Variable Emissivity Through MEMS Technology PDF
Preview Variable Emissivity Through MEMS Technology
\ \ VARIABLE EMISSIVITY THROUGH MEMS TECHNOLOGY Ann Garrison Damn, Robert Osiander and John Champion _, Ted Swanson and Donya Douglas:, and Lisa M Grob3 ITheJohns Hopkins Universi(y Applied Physics Laborato_., 111019Johns Hopkins Road Laurel, All) 20723 2NASA Goddard Space Flight Center, Thermal Branch, Greenbelt, MD, 20771 3Swales Aerospace, Beltsville MD 20705 ABSTRACT performances of the new vari-e technology and standard thermal control systems are also presented inthis paper. This paper discusses a new technology for variable emissivity (vari-e) radiator surfaces, which has KEY WORDS : Satellite, Thermal Control, significant advantages over traditional radiators and Microlechnology promises an alternative design technique for future spacecraft thermal control systems. NOMENCLATURE All spacecraft rely on radiative surfaces to dissipate -Z zenith radiator waste heat. These radiators have s_'ial coatings, +Z nadirside radiator typically with a low solar abso_ and a high +Y cold side radiator infrared-red emissivity, that are intended to optimize Greek symbols performance under the expected heat load and thermal o_= absoptivity sink environment The dynamics of the heat loads and _3 =emissivity thermal environment make ita challenge to goperly size the radiator and often require some means of regulating the heat rejection rate of the radiators in ord_ to achieve INTRODUCTION proper thermal balance. Specialized thermal control coatings, which can passively or actively adjust their All spacecraft and the insmtments they support emissivity offer an attractive solution to these design require an effective thermal control mechanism in order challenges. Such systems would allow intelligent control to operate as designed and achieve their ex-pected of the rate of heat loss from aradiator in response to heat lifetimes. In an increasing number of satellites, optical load and thermal environmental variations. Intelligent alignmem and calibration require a strict temperature thermal control through variable emissivity systems is control. Traditionally, the thermal design is part of the well suited for nano and pico spacecraft applications spacecraft layout determined by all subsystems and instruments. Heat load levels and their location on the where large thermal fluctuations areexpected due to the small thermal mass and limited electric resources. spacocr_ equipment temperattge tolerances, available power for heaters, view to space, and other such factors Presently there are three different types of vari-e arecritical to the design process. Smaller _ with technologies under development: Micro Electro- much shorter design cycles and fewer resourc_ such as Mecbxa'tical Systems (MEMS) louvers, Electrochromic heater power, volume, and surface, require an active devices, and Electrophoretic devices. This paper will approach. Commonly used active methods such as describe several lXOtotypes of micrornachined (MEMS) electric make-up heaters, heat pipes or mechanical louvers and experimental results for the emissivity louvers incur additional electric power requirements, variations m_ on theses ixototypes. It will further weight, and bulkiness to the system. New, more flexible discuss possible actuation mechanisms and space _hes are desired to address these conccms. One reliability aspects for different designs. Finally, for solution is a radiator coating with a variable infrared comparison, parametric evaluations of the thermal emissivity which can actively be adjusted in response to variations in load and environmental conditions. Such variableemissivity thermal control coatings, which under the louvers has been removed using deep reactive include electrochromic and electrophoretic devices, have ion etching (DRIE). Figure 2 shows images taken from been under development at NASA_ Space videos of the prototypes for each design when actuated. Hight Center (GSFC) since the mid-1990's. More in the closed, semi-open, and open position. Figure 3isa recently, anew technique has been investigated in which visible image of a louver array closed, and with some micromachined mechanical louvers can be opened or louversopen. Where the louversare open, the dosed to vary the emissivity of a radiator surface. backgroundcanbcseenthroughtheetchholes. (Champion 1999) While very similar to macro scale louvers, cunent micromachining techniques allow for EMISSIVITY MEASUREMENTS fabrication of devices with feature sizes on the order of micrometers (Gilmore 1994, Helvajian 1997). These Infrared images taken at room temperature allow for a micro louvers can be fabricated on suitable substmtes reasonable estimate of the performance of the louvers to and used inplace of traditional tad/atom. be made. The die containing the louvers was mounted onto a radiator held at 40 °(7. Infrared images in the MEMS LOUVER DESIGN AND FABRICATION wavelength range between 2.5 _xmand 5 panwere taken with an Amber Radiance Infrared Camera. Using a Micro-electro-mechanical (MEMS) louvers are similar dose-focus attachment, the resolution was on the order in design to traditional macroscopic louvers in that they of 10 pm per pixel. Calil:ration was performed on a can be opened or closed to expose an underlying high window etched in the sul:_strate showing the radiator. emissivity radiator. Their small size, a few hundreds of (emissivity = l) and on agold-<x)ated area (emissivity = micrometers, allows for different mode of operation. 0.02) of the _. SuR_est deletin_ this next They area attached tothe radiator surface and do not add sentence..seems to conflict _vith the t_O.02 of abo_e substantial weight or bulk to the spacecraft. The large --Vapor dqyositcd gold tuJsan absorplion efficicnc_ of number of louvers, which ideally can be actuated •1!}-. 3;H_Ian cmissivity rating of •{)3-.1,---- Images of independently, also allows control of the number of open the calculated emissivities of the MEMS louvers ate and closed louvers and does not require any iraermediate shown in Figure 4 for different numbers of closed position.ThreedifferenMtEMS designs have been louvers together. The louvers were opened manually investigatedand fabricatedT.he threedesignsare using a D'obe tip. The average emissivity, t; for the depictedinFigureI,showingeachprototypeinthcopen louver area is 0.5 for the number n=O of open louvers, andclosedpositionT.hesimplestdesignisalouverthat 0.75 for n=6, and 0.88 for n= 12, respectively. Note that canbeopenedtoaverticaplositiotnocxposeanareaof a sizeable fraction of the area over which this the high emissivity substmte to space. In a second measurem_ was made is devoted to mechanical design, multiple levels of sliders move across each other. supporting the device opemtiort Through In this case, the total area, which can be e_ design modifications, we believe that the ratio of louver depends on the number, n, of laycrs available in the area to support structure area can be increasedw,hich fabrieation process and is about (l-l/n) times the slider would increase the variation in effective emissivity up to area.. Advantages of this approach include the two- the order of eighty percent. dimensionaldesign,the linearvariabilitoyf exposed area.and the stttrdinesosf the dcsigrt The third LOUVER ACTUATION prototype mimics a bi-fold door. This design is more complicated thanthe other two since ituses more hinges, For a successful application of the MEMS louvers for but we cxpectit to be more ragged than the single spacecraft thermal control, an actuation mechanism has louvers while goviding the same active area. to be identified which allows a high level of individual Preliminary experience has shown that these devices are control, rugged Ol)em_tion, and a minimum of space. less likely to break during release and operation. The Note that the area covered by the actuator reduces that MEMS Ixototypes have been designed at the Applied available to the vari-e radiator and presents an emissivity Physics Laboratory (APL), fabricated at the MCNC bias. Highly individual louver control provides the best Technology Applications Center using the MCNC accuracy in setting the emissivity and further allows Multi-User MEMS Process (MUMPs), and subsequently increased control of the spatial emissivity variations and released and tested at APL. The base material for the operational redundancy. Further, low power current devices is polysilicon and the exposed, top consumptionand zero power ina static condition are surface is coated with gold, which has a very low required for small srmecrafl alaclieatiom. Several emissivity of 0.02. For the section of the silicon chip, actuation mechanisms have been investigated including which underwent emissivity testing the silicon sul_wate anelectrostactoicmbdrive.WhilethisstandaMrdEMS to 50,000 cycles. Various coatings and design actuatoisra low power and reliable device, the large area modifications to minimize friction are being considered. required and, from a space-craft perspective, the Similarly, the effects of fatigue on hinges and actuators relatively high dri_4ng voltages (tens of volts) are also being examinextFinally, ma space environment, necessa_, ;rod the tendency of a coating exlmSed to the MEMS louvers will be subjected to high-energy the space emironment to build up a static charge, irradialiorc As a result, charge buildup in dielectric are limitations. Another mechamm_ a layers could occur which may lead to inconsistent or "heatuator"(Butler, 1999), which occupies less area on degraded operation of either the louvers or deetrostatic the die has also been investigated with prototype louvers. actuators. It does not require high voltages but will consume power in the form of current to heat the actuator structure. PARAblETRIC ANALYSIS Similarly, other actuation mechanisms which involve thermal expansion, including paraffin actuators, An incremental study aPl_oach was used to bimetallics or shape-memory alloys such as Nitanol TM progressively assess the thermal performance of different (Seguin, 1999) require current for the actuation. vari-e radiators (Grob, 2000). The orbits and Therefore, the design will ensure that current is only environments used in all phases of the study were required to change the emissivity. selected to reflect the TERRA s'pacece_ design, with specific emphasis on the spacecraft's batteric_ RELIABILITY ASPECTS TERRA's orbit is a sun-synchronous 10:30 descending node, 98.2° inc"lmationpolar orbit at an altitude of 705 No discussion of the use of MEMs incritical space flight km. For the dynamic study, TRASYS and SINDA would be complete with out menlioning the unique Thermal Math Models (TMMs) of a simplified cube concerns specific to MEMs. The louvers must survive provided heater power requirements forthe zenith, nadir through the launch and operate in the harsh environment and cold side radiators surfaced with traditional and vari- of space. In addition, the effects of pre-launch storage e coatings. The parameters used in the transient analysis must also be taken into consideration. A non-exhattstive are listed in Table 1. Texture set points of the list of the of MEMS reliability concerns includes: proportional heater modeled m SINDA, were set below stietion, wear, fatigue, comamination, and radiation the vari-e light to dark set point in an attempt to effects. (Stark, 1999). minimize the number of cycles the variable emissivity surfaces would experience. Of particular interest are the Although stiction has not been observed in the lxototype spacecraft "cold side" (+Y) radiators where most of the devices, the MEMS louvers are probably susceptible to housekeeping equipment (including the nickel hydrogen this failure mechanism as a result of electrostatic batteries) resides and the insmnnent deck on the nadir interactions, capillary forces, or even localized cold side (+Z) for the typical LEO science missiort Both the welding (Patton, 1999). These concerns can be addressed nadir and cold side radiators exlxxience mainly albedo in several ways. For example, proper ground design and IR fluxes. The albedo is significant, however, in the should minimize the potential mechanical seizure due to low beta orbits. During the year, the TERRA spacecraft electrostatic clamping Excessive condensation of has a beta range of 13° to 32°. The zenith (-Z) radiator moisture, especially during lXe-launch storage, can be was also studied since a significant portion of the mitigated through the use of hydrophobic coatings and TERRA housekeeping equipment is accommodated outgassing techniques. Furlhennore, aplxolxiate there. The zenith radiator experiences direct solar fluxes, packaging could be employed to txevent the which emphasizes the significance and/or limitations of accumulation of water and other contaminants on critical the current vari-e technologies' absorptivities. The surfaces ofthe &'vices. transient results from the simple cube model are listed in Tables 3. Note, the effects of the UV environment and While relative humidity (RH) levels in excess of 70% radiator absorptivity on some of the radiators. On all have been associated with been degraded mechanical vari-e-mdiators, the heater savings are in the order of performance attributed to increased stictio_ elevated 60% to 80 % compared to the fixed radiators. In an frictional wear between contaclmg parts has been additional calculation modeling the TERRA batteries observed In extremely low RH environments (Tanner, with an MEMS battery radiator the orbital average heater 1999). Due to the negligible RH of the intended power savings was 54 watts, essentially all the power operational environment, the possible degradation of the needed for the batteries to maintain their minimum hinge joints over the device lifetime is an important allowable temperature in the original configuration. Inall issue. Minimnm lifetimes will be on the order of 10,000 calculations, the power needed to control the vari-e surfaces wasconsiderednegligible. TABLE 1. Transient Analysis Assumptions and Parameters. Typical sun-synchronous LEO orbit / Material Radiator Properties beta range 13°and 32° Hot Case Models: EOL environment and properties on all cube surfaces, Panel BOL EOL temperatures maintained @ 30°C AgTeflon 0.08 0.18 0.79 Cold Case Modds OSRs 0.08 0.14 0.80 BOL environment and D,-operlies on all cube surfaces, Panel White Paint 0.15 0.40 0.85 Electrochromic 0.30 0.30 0.2001 0.70I21 tempemttm_ maintained @ -IO°C Set Points: IVlEMS 0.20 0.20 0.20I110.80[21 -8°C to-10°C forprotx_onal heate_ Eleclrophoretic 0.20 0.20 0.291_1 0.80121 -5°C for vari-e stwfaces [1]light/closed/activated state [2] dark/open/base state TABLE 2. Results for orbital average heater power [W] from transient cube model for different radiator positions Orbit Ag Teflon OSRs White Paint I Electro- Electro- MEMS I chromics phoretics Zenith Radiator 13° EOL 32 37 28 9 10 6 Tmax/Tave [°C] -8/-8 -8/-8 4415" 34/3* 01-6 0/-6 32° BOL 43 44 41 10 10 7 Tmax/Tave [°C] -8/-8 -8/-8 -8/-8 19/-2" -5/-7 -5/-7 Nadir Radiator 13° EOL 10 11 Tma:CI'ave [°C] -8/-8 -8/-8 4/-5 1/-5 -41-6 -41-6 32° BOL 14 14 13 Tmax/Tave [°C] -8/-8 -8/-8 -8/-8 01-6 -41-7 -4I-8 Cold Side Radiator 31230"BEOOLL'"'* 4309 4400 4413 1134 1144 910 * Elevated temperature due to higher _ and non-optimized area **Temperatures never go above heater set point temperatures Fig. 1:Different design concepts for athermal control structure: louver (top), slider (middle), and folder (bottom).
The list of books you might like
