L. GERLACH, A. FOURNIER-SICRE, A FROMBERG S. KROEHNERT ESNESTEC BRITISH AEROSPACE TELEFUNKEN SYSTEM TECHNIK NOORDWMK BRISTOL WEDU THE NETHERLANDS GREAT BRITAIN GERMANY ABSTRACT The Space Telescope Solar Array (STSA), which is supplying t he power for the joint NASA / ESA Hubble Space Telescope (HST) is the largest flexible solar array built to date, carrying 48760 BSFR silicon cells. The double roll-out solar array was successfully deployed in orbit on 25th April 1990. STSA is designed to survive at least 5 years in low earth orbit (30,000 thermal cycles) and will supply at least 4400W at 34V after four years. STSA had to, be designed to survive the aggressive ATomic Oxygen (ATOX) environment. This was achieved by t he development of an ATOX resistant carrier substrate and of two different types of ATOX resistant interconnectors. A 10 Rcm BSFR solar cell with an extremely smooth surface was developed to ensure a good quality of the interconnector weld. The array is protected against shadowing and hot spots by solar cell shunt diodes. Thermal cycling was performed up to 65,000 cycl es which demonstrated that the lifetime of t he array should be in the order of ten years. The solar array deployment, which took place on 25th April 1990, consists of two phases. Firstly, the solar arrays (initially latched parallel to the body of the HST) are rotated by magnetic stepper motors so that they lie perpendicular to the HST (primary deployment). Secondly, the SAB's are unrolled from the storage drum by bi-stem booms at either side of each blanket which drive a spreader bar at the end of the SAB's. The storage drum and cushion roller (on which the cushion is stowed during deployment) are also driven during secondary deployment. STSA is the first solar array designed for in-orbit maintenance and replacement. Replacement of t he solar arrays will have to take place at least once during the fifteen year lifetime of the HST. The solar arrays are secured to the telescope by a manacle clamp which will be separated by astronauts at the end of the solar array lifetime; if necessary, manual overrides can be used for stowing the secondary and primary deployment mechanisms (SDM and PDM). The Second Space Telescope Solar Array (STSA 2) which will be UPPER LOWER 1 . INTRODUCTION The Space Telescope Solar Array (STSA) is supplying the power for the joint NASAIESA Hubble Space Telescope (HST). The STSA is a unique European development. Under the prime responsibility of the European Space Agency, British Aerospace and Telefunken Systemtechnik have developed and built the largest flexible solar generator to date. It consists of two double roll-out solar array wings (Fig.1) which are deployable and retractable with 48760 BSFR silicon solar cells interconnected. Each wing is equipped with two solar array blankets carrying the solar cells which are protected from each other by an embossed Kapton cushion whilst they are stowed on a common storage drum. + v2 WING LOWER BLANKET Figure 1: HST - Solar Array Config LOWER BLANKEl su" - v2 WING UPPER BLANKET UPPER BLANKET uration 1308 0160-8371/90/0000-1308 $1.00 1990 IEEE attached i n place of the existing array, is currently under assembly and test in Europe. STSA is designed to survive intact for at least five years i n a 590km low earth orbit. This means the solar arrays will have to withstand 30,000 thermal cycles between about +/- 100C in t he presence of our aggressive atomic oxygen environment (ATOX). STSA is required to deliver at least 4.4kW of electrical power at 34 volts after two years i n orbit. This paper will first describe the layout of the solar array blankets and substrate, then explain tlie way i n which the design has been developed to protect the SAB's against ATOX attack. The solar cells and interconnectors ar e then described, followed by the way in which the array is protected against shadowing and finally tlie lifetime verification of power output and mechanical integrity is presented. 2 . FLEXIBLE BLANKET DESIGN AND ELECTRICAL LAYOUT Hach of the sol& array blankets is made up of 5 identical power generating elements, t he so- called Solar Panel Assemblies (SPA), an outer and an inner SBA OBA SPA buffer (OBAJBA), a Drum Buffer Assembly (DBA) for t he mech- anical fixation to the stowage drum, and a Spreader Bar Assembly (SBA) for the mechan- ical interface to the spreader bar used to draw the blanket out of 1. tlie drum. The basic substrate for the flexible blanket consists of a glass fibre/Kapton com- pound. 48760 BSFR solar cells of 2cm x 4cm are bonded onto the 20 SPAs and provide the re- quired power to the HST. Each of the power-generating SPAs are equipped with 3 solar cell strings, each having 106 solar cells in series, two of them with 8 single solar-cell rows in parallel and one of them with 7 single cell rows in parallel. The single cell rows for each string are connected via MoAg tapping bars i n groups of 14, 15 or 16 cells. Al l of these groups are protected by shunt diodes which are required to limit hot spot temperatures due to undefined shadows and possible broken solar cells to temperatures below 140 deg.C To allow easy integration and replacement if necessary, all panel assemblies are interconnected via a piano hinge technique using a stainless steel hinge rod. The IBA acts also as electrical interface between t he blanket and the solar array deployment mechanism. On the IBA, the individual solar cell strings are electrically connected by means of 75 micron thick silver foil strips to main and redundant connections per SPA and are routed to a flexible printed circuit board which serves as interface to the harness attached to the deployment mechanism. For power transfer from the electrical strings to the IBA 6.3 cm wide silver mesh strips are used. They are contained within the flexible blanket substrate and are connected through 'windows' to the bus bars from the solar cell strings. Bridge pieces are used for the electrical connections from SPA to SPA across the hinge. Data lines and power harnesses from the SPA'S pass through the stowage drum via a flexible harness which uncoils and then coils so that it is wound in the opposite direction following SPA SPA SPA SPA IBA DBA Solar A m y Blanker Configuration (honl Side) 4 b \ warCeHsbing hangemen1 6.45 m ' &id@ Piece Wring Concept of Solar h y Blankel Data Lines Figure 2: Details of the solar array blanket 1309 deployment. Conventional wi r e harnesses conduct power along the PDM and solar array drive. Main and redundant paths are cross strapped to improve reliability. During eclipse, t he SAB's are protected from the HST by blocking power diodes. The STSA design is such that magnetic moments from SPA's on opposing SAB's cancel each other out. Power from the 20 SPA's are grouped into seven electrical sections. Three SPA's are allocated to each of the six HST batteries. The remaining two SPA's drive the HST bus directly. 3. ATOMIC OXYGEN PROTECTION Results from early space shuttle flights and more recently from the recovered LDEF spacecraft indicated that many materials used i n the initial STSA desigh were susceptible to ATOX i n low earth orbit. Of particular relevance, Kapton H film showed a change in surface properties and mass reduction and silver was shown to be ex- tremely ATOX susceptible. Other materials, such as silicone adhesives, were found not to be ATOX susceptible.The SPA substrate is shown in t he Figure 3 below. In essence, the 50 micron Ag- Glass flbm fllled with DC93500 Adhesive (DP46971) Kapton H Adhesive (DP46971) Silver Mesh Adheslve (DP46971) Ka tonH Adieslve (DP46971 Glass fibre filled wi d DC935W II thickness: 210 micron Figure 3: Cross-section of SPA Substrate mesh rear-side wiring is surrounded by a symmetric sandwich of silicone adhesive, Kapton H and glass fibre filled with DC93500 silicone adhesive which makes it safe against ATOX attacks. Special processes ensure that the 50 micron thin layers of DC93500 are completely cured which was proven by thermal and spectrometric investigations. The outermost part of the cushion which would be exposed to ATOX was protected by spraying with a silicon adhesive. The treatment with silicone introduced a further risk of sticking of the blanket rearsides to each other and t he blanket frontsides to the cushion. However, substrate and cushion samples with various interfaces were blocked for up to 1 year and at temperatures up to 112OC under blanket tension without measurable sticking. STSA deployed in space, after 14 months in the stowed configuration, with no sticking anomalies observed. Two types of solar cell interconnectors were developed and qualified in the course of the STSA program: (i) a 20 micron silver interconnector with an in-plane stress relief loop. (protected from ATOX by a 1.5 micron gold layer outside the welding area) (ii) a silver plated molybdenum interconnector with an out-of-plane stress relief loop (5 micron Ag/0.5 micron Pt/ 15 micron MoI0.5 micron PtJ5 micron Ag). The AgAu interconnector was chosen for t he shunt diodes. Molybdenum is not ATOX suscep- tible, however, the silver and platinum is required to provide a weldable interface for the solar cell interconnection. 4 . SOLAR CELLS ASSEMBLY DESIGN An extensive investigation has been performed for the selection of the basic element, the solar cell assembly (SCA). The SCA is a solar cell equipped with an interconnector 100% covered by a cover glass. Base material Base resistivity Junct i on Back surface field Back side reflector Contact system Antireflective layer Surface roughness Silicon thickness Di mensi ons BOL SCA efficiency SCA absorbtivity SCA emissivity - Crucible crown silicon - 10 Ohm cm - n-on-p shallow diffused - p+ doping (boron ion - Aluminium layer - Ti(Pd)Ag multifinger implantation) grid - TIOX - < = I 5 microns - 250 microns - 20.8 mm x 40.2 mm - 0.75 (unloaded) - 0.83 (hcmispherical) - 14% (IAMO, 25OC) The surface roughness of 1.5 microns (0.0015mm) is rcquired to ensure a good electrical contact to the solar cell interconnector during the mission. Tabl e 1: Solar Cell Characteristics Material - Soft Boro-silicate base glass (CMX) with cerium dioxide Antireflective layer - MgF2 (single layer) Adhesive for cell bonding Silicon, DC93500 + Pri mer Di mensi ons - 21.0 mm x 40.5 mm Thi cknes s - 150 microns Tabl e 2: Cover Slide Characteristics 1310 Apart from ATOX resistivity and the required electrical power (the solar cell has an efficiency of 14%. BOL, 1 AMO, 2 5 T ) the SCA compound is requested to survive at least 30,000 thermal cycles between k 100C. r , a ! welding relief joints LJLU I. .. welding joints Figure 4: Molybdenum Sol ar Cell I nt er - connector (not to scale) The thermal environment induces maximum stress to ' the solar cell interconnectors and welding joints on solar cell contacts. The effects on interconnectors and on welds are quite different in nature and must be considered separately. The stress on solar cell interconnectors is imposed by the thermal movement of t he individual components with different coefficients of thermal expansion leading to variations, in the solar cell/solar cell-gap with resultant 'kycling stresses t he interconnector stress relief loop area. A suitable interconnector design has been developed. 0 25mm silicon solar cell Figure 5: Cross-section through I nt er - connected Solar Cell Assembly Bonded onto t he Flexible S u b s t r a t e The stresses induced to welding joints on solar cell P- and n- contacts are volume stresses created by t he thermal interaction of different materials, i.e. silicon, silver and/or molybdenum. Stresses will be minimized by an adequate thermal match (expansion coefficients) and by a minimization of material mass, i.e. material thickness and area of welding spot. The extremely smooth solar cell surface with a roughness of Rz $1.5 micron and a specially optimized parallel-gap welding technology lead to a very good welding integrity. A Scanning Laser Acoustic Microscope (SLAM) has been proven to be the best tool for examinating or evaluating the quality of welding spots and their degradation due to thermal cycling. SLAM images before and after 30,000 thermal cycles have proven that there is a si gni f i cant margin i n the EOL MoAg interconnector/p-cell side welding area. Figure 6: SLAM image of a molybdenum i nt er connect or welded on a bus b a r (uncycled) 5 . PROTECTION AGAINST SHADOWING The Space Telescope Solar Arrays have to be designed to survive shadowing, which can occur due to external vehicle activities or by the high gain antenna during Telescope roll manoeuvres. Shadowing can lead to unacceptably high hot- spot temperatures due to electrical power consumption in the shadowed solar cells when a solar cell has to absorb instead of producing energy. This occurs during short circuiting of the electrical strings by the power-conditioning unit, as well as under normal operating conditions (36 Vlsolar cell string level) when there is a critical shadow on the array. So-called "shunt diodes" now limit power con- sumption in the shadowed cells by reducing the max. possible cell reverse operating voltage. It must be ensured that the maximum possible voltage within a shunt interval will not cause an electrical breakdown, destroying the solar cells affected, and that t he maximum power 1311 consumption in these cells will not lead to hot spot temperatures that are unacceptable for structural-integrity reasons. It was found that a maximum of 16 solar cells can be tolerated in series. Due to geometrical constraints, t he string arrangement selected and the required 106 solar cells in series, the single cell rows of each string are connected in parallel at intervals of 14, 15 or 16 cells in series. Each of these intervals i s protected by a shadow protection diode. A further design aspect was to limit t he diode temperatures i n case the diode is switched into operation whilst conducting the worst case string current due to a shadow or a component failure of some sort. The current density reduction in the diodes was achieved by using a flat solar cell shunt diode, this having the advantages that the structural integrity of the flexible roll-up is not affected (as it would be the case if a discrete diode type would have been used) and that the same type of ATOX resistant interconnection technique could be used as applied for the solar cell compartments. For protecting the flat solar cell shunt diodes against the ATOX environment and to support the emissivity in this are4 the same CMX coverglass type as used for t he solar cells was bonded onto the diode. base material base resistivity dopi ng/ j unc t i on c ont a c t s di me ns i ons cover gl as s emi ssi vi t y abs or pt i vi t y i n t e r c o n n e c t o r - crucible crown silicon - > 50 ohmcm - n+ pp+ type - Ti/Pd/Ag - 20 mm x 20 mm x 0.2 mm - 150 micron CMX - 0.83 (hemispherical) - slotted in welding areas - in-plane stress relief loop - 20 micron Ag coated on front and rear side with 1.5 micron gold (except welding area) - 0.15 Tabl e 3. Connect or I nt egr at ed Sol ar Cell Shunt Diode (C I C -d iod e) Char act er is t ic 6. LIFETIME PREDICTION 6 . 1 STRUCTURAL INTEGRITY Aside from the design selection and related developmental testing activities it was necessary to accrue definitive analytical and test evidence that the HST solar generator and all its components and mechanisms can fulfil t he planned mission. For the solar array blankets the qualification or life tests were carried out on several flight representative units and modules, considering all critical areas of the design. The materials and components for the qualification samples were selected from flight-production lots. The very severe qualification test sequence is summarized in Table 4. Based on in-orbit measurements the the solar array temperatures will be in the range between +/-7OoC only. Configuration Control Visual lnspection Electrical Performance Humidity Test Visual Inspection Electrical Performance Thermal Vacuum Continuity Check 11,500 Rapid Thermal Cycles (+85/-100) Visual Inspection 3,500 Rapid Thermal Cycles (+95/-85) Thcrmal Vacuum Continuity Check Visual Inspection Electrical Performance 7,000 Rapid Thermal Cycles (+75/-105) 8,000 Rapid Thermal Cycles (+75/-113) Thermal Vacuum Continuity Check Visual Inspection Electrical Performance 25 Roll-Ups 5 Roll-Ups Tabl e 4. Qual i f i cat i on t est sequence f or sol ar a r r a y sampl es (all t emper at ur es in "C) The solar array blanket samples passed all of these tests satisfactorily. To establish the operating margins of the blankets, one solar array module was subjected to extended thermal cycling testing, up to 65 000 cycles without failure. This indicates that the array's structural integrity is far better than specification, which will allow it to stay much longer in orbit than planned, and that the power output will dictate the maintenance schedule. 6. 2 ELECTRICAL PERFORMANCE I n order to predict the solar array's power delivery as accurately as possible and be sure that the solar generator will deliver the required power to the spacecraft for at least 5 years any negative effects on the electrical power output and their origin must be studied in detail. Extensive radiation studies and tests have been performed to predict t he total accumulated fluence and evaluate the effects of electron, 1312 proton and photon radiation on the solar cell assemblies. For the Space Telescope cell type in combination with the 150 micron thick cover glass t he resulting degradation characteristic is given in figure 5 for the most interesting parameters. YEARS INORBIT 1 2 3 4 6 75 px: P M U Ilowever, if only the effects of radiation are used to evaluate t he power during lifetime t he resulting total degradation would be under- est i in a ted . I n addition to the natural environment (radiation temperature, ATOX) further potential impacts on the power output, such as solar cell orientation errors, random cell failure and mis- match, micrometeorite and space debr i s bombardment, have to be considered. These effects are considered in form of current and voltage loss factors applied to the solar cell I/V characteristic. As a result of these degradation studies, and the initial power readings after the solar arrays were deployed in space on 25 April 1990, the Telescope's solar generator will have an expected performance as indicated i n Figure 6, resulting in a minimum power supply of 4.6 kW after two years, and 4.4 kW after four years in orbit. This is far better than the specified requirement of 4.4 kW after two years in orbit and, together with the very good structural integrity of the arrays after deployment fully supports the goal of significantly extending the lifetime of the HST solar generator. This excess power gives the astronomers and scientists much greater flexibility i n scheduling their observation times. 4 2 0 0 4 . . . , . . . , . . . . . . . , . . . , . 0 385 730 1096 1400 l l P S T l m In Orblt (&ye) Figure 8. Predicted solar generator power p r o f i I e 7 . CONCLUSIONS The concept chosen for the Space Telescope's solar arrays is a sound solution for this five-year low-Earth-orbit mission, with its inherent stringent requirements on resistance to atomic oxygen and thermal cycling effects. Based on the successful testing of qualification sample over 65000 thermal cycles, t he European provided HST solar array should survive 10 years in its orbit and supply sufficient power to the Telescope throughout that lifetime. The new technologies developed in Europe for the HST solar array blankets are also an excellent basis for meeting the challenge of future projects such as Columbus which will have even more stringent power demands and increasing commercial constraints. REFERENCES L. Gerlach, Improved Solar Generator Technology for t he Eureca Low Earth Orbit, Proc. 5th. European Symp. on Photovoltaic Generators in Space, Scheveningen, ESA SP-267, Nov.1986. D. Poeck, Design Development, Qualification and Flight Production of an ATOX Resistant Flexible Solar Array for LEO Mission, Proc. European Space Power Conference, Madrid, Oct. 1989 (ESA SP-294). 1313