UV EXPOSURE

Candidate sail materials were characterized, prior to UV exposure, to determine the baseline thermo-optical properties (e.g., solar absorptance, emittance, and reflectivity). These samples are exposed to 2 suns of UV radiation and periodically removed from the vacuum test chamber to characterize thermo-optical properties. The solar absorptance (α s ) and the thermal emittance (ε) are determined by measurement instrumentation. The Laboratory Portable Spectroreflectometer (LPSR) measures the spectral reflectance of material surfaces over the wavelength range from 200 to 2800 nm. The LPSR calculates α s from this reflectance spectral data.

240

DAVID L. EDWARDS ET AL.

The Reflectivity is determined by using the simple relationship: T +R+A=1

(4)

where T is the transmittance, R is the reflectance, and A is the absorptance of the sail material. For our case, we assume that no light is transmitted through the sail material, thus reducing eq. (4) to: R =1− A

(5)

Thermal emittance is measured using the TEMP 2000 and the Laboratory Portable Infrared Reflectometer (LPIR) instruments. The TEMP 2000 measures thermal emittance over an integrated wavelength range from 2.5 to 20 μm, while the LPIR provide spectral data over the wavelength range from 2 to 20 μm. 3.3. HYPERVELOCITY IMPACT (HVI)

Sail materials characterized by HVI, to-date, are aluminized Mylar (Al/Mylar) and aluminized Teonex (Al/Teonex). These materials were exposed to a uniform dose of electron radiation. The dose levels were 100 Mrad and 1 Grad. Hypervelocity impact (HVI) tests were performed on control sail material. These tests established the baseline response of the material. Irradiated samples were exposed to 100 keV electrons. Four HVI samples were placed in the test chamber such that the electron flux would pass sequentially through each sample. When the desired dose levels in the materials were achieved, the HVI samples were removed from the vacuum chamber and delivered for HVI testing. Each sail sample was positioned in the HVI test chamber and individually impact characterized with a single projectile. Photographs of each HVI sample were taken subsequent to impact testing. The photographs indicated that several samples exhibited micro-cracks extending radially from the penetration site. To account for the damage presented by these micro-cracks a damage area was defined. Figure 6 shows a typical Penetration diameter for HVI on sail material. Measurements were obtained of the impact site including penetration diameter and total damage area surrounding the penetration site. The HVI analysis was initiated after the sail material sample was removed from the chamber. After the HVI test, a photograph was taken of the impact site. A microscope equipped with two micrometers was used for impact site measurements. The penetration diameter was measured as well as the surrounding damage diameter. These measurements were then compared to the projectile size to determine if the damage from HVI changes with radiation exposure. The ratios of Penetration area to projectile size and damage area to projectile size are referred

STATUS OF SOLAR SAIL MATERIAL CHARACTERIZATION

241

Penetration Area

Damage Area Figure 6. Penetration site of HVI of a 1.0 mm diameter projectile with velocity of 7 km·s−1 . The photograph shows the penetration area and the surrounding damage area induced by the impact. Sail material was Al/Mylar

to as “Beta factors.” These Beta factors are defined in eqs. (6) and (7). Beta Penetration =

(Area of Penetration) (Cross-Sectional Area of Projectile)

Beta Damage Area =

(Damage Area) (Cross-Sectional Area of Projectile)

(6) (7)

4. Results Preliminary results on the effects of charged particle exposure have yielded some interesting, but not unexpected results. The mechanical properties, including, ultimate stress and ultimate strain, decrease with increasing radiation dose. Stressstrain data was obtained for each exposure condition. Three to five tensile coupons of each material were exposed to each exposure condition. The results, to date, are shown in figures 7 and 8. In addition to mechanical properties, thermo-optical properties are measured, when practical. Figure 9 shows the effect of electron radiation on the reflectance property of aluminized Mylar as a function of increasing dose. The solar absorptance values, shown in table 1, indicate little change with radiation exposure. The spectral data, shown in figure 9, shows a slight decrease in reflection with increased radiation exposure. This reflection loss occurs in the low wavelength region of the reflectance spectrum. Table 1 details the thermo-optical properties obtained to date for materials exposed to long term Near Ultraviolet Radiation and other materials exposed to high doses of electron radiation.

242

DAVID L. EDWARDS ET AL. 5

Ultimate Tensile Strength (psi)

10

4

10

Al/Teonex (e- exposure) Al/Mylar (e- exposure) Al/CP1 (e- exposure) Al/CP1 (p+ exposure) Al/CP1 (e- & p+ & UV exposure)

3

10 0.001

0.01

0.1

1

10

2

10

10

3

4

10

10

5

Dose (Mrad)

Figure 7. Relationship between ultimate strength of candidate sail materials and radiation dose

The data, shown in table 1, indicates the solar absorptance of the Aluminum surface of the sail material has not significantly changed with either UV exposure or electron exposure. The back surface (e.g., polymer surface) emittance has also remained relatively stable, showing, at most, a 4% change after 1 Grad of electron exposure. 1

Strain to Failure

0.1

Al/Teonex (e- exposure) Al/Mylar (e- exposure) Al/CP1 (e- exposure) Al/CP1 (p+ exposure) Al/CP1 (e- & p+ & UV exposure)

0.01

0.001 0.001

0.01

0.1

1

10

10

2

10

3

10

4

5

10

Dose (Mrad)

Figure 8. Relationship between strain to failure of candidate sail materials and radiation dose

STATUS OF SOLAR SAIL MATERIAL CHARACTERIZATION

243

TABLE 1. Sail reflectance as a function of exposure A1/Mylar UV ESH 0 250 500 750 1000

AL/Teonex

A1/CP1

Alpha Emmitance Reflectance Alpha Emmitance Reflectance Alpha Emmitance Reflectance 0.08 0.08 0.08

0.23 0.24 0.25

0.92 0.92 0.92

0.08 0.08 0.08

0.25 0.25 0.25

0.92 0.92 0.92

Electron Dose 100 Mrad 0.09 1Grad 0.09 5Grad

0.24 0.24

0.91 0.91

0.09 0.09

0.25 0.29

0.91 0.91

0.09 0.09 0.09

0.27 0.27 0.27

0.91 0.91 0.91

The Hypervelocity Impact (HVI) testing was designed to determine the effects of HVI over the functional lifetime of the sail material. Impact testing was performed at dose levels of 100 Mrad and 1 Grad. Sail material specimens were first exposed to electron radiation, then subjected to HVI. Beta factors for penetration and damage areas for the candidate materials Al/ Teonex and Al/Mylar after radiation exposure and HVI characterization are shown in figures 10 and 11. These materials were exposed to radiation doses of 100 Mrad and 1 Grad and then subjected to HVI with projectile of 3 km·s−1 and diameters of 0.4 and 1.0 mm. The complete data set for tests completed to date is shown in table 2.

0.98 Pre-Exposure

0.96

100 Mrad 1 Grad

Ref lectance

0.94 0.92 0.9 0.88 0.86 0.84 0.82 200

300

400

500

600

700

800

900

1000

Wavelength (nm)

Figure 9. Reflectance of aluminized Mylar solar sail material as a function of electron radiation dosage

244

DAVID L. EDWARDS ET AL. 1.6 1.5

Beta Penetration

1.4 Projectile Velocity 3 km/s

1.3

Al/Teonex 0.4 mm projectile Al/Teonex 1.0 mm projectile Al/Mylar 0.4 mm projectile Al/Mylar 1.0 mm projectile

1.2 1.1 1 0.9 0.001

0.01

0.1

1

10

2

3

10

10

Dose (Mrad)

Figure 10. Beta factor (penetration area) relationship with radiation dose in candidate sail material

This data does not constitute a statistically significant data set. This data does consistently indicate the damage area is no larger than a factor of 2 of the projectile diameter. The tests also indicated that the impacts did not result in rip propagation. The data indicates little change in hypervelocity impact response with increased radiation exposure of these materials.

1.8 1.7

Beta Damage

1.6 1.5 1.4

Projectile Velocity 3 km/s

1.3

Al/Teonex 0.4 mm projectile Al/Teonex 1.0 mm projectile Al/Mylar 0.4 mm projectile Al/Mylar 1.0 mm projectile

1.2 1.1 0.001

0.01

0.1

1

10

2

10

3

10

Dose (Mrada)

Figure 11. Beta factor (damage area) relationship with radiation dose in candidate sail material

STATUS OF SOLAR SAIL MATERIAL CHARACTERIZATION

245

TABLE 2. Beta factors of candidate solar sail materials Exposure (Mard)

Projectile size (mm)

Projectile velocity (km/s)

Beta factor penetration

Beta factor damage

Al/Teonex Al/Teonex Al/Teonex Al/Teonex Al/Teonex Al/Teonex Al/Teonex Al/Teonex Al/Teonex Al/Teonex Al/Teonex Al/Teonex

0 0 0 0 100 100 100 100 1000 1000 1000 1000

0.4 1 0.4 1 0.4 1 0.4 1 0.4 1 0.4 1

7 7 3 3 7 7 3 3 7 7 3 3

1.35 1.11 1.49 1.11 1.08 1.07 1.22 1.03 1.73 1.05 1.36 1.06

1.53 1.11 1.49 1.14 1.59 1.07 1.68 1.2 1.97 1.11 1.74 1.13

Al/Mylar Al/Mylar Al/Mylar Al/Mylar Al/Mylar Al/Mylar Al/Mylar Al/Mylar Al/Mylar Al/Mylar Al/Mylar Al/Mylar

0 0 0 0 100 100 100 100 1000 1000 1000 1000

0.4 1 0.4 1 0.4 1 0.4 1 0.4 1 0.4 1

7 7 3 3 7 7 3 3 7 7 3 3

1.35 1.27 1.58 1

1.9

1.33 1.02

1.78 1.18

1.02

1.15

Material

1.58 1.18

5. Future Plans Upcoming testing on the candidate solar sail material will focus on missionspecific applications. Test plans are being developed to expose the material to emulated Solar Polar Imager (SPI) and L1 Diamond radiation environments. The mechanical and thermo-optical properties will be evaluated over mission lifetime dose exposures. Hypervelocity impact tests will be conducted on sail materials exposed to mission middle and end-of-life radiation doses. The goal of this research is to select the optimum material for these specialized missions. References 1. Solar Polar Imager, http://umbra.nascom.nasa.gov/SEC/secr/missions/polarimg.html, January 2004. 2. Personal communication with Dr. Greg Garbe/NASA/MSFC.

246

DAVID L. EDWARDS ET AL.

3. Edwards, D., Hubbs, W., Gray, P. Wertz, G., Hoppe, D., Nehls, M., Semmel, C., Albarado, T., and Hollerman, W. (2003) In Proceedings of the 9th International Symposium on Material in a Space Environment, Noordwijk, The Netherlands, ESA Publications, Noordwijk, The Netherlands June 2003, pp. 16–20. 4. Albarado, T., Hollerman, W., Edwards, D., Hubbs, W., and Semmel, C. (2003) In Proceedings of ISEC 2003: 2003 International Solar Energy Conference, Hawaii, 15–18 March 2003. 5. Edwards, D., Hubbs, W., Stanaland, T., Hollerman, A., and Altstatt, R. (2002) In Proceedings of SPIE Photonics for Space Environments VIII, Vol. 4823, 2002. 6. Tech ICP-Solar Sails: Potential Mission Applications. www. inspacepropulsion.com/tech/ sails missionapps.html 7. McInnes, C. R. (1999) Solar Sailing: Technology, Dynamics, and Mission Applications, Praxis Publishing, Chichester, UK, pp. 1–50. 8. ASTM E-490-00a. (2000) Standard Solar Constant and Air Mass Zero solar Spectral Irradiance Tables, 200. 9. Halbleib, J. A., Kensek, R. P., Mehlhorn, T. A., Valdez, G. D., Seltzer, S. M., Berger, M. J. “ITS Version 3.0: The Integrated TIGER Series of Coupled Electron/Photon Monte Carlo Transport Codes”, SAND91-1634, (March 1992). 10. Ziegler, J. F. (2000) The Stopping and Range of Ions in Matter, SRIM-2000.40, www.SRIM.org.

ATOMIC OXYGEN DURABILITY EVALUATION OF A UV CURABLE CERAMER PROTECTIVE COATING BRUCE A. BANKS,1 CHRISTINA A. KARNIOTIS,2 DAVID DWORAK,3 AND MARK SOUCEK3 1 NASA Lewis Research Center, 21000 Brookpark Rd., M. S. 309-2 Cleveland, Ohio 44135 2 QSS Group, NASA Glenn Research Center, Mail Stop 309-2, 21000 Brookpark Road, Cleveland, OH 44135 3 University of Akron, Dept. of Polymer Engineering, 250 South Forge St., Akron OH 44325-0301

Abstract. The exposure of most silicones to atomic oxygen in low Earth orbit (LEO) results in the oxidative loss of methyl groups with a gradual conversion to oxides of silicon. Typically, there is surface shrinkage of the oxidized silicone protective coatings which leads to cracking of the partially oxidized brittle surface. Such cracks widen, branch and can propagate with continued atomic oxygen exposure ultimately allowing atomic oxygen to reach any hydrocarbon polymers under the silicone coating. A need exists for a paintable silicone coating that is free from such surface cracking and can be effectively used for protection of polymers and composites in LEO. A new type of silicone-based protective coating holding such potential was evaluated for atomic oxygen durability in an RF atomic oxygen plasma exposure facility. The coating consisted of a UV curable inorganic/organic hybrid coating, known as a ceramer, which was fabricated using a methyl substituted polysiloxane binder and nanophase silicon-oxo-clusters derived from sol-gel precursors. The polysiloxane was functionalized with a cycloaliphatic epoxide in order to be cured at ambient temperature via a cationic UV induced curing mechanism. Alkoxy silane groups were also grafted onto the polysiloxane chain, through hydrosilation, in order to form a network with the incorporated silicon-oxo-clusters. The prepared polymer was characterized by 1 H and 29 Si NMR, FT-IR, and electrospray ionization mass spectroscopy. The paper will present the results of atomic oxygen protection ability of thin ceramer coatings on Kapton H as evaluated over a range of atomic oxygen fluence levels. Key words: atomic oxygen, silicones

247 J.I. Kleiman (ed.), Protection of Materials and Structures from Space Environment, 247–263.  C 2006 Springer. Printed in the Netherlands.

248

BRUCE A. BANKS ET AL.

1. Introduction Silicones are one of the few polymers that can be applied by painting or spraying over composite or other organic spacecraft materials which have afforded reasonable protection from low Earth orbital (LEO) atomic oxygen attack. The gradual oxidation of the silicones in LEO results in an oxidized silicone surface which becomes a silicate or silica [1–3]. This surface conversion from silicone to silica also tends to produce tensile stresses in the surface of the oxidized silicone. An increase in the surface microhardness also results due to the atomic oxygen conversion reaction with the silicone [4]. A variety of approaches have been or are now being explored to identify silicones, silicone copolymers or silicone-hydrocarbon blends that could provide flexibility as well as atomic oxygen protection [5–8]. Results to date indicate that hybrid polymers composed of inorganic and organic polymers hold potential to survive LEO atomic oxygen attack. The silicones which are dominated by a oxygen-to-silicon ratio of 1.5 have shown greater resistance to atomic oxygen attack than the silicones with a ratio of 1.0. Polyhedral oligomeric silsesquioxane (POSS) contains covalently bonded reactive functionalities appropriate for polymerization or grafting. It can be blended or copolymerized with many aerospace polymers and is being considered for atomic oxygen durability [8]. The resistance to atomic oxygen attack of silicone blended or copolymerized polymers has been dependent not only on the oxygen-to-silicone ratio but the fractional fill of the silicone. The challenge to make functional use of such blends has been to find an adequately silicone-filled polymer that contains the appropriate protective silicone such that it has mission dependent properties that are acceptable. Examples of some of these are atomic oxygen durability, volatility, optical, thermal, mechanical and ease of application. Because of their ability to provide atomic oxygen protection, thermal stability, flexibility, and stability; polysiloxanes are an attractive candidate solution to achieving ideal protection from the elements of space. However, this is just part of the solution. The vacuum ultraviolet (VUV) radiation and high energy particles can still damage and degrade the composite material. Therefore, to incorporate protection from those components as well, ceramer coatings; which are inorganic/organic hybrid materials, can be utilized. Ceramers are part ceramic (inorganic) and part polymer (organic) and can offer protection from atomic oxygen as well as UV radiation and high energy particles via the in situ fabrication of nanophase silicon-oxo-clusters [9, 10]. The silicon-oxo-clusters are formed through a series of hydrolysis and condensation reactions between sol-gel precursors. The intention of a ceramer approach is to acquire a synergistic effect between the inorganic and organic phases on a nanoscale through the use of phase coupling agents, which for this system are alkoxy silanes pendant from the polysiloxane chain. There is confirmation of a synergy between the phases and this approach affords a uniformly distributed nanophase within a continuous organic phase [11].

249

ATOMIC OXYGEN DURABILITY O O

OH O

O O

O

OH

Atomic Oxygen O O O O O

O

HO

O Si

O

DUV

O Si

O

O Si

O

OH

Dissipated Heat

Si O

High Energy Particle

Metal Oxide

O

– SiO2 Layer Coating Composite Substrate

Figure 1. Depiction of the formation and function of protective silicon oxide layer and silicon-oxoclusters

Once the coating is exposed to atomic oxygen, a protective layer of silicon oxide is formed and, with the incorporation of silicon-oxo-clusters, the coating should protect the composite material against further atomic oxygen erosion, high energy particles, and VUV radiation. Figure 1 is an overall diagram of the coating’s function [12]. This paper summarizes an investigation of the high fluence atomic oxygen durability of ultraviolet radiation curable ceramer protective coating consisting of methyl substituted polysiloxane and compares the results with the commonly used silicone coating DC93-500. 2. Materials: Methyl Substituted Polysiloxane 2.1. SYNTHESIS

2.1.1. Experimental Materials. Octamethylcyclotetrasiloxane,1,3,5,7-tetramethylcyclotetrasil-oxane, 1,1,3,3-tetramethyldisiloxane, and vinyl triethoxysilane were purchased from Gelest, Inc. and used as supplied. Wilkinson’s catalyst (chlorotris (triphenylphosphine) rhodium(I), 99.99%), Amberlyst 15 ion-exchange resin, and 4-vinyl-1-cyclohexene 1,2-epoxide were purchased from Aldrich and used as supplied. Toluene, supplied by Aldrich Chemical Co., was distilled in order to eliminate any impurities. Irgacure 250 was supplied by Ciba Specialty Chemicals and used as received. Air sensitive materials were transferred and weighed in an inert atmosphere dry box under argon. Synthesis of poly(dimethylsiloxane-co-methylhydrosiloxane), hydride terminated. To a three neck, round bottom flask, equipped with a reflux condenser and

250

BRUCE A. BANKS ET AL. H3C O

H3C

O

CH3

O H3C

Si

+

H Si

CH3

O

Ion Exchange Resin

CH3

CH3

Si Si

CH3

H3C +

O H

CH3

HSi O SiH CH3

H3C

CH3

H

H

O

Si H3C

CH3

O

O

H

Si

Si H3C

H3C

CH3 Si

N2 Blanket

CH3 CH3

SiO (SiO)n (SiO)m Si

H

CH3 H CH3 CH3 m >> n

O H3C == CHCH3Si(OEt)3

CH3 CH3 O

Rh Catalyst N2 Blanket

CH3

CH3

SiO (SiO)n (SiO)m Si CH3 (CH2)3 CH3

O

CH3

Si(OEt)3

Figure 2. Synthesis of polydimethylsiloxane-co-methylhydrosiloxane, hydride terminated

nitrogen inlet/outlet, was added octamethylcyclo-tetrasiloxane (90.00 g), 1,3,5,7tetramethylcyclotetrasiloxane (5.33 g), 1,1,3,3-tetramethyldisiloxane (0.67 g), and Amberlyst 15 (20 wt%). The mixture was stirred at 90◦ C, under nitrogen, for 15 h. The viscous solution was then filtered to obtain poly(dimethylsiloxane-comethylhydrosiloxane), hydride terminated of various molecular weight ranges. Vacuum filtration was performed (