PROJECT MEDUSA Michigan EDucational and Utility Satellite for Alaska Cover designed by -John Van Roekel and generated by the MTS/IBM 360 A Student Design Project Department of Aerospace Engineering The University of Michigan Winter Term 1971

TABLE OF CONTENTS SUMMAR Y 1. GENERAL INFORMATION 1 1. 1 Need for a Satellite Communications System for Alaska 1 1. 2 Project Definition 1 1.3 Overall System Description 2 2. COMMUNICATIONS 6 2. 1 Introduction 6 2. 2 Geometry 6 2. 3 Downlink 7 2.4 Uplink 10 2.5 Transponder 12 2.6 Traveling Wave Tubes 15 2. 7 Ground Receiver Equipment 17 2. 8 Telemetry and Command 20 2.9 References 24 3. SATELLITE POWER SUPPLY 25 3. 1 Introduction 25 3. 2 Solar Arrays 27 3.3 Battery System 29 3.4 Power Distribution 30 3.5 Weight Breakdown 32 3.6 Reference 34 4. ATTITUDE CONTROL 36 4. 1 Introduction 36 4. 2 Definition of Axes 36 4. 3 System Requirements 38 4.4 System Description 38 4. 5 Attitude Sensing 39 4. 6 Onboard Computer 41 4. 7 Momentum Wheels 41 4.8 Redundancy 43 4.9 Propulsion System 44 4. 9. 1 System Description 44 4.9. 2 Fuel Use 44 4. 9.3 Propellant and Hardware 44 4. 10 References 48 5. THERMAL CONTROL 50 5. 1 Introduction 50 5. 2 Satellite Main Body 50 5.2.1 Radiators 50 5. 2. 2 Shunt Resistoirs 50

5. 2.3 Insulation 50 5. 2. 3 Thermal Coaiings 51 5. 3 Satellite Feed 51 5. 3. 1 Radiators 51 5.3. 2 Thermal Coatings 51 5.4 Thermal Component Weight 5t 5. 5 References 6. STRUCTURES 53 6. 1 Introduction 53 6. 2 Overall Satellite Configuration 53 6.3 Spacecraft Design Features 55 6. 3. 1 Launch Loads 55 6.3. 2 Solar Cell Array Mount and Drive 55 6.3.3 Equipment Panels 58 6. 3.4 Propulsion and Control Mounting 58 6.3.5 Reflector Support Columns 62 6. 4 Reflector De sign 62 6.4.1 Design Requirements 62 6.4. 2 Reflector Selection 65 6.4.3 Operation 67 6.4.4 Construction 67 6.4.5 Reflector Support Structure 67 6.5 Feed Assembly Design 74 6. 6 Feed Support Structure 74 6. 7 Weight Budget 76 6.8 References 79 7. LAUNCH VEHICLE 79 7. 1 Ifitroduction 79 7. 2 Launch Site and Window 79 7.3 Launch Vehicle Weights 81 7.4 Launch Vehicle Systems 81 7.4. 1 Guidance and Control 81 7.4. 2 Electrical Power Systems 82 7.4.3 Telemetry Systems 82 7.5 Loading Factors 82 7.6 Fairing 83 7. 7 Spacecraft Attach Fi~'t'ing 83 7,8 Burner II 83 7.9 Useful Payload 86 7. 10 References 86 8. ORBITAL ANALYSIS 87 8. 1 Final Orbital Requirements 87 8. 2 Mission Tlmetable 87 iii

8. 3 Launch Phase 89 8.4 Transfer Orbit Injection and Transfer Orbit 90 8. 5 Geostationary Orbit Circularization 90 8. 6 Rendezvous Maneuver 92 8. 7 Walking Orbit 92 8. 8 Perturbations 92 8. 8. 1 Introduction 92 8. 8. 2 Corrected Geostationary Radius 95 8. 8. 3 In-plane Perturbations 95 8.8.4 Out-of-plane Perturbations 95 8.9 References 96 9. PROGRAM PLANS AND COST 97 9. 1 Summary Schedule 97 9. 2 Co st Estimate 97 9.3 References 98 APPENDIX A - Communications 100 APPENDIX B - Power 114 APPENDIX C - Attitude Control 116 APPENDIX D - Thermal 124 APPENDIX E - Structures 133 A.PPENDIX F - Launch Vehicle 143 APPENDIX G - Orbital Analysis 144 ACKNOWLEDGEMENTS 15 2 PERSONNEL ASSIGNMENT CHAR T 155

SUMMARY Project MlEDUSA is a satellite co:nmmunications system providing Education Television (ETV) and General Purpose Television (GTV) to the residents of Alaska. Of primary concer.n is providing this service at a cost of about $120. This cost covers the receiving antenna, installation, and the adaptor unit which is compatilble wiDth an ordinary TV set. Launch of the satellite will take place October 16, 1974 from the ETR using the SLV-3A./Agena D/Burner II launch vehicle configuration. The satellite will be placed in geostation~ary orbit at 1700 W longitude and will provide 5 years of service. Once in geostationary orbit, the 783 pound satellite will first deploy its 30' diameter flex-rib reflector, then its two solar arrays in the north and south directions. The satellite will then use its on board propulsion system to get to the desired on station position. Television transmission which receives an uplink from Alaska or the West Coast can begin within 18 days from the date of launch. Three TV channels with a center frequency of 800 MHz will be provided from 8:00 am to 12:45 am every day of the year. The power required for all subsystems on board (725 watts EOL), will be provided by the solar arrays and batteries for five years. The solar arrays provide the bulk of the power and are supplemented by the batteries during the evening hours when the arrays are partially shadowed by the reflector. The batteries also provide the power necessary during eclipse. The thermal subsystem is primarily passive and consists of radiating surfaces, power shunts, insulation and heat pipes. Attitude is controlled by reaction wheels which are periodieally activated through an on board computer by sun sensors, earth sensors, and rate gyros. Hydrazine thrusters provide for unloading the momentum wheels and for east-west stationkeeping.

I I. 211 NEED FOR A SATELLITE COMMUNICA4 TIOrONS SYSTEM FOR ALJASKA Alaska wigth a promise of oil and m ineirl depo sit mtmay be oi t1;he ve -ge of rapid expansion. Many people will be living and rworking in hne ]N>qoreth slope oil fields w.v-ith virtually no means of commlnunication w\ith the rest of the State or United States. A. ground network for a communication system is an economic and physical impossibility in the foreseeable future. The population is too sparce, the distances too great, and the terrain and climate too severe..Live, coverage of the Apollo 11 mnission provided in Ajn- chotrage via the Department of Defense's TACSAT xwas ather fiirst timre in the hisgtory of Alaskan television that a live vo eracre of a di.stant ne-ws eve-nt was made available, A satlle t communications system would solve all the above problems. Distance is no barr ier aid cost is comparatively loew. The benefits of the system wourr ld be en srrnous, To both the urban and rural Alaskan, it means the end of isolatios fronm the rest ofi thee wsorld. It provides live television, insCant access to vital news and!/or tech.nical data. It offers the advantages of advanced educational and nmedical technjiques. And also, it would offer the Ala skan reside nt the pleasures of modern te evision en tAertainm9eixtL,. 1, 2 PROJECT DEFI!NITIONQ Project MEDUSA (Michigan ED_ xucational, and Utility Satellite for Alaska) is a prelml ary design; st dy n-vestigatiug the possibility of developing a light woeight com-rimunications satellite systemn providing ETV and GTV to the residexnts cf Alaska. The following constraints have been placed on the systenm: Priima-rily stateof.-thearl deveiLopm.ent wit7h limited technical de'v e opment s. G riound receive r must cost less ',han $ 50. 00 and be adaptible to present day TV se'K. 3. Filve yea(zr minirnmurn lifaetime]e4 4. Useful payload i'veigh't less thasa 800 pounds. 5. EOL sat'elIl.te powr3C less tcha2: one kilowatt. 'Thes7,L$e cons vrainr 6s -i.mpo se td h'2! rh lffI. i otsisn on the entire system. Some of the mcore imspor stantc ones are refle-tor a design du e to limi+tations on shroud size, travelinlg wave tube technology, anud the necessity of a good i.au:ach guidance systemo

1. 3 OVERALL SYSTEM DESCRIPTION' Project MEDUSA is a iaeliLe comminnni atinons system deslgned to1 providre Educati-onal Television (ETV) an Geine ral P-aupose Television (GTV) to the residents `s 'of A.laska. Th ree tele-vision dhaijnel s (TASO grade #2), will be available so 4hat progranmmng is flexible. The satellie will provide this service from 8:~00 am un)til 12945 ami- eve ry day of the year. Duri-ng the day, the channels could be used for educational purposes, supplementing existing faciilities,, The evening hours could be used primarily for entertainment with the option of adult educatlion and ~n.ews broadcasts. This service would be made available at a cost of $e120 to all of the citizens of Alaska except those living in a small area in the southern part of the panhandle and the area near the Bering Strailt These areas would require a. 6 85 foot ground receivilNg antenna insteadd of a 5. 2 foot antlenna at an increase in cost of about $18. The television transmilssion could receive uplinks from both the West coast and from Alaska. The Alaskan uplink could provide the educational programs while the West coast uplink vwould serve as the link between Alaska and the rest of the United States. Alaska would essentially have any of the ser.Lces now available to California residents, including live news broadcast and modern ente rtainmrent o A 30 beamwidth was necessary-to adequately cover the region from Ketchi.z.an to Nor ae. 'This r-esul'ted i. a 30. realeketo r.- Thr"ee FM (ch annels are provided, each usi~.g 35 LMHz band-w7idth and requiring 59 watf;s of RF pov er. A total of 177 waltts RIF power is require'd for TV t' ansmi ssion. Since presenlt day 'TV sets use iiAM, a zmodulation co:nverter is neLcessary on the grouAnd recenivlng aniLto The entire receiving ui -.can be instaledd on most TV sets for a total cost of ab ut $120. MED3USA wvill be laun-hed in late 1.974 f rom the Eastern Test Range (ETR). The SLV-3A/Agena D/Burne(.r II launch hvehicle configurg:htono -wvill be used to pult hb.e 783 pound satellL.te irmb'o geostlati-.-:onary orbit at 170 W longitude. The B-II was chosen beca'se it can pro-vide the accuracy and stability requir ed by the satellite 'coniguration. L the stowed position, the satellite is 109 inches long and requires a 76 inch inside diarneeter shroud.

The 76 inch inside diamrte"se -i, necessary to stow the Lockheed flexrib 30 foot diamete r eflector. This refiea.oor is used for both television tran smission and:receptiorn. Am of-et feed enables the satellite to see the West G tast. The po inting ac cries pro-vided by the at'titude con-trol subsystem are + 0. I~ in the pitch axis, + 0. 2~ in lhce roll axis, and + 1. 0 in the ya w axis according to the axis coordinate sys~tern set up for the satellite Si.; sun sensors, one earth sensor, and rate gyros feed iziorrnat-ioni to the on board cnomputer which in aturn activates either the mcaementLiurn wvheeis or the hydrazine propulsion system. The hydrazine threusters (0, I Ib) a(ep use)d to unlo.ad the mnormnentum Nl eels and to east.- (est stationkeepo No north-south stationkeeping is used. The orbital plane of the satellite illl be initially inclined 2O 10 off the equatorial plane and will drift at a rate such that it gill coincide vtith the equatorial plane at 2 1/2 years and will again be inclined 2. 1 off the equatorial plane at the end of five years. Communications, telemetry, and all other satellite subsystemm require 7 25 EDL watts of powver. TTwo 2. 8 x 32. 7 ft solar arrays are deployed immnediately following the deployment of the antenna. These solar arrays provide the 725 watts requlred wvhile the satellite is transrmitting (fromn 8:00 am until 12:45 am). Two 43. 7 pounzd Nickel-Cad mium bateics conp peise ntrh secoadary ysstenm which is used during the launch phase, the e-.lipse peri-od, and durzing Shadowing oil' the arrpays by the refle tc1tor. The thermal s-ubsystem is primarily' passiv e. Its compoznents irnclude radiating surfaces, powver shunts insulation, and hea~t pipes. The spacecraft str.ucture consists of tvwo main support sections. A. hub strui-ture (feed) supporsted by a K-truss will concentrate the feed a6ssembly weight at the support points while still prsoviding a sample -way of mou:L -ting the complete feed system. The main spaceccra..f`t module will use a s-trong base structuire to allow veasy mou rting 'thr oughout the module witEhout coencet 'aating individual loads only at the support points. The interface between the spacecraft body and the Burner II is a three pointb support. This design was necessary to mreet B-I requirements. Alunaiinum 7075-T6 was chosen as the mrain structural material because of its tveight, reDlative ease of fabrication, aind high thermal conductiv ity The total cost (oe the systerm i estimrated to be $97 nillion. The developrment of this type of satellite ystemis likely to be,borni e by the United Stalts a-ad Alaska, the ground system- insalaltinn borne by the u er, anld the operat.ion conducted by a broadcasting company. The worth of a satellite broadcastling system becomes evide dt ews.hen the satellite system is compared t- o the cost of impleme1i'ng a e peracLing a tesaentrial system providing t the 3- channel service to all of Alaska. The ground syste rm requires 60 stations wil'th I00 nmile spacing to (over the 0. 6 milli,on square m.ile area.. The _total costs for operating the systemn for five years wv-;ould be $207 tniilion- The t erestrial mnethod w-eould be 2. 1 tisme~s the cost c- f the s~aitelite,syst$eme approfa(ch. A. e5ieght and powei' budget fors all the &ub;,syst' ms is s howniS@' 'in Table. 3. 1. Figure 1o 3 o show, the st'ellite e'"iguate.'on l af it ifs on taStisn.

Reflector Torus Feed Assembly Main Body Support Structure 1 of 24 reflector ribs Deployed Solar Array Figure 1.3. 1 Overall Configuration: On Station

Table 1. 3. 1 Weight and Power Budget Weight (Ibs) Power requirements (watts) Communications Transponder 43. 0 585. 1 Cable 4. 2 -- Monioring and Switching 25. 0 Telemetry -2. i 27. 0 (1 04. 3) (612. 1) Reflector (100. 0) Power Drive and Deployment 61. 0 - Solar Panels 60. 2 Array Drives 25. 0 8. 1 Batteries 87.4 10. 0 Power Conditioning Unit 24. 9 50. 0 (258.5) (68.1) Attitude Control Fuel 60. 0 - Tank and Valves 23. 0 Thrusters 3.6 6. 0 Computer 4. 0 5. 0 Sensors 17.0 6.5 Gyros 3.6 13.5 Momentum Wheels 30.3 13. 8 (141.5) (44.8) Thermal Heat Pipes 9. 0 Insulation 6. 0 Thezrmal Coatings 1. 0 (6. 0) Structures (162. 7) Tota.l Spacecraft Wt 783. 0 Margin for Grc oi-h ll11117.0 Total Power 725. 0 Watts Total Capability 900. 0

C OM.M U.,N.,JIC AIL. N OS 2. 1 IN TRODUCTION The comrnuynistmLat ons systmf t~he presenrt projEct isL required to prvide o itPdee TV channrel that ca r be received by cconventional TV r ece-vers equipped 'wil.h inexpensive ant enras and co.:~ver tersO Conventional TV is amplitude modulated (except for voice). 'This system conserves bandwiwdth at the expense of pow\ver. For the application at hand, however, i-t appears tihat bandwdl&dh is no+t scarce, while power on board the satellllte is seve:ely limited. 'Therefore, frequency modulation (FM) was chosen. The mcoddlation consverter that becromes necessary at each receiver is inexpensive (see Section 2, 7). The choice of frequency is do mirnated by the cost of ground receiving equipment. A 5. 2' parabolic dish is used as the ground receiver a nt-enna (except in certain fringe areas where a larger dish is indicated). In the UHF band this can be a $22 -wire mesh; the S-band or X-band would require a solid reflector tha t cos's $200~1 The costs just quoted are based on a production quantity of 105 Hows-Uever, t:he population of Alaska is only about three times that. The $200 figur, is sure to drive the audience well below 105 and the price of a dish well above $200. It thus appears that a truae broadcast service for Alaska is possible at present only in the UHF band. Accordingly a center frequency of 800 MHz is specified. The price to pay for this choice is the large on board reflector that becomes necessary. We use a 30~ parabolic reflector as the main satellUte antennao. Mo'1'& detail on the compafrison of different requencies appear in Appendix A. 5. The FM system specified follows the one studied by GE for their India system2vwith a Qmodulationr index of ], 5. The required C/N ratio is 7.5dB corresponding to a S/N ratio of 39. 5dB (TASO grade 2) after demodlulation and reenmphasis. (This 7 5dB figure should be contrasted with 33. — dB for AM - a factor of 400 in poer.~ ) The baseband (video + audio) is 5MHz. By Carson's rule this implies a noise bandwidth,of 25 MHLzs/channel A spectral occupancy of 35 MHz/chaannel is allowevd to provide for channel separation. 2, 2 GEOEMETRY A snCch onous sateie is located above the eq uato.: at 22, "67 x 10 nm from thhe earths cerizatero Tthe earthS grid of~ ongitude and rlatbiude as seen fromrc the sa~elIlte 1~5 shown i:n Appenodimx A. 1. Alaska, located in laatiuades

54 70 7Ns appears flat.sened out near the r.rm of the earth. A. circd ar beam that covers Alaska has to be of a diameter Of abotut 3. Less than one third of the beam sectioon actually illuminates Alaska. The balance its the ocean or goes ove r the riLm of the earth. Du.e to its particulaPr shape Alaska appears narrow er wheln viewred from the east thanz from the ',west. TThis. wTould favor an easterly positi ord.g of the satellite (eo g. 1400W with a beam of 3 covering all of Alaska). Horitever, con1sidesratjiorsns o-f thie timrrme of the ellpjpse th at occurs durin.g certL1ainr pe riods in lhe spring and fall and of ch'd ows cast by the antenna on the solar arrays favor a westerly location (the furt'her west the s atellite is, the later in,the day the shadows and the eclipse occur). The project ma-nagement consider ed the ability to pro'vide service t~hrough 12W00 am. Ajatska tinme year round to be the overriding factor, a.:id a westerly 1location was decided upon. The locatio:i actually chosen is 70 0W. This delays the beg'i:nLing of the eclipse at equinox to 12:45 am. The beam mte rsection wit-h the surface of the earth is mapped out in Figure 2, 2, 1 wvhich shows contours of beam diameter and of satellite elevation, It wiill be noticed that coverage of all of Alaska (excludinag the Aleutia- Islands) requires a beam width of 3. 5-. The Bowest elevation is 9 and it occurs at the -rorth-,east;: corner of Alaska, A.. mnargin on this elevation is the factor thaot places a limit% on the,weste rly positioning of the satellite. (The satellit.e is expected to vary in eleyvation by up to + 20 in a daily cycle. ) For this reason yve use an elevation of only 70 in the link calculations~ Naotie. that the locatlon of the sateillite is wivthin easy vewofthe West Coast of the continental UO S. A. San Franc.isco, here the West Coast uplink station is located is 4. 20 off the center of the beam. 20 3 DOWNLINK The dow~nlnk calculation in wvhich the n-oise at the receiver determines the RF output required of the satellite is presented in Table 2. 3. 1. It uses the frequenc y and mrOdGalation system specified in Section 2. 1. 'The neessary RF pow-e-r is fixed at 59 w/Fcha-n.el or iT7 7w for all three channels. The satellite transmi;er ante ana irs a 3 reflector described i-n the structures sectio'na of the report (Sectilo' 6.4). The reflector is illuminated by a helixin-a-hol nn feed -with thle rim of the refIector being at the IOdB angle of the feed. The details of the feed desig appea i Appe Appendix A. 2 The ground rec eiver anaterna is a 5.2' wir e mesh pa rabolic dish. It i.s further de scribed in Section_ 2. 7. $,,4~c,, ~NI"~V~~k~hB~ "Ji~ IF~P~O~20 ~7

10' 165'~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~' unte2 875' -~~~~~~~~~~~~ 0~~~~~~~~~~~~~~~~~~~~~~~. I~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 0P. Linroln Sac....... 0!., '~; ~~~~~~~~ ~ ~ ~ ~ ~ ~~~~~~~,. C, LE.. - E ~~~~~~~~~~~~ ~ i,""~"~~?.~ Fgur.22.1Mapof he eam:- ' ~' 8 '~m:::3~ '~ r~'N~~~~~~~u~P~, 0~~~~~~~~~~~~~~~~~~~~'""! i ~ ~..::::~~ I~:,, -_ "~-,~,.;~..;. ~ ~ ~ ~.~ "'~.;~.;,'?~."'...'.:[~i:,i~,~7,rrF,~"::'".': ':"'.:-':.."".".;".'l:':'n 5 0 b,..,-.j~....,Z.,,,\,,.............,n.... ~..\~.~..,-.......~. sffheff, Nunivak 1,44-'r~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~!',';.?:~-.,,.%~~'.........:.. R~~~~~~..:....-,-................... '.... ". /'"''."' 8? —.'""0:..:: 00~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Pribi~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~........ Is.a _............ I-.,..t... ~....~ Ft~~~~~~~~~~~~~~~~~, ~.-" Ne B ' 'IN'..UD.... '},'~,:Jt, BAY''. ak~~~~~~~~~~~~~~~....;....h-.o.ur;hil 1 o~tr Z" 5 R. hi~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ G~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~- Is~~~~~~~~~~~~~. Is~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~'. leu~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~f am pins.: Ct -,i, '-t th~~~~~~~~~~~~~~~~~~._UR',- St.. '.: 45* S~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~. Sa~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~,~v..~~r A~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.' s. "%~~~~~~~~~I ooel S a-In~~~~~~~~~~~~~~~~~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~. =<35, —, rez -.0111~~~~~~~~~~~~~~~~ (M k~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~~~"ak "-::::.0-1 ca PIag, WM 4W~~~~~~~~~~~~~~~~~~~s f Qv~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ '"lg " ' Ta:~ gp' 1 q.'211 Ville~~~~~~~~~~~~~~~~~~~~~~~~~~~:r Key~~~ HonoluW0,~~~~~~~ ~~ ~~~~~~~~~~~~~~~~~~~~~~~ s-Pot i.. 110~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~]o':1 too Mdr e' ruz Santiagor; dew or 1z~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ (US. PI' tJAMA Johna...,.~ AaN C.~0ke", (M Ity Pot.5~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ I5* iX' ~'~ ~t~~~~ ~~~ BOND -"'- ~"-L"'~ - ~ - Prnc.:o,- '';..k

Ce::e r faer 'ee,y q:u 8y IH Me d 'ai -on FMA Spe ia ',!. o C capai yv/cJhanne MHz 3H5 Nois'1 se 'blan*:dtwld'th /chnarnel 'MHz - t - n a S ~'N r a, db 9.r 9 ReSq'.red Cr/N ra-c, d B M, rr g.. 'b'' dB 5 Actual C, " N':aio dB 9. 0 Nu'oise Tei-npe at)Ixre K 0 Noise dBW 4 i 24'. Required Carrier a C receiver dB W' '1. 5 (h-alf,- pow' _e r ) Re tquired c-a r.ier a, treceiv e r. dW vr - 2Bo 4 (center ef Lbeam) Free space CaceiL dB 0 18 8 2. Propagati~on losses dR B Po,4:in: g errcr 1(gad) dB -d?O Receiv. a.rc- enea gain dB I7.5 ERP/'chan&el dBW d:B2.: Tra-sm. a~enn gai dB j: —3 d W7'a': - 9 RF pow-,er-.-" L. chahinnels W _,.W a. 7'17

Ihe eijlmes -- s of los s e e s an of ~* 2 n,Er)d i; t'he link c alc lti n are pres ented iz Appe dic es A. 3, A. r es5pecev1. eY. Tho e link iauplatn s ca-r 5hed out f aO poi on ithe half po-wxer beam angie aund at an assuTmed elevavtiin ag,-Le of 7. At 800 MHz the haf-powve r bearnmwidth is 30 At the highest f:reqleny irn the band - 852. 5 MHIz - the halspoa eri beand i elth is 2. 8?0 A poi, ig errTo of + 0. of the i sat litet refle to (Sectio 4. 3 eff el e, ci7u-s tChis ton t 2. Figure. 2. 1 showrs that some areas at the t1ip of. the panhandLe (raround Ane ttle IslaEn) and on the wester-n tip (arou. nd NcXrne) a re at 3. SJ For this condilon the reduction in poweri is K 4dB relative to the oe' -e.. of t hre beam or 24, dB relative to the half-pcw-er ievel. For thisL condltlin a dseceive dish o:. 859 is required to make up for the reduc(s:tios in beam-n powe.r densi-ty This enlarg ed dish is also des>ctbed in Sectioon 2. 7. The populatio2- involved is estimated at 22, 000. 2.4 UPLINK Link edlculat9ions are reversibe aes regards geometr:ical factors and losses (enise aed C/N requiremenrts are not reversibl e). Our basic uplink concept is tailored around an mraiviert(ed d3ownLink wilth a 5. 21 transmitter dish and the, 30Q re flector ori t the aLite atiirg as thee teneive a:tE. This arr'angmer dispenses with a rece-iving an1tearnna orn the satellite; it aLlows the uIse of a rela'tiveldy smr-Ell o5. 22)l dish as the grouand transmitter antenna; ra and it~ aliovJs the placeme t of an uplink taelon ang~sywhere withil the dowvnli:k beam, i. e. anaywhere in Alaska. The (son2ipt outlined above murtsr be rmodified in order to aillowv frequency sepa tira —ion boetwbvee- the upli>ra aud downvslink. A frequency sep'aAation of about 1200 MHZ.H bet feen the edges of the tiwo beams - ras.onsizdered n ecessary. The uplink c enter Krequency w as the refore skhift(ed up to 1 00 MHz (;wh-;ivch tur ned oult nlore advaontageo's than shiifing do'wVn). This increases the refle.tor gain at the c enter of the beam to 38dB but ctaous it dow;t,iioX to 31dB at 3. 3 3 (The 3 30 includes ain allowIance for point error of' tghe satellite reflector. ) 'The Cl/N rtlatio fop the uplinrk mus-^tB} C.e conr.isiderably highel ChanL LOF the dow'nilink. It' was made 14dB higher. Also 'tJhe rn.oise seen by the sat~eiLfe antenna f aci.g the eeeth is higher than that s yee 1 th ground antenn:a facing the sky. (For de tails on noise estima. es see, Appenrd-,d iX 4A ) All the abo ve facotors are. no-rperated i:n the uplink calcu la.;ion, TaWpl~e 2. 4 1 owhich show's that "'he poweir necessary for the uplink varies f~rom 350wZ %L beat Pm cner So 1650mg it 3 30 '7clEtchik:a. n NomPe). A groi da sitaisos~n ubeiing a 2. ) antieLn ae n 1d% Xt50"r RF output cana easily be ml.ade c.'bin)le, '.nd lilve TVr cn be.init.iated 'ywh'1re in 1 skla.ka.,

Table 2. 4. 1 Upli f.k Caica'tio Bea-;Mlm entr, 3. 3~ Center Frequency 1100 MHz No i se b andwldt /channel 25: MHz Noise Temperauroe K 1487 Noise dBW - i 22.. 9 C/N ateio dB 23 Ca rrier dBW -99.9.. Receiver gain dB 37d 8 31. 0 Free spa(e space at enua tionr dB -18 8 8 Propagraion loss dB -1. 9 Poainting error dB -0. 4 ERP/chanrinel dBW ~4? 7. 4- I Z Transmitter gain dB 2. 0 RF/channel dB W 25... 4' 3 21 Watt 35 0 i 650 Apart from these moobile units, a.ceniral uplink station is specified in Fairbars.o This station must be equipped with a 1 0u aantienna to handle telemetry and command (Section 2. 9. The same 10) an.Ienna will be used also for uplink TV tran smission. With it-s 10e diameter and cent-rcal location wi requires olny 100 W/channel in RF output. A West Coast uplink sta.tio n is;o be lc.cated near Sa.n Franc:isco. Its a:cress 'tao the sa-ellite is via a displaced feed o f t-he 30' ref'elto The displaced feed is a helix~ It only -sees part of the 302 reflectsor, and the gain is co.rrespondingly decreased from 38. 0dB to 34. 8dB 'This is more than made up for by t1 uhe use of a 101 'tra-sm.itter axntenna. The RF power output required is ZOO W/channel. See Appenpdi: Ai. 2 for the d etails of $the displaced feed. _~~~~~~~~~~~~~~

2.5 TRANSPONDER The fuction of the transponder is to armnplify the signal received from the ground stations and to shift its frequency for the transmission back to earth. The frequency shift is from the uplink center frequency of 1. 1 GHz to the downlink center frequency of 0. 8 GHz. The amplification is by 116. 2dB, from the uplink received carrier power of -98. 5dBW/channel to the- downlink RF output per channel of 59 watts (see uplink calculation, Table 2. 4. 1 and downlink calculation Table 2. 3. 1)> In principle the three channels covering 105 MHz of bandwidth could be treated as one composite band; they could be amplified and retransmitted together without being separated. This arrangement would entail certain problems relating to the divisiozn of power among the three channels. Also it would require a power amplifier with an output of nearly 200w. As our power amplifier is a TWT (for details see next section) operating at 800 MHz which will have to be developed, it was felt that specifying a 200w output was pushing the limits of present day technology. For these reasons it was decided to separate the beam and amplify the three channels separately. Figure 2. 5. 1 is a block diagram of the transponder system. The gain (or loss) and power consumption of each element - is. given in each block. The level of RF power is shown for each line. Note that the system is made up of two parts, the TWTs, a multiplexer and the first stage of amplification being located near the feed, and the rest of the system being in the satellite body. The two parts are connected by coaaxial cables about 15' long (the focal length of the reflector is 13. 2') inkvol vLing a 11. 1dB loss (the cable is.335" coaxial with silver plated copper conductor4.) The reason for splitting the system in this way is to avoid this line loss after the TWTs where it would cause a 2Z.5%iloss: in RF power, or before the first stage of amplification where it would hurt the C/N ratio. The Alaska feed (in the horn) serves for both reception and transmission. It therefore connects into a mul.tiplexer that prevents mixing of the received and the transmitted signals. The displaced feed is used for reception only. It therefore goes to the line of the received signal via a T junction. The received signal goes through four stages of amplification and through frequency shifting (down by 300 MHz) before it is split into the three channels defined by filters hacnh chanel Ais put >*tfhro tg.h;' a povw;ir limifted ani.plifier to inisure a unifornm pow:er level befo iet going — into: it.s TW'TWT for.the:final -Stage of a, plifi cation. The mutiplexaer parameters follsw data in Section 5i 2. 5 of Reference 1, except that a higher loss is allo w ed in the present system.

West Coast Alaska Feed.Feed ~ input -98.5dBW -99.9dBW I output 177 Watt 177W rasi Stor 0 -99*9dBW Amp;ifier _ M)- ultipLexer '25dBg 0*5W|3Y -W1*4dB ~ 881W z 9. IdBW 81W. KIN 19. ldBW 19.IdBW -77*9dBW | 33dBj 193W 33dB, 193I | 33dB, 1T33l -13w. 9dBW -_d B'.dB 13 ~ 9dB]W -1l1dB Feed to Satellite LddB -1.dB KldB '-l.ldB * CableldB - 12.8dB 8d-12.8dBW -12dBW -79.0OdBW |Translstor r ran. Amp. rans A m, ranS _Ai. Amplifier (Limited) (Li/ited) (Limited)P 25dB, 0.5W u'o to 25d:B up to 25dB up to 25dB o, 5v. 5 5. 0 -54. OdBW -20. 8dBW 208dBW' -20.8dBW | TAplra iesr |-6dBFilt e r Filter e x Filter iAmpI ifie r 25dB, 0.5-dB -6dB -6dB -29. OdBW - 14.8dBW -14 8dBJW -14. 8dBW............t I t [ Transistor.Ampplifie r 25dB, o.5W | l.*Odw..-4... OdBW O.idBo os o-in-ator 13~ur 2 1G2 6WE 13

Table 2 5.1 Transponder Components Item Number carried Weight Size Gain Voltage Current Power Heat Available (Number powered) lbs in dB Volts mA W W Multiplexer 1 8 1 Oxl 0x4 -1. 4 - - - 66 TWT 6 (3) 4 20x4x3 33 see Bection 2, 6 193 112 Transistor ampliifier y 8 (4) 0. 2 3xl. 3xl. 1 25 -15 333 0.5 0.5 WJ-737| Transistor amplifier 3 up to ((limited) 6 (3) 0.4 3 25 - 15 333 0.5 0.5 Mixer 2 (i ) 0.5 4x2x2 - 6 - - - 0 | _ _ _............. _,,._, _ _ __ __ Fil__erj 6 (3) 0.8 1.453 -6 - 6- 0 WJ-0 OsciEllator 2 () 0 6!. 145 24, i 2o 62 | | | | I; - 4 2R I iWJ-57i -24 i. 5 Cable 50 f' 4. 2 S-witching and Monitor ing 25 Total ___ _I_72. 20 _ _ _ _ 585.1 408. i

The system incluides backups foor all itemas other than the iihultiplexer and cables. These are not show..'n in the block diagram. The swi",;ching between compornzaen~ts and their backups is discussed in Appendix A. 6. A list of all components witlh their respective sizes, weights, power consumption, and heat dissipation is given in Table 2o 5. 1. The weights and powers given are for a single item. Ho wever, the totals are computed on the basis of the na mber cf itemls ied or the umber powered at one time. Commercially available eqpl m-zie ntr tLhat either meets the requirements or comes qclose to meeting therwase the basis opf some of the lear!s In these cases, the commercially available items are identified in the la st column of the table. The numbers refer to the Watkins Johnson Catalog. The weights and losses of multiplexers come from paramnetric data in the Section 5. 25 of Reference 1. Detailig of the TWTs are given in Section 2. 6. 2 6 TRAVELLING WAVE TUBES In the search for a power amplifier it was decided against solid state amplifiers because of their loPw eficiency and against gridded tubes because sof their questionable lifetime. Traveling t wrave tulbeswfre the final choice. Traveling wave tubes (TWT) have been space qualified and a 9n0Te have been operating over 3 years in space and are still operating. There exist space qualified TWTs with a power output of 100w, namely the Watkins-Johnson WJ-395 I and WJ-395-3. Unfortunately, these tubes operate in the S-band. High power TWTs at UHF frequencies for space use have engot yet been develoed Pjet MEDUS requ s hahe ewdevelop 800 MHment ofaup spa e qualifi39ed TWT th a satperated HoRF utput f 85 watt operating between sizens, weights, gains, and efficiencies of TWTs do not vary appreciably with wavele gth. Therefore, it is eunpected that the new 800 MHz will measure up to the WJ-395 IS in these respee ts. Hovrever, to be conservative, allowances for incrzeases in size and weight were made and are shown in the figures that appear in Table 2. 5. 1. The efficiency specified is 42%, equal to the guaranteed efficiency of the WJ-3951 (although i typical efficiency is quoted as 47%0). The voltage and current i puts,uin the TWT were obtained by interpolation between the WJ-395-1 (100wj and the WJ-448-1 (50w). Figi re 2. 6.! is a schematle diag amn of the TWT wnrhich shows the connections of the various inputso To tu rn the TWT on, the heater voltage must be applied abo tt wo minu.tes ahead of the other voltages. All voltages must be cut off within a period of tkhe order of 0. 1 mse( if the helix current exc eeds its allovable limit. 1, $

RF IN RF OUT 0.46mA l2mA 89mA ANODE - 1HELIX 12.5mA HEATER | | | COLLECTOR 3.4 voltts volts 0.9A l l | 101.5mmA 101.5mA FIGURE 2.6-1 THE 85 WATT T.WL.T. (operating at 81 watts) 16

2, 7 GROUND RECEIVER EQUIPMENT At this time off the,,.el reaceivers p2 not exist for our system. Information from GE TVBS and TRW TVBS indicates that an adaptor could be built for less than $120 (at a production volume of 105). The basic taten-a employed for reception of the satellite TV broadc.st is a 5. 21 diameter wire mesh para bolic dish, In some fringe areas (section 2, 3) a 6. 85I dish is required. A l ~'mesh should be sufficient at 800 MHz. The dish must be mounted to withstaznd 125 mph winds. Accumulation of snow on the di'sh could be brushed off. A -tentative scheme of the electronics of the adaptor is shown in Figure 20 7. 1. This scheme employs a preamplifier at the antenna. This preamplifier must have a 'nc:ise figure of no more than 4.9dB to maintain the syst-em noise estimate assLmed in the downlink ealculation (Table 2. 3. 1). It also must have enough gain to fix the C/N ratio. The rest of the electronics translates the signal to an inl ermediate frequency, where it is further amplified. Finally the signal is demodulated, remodulated AM, and recombined vwith the audio signal in the usual format of AM-VSB. The demodulation and remodulation scheme shown in Figure 20 7. 1 is the inverse of the one to be used in the uplink stations in which the audio signal frequency modulates the video signal which in turn modulates the carrier. A block diagram of this process's is shown in Figure 2. 7. 2. The adaptar described in Figure 2. 7. 1 assumes a common intermediate frequency for all three satellite channels. To switch from one channel to another one chainges the frequency' of the local oscillator, This could be achieved inx a crystal conitrolled, oscillator by shifting to another crystal. A three-way switch on the adapter could be used for this purpose. Another switch should allow the custro)rner to change between the adapter and an ordinary antenna for terrestrial TV broadcast. In this scheme the same receiver channel is u ed to rece ive o-ry cne of the three satellite channels. 5 An audience of 10 is planned. The cost of the adapt.r therefore appears in the cost of the system mu tiplied by 105 The need to keep the cost of the adapt r low is further emphasized by the consideration that a high cost to be palid by the individual user is sure to drive the audience far below 10- and make a true broadcast system impossible. Accurate estimates of the price of the adapter, once it is in demand in large quantities, are hard to make. Data from GE TVBS leads to the folloPwing co st breakdow-n: _,~tenna $22 Ele ctonics $ 8. 15 (noise figure of 3. OdB) 'Iotal $90. 17

F ---........... - _ INDIOOR UNIT|| Antenna T o TV Switch Recei Antenna, I?T Signa~l over Supply ISitch IAud. io RF...ubr r.rier | Combiner I -1?1 1 1 | |^r?,nsl rter j re~ I I,' 111 Outd oo Unit _ --—, ---Iier |.i:ffi~ie) - -7,,, _lT-eo eI~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ I IIF |Video I [r Dilifier Limiter criminator Am-Pifier I _ _, __ _. I -, - -' - - -o - - - -

SIGNcAL M ODULATOR AUDIO MODULJATING SI GAL AMPLIFIER FIGURE 2.7.2: Sign:al proce&sin-r iln u )l.ink stations. 19

This figure is probably an underesthima-e. TRW TVBAquotes a comparable figure except that ~e. nioise figure is 4. 05 dB, w-hich is stB l sufFicient for the prese nt y-stem. A mrore recenttGE study (relating to S-band freluenoies proposes an adaptor system using a low noise IF amplifier and no preamplifier; a detailed cost breakdown leads to a price of $27. 50 for the electronics. That system involves a 6 dB loss in the mixer at the front end and consequently can not be used. Consequently the low noise IF amplifier is also nco,- needed. The above survey tends to indicate that considerable uncertainty exists concerning the coslt of low' noise electronicso In the face of this uncertainty, $38 was allovwed for the cost of electronics. The receiver adapter cost breakdown is specified as follows: At enna $22 Installation $60 Electroni Lcs (>F of -o9 dB) $38 To]al $120. The cost of the larger 6. 857 a ntena,. that is needed in fringe areas is obtained from GE parametric data as $40, The remainder of the unit cost stays the same so the total cost for.he larger antenna is $138. 2. 8 TELEMETRY AND COMMAND The function of the tetdJmcetry system is monitoring and command of the satellite. TIeh s tem mvolives no technological advance over previous telemetry systems t uses PSK od cdation, The ground antenna is a I0O parabolic reflector located at the central uplink station in Fairbanks. The grouend transmitter powver is 50w and the satellite transmitter putrs oult ).V;7' (ohf wrhich 0. 1w is dissipated in the diplexer). The sa tellite antenrna system ms1 t 1 e isotropic and must have the uplink station in ei v regardless of the orientation of the satellite. Its gain correspondingly is 0dB. A prc.bability of error not to exceed 10-5 on the downlink and 106 6 ong the uplink is spe cififed. All this data together with the estimrate s of lo sses and of no s e i n Appendi e s A. 3 and A. 4 go e s into the telemetry uplinik and dowunllunk ceaulati1ons in Tables 2. 8. 1 and 2. 8. 2. The link calculations in turn determine;the allowable bit rates as 2800 bits per second On the downlink an.d 42300 bps on the uplink. The last figure is the link capability; the actual bit rate to be used will be limited by the capability of the command decoder. A black diagramn of the te zlemetry system is s-hown in Figure 2. 8. 1. The prime c.ornamutator passues 640 bits p'r cycle. This represents 32 wo.rds of 20 bits each. At the down-link bit rate of 2800 bps t;his corresponds to a commut.ator samplng,ra+e of o4. 4 cycles per sec. Twenty-four channels are fed direcly lintS;o the prime commutaltor 203

DATA ANTENNA SUI B-I' SUB- 0 COTHl}JTATOR ' -\ t RFigure 2.8. 1 Telemetry System Block Diagram S-BAND DZSTRZBETO R DECODi TR AMIST7IT E 'R STOPCOGE Figure 2.8 1 Telemetry SYStem Block Diagram

from the sensing equipment, one is a clock reading, 2 channels come from the two subcommutators, and the remaining 5 channels are reserved for command verification. Inrformatlion that is monitored through the subcommutators is transmitted only once in every 6/4 cycles of the prime commutator, i. e. about once in 15 seconds. The different functions are identified by their location in the cycle. Each cycle starts wvith a code word. The full 20 bit word.- corresponding to an accuracy of 5 decimal digits is necessary for only a few functions. In the case of others different functions share the same word. The list of functions monitored and their tentative arrangement into words and assignment to the prime commutator or the subcommutators is given in Appendix A. 7. Transmission of a command word enables the command distributor. The code for execution is transmitted only after the original command is correctly repeated. The system is capable of distinguishing 128 different commands. The allocation of these is given in Appendix A. 8. A list of the components of the telemetry system including breakdown of weight and power is given in Table 2. 8. 3. During launch and prior to arrival on station, the telemetry system can either be fully powered, or have the command receiver powered with the telemetry system shut down. Table 2. 8. 1 Telemetry Uplink Calculation (2. 2 GHz) Trans. Antenna Gain 4B 34.4 Rec. Antenna Gain dB 0. Free Space Attenuation dB -191.3 Propagation Loss dB -1. 7 Coupling Loss dB -1.3 Power ratio -159.9 Transmitter power 50 watts 17 dBw Received power 17 6 -142.9 E/No requirement (10 error) 13. 0 Allowable noise - 155.9 dBw Noise temperature 5040K Bit rate 42300 b. p. s. 22

Table 2. 8. 2 Telemetry Downlink Calculation (2. 3 GHz) Trans. Antenna Gain 0 db Reco Antenna Gain -34. 7 db Free Space Attenuation 1,9.1.. 8 db Loss 1. 7 db Coupling Loss 1. 0 db Power Ratio 159. 8 db Transmitter RF Power -. 5 dbW Received RF Pover -160. 3 dbW E/No Requiremen7(10 -5 error) 11.8 db Allowable Noise -172. 1 dbW Noise Temperature 273~K The maximum bit rate for the allow able noise is 2800 bps Table 2. 8. 3 Telemetry System Components Equipment Power Weight Volume Heat (Watts) (lb s.) (ins. ) (Watt Subcommutator 1. 0 ea 1. 0 ea 3x2x2 ea 1. 0 ea (2) Commutator.5. 5 2x2xl. 5 Encoder 2. 0. O0 4x3x2 2.0 Decoder " 2. 0 2A 0 8x4x. 5 2. Sto rage Diplexe -.o 5 4x3x. 5. 1 Transmitter"!0. 0 o 8 3x2x3 9. 0 Receiver' 3.5 4.0 7x5x3 3.5 Command 3.0 2. 0 8x4xl 3.0 Distributor' Clock 2.0 2. 0 5x4x2 2.0 Demodulator. 5 1 o 0 2x2xl. 5 Cable & Misc. - 0 - Oscillator 1o 5 1o 0 2x2xl 1.5 Antenna 4.0 - Tota l Weight 32. 1 bs Total Operational Power 27. 0 Watts Tetal Heat 26.![ Watts *6 Two carried for redunda. ncy23

2. 9 REFERENCES 1. Television Broadcast Satellite (T/BS) Study, TVBS Technical Report, Volume III prepared for NASA unlder cox:tract NAS3-9708 by the General Electric Company Space Systems Orga.nization, Valley Forge Space Center, Section 4. 2. Chapter 7 of Reference 1. 3. Reference Data for Radio Engineers" ITdT 1956, Chapter 1. 4. Catalog, Alpha Wire Corporation, New York, N. Y. 5. "Microwave Devices - Caalog No. one", Watkins Johnson Company. "Watkins Johnson Elec.tron Devices Catalog", September 1969. 6. Herman Jankoowski, "High Power Transmitters for Space", AIAA Paper No. 70-4;36 p. 7. 7. Robert S. Hughes, "Spacecraft S-band 10-100W RF amplifiers tubes" AIAA Paper No. 70-506. 8. Robert G, Clabaugh and Capt. Everrett B. Richardson, "Large Population Orbital Experience with Long-Life Travelling Wave Tubes", AIAA Paper No. 70-507. 9. Hughes Electron Dynamics Division Catalog. 10. Catalog of the Alto Scientific Company, Palo Alto, Calif. 11. Section 4. 2 of Reference 1. 12. J.O an.ser, P. L. Jordan et al., "Television Broadcast Satellite Study", prepared for NASA under c(.ntract NAS3-9707 by the TRW Systems Group. Section 10. 7 13. "Ground Signal Processing Systems, Summary Reports on Analysis, Design, and Cost Est-imating", prepared for NASA under contract NAS3-11520 by the General Electric Compainy Space Systems Organization, Valley Forge Space Center and the General Electric Company Electronics Laboratory, Syracuse, N. Y., Section 6O 14. Project MISSAC, Department of Aerospace Engineering, University of Michigan, Winter Term, 1968. 15. Project SCANNAR, Department of Aerospace Engineering, University of Michigan, Winter Term, 1970. 16. "Telemetry, Command, and Tracking Subsystem", General Electric Company, DBS-MS-02, Novem.'ber 1966. 17. Myron H. Nichols and Lawrence L. Rauch, "Telemetry" CICE 520 course notes, 1966, Figure 6. 1. 1.

SATELLIT E POWER SUPPLY 3. t INTROD UC TION To provide the necessary 725 wat's of electrical power for a five year lifetime, various pow,7er supply methods were investigated. Because of the long duration of the mission, fuel cells and complete battery power were rejected. Extr"eme weight wo.ould also be encountered using a batteryonly system. ARadiois~ostope Thermoelectric 'Generator was also conrsidered, but it poses the problems of weight, cost, and radiation hazards. The system chosen provides comrplete poewer for MEDUSA over the five year lifetime of the satelliteo It consists of: 2, Two 2. 80 ft by 32. 71 ft solar arrays 2) Tvo nickel cadmium batteries. The solar panels provide a maximum of 1545 watts of power at the beginning of the life and supply the required 725 watts at the end of life. (This assumes degradatlion of 530/o including a maximum angle of incidence of 31. 60)` Normal broadcast time will be from 8:00 am until 12:45 am. From 8-10 pm until beyond the end of broadcasting at 12:45 am, Alaska time, a portion of the solar array is shadowed by the 30 ft reflector. During the fourth year of operatio n, tshe shadowing effects and panel degradation com-bine to reduce the panel output to below 725 watts. Since it is desired to broadcast from 8:10 prn to 12:45 alm over the entire lifetime of the satellite, supplemertary battley power is required. Also taken into consideration wasthe fact.that the satellite would encountelr eclipse periods of up to 1. 2 hours duration. W7ith this in mi-rnd.i, arel cadmium batteries were sized to provide supplemerntary povwer for thrree channel capacity at the end of five year-s and also provide sufficient satellite sysltels power during a time of maximum eclipse. The maximum powver ever rerquired from the batteries is 5133' watt-hours. This is more than is required duri.ng the lau.nch to orbit phase which is before the solar panels are deployed~. Because of this fact and the problem of limited body area on the satellite@ i1t wves de'ided to reject the use of body-mounted cells for powzer during the laur:ch phase. Thus, MEDUSA is equipped viith a power system -which will provide all necessary po'wer for the envisioned five-year lifetime, A complete breakdown of power required by various subsystems of the satellite is provided in Table 3. 2. 1i See Appendix B. o See Figure 3a 1. 1.

FIGURE 3.1.1 POWER CURVE 156'00 1545; WAITS BOL 1400 1300(~ 1200 1000 600 N800 725 TTQ[SCw5;- YEAL~RS 700 OPICRATINGwca P OWIMs~ 725;"WATO;~s BATTERY~P~ 600 50 BATTERY CHARGE 40 MAXIMM 347.5 Watts UMBRA 300 1.2 HOURS 0NON BROADCAST MODE 200 100 75.8 WATTS 1RkYXRY 0

Table 3.. 1. Power Required Launch Eclipse Ombit Communications 0. 0 0. 0 585. 1 Attitude Cont-trol O., 0 4. 8 44~ 8 Telemetry and Command 27. 0 27. 0 27. Q Power Conditioning 2 0 7. 0 50. 0 Array Drive 0. 0 0. 0 8. 1 Battery Charging 0.0 0. 0 10.0 29 l0 75. 8 725.0 3. 2 SOELAR ARRAYS For ease of storage before deplcyemnt and lightweight qualitites, a roll up design was chosen for the satelliite's arrays. A rigidly mounted solar panel was considered for the satelli<e, bu5at such an array could not supply the power required for the durationn of 'the mission. The array system consi.sts of two 2. 80 ft by 32. 71 ft pan:ls (dirre nsirons include extendable booms and endpieces). Twvo feet of dummy array are utilized to extend the panels as far as possible beyond the outer edge of the reflector. The dimensions cited give an aspect ratio of 11. 68, The panels and drums are rotated at the rate of one revolution per day so that the panel face w-7,illl follow the sun for full illumination. Sun_ sensors guide the panel for orientation, To convert solar energy into use2ul electrical power, 2 cm by 2 cm N-on-P silicon solar cells of 8 mnil 'Ehickn4ess were chosen. A 6 mil quartz cover glass is provided fo.r cell prc'tect',~ion from radiation damage. Using this coverglass dictates a pan el ternperature of 1 l5s Farenheit. To achieve the desired 28 volts, 68 cells are placed -i'n.eries, For 1545 watts at the beginning of life, 456 cells are required in parallel* The panel width requires that the 68 cell strings be split in half, vwith 34A cells across the array. Two half strings are series connected to provide 28 volts, Wider panels, not requ'iring the splitting of strings, - w>ere prohibited by structural dimensions. The 456 strings are divided into 76. six st'-riag modules, wvith 38 per panel. (Diodes are connected in series with each cilar cell string to prevent current reversal. This will protect all cells f:om danage during reflector shadowing. )A diagram of a solar panel is given in Figure 3. 2 1o For roll up purposes, an 8" dia.m'cter al uminum drum was selected. The solar cells are mounted on flie:Sibal KIapton substrate with RTV 577 adhesive. A lightweight foam cushion is wrapped auround the cells for protection during storage and deploymnent. This cushion is removed from the panel surface during See Appendix B. 1. 2:7

DRUM ASSEMBLY: PARTIALLY DEPLOYED One Module of 38 on Each Panel Drum As sembly Drive Shaft Drive Shaft FIGURE 3.2.1 28

deployment via a take-ulp roller. For,:_-., ' the DeHavilland BISTEM tubular boom extension rnechanism ics,uii-azed. The booms are wvrapped on the drum and unwind as the pancels are d-ployed. A small mottor placed inside each drum, and connected by spur gea's and a shafIt to the stems, drives the booms out to deploy the array. The arrays are extended at a rate of four feet per minute, To provide structural support for the arrays, the drums are mounted on bulkheads at Lhe ends~ The bulkheads are co-rn.20:ected by U-bars and, with the use of a mount-ing plate, are attached to the rotat'inag shaft in the satelline body. Power is rem9oved frsomX th pt a le by means of a wirDing harness and transferred to the satellite powze cond-utioing qu ipmeant across slip rings. The harness is coiled inside the -drum and nrinds as the arrays are deployed. Because the arrays rotate to followv/ the su.n., some reVolving mechanism for power transfer is required.~ Slip rings vere chosen, since it was felt the rings would be more reliable than a twistling c(able sys-tem wi;hich would require a daily reorientation of the panels to unw.gind the cable. Daily reorientation would require that large amounts of anrgular ornrioent.,ui be renmoved via the momentum wheels. TwO l- irng assemblies are specified so -hat even at the begnni ng of life only slightly more tha-n two armperes w-illl cross an individual ring. Since the current across a ring is loxw, light brush pressure can be used. Because of the low rotation rate and low brush pressuTre, slip ring wear should be minimal and not pose any problems. 3.3 BATTERY SYSTEM The batteries. were sized t fulfill the following requirements: 1) Launch to orbit power 2) Supplementa ry powver at t-ir nes of shadowing during the final twvo years of operation (see Appendix B. 3). 3) Power during periods of eeclipse~ (T and C and AC) Silver zinc, silver cadmium, and nickel cadrnum bateeries were considered. Silver zinc batteries were rejected du'.e to lack of cyclic ability. Silver cadmium batteries, aLthough having a high pW7e r density, w7 Vere rejected due to their short life relative to nickel cadmiumn batteries. Thus, nickel cadmium batteries were chosen as the source of secoindary po1v r "sor their high reliability and cyclic capability. To fulfill the requiremeent.-s of -LChe mr~iission, a toa.l battery capacity- of 30 ampere-hours was fouwrd to b necessa r y~ I't rwas decided to provide this ~ 9capacity wi4ith wio i7 I5 aAm T pe- ve-hou% battre-ies -%on,-n c)tIed -inE par'alel. Each battery has its own casing, heates, an.dr charge control. This system wvaSas selected to prov-de rsedundancy in the h ev~ent that one battrey should ftil. If this should occur the pir-Rary rnin:Ez.)micn ob3ectiv- s s eou]Lld siEll be fulfilled. 29

Complete power for the eciipse pe'l, i 1KilI b p To, vided by the other urnit. During the lanunch to or1bit period the b M; ieR1s will! be. discharged ts 44% of their total capacity. This correspo)nds to:~ -68 watt-hours which includes deployment. ILna the first tc third year of opetraioin, the batteries will be used only during eclipse and will not exceed i0% depth of discharge. In the third to fifth year of operation the batte ries will be increasingly called upon tco supplement the solar array durlng periods of shadowing. At the end of five years during shadowing and the ~w)orst p ossible eclipse, the battenie s -wil not exceed 64%0 depth of discharge. This depth of discharge is not critical since this is at the end of life of the satellite and for the first three years of operation fle batteries are discharged only slightly. Because of the large amount of power required from the batteries during retlec1tcr shadowing, there is iot complete redundancy provided for this period. However, it was felt that the penalty incurred by dropping one or twJo telev;ision channe ls for the shadowed pertod should one battery fail was not so great as to require a totally redundant battery system with a larger weight than the system specified. Each battery ccansists csf 24 celsx in series, each with a nominal voltage of 1.2 volts and a 15 ampere-hour capacity. This scheme provides the 30 ampere-hour capacity at 28 volts vhich is in the voltage range required by most of the satellite subsystems. This also nmatches the voltage at the solar array bus. The batteries are to be charged during perrods of no TV transmission, namely the early morne ing hours bet-weenz 2 am and 8 am, Alaska; t&ime. During this time they will be charged at the c/8 rate. Battery voltage, temperature, and pressure will be monitored to prominde rel iability. If the temperature or pressure becomes toos high, battery charging will ' -c re(duced to the c/20 rate urtil such time as these values return to -within operationall tolerances. After the batteries are fully charged, they will be trickle charged for the rest of the day to maintain their capacity. Each string of the psolar arraeny is plaed in series with a diode to prevent discharge of the batteries into vtae soeatr array during shadowing and eclipse. Both a diode and a gat aee re placed s seres Nrith the negative lead of each batte y. This is done to prevent discharge -f the baLezry during handling and to provide the ability to isolatle each ba-'e. tier in case of failure of one of them. These dioedes do rot interfere i-ith h it ',r-Lr charging by the solar arrPays or by ground personnel before laurnch. 3.4 POWER: DISTRIBUTIONii A block diagranm of the povier s ystms is given in Figure 3. 5. 1. The pver distributer routes ppimary7 poier frorom the arrays to the low and high vsltage regulators. The lowv voltage regulator con rols the anregulatedL ar ay output and furnishes th. $at' SllSiysusy stemswth 28 volts de. The high voltage See Appendix B. 2.

HIGH ICATHODE VOLTAGE --- | ANODE q(i) SOLAR ARRAY REGULATOR HELIX #1 COLLECTOR SOLAR ARRAY LOW #2 L #2.~~~~ r |VOLTAGE l.. REGULATOR MODE SHUNT POWER -r LOAD CONTROL OUTPUT SUBSYSTEM (28 volts) DISTRIBUTION CHARGE CONTROLLER BATTERY #2 # 2 CHARGE CONTROLLER FIGURE 3,5.1 31

r egulator provides the comm. un icaic,,:!~,s eq ipmnent with the voltages required. Theo T W;T requires inhputs of 0. '46 rnA an t 990 v, 12. 5' m-A at 56-0 v,, 101:,. 5 mA at 1800 v,: and 0. 9A at 3.4v (F.igure 2.,6. 1).. Both regulaiors co::treol the battery as well as solar array power and convert it to the required voltages. (Battery voltage may vary with temperature and depth of discharge, but the regulators will compensate for this effect. ) The high voltage regulat'of ias locazed on the fee, as near the point of use as possible to prevenzt high voltage breakdowns. (A current limiter is included in the circzuit to automatically disconnect& a T WT should a failure occur. ) Before complete degradation of the arrays occu-rq excess power will be delivered by the arrays. This extra power will be "dumped" t1hrough the shunt resistors under control of the power distribution unit. The mode control governs use of the batteries and receives commands from the power distribution unit. It regulates charging or discharging of the batteries. During the on-station period of orbit, before the system requires supplementary battery power due. to reflector shadowing, the normal cycle of operation will be as follows. F n m 8o:00 am to 2:45 am, Alaska: time, 725 watts of power will be supplied b1y the solar arrays for satellite operation and broadcasting. After broadcasting, an eclipse period may occur, during which all system pcwer is supplied by battery, The power distribution unit -will sense a drop in panel voltage, and comsmand the mode control to discharge the batteries. After the eclipse has passed, s 'he pow-er distributor senses a resurgence of array vsoltage and the batternes will be signaled to cease discharging. After eclipse, the arrays are again illuminated. Batteries will be charged at the c/8 rate until they are recharged. At 8:00 am broadcast resumes, with the arrays providing full power. During the communications period the batteries are trickle charged. During the fourth year of operation, when reflector shadowing prevents the arrays from delivering the required 725,watts, supplementary power will be drawn from the batteries. Durifng the daily shadowing period, the power distribution unit-i"will detect the armount of battery po!wer required and regulate battery discharge accordingly, (see Figure 3. I. 1). 3.5 WEIGHT BREAKDOWN Solar panel area (booms, spar, durnmy array) (30. 71 ft + 2 ft) 2.80 I't + (30. 7 f1 t + 2 t) 2. 80 ft - 91. 6 ft2 + 91.6 fit2 183. 2 ft2 Effective cell area (30.5 ft) ~ (2. 33 ft) + (30.5 i) t (2, 33 ft) = 711 ftZ + 71.1 e t2 = 142. 2 ft2 t32

Aspect Ratio of array 11.68 Solar cell weight (grams/cell): 8 mil cell -;/solder. 350 6 rnil cover glass. 130 adhesive 100 R. To Vo. 075 Connection Tabs. 071 726 grams/cell Number of Cells: 68 cell (series) o 456 cells (parallel) = 319008 cells Total Weight of Cells (lbmn) (3T, 008 cells) (o 726 grams/cell) 1 lbm/454 grams = —49.6 bm Substrate Weight (2. 58 ft wide) 2 (.0425 lbm/ft2) (167. 6 ft) = 7, 1 lbm Cushion Weight = 3.5 Ibm Solar Cell Weight: (ibm) cells 49, 6 substrate 7. 1 cushion 3. 5 60. 2 Ibm Drum Assembly Weight Breakdown (each): (ibm) Drum (40% holes) i.6, U bars 2.0 Moto r 1. 0 Boom (2) 11.5 Spar 0. 9 Take up Rolle-r 2.5 Bulkhead 6. 0 Bearings 2.5 Transfer Shaft 2. 5 30.5 lbm Total for Satellite (2 Drum Assemrblie s) z= 6.1'0 lbm 33

Drive System: (ibm) Slip Rings (2) 8. 0 Drive Molor 3. 0 Shaft 4. 0 Drum Plate 3. 0 Bearings (2) 4. 0 Wire s 3.0 O 25.0 ibm Power Conditioning equipment weights: Transformer 1 0. 0 Regulator 3. 9 Battery Charger (2) 4. 0 Power Control 5.0 Shunt Resis"o r 2. 0 24. 9 ibm Battery (30 amp- hr) = 87.4 Ibm Total Power System Weights: (lbf) Deployment 61.. 0 Cells 60. 2 Drive Assembly 25. 0 P. Ca U. 24. 9 Battery 8 7.4 258.5' ibm Table 3. 4. System Parameters System Watts/Pound = 5. 98 watts/pound Array Watts/Pound - 27. Z watts/pound Array Watts/Square Foot = 10o9 watts/square foot 3. 6 REFERENCES Airflight Company, -Basic Slip Ring Design. Berry, L. B., and W. D, Browln and W. P. Dawson, Flexible Integrated Solar Array, Air Force Co-tract AF 33(615)-2750, August 1967. General Electric Company, General Electric Hermetically Sealed Nickel-Cadmiurn Aerospace Cells, Technical Report 4-66(3M)6320. Gibson, Richard, DeignData fo_ SpaycteSyt, Bendix Aerospace Div. Feb. 197 Ralph, E. L., Performance of Very Thin Silica Solar Cells, Heliotek Company, Research Paper Bi18A, March 1967. 34

Ryan Aeronautical Company, Design, Fabrication and Demonstration of a Deployable Large Area Solar Array Structure for Interplanetary Space Probes, March 1967. Ryan Aeronautical Company, Roll Up Solar Array, Technical Report 5856, August 1968. Sonotone Corporation, Sonotone Battery Instruction Manual, Technical Report BA-89, 1962. Robdt and Daspet, Photovoltaic Devices and Systems Abbot Company, Power Supply Catalog 1971, 1971. 35

ATTIT. UDE CONTROL 4. 1 INTRODUCTION The purpose of an attiltude control system is to maintain a predescribed orientation for the satellite. The need for such a system is very important. It is an essential system for a high pover ed communications satellite, for it maintains an attitude so an efficient directional reflector can be used, and the needed solar panels can be pointed towiward the sun. This system is also needed on this satellite for the purpose of guiding itself to and maintaining the geostationary orbit. This orbit cannot be attained by the individual stages of the launch vehicle alone. The satellite's attitude control and propulsion systems must accomplish the final rendezvous phase to the intended position of orbit. While on station over a period of time the satellite will drift in an eastzard direction. These stationkeeping problems must be corrected by the combin ed attitude control and propulsion systems. 4. 2 DEFINITION OF AXES Before a detailed description of the system is discussed, a reference cooraiiiate system should be defined. For this satellite there will be two sets of coordinates; one for the purpose of defining lhe moments of inertia (body fixed) and the other for the purpose of calcTulating disturbances and describing the control system (earth fixed). The body axis is a rightl handed (o_9ordinrate system and is defilned as follows: the y-axis runs along the axis of the solar panels, and the z-axis is the axis of symmetry of the reflecttor. This system has the minimum moment of inertia on the z-axis, aind the maximum moment of inertia on the x-axis. The earth-fixed coordinate systefm tdesc ribing the attitude of the satellite) is as follows: the yaw axis is desc(ribed as the local vertical to the earth, the roll axis is along the direction of flight, and the pitch axis is the local horizontal, in the nsorth-south dixretion {see Figure.4. 2. 1). These two systems are nearly coinzcident, except for the fact that the satellite is pointed at Alaska. This meawns the satellite is rotated rnorth 8. 10 about the x-axis, and from the saltell'.te s position of 1700 west lo.ngitude, it is rotated 3. 500 east toward Alaska.

3.50 YAW X;' ROLL z ( Direction of Fllght3) ~ W~~~~ i; ~~ ~(Local.4~~~~~ 1 u IVertical).CH Y * y Pigure gol Definition of Axes 37'

4.3 SYSTEM REQUIREUIREMENTS The system is designed to meelt several requirements. First, the b-est po:ssible pointing accu.racy- was desi:ra. ix e fo:r the comnmunications system. They are-then de'fined to be + 0. IC) on, the, pitch axis, + 0.. 2 on the roll axis, and + 1.:00 on yaw axis. Second', the syslern is to provide the solar array drive mortoor with pointing L.:n'%ormation. Third, the systerh is to provide the propulsion sysem writ;h data as to when and how to fire its thrusters. Lastly, for a satellite of this size, the system should be simple, lightweight, low in power consumptfion. aid above all, reliable. Four different systems -were considered. The following is that analysis. 1. Gravity-gradient systerms a$re Vey reliable since they are completely passive and requi re no s wnsrs or elecl-t ronic s, but the accuracy of these systems iis notI good enough to meet the system requirements. 2. Spin-stabilized systems are acorurate enough and are fairly simple. Our satellite requires that it have solar arrays mounted out from the body, so this system is not physically possible. 3. A mass expulsion systme has good accuracy, but with a satellite that has a lifetime of 5 years the fuel loads needed eliminates this system. 4. Three-axis stabilization with nmomerturn wheels has good accuracy and low fuel loads. The onlv limita2ion_ involved is the sensor accuracy, but this is good enough to meet the satellite's requirements On the basis of this arialysis the three-axis control system has been determined to be the best system for tghe satelli"te. 4.4 SYSTEM DESCRI:PTION In this active, three-asis attitude co:troll system, the sensors, nmomentun wheels, gyros, computer, and propulsiorn systerm work together to maintain a predescribed orientatiLon with the earth. This is done by sensing and couLnteracting any disturbing force or torque t'hat acts o n the satellite. The sensing for the initial ofien~tetioz o: the satellite will be provided by the sun sensors. These sensors have a cornplete field of view, and will always be able to take a fix on the sun.o The on station sensing will be provided by a synchronous altilude hosizon sensor. i't w'ill supply all of the pitch and roll information, while the s~ sensors w:vill supply the yaw data.

There is a rate gyro package that.will supply data on the rates of deviation. The system also includes a r.ae.-integrating gyro for the yaw axis, which augments the su, s nso s nO d takes over completely when there is a solar eclipse. All signals fromn the sensors a-d gyros vwill be processed by the onboard computer. The comnputr will then send an appr;priate signal to the momentum wheels and/or the propulsion system. There are three nmomentum whneels, one mnounted along each of the yaw, roll and pitch axes. The momentum wheels maint&ain an attitude by transferring momentum to the spacecraft2 This is don)e (a-rhen the spacecraft is disrupted for some reason), by changing the speed of the twheel. All of this is controlled by the computer. When the speed of the vwheel reaches its maximum, an appropriate signal is sent to the propuision syste em, -n^hiich fires its thrusters in such a way as to unload the wheel. The propulsion system per foreis several f anctions. It serves to place the satellite in the proper orbit, o rentl the satellite, unload the momen.tu;m _Sheels, and to function for stationkeeping. A block diagram of the system is show-n in Figure 4. 4. 1. 4.5 ATTITUDE SENSING To maintain tb~desi>rd. orienitatiion for a spacecraf't, information about its attitude and any deviation from tha.t said at;titude matist be supplied to the system. In this system visual and inf reds ensors, and gyros will supply this needed info rmation. There wi]l be two different systemns of sensors; earth and sun sensors. The sun sensor system, will provide complete spherical coverage. This complete coverage could be obtained with five senrsocrs, buu beeaus of placement. and shadowing effects this system will obtain athe spherical coverage with six sensors (see Figure 4. 5. o ). Each onFe of these sensors has a field of view of 1280 x 1280 and because of their p1ace iefl they overlap each other's coverage area. Since osnly one of the senzsors will be ~wtorkirig:at any one time, the included amplifiers and storage registers ar e shared by all of -he sensors, thus keeping the total weight and powiez of the system to- a minimum. During the rendezvous phase of thhe flight bhese sun sensors will provide all the attitude informationo They iw;l also help orient the satellite as it proceeds to station. The su'. sensors, except- during the _ eAipse, will always be able~ sight the sun, so att-i"ude infocrrna.tion wzill alw -ays be available. During the eclipse perliod a rateiuntegraU riag gyr o; about the yaw axi s wil3l provide all the needed information~ This gyro vi.ll also compliment the system by giving it greater accuracy and a redi0ndancy factsoro 39

Solar Array Drive. Motor Data Sensors Sensor t Pi o Electron Wheel Wheel Wheel ~TLM. | e | C XT CU,.Earth D' e n A ens or r t ropulsion a r S J... t o TL TLM o an Cm d n e T ~ P r Pc a r hD C R R ac tt ystee TLMet e r 1 r I Gyro Computer TL T Be ToM A Compensation and Override Commands ' Override Thruster Commands C Primary Yaw, Secondary Pitch & Roll Data D.' Primary Pitch & Holl Data E Pitch, Roll & Yaw Rates F Seeondary Yaw Data G: Uncage H Drift Compensation I Momentum Wheel Drive J Tachometer TLM Telemetry 4O

When on station the sun s-risosrs v-~i1L prcvide the system with all the yaw attitude information, anrd vill provide s pplemerz.tary pitch and roll attitude info rmation. The major contributtor of pit;ch and roll Informat-inc: whe. on station will be the infrared synchronous orbit horizon sensor. It will always view the earth and has a high degree of accuracy~. (Th:is is shown in Figure 4. 5. 2. ) There is also included in the systern a three-axis rate gyro package. These are primarily used to obtain the raatLes of the de lations and to supply this information to the computer, They also improve the accuracy of the system, along with decreasin-g the steady statLe eor and increasing the total response of the system. 4. 6 ONBOARD COMPUTER All sensor data, ground commrrrands, anad other input data must be processed in a logical manner so that proper acstion can be taken. This must be done by a computer. In order not t o overburden the telemetry system, an on-board computer is needed. The computer will process all digital and analog signals from both rate and position sensors. This will provide compensation, and a logic in sending the proper signals to the momentum wheels and control thrusters. The computer must also be able to receive commands from the ground, and execute them either immediately or have the ability to execute them when needed. This computer will have to be developed to satisfy the missions specific requirements. The present state of the art indicates that this will not pose any great difficulties. 4. 7 MOMENTUM WHEELS The momentum wheel is basic- ally a momtentum exchange device, Even though many of the satellite's disturbing tsorques are random in orientation aand magnitude, the wheels solve the problems and p6rovide ari accurpate atititude csontrel system. The wheels are fairly small, lcow in wvxeight arid consume very little electrical power when compared to other systems. The wheel itself is simply a mot-or z-ith a large inertia rotor. When s~ignaed'-, the motor accelerates or decJelerates the rotor. The resulting torque is transmitted through the unit to the vehicleIs structure. This torque counteracts the disturbing torque which is aceing orn the spacecraft. The wheel can supply additional torque until the motor reaches its maximum speed, whereupon the wvheel has reached the maximum momentum that it can store, The wheel rmust then be unloaded, where the speed of the

i sensor mounted on rront of reed 120~ 2 sensors mounted 3 sensors on back of antenna mounted 120o9 torus apart around the feed With these 6 sensors complete 3 B~ I spherical coverage is achieved. FigureS14 Sun Sensor Location & Coverage Pitch and Roll sensing provided by two pairs of tangent fields. Figure4.5.2 Earth Sensor Coverage 42

wheel is returned -to near zero. This u.loa.ding is accornplished by an opposite torquing effect provided by the propuilsion syst'em. The speeds and direction of the wheels are sensed by a ta.chometer, and this data is se'g~ to the computer, which takes actlion if andL d when the wheels need unroaditgg, The sizing of the wheel is a function of the disturbing torques' magnitude and frequency. It is essential, lhat if the torques are cyclic, the wheel be chosen so it does not saturate due to these torques. The sizing calculations, the calculation of the disLturbing toiques and wheel uiaoadings are found in Appendix C. 3, 4, and 5. 4.8 REDUNDANCY The telemetry uplink to the satellite will make possible the operation of the system if the computer fails to funcction. In this case all data from the sensors will be fed to the ground, and the appropriate commands will be fed up to the momentum wheels and/or thrusters, to make the necessary corrections. In case of earth sensor failure the sun sensors can provide limited pitch and roll information to the systemo The loss of a sun sensor will resuit in a larger' atitiude e&.r0r, but the yaw sensing could be performed by the rateintegrating gyro which is intended to do that job during solar eclipse. The loss of a rate gyro will result in a larger error also. Howvever, at synchronous orbit the encountered rates should be slcow enough for the position sensors and the computer to compensate for them. The reliability for some of the system components are: Momentum Wheels Life requirement: 5 years at T000 rpm Reliability:.999 for 3 years This would be reduced slightly for 5 years Rate Gyros Life Tim]e: Greater than 5 years Earth Sensor Reliabili~tyo High reliability components can be furnished with life 7 3 years, 5 years does not seem to pose any problems

4.9 PROPULSION SYSTEM 4. 9. 1 System Description MEDUSA's propulsion system will be used for rendezvo us.and. walking:o rbit rn-ia naarve r s stationkeeping, and momen-tum wvheel unaloadil g. A monopropellant system will be utilized, vwith hydrazite (N2H ) as the fuel and gaseous nitrogen (GN2) as a pressurant. 4. 9 2 Fuel Use The pre- station mraneua-vers will involve relatively large translatioe al velocities, so two ten pound thrusters vill be used for these maneuvers. For rotational maneuvers during acquLisition and s atlionkeeping, which includes iast-west stationkeeping and unloading the mome ntum wheels, sixteen 0. 1 pound thrusters will be used, Positioning of the thrusters is shown in Figure 4. 9. 1, followed by a table of maneuvers and the thrusters used for these maneuvers (Table 4. 9. 1). The ten pound thrusters will be fired in a continuous mode, giving a specific impulse of 210 sec. The 0. I pound thrusters will be operated in a pulsed mode, givilng an I of 150 seconds. sp Table 4. 9. 2 gives a fuel wneight a-nd impulse breakdown for the mission. A derivation of each of these figures ill be found in Appendix C. 9. There will be 13. 6'lbs of reserve fuel, w hich should be enough to take care of any unforseen maneuvers and emergency fuel uses that might arise. As can be seen from TS able 4. '9. 1. certain maneuvers have redundancy. The most critical maneuver is rotation about the pitch axis, where the highest pointing accuracy is required. Therefore, this maneuver has the greatest redundancy. The loss of a thruster for one of the other maneuvers which is not redundant will have harmful effecis on the mission. The system will either lose accuracy or wiil use mo.re fuel to correct for the loss, thus shortening the lifetime of the mission~ 4.9.3 Propellant and Hardware The hydrazine will be stored >n bladders in two cylindrical tanks, the remainder of each tanrik being filled wvith L-he nitrogen under pressure. The hydrazine will be fed to the thruster using a blowdown pressurized system. The hydrazine is then decomposed by a c atalystL into gases at 1800 degrnees F, which then are expelled, producing thrust. ~~~l~AT

150angle 0 1 109 4 36' l ~-~ x x XI roll roll roll yawZ -y pitc 48" -Y pitch 150 Side I5O~T-; lO 5 ~~~~10B3- l 301" IS1 Antenna End Burner II End Thrusters 2, 6, 10, 14 are inclined 150 to the horizontal for protection of the craft from the hot exhaust, Thrusters IOA and lOB are inclined 150 to the 4 1 1 horizontal in order to clear the z 1 10 Burner IL yaw -x 4 TP pitch Figure 4. 9. 1 Thrustex lcmn

Table 4. 9.1 Maneuvers and Thrusters Used (Refer to Figure 4. 9. 1 for thruster numbers) Total Maneuver Thrusters Thrust (lb s) Moment Arm (ft) ITorg ue (f- b s) + Y 3,7(11,15):" 0. 2000 - Y 4,8(12,16) 0. 2000 + X 1 (9)" 0. 1000 - X 5, (13)' 0. 1000 + Z 2, 6 0.1932 BOA, lOB 20.0000 - Z 10, 14 0.1932 + Pitch 2, 14 4. 502 0o4502 1, 13 1.75 0.1750 - Pitch 6, 10 4. 502 0.4502 5, 9 1.75 0.1750 + Roll 4-8,11-15 1.75 0.3500 - Roll 3-7,12-16 1.75 0. 3500 + Yaw 3, 8 4.00 0.4000 - Yaw 4, 7 4. 00 0.4000 These thrusters will be backup thrusters, not normaly used because of the undesirable side torques they produce. Table 4. 9. 2 Maneuver Impulse (lb- sec) Fuel Weight (lbs) Injection errors (AV=127 fps) 46:,I8 21. 99 Walking orbit (AV=20 fps) 730 3'. 460 E-W stationkeeping (AV= 16.9 fps/yr or AV= 84.:5 fps /5S yr s) 3 073 20. 420 Reaction wheel unloading 2. 7 0. 085 Thruster misalignment correction 66.5 0. 443 8500o 2: 46. 398 Reserve (fired in pulsed mode) 2040. 0 1L3. 602 TOTAL 1 0540-.2' 60. 000 46

N2 Fill and Drain Valve N2H4 H || | N2H4 10"? 2H4 Fill and Drain Valve Filter Latching Valve Pressure Drop Latching Valve Latching Valve I- Latching Valve s _ 10 lb Thrusters 0. 1 lb Thrusters Figure 4 9. 2 Propellant System 47

Each tfnk will have a volume of 1309 in3, of which the hydrazine will occupy 827 in, and the nitrogen 482 in3f The initial pressure in the tanks will be 250 psi, and the pressure will decrease as the hydrazine is expelled to a final pressure between 92 psi awad 107 psi, depending on whether or not the reserve fuel has been used up. As the pressure drops, the thrusters will have to be "'on" longer to make up for the lower thrust levels which result. This is done by the onboard computer, using a pressure gage to feed the pressure information to it. Latching valves are provided for stopping fuel leakage from a thruster. A schematic draywing of the propellant system is shown in Figure 4. 9. 2. Attitude Control System Weight and Power Breakdown Component WWeightr lbs Power watts 3 Momentum Wheels 30. 3 at 10. 1 ea. 13.8 at 4. 6 ea. I Earth Sensor 9. 0 3.5 6 Sun Sensors plus Electronics 8. 0 3. 0 3 Rate Gyros in Package 2. 6 10. 5 at 3. 5 ea. 1 Integrating Gyro 1o 0 3.0 1 Onboard Computer 4.0 5.0 16. 10 lb Thrusters 2. 24 6. 0 210 lb Thruster 1.4 Valves, Tanks, and Plumbing 23. 0 - Total Fuel 60. 0 - Totals 141.54 lbs 44.8 watts 4. 10 REFERENCES Beusch, J. U., "Three Axis Attitude Co..:rol of a Synchronous Communications Satellite", AIAA Paper 70-456, AIAA 3rd Communications Satellite Systems Conference; Los Angeles, Calif. April 6-8, 1970. Braga- Illa, Alvise A., "The. Fututre of Self-Contained Control of Synchronous Orbits", AIAA Paper 70-479, AIAA 3rd Communications Satellite Systems Conference. Cannon, Robert H., Jr., '"Some Basic Response Relations for Reactison-Wheel Attitude Control", ARS Journal, Janauary 1962, pp 61-74.

Dinter, Henry A., Inertial Sensors Theo,.y a —,d Application, Ho xeywvvel Report AM-62-2, July 1967. Froelich, Ronald W., and Pata.poff, Harry, Reaction Wheel Attitude Cont', rol for Space Vehicles, Space Technology 1aboratories, Inc., Los Angele-s, Calif. Greene, R. H., Ross, M. S. o and West, A. H., "An Efficiency Evaluation of the ATS-3 Hydrazine Orbit Control System", AIAA Paper No. 70-460. Greensite, A. L., "Analysis and Design of Space Vehicle Flight Control Systems", Vol. XI.I, Attitde Control in Spae, Geneal Dynamics/Convair Report GDC-DDE67-001., August 19670 Haloulakos, V. E., "Thrust a.d Impulse Requirements for Jet Attitude-Control Systems", Journal of Spacecraf, Vol. 1, No. 1, January 1964. MISSAC - Michigan Instructioanc al Satellite for South American Countries, Feasibility Design Repof t, The University of Michigan Department of Aerospace Engineering, April 1968. Sabroff, A. E., "Advanced Spacecraft Stabilization and Control Techniques", AIAA Paper 67-878, October 1967o Project SCANNAR-Satellite Communications and Aircraft Navigation for the North Atlantic Region, A Student Design Project, The University of Michigan Department of Aerospace Enginee-ring, April 1970. Sawyer, R. H., "Secondary Propulsion System Capabilities as Compared with Flight Control Requirements", AIAA Unmanned Spacecraft Meeting, Los Angeles, California, March 14, 1965. Singer, S. Fred, Torques and Attitude Sensing in Earth Satellites, Academic Press, New York, 1964. Trudeau, N. R., Sarles, F. W,, Jr., and Howland, B., "Visible Light Sensors for Circular, Near-Equatorial Orbits", AIAA Paper 70-477,;. April, 1970. U. S. Rocket Motors", Aviation Week aid Space Technology, Inventory and Forecast Issue, March 9, 1970. 49... f.......

THERMAL CON4TROL 5. 1 INTRODUCTION If no provisions were made for thermal control, a satellite in synchronou orbit wvould be subjected to temperature extremes well beyond the survival limit of its components as the sun would shine on the satellite from different angles, a as the satellite would pass through earl hls shado-w. A completely passive systemn is the most desirable because of its simplicity and inherent reliability. Because of MEDUSA nearly constant heat output, it w7as initially assumed that a totally passive system could be used. As the design progressed, however, it became apparent that the feed temperatur could not be passively controlled. Because of the high heat dissipation of the TWT's, the need to run any three of six tubes, and the need to keep weight as low as possible, it becamne necessary to use heat pipes (Figure 6. 5. 1 ) 5. 2 SATELLITE MAIN BODY 5. 2.1 Radiators The operation of the communicatiozns system requires three-axis control, and therefore, there will be two faces of the satellite body which will always face north and south (Figure 6. 2. 1). These surfaces are used as radiators since they will never face directly into the sun. All heat producing components within the body (except power shunt resistors) will be mounted directly on the radiators. Some difficulty was encountiered in calculating the radiator temperature because of the radiatioyzr coupliaig between the radiator, reflector, and solar array. Preliminary calculations (see Appendix D) show that the radiator temperature should range between 80 F and 8.60F, but a more detailed analysis should be carried out. See Table D. 2 for heat dissipated and temperature limits of all components. 5. 2. 2 Shunt Resistors Resistors will be placed on the east-vest ewalls of the body to dissipate excess power developed by the solar arrays. 5. 2, 3 -Insulation All internal surfaces, exscept "the.aosrth. south faces, will be covered by insulation to minimize heat input and radattion from these surfaces to the internal componePnts of the satellite. The i-sulation will be a blanket of 40 layers of Al/ mil M~ylar stacked at 60 ltaye s pesr inrch (Aluminrum side out).

5. 2.4 Thermal Coatings All external surfaces will be coated with a white paint, zinc oxide/ potassium silicate (Z-93), with the exception of the body-reflector interface. All uninsulated internal surfaces will be coated with a stable black paint (Cat-a-Lac) to help equalize internal temperatures. 5.3 SATELLITE FEED 5.3. 1 Radiators The entire outer structure of the feed will be used as radiator area (Figure 6. 5. 1)..1 Around the cylindrical portion are the 6 TWT's and the power conditioning unit. It proved necessary to use a heat pipe network to distribute the heat enough to allow any combination of three TWT's to run simultaneously. One heat pipe completely surrounds the cylinder, and six others run normal to it over the full length of the cylinder. The multiplexer, a transistor amplifier, and an earth sensor will be mounted on the end plate facing the earth. 5.3.2 Thermal Coatings The entire outer surface of the feed body is coated with Z-93 white paint to minimize solar absorption and provide good radiating qualities. All internal surfaces are coated with Cat -a-Lac black paint to minimize temperature variations among componeits. No insulation is used on the feed. With this configuration,9 the feed will have a maximum temperature of not more than 1030F and a minimum of not less than 180F as the satellite emerges from the maximum 1. 2 hour eclipse. 5.4 THERMAL COMPONENT WEIGHT Component WVeight Thermal Coatings 1. 0 lbs Insulation 6. 0 lbs Heat Pipes 9. 0 lbs To tal 6. 0 lbs 51

5.5 REFERENCES Hamilton, D. C., and Morgan, W. R, Radiant Interchange Configuration Factors. NACA TN 2836 (1952). Heller, Gerhard B., Thermo physics and Temperature Control of Spacecraft and Entry Vehicles, New York and London: Academic Press, 1966. Krieth, Frank, Principles of Heat Transfer, Scranton, Pa. International Textbook Company, 1958. Krieth, Frank, Radiation Heat Transfer of Spacecraft and Solar Power Plant Design, Scranton, Pa. International Textbook Company, 1963. Sparrow, E. M., Cess, R. D., Radiation Heat Transfer, Belmont, California, Brooks/Cole Publishing Company, 1967.

STR UC T URES 6. 1 INTRODUCTION The structural design of the MEDUSA spacecraft is based on the idea of bench made components being assenmbled into a final form that is both reliable and able to meet the consditiosns of he launch environment. The interface betweeha the spacecraft and the Burner II is a three point support. It is, therefore, necessary for the loads to be concentrated at these points. To accomplish these tasks, the spacecraft consists of two main support sections. A hub structure supported by a K-truss will concentrate the feed assembly weight at the support poi.nts rhile still providing a way of mounting the complete feed system. The main spacecraft module will use a strong base structure to allow easy mounting throsughout the module without concentrating individual loads only at the support poiints. With these arrangements, the mournting of prealigned component systems is considered feasible. Aluminum 7075-T6 was chosen as the main structural material, because of its light. weight, relative ease of fabrication, and high thermal conductivity. Since a semi-passive thermal control system is to be maintained, the high thermal conductivity was an important factor. 6. 2 OVERALL SATELLITE CONFIGURATION The recommended satellite configu~rataon is shown in Figure 6. 2. 1. This configuration allows for the mounti-zig of TWT's in the feed structure. This will eliminate, or at least miniir'ze the po wer loss between the communications components. The decision to support the feed out from the main spacecraft module, rather than sxlpport the reflector out from a main spacecraft module that contains the feed, was made to reduce the shroud length and diameter requirements and keep the extended portion of xeight to a minimum. If the reflector had been supported out from the spacecraft, the maximum diameter would have been needed at the top of the shroud and a way of getting the solar arrays out beyond it wTould have been necessary. This would have thus increased the length, diameter, and weight over the system chosen. With solar cell arrays stored h izorLntally on the north and south sides, both easy access to equipment panels a.n simnple single axis rotation (for maintaining sun orientation) cai be obti.aied. There is, however, ample space available for the soar els to be gnd ically durin the launch phase since they do clear the Burner II.

Figure 6. 2.1 Launch configuration of satellite = = Earth... _' —..x Sens or Sun Sensors 120~ apart Heat Pipe Multiplexer Upper Mounting / Hub Suppor Panel Feed - Sun As sembl, Sensors - Upper Fittings for K-truss We st Coast Feed Upper Diagonal Center Fitting Refl e ctor (stowed) Diagonal Sun Stowed Psor sition

This clearance is provided so that' the Buzrner II could remain with the spacecraft in geostationary orbit~ The decision to keep the Burner II was made so that close alignment of the enteatlr of pressure and center of gravity could be obtained without carryLn ig: useless exmtrs weight into orbit or using some deployable mechanism. With this close alignment of center of pressure and center of gravityithe fuel requirement for thLe attitude control system is minimal. The clearance of the solar arrays w-ith the Burner II alsb eliminates the mission critical phase of separation $see Appendix Ea, I for change in center of gravity). The in-orbit configuration is shov-n in Figure 6. 2. 2 (minus the Burner II). 6.3 SPACECRAFT DESIGN FEATURES 6.3.1 Launch Loads The spacecraft structure ias desitgned to withstand the loads created by the Atlas-Agena launch vehicle, The total load factors seem to reach their greatest values during this phase. Putting in a safety factor of 1.5, the maximum loading conditions are: Long itudi aal Lateral Loading Condition Load Load Max. Longitudinal 7. 34 x 1. 5 = 11. 05 g 2. 35 x 1.5 = 3. 525 g Max. Lateral 3.0 x 1.5 - 4.5 g 2. 7 x 1. 5 = 4. 05 g These load factors include, an allow ance for vibrational effects. Informa-tion on the minimum natural frequen.cy allowable for the launch vehicle was not available, but the natural frequency oif the structural members used in MEDUSA seem to be high enough so that they shoulld not present any real problem (see Appendix E. 6 for actual values)., Further study of the structure as a whole, with respect to frequency response and stiffness, with a full scale mockup is necessary to verify this. 6.3. 2 Solar Cell Array Mount and Drive Since the solar cell arrays are mrnouited nforma]l to both the north and south sides, they are only required to rota;e about one axis; a common one. Thus only one shaft, passing through the spacecraft and connecting both solar cell arrays, is necessary. The entire solar cell array drive mechanism can thus be completely assembled and alligned wifIChim A As ow n fram;Ae before be ng mounted on the satelite. The mounting brakets on the spacecraft will then simply lock that frame into postlion (see Figure 6o 3 io).

Deployed Solar Arrays Unfurled Reflector Alaska Feed Assembly Burner II Attaching. Points 4 Ribs Figure 6. 2. 2 On Station Configuration Scale 1/8" = 1' 56

Battery Chargers Batteries slip ring B earing Scale 1 1/2" = 1' Figure 6. 3. 1 Solar Array Drive and Battery Position 57

Proceeding in this manner myeans that tAhe only emaining duty after mounting this frame is to connect the actual solai arrays and deployrmienEt mechanism to the shaft (see Appendix E. 2 for details of frame and support). 6.3. 3 Equipment Panels In order to maintain a completely passive thermal control system, it is necessary to mount heat dissipating equipment on the north and south sides of the spacecraft. These sides are, therefore, the equipment mounting panels, Since the solar panels are also on the north.- souh sides, there is a limited amount of space available for this. A problem- thus develops in keeping systems together, so that extra wiring is elirminataied, and weight and heat dissipated are balanced in the space available. To obtain a heat balance, the power control equipment, gyro box, and cormmnurnicL.atioons equipment are mounted on one of the faces, while the telemetry equipment is mounted on the other. Except for power regulating equipmnent this position allocation provides good system continuity (see details of mounthig Figure 60 3.2 and 6. 3.3) Continuing with the idea of bench made compoznents, all of this equipment can be mounted along with the appropriate circuitry on the panels before being attached to the satellite frame. These north-south panels are made of 08 inch sheet aluminunm to insure thermal conductivity along the panel and to provide a solid mounting surface. The other four panels are not needed for equipment mounting so they are made of thinner. 02 inch aluminum thabt is thermally insulated from the rest of the satellite. 6.3.4 Propulsion and Control Mounting The propulsion system uses a monopropelle.nt: hydrazine. Thus, there is, a need for only one fuel tanko To give the system added redunndawcy, two separate tanks are used rather than just one. These t.anks can be rmade, equipped, and filled before being placed in -"t-he satellite. When placed on the platform provided for them in the satellite, the tanks can be connected inLto the plumbing system that has already bee n instalLed. The actual thruster system is mounted entirely on the east and west ref'lector support columns. The plum bing system is thus, concentrated i - these two areas. The thrusters are positioned to meet the requirements set by attitude and control for alignment and station keeping. Since the fuel tank platform is a perm.anent part of the struct.re, the momentum wheels can be aligned and mounted directly on it. They will therefore be securely fastened near the cent-er of gravity. This operation can be done befoe other components are installed avith no pesultng installation interference (see Figure 60 3e4 and Appendi Ea 3). 58

r. ~ - - - - - - -m - r r- — 1 i a 1 t~~~~~~1 Transistor - xAmplifiers 2 Transistor Transistor Transistor fT 1 Amplifiers Amplifier Amplifiers I.I ~~~~Transistor _______________j (2) _____________ (2) (Limited) (Limited) (Limited) T ran si storr Amplifiers (2) Oscillator Filter ilter Filte er Power Controller Mixers Scal / 1 ilter Filter Filte (2) I Lfl (2) indicates a s I II ( I I I I i I I~~~~~~~~~~~~~~~ stacked configurati 0 s.clla~to Power Sun Sensors Regulator Electronics Box I N INorth W-all -mm" ___ _ME_ -C_==- _0=6=W_=som=- 1_22M=- W = - -zmmw -Z dl Figure 63

(2) sub Sub Sub commutators Prime i Prime A/D Commutatol Commutato'.. Encoder,,_______________..., _...~....,_;O scillator Decoder Storage, Command Command Clock Distributor Distributo r Diplexer 0 S-band S-band Transmitterp.. Trahsmittel Decoder Storage Receiver Receiver Demodulation Figure 6.3.3 South Wall

Figure 6. 3. 4 Propulsion System ~ ---.... k Rate Gyro Control Computer Gyro Box Thrusters Fuel Thrusters Tanks Momentum

The only sensing devices that are required on the main spacecraft body are sun sensors that view the area around the bottom of the satellite. With the Burner II permanently attached to the satellite two sensors, mouunted out from the satellite body, are needed to view space behindthI Burner II. These sensors will be fixed to the bottom of the reflector torus. All other sensing devicees are mounted on the feed and will be discussed in SectioIn 6 5. 6. 3. 5 Reflector Support Columns The six main support columns for the reflector are evenly spaced around the perimeter and allow for alignment of the reflector by shimmi.ng at a limited number of support points. The actual interface between these columns and the reflector torus will be described in the folllowing section (for strength analysis see Appendix E. 4). These columns will also be used as the wall supports, by simply adding flanges between them along the top edge and down the sides of the ones supporting the north-south side panels (for basic structural member positioning, see Figure 6. 3.5 and for total internal spacecraft system, see Figure 6. 3. 6). 6. 4 REFLECTOR DESIGN 6. 4. 1 Design Requirements Inherent in the design of any commu. nications satellite- -which is to serve only a small partion of the earth's surface —is the design of a downlink parabolic reflector. The purpose of this reflector is to shape the signal emitted by the feed into a beam only a few degrees wide. This concentrates the signal in the area which is to be served. It has been determined that the optimum half-power beamr width fo r coverage of Alaska from geostationiary orbit is approximately 30~, The relation between reflector dianmeter, d, and the half-power beam width is: 730 d N where X= wavelength = 1.23 ft 0 - half-power beam width ~ 3. 00 Thus, d 30 ft. Si es th-ere has 'been c~o;nsiderablework done on reflectors with diamneters of 30.0 ft, this diameter was chosen. This fixes the half-power beam width at 3.0~ 62

Vertical Equipment Strut Panel Main LoBaser Diagonal Strut tSurupport Columnss Fitting Fuel Tank and M.. W.Mount' -Main Base Structure Reflector Scale 1 1/2" = 1 Figure 6. 3.5 Structural Member Positioning 63

Figure 6. 3. 6 Total Internal Spacecraft System Positioning 64

6. 4. 2 Reflector Selection The flex-rib reflector chosen is similar to the type being developed by Lockheed Missiles and Space Co. This particular design was picked for the following reasons: 1) Furled volume 2) RF characteristics 3) Weight 4) Dependability. Furled Volume. In the stowed configuration, the reflector fits within a torus which has an outside diameter of 76 in, an inside diameter of 59 in and a depth of 7. 5 in. As can be seen from Figure 6. 4. 1 this amounts to a very small fraction of the total volume of the spacecraft. RF Characteristics. The design chosen uses 24 ribs. As a result, the erected reflector only approximates a paraboloid; it is actually made up of 24 parabolic cylinders. The degradation in gain due to this difference in geometry is: IAG 0. 048 db. The gain of a true paraboloid is: (Appendix E. 5) Gt 35. 08 dbd Thus the actual gain of the reflector is: G = Gt - JAG I G 3 5. 03 db which is acceptable. Weight. The weight of the reflector, including the hub, ribs, mesh and interface is 100 lbs. Dependability. One of the most attractive aspects of this design is its dependability. Since all the energy required to erect the reflector is stored in the ribs, complicated mechanical erection equipment is not needed. LMSC has estimated the reliability to be 0. 99964. It should be noted that a reflector of this type Till lbe flowrn on the General Electric AST - F communications satellite in 9 7 3. 65

Figure 6,4. 1 Reflector in Furled Configuration Reflecto r Hub 66

6. 4. 3 p eration The reflector is stowed by wrapping eachk of its 24 ribs about the central hub. The flexible mesh is folded between adjacent ribs. The resulting wrap is contained by 24 spring-loaded doors and a rest raniing cable. When the appropriate command is given, the cable is cut by a pyrotechnic device. The doors spring open and the ribs unwrap and lock in their open position. This takes approximartely 2 seconds and the reflector is completely stabized,within 15 seconds. The maximum torque experienced by the spacecraft during deployment is about 6000 ft-lbs. This is not critical as the solar arrays would not be deployed at this time. It is important to note that there is no total angular momentum stored in the furled reflector, i. e., the spacecraft will have approsximately the same rotational rate before and after deployment. (If there is any change there will be a slower rate because the moment of inertia will have increased. ) 6. 4. 4 Construction Hub. The hub is shown in Figures6.4o.2t 3, and 4. It is basically a torus with a U-shaped cross-section. The spring loaded doors cover the open side of the U. It is constructed mainly of magnesium to reduce thermal distortions Each rib is mounted on a spring loaded hinge and is equipped wiith a lock to hold it in the open position. There are also adjustmener bolts on each hinge mount which are used to position the rib. Ribs. The ribs are each 12 ft long and taper from 6 in at the hub to 3 in at the tip. In its deployed position the cross-section of the rib is an arc. When the rib is wrapped about the hiub, this c amber is flatiened out. Along the inside edge of each rib are holes to which the mesh is se-wnr.. The ribs will be constructed of 0. 0~5 in 6061 Alumninum. This Nwill result in lower thermal gradients than other slightly stronger alloys. Mesh. The reflective mesh will be made of Chromel R, a chromium material which can be knitted into a fine mesh, After knit'ing, the mesh is gold plated to make it cond'uctive. This matLerial has exhibited a 99. 6% reflection. 6o 4.5 Reflector Support Structure The reflector is supported at six equally spaced locations. The supports are made of magnesium alloy (A% 3l-H24). The magnesium alloy is used to rminimize reflector hub/support s.tru$ctu1 distortionss. The supports where designed to withstand launch loadings of 11. 0 g vertical and 3. 5 g 67

Figure 6. 4. 2 Reflector Hub - Top and Side View Cable Cutter Re straining Cable,. -..76 ".. 68 -7 1 1 -\

Totrus Upper $Surface " " (Partial) Door Hinge Door igUre. 3 Reflib and Doo Stifee Ild Doo-r Configuration -Docsir,

Door Stiffer Spring Re straining UpI~per \\ t~r~ ~ \\\ tr~~J Cable Reflector Support 0 C able Guaild Spa e craft \ Rib Adjustment B'olt Door King Reflector Support

lateral, as -well as a. torque of approximately 6000 foot-pounds (note: this is for a reflector with forty-eight ribs) produced during the refleClor deployment. A first approximation consideration Tas also given to thermal induced deflections in an attempt to see if these deflectiosns were wi-thina acceptable limits. (The reflector support calculations are in Appendix E.6.) Support/Reflector Interface The two parts of the support were designed to a-ccommodate the six degree tilt of the reflector torus with respect to the x-y plane of the satellite. The lower support is a permanent spacecraft comnponent on which the reflector torus is set and bolted (see Figure 6o4.5 ). This restrains the torus in six directions. Pre-launch alignment of the reflector is accomplished by placing -shirhs. between the support and the torus at the appropriate locations. The upper part is slipped over the shear pin and bolted to the reflector support column after the reflector is in place. This part allows for prelaunch alignment and restrains the motion in two directions, especially during the launch phase. The maximum stresses (1272 psi for the lower support and 11, 707. 2 psi on the upper support for vertieal plus lateral loading, and 15, 900 psi on the lower support during the reflector deployment) are well under the yield strength of 29, 000 psi for the alloy used. The large safety margin assures the rigidity of the structure. Sup port/Spacecraft Interface The two structural parts are connected to a collar made of the same magnesium alloy. This collar more evenly distributes the load to the enclosed reflector support column. The lo wer support is permanently mounted by use of titanium connectors. The upper support is bolted to the spacecraft also by means of titanium connectors. The exact method of connection (i. e. rivets, bolts, etc) will have to be determined by e.zperimental study of thermal and mechanical trade-off to best minimize spacecr aft/support distortions. It is important to note that to a first approximation, the maximum tip deflection due to the support is 0. 0563 inch, wKihich is within presen system requirements. Electrical Interface The electrical connection of the spacecraft and the reflector (for reflector deployment) will be made at a later phase of the satelite's development. 71

Upper Support Sec AA 29. 15 in ---to Ref Reflecto r Torus RfcoSp rte6. S 1 R is Sate l liatel Body 29 85 in to Ref Lower Support to Ref Sec BB Reflector Support Scale 1'4 Ref is Column Satellite Body Figure 6.4.5 Reflector Supports 72

Super Insulation Placed in this Area Upper Support Reflector Torus Satellite Body Reflector Lower Support Pigure 6. 4. 6 Reflector Insulation

Therrmal Interface It is desirable to isolate the spacecraft body from the reflector from the point of view of possible warpage due to spacecraft/reflector coupling. By assuming that the heat transmitted by conduction through the members is small compared to that of the radiated heat, it is proposed that super-insulation be placed as is shown in Figure 6.4.6. This ~would then minimize the effects of the radiated heat. The problem would be further studied in a later phase. 6.5 FEED ASSEMBLY DESIGN (see Figure 6.5.!) The structural design of the feed is again designed for bench assembly with alignment on the support hub later. Since the feed holds orrlst of the equipment where great heat lose occurs, the thermal gradients and amount of radiating area became the major design criteria. The addition of a mounting cylinder around, and a mounting plate above the main feed cone gave the necessary radiating area. Heat pipes were added to decrease thermal gradients that could arise, With a section cut out of the outer cylinder, a position is created for the west coast feed. Since the inside of the main feed must be smooth, it can be mounted only by the use of a fange. With this flange around the top of the feed it can be mounted directly to the top of the feed struct-re0 The power conditioner and TWT's are mounted along the inside (of the outer cylinder. The earth sensor, multiplexer, telemetry antennas, and one sun sensor are mounted on the top mounting plate, The nultiplexer is slightly indented into the top mounting plate to insure ful covera ge by the earth sensor. Three sun sensors are mounted on the outside of the structural hub, directly above the support fittings. This arrangement along with the two mounte on the torus will give complete sun sensor coverage of space (for actual positioning of these components see Figure 6.2.i )1 6.6 FEED SUPPORT STRUCTURE The recommended feed support structure is composed of a top hub assembly connected directly to the base by.f three see tion K-truss. Thiis type of support was chosen mainly because of the research done by Gene-ral Electric in their ATS F and G studies. In these studies, it was disovrerred that the K-truss interfered with signal transmission much less than the coinveationalI A-frame type design. The only way to determine what the actual interference effects of the K-truss are, would be to build a model of the truss and physically test it. Along with these tests, the s$tuctue oiuld also have to be tested for random vibration input and thermal effects. The thermal effects can be 74

Outer Heat Pipe Earth Sensor Multiplexer Outer Heat Heat Pipe (runs length of / _ TWT) West / Power ransformer Coast Feed (above west coast feed) Main Feed Cone Main Feed Helix Scale 1" = I' Figure 6. 5. 1 Feed Assembly Design 75

held to a minimum by the use of a series emittance coating (a =. 05, E =. 1). This will keep the struts within a temperature range and temperature gradient that will limit distortions of focal positioning to acceptable levels. To insure this, the struts will be insulated from the rest of the structure (for breakdown and sizing of members see Appendix E. 7). A change to a dielectic support cone may be warranted if the development of such a cone proves to be both reliable and light weight. As far as RF interference is conce-rrd this would be the optimum support structure. 6.7 WEIGHT BUDGET Component weights (lbs) Section totals (1 Feed 114.45 6 TWTI's 24. 0 1 multiplexer 8. 0 2 transister amps 0. 4 main feed 8. 18 west coast feed 1.82 2 telemetry antennas 2. 0 switching equipment 20. 0 1 earth sensor 9. 0 4 sun sensors 3.4 1 power transformer 10. 0 7 heat pipes 9.0 structure 18. 65. Feed support 90. 93 K-truss 60. 35 fixtur e s 19. 0 thermal coating 1.0 hub 6. 38 cable (throughout satellite) 4. 2 Reflector system 101. 7 reflector and interface 100. 0 sun sensors 1.7 Spacecraft North wall 40, 33 1 ratei:integrating gyro 1..0 12 transister amps 3. 6 2 mixers 1. 0 2 oscillators 1. 2 6 filters 4. 8 switching 5. 0 1 computer 4.0 1 gyro set 2.6 1 sun sensor elec. box 2, 9 1 power controller 5. 0 1 power regulator 3. 9 structur e s 5. 33

South wall 33.43 1 commutator 1.0 3 sub commutators 2. 0 1 encoder 1. 0 2 decoder/storage 4.0 1 diplexer 0.5 2 transmitters 1.6 2 receivers 8. 0 2 command distributors 4. o 1 clock 2. 0 1 demodulator 1.0 1 oscillator 1.0 cables and misc. 2.0 structures 5, 33 Interior of the spacecraft 402. 16 2 batteries' 87.4 2 battery chargers 4. 0 solar array drive 25. 0 2 solar array mounts 61.0 2 solar arrays 60.2 2 fuel tanks 16. 0 fuel 60. 0 3 momentum wheels 30. 3 plumbing and valves 7. 0 2 10# thrusters 1.4 16 0. 1# thrusters 2. 2 -insulation 6. 0 2 telemetry antennas 2.0 shunt resistors 2. 0 structure 37.66 Total 7.83. 0 pounds 6.8 REFERENCES Crandall, S. H., ed., Mechanics of Solids, McGraw Hill, New York, 1959. "Feasibility Study for a Lightweight TV Broadcast and Communications Satellite", Doc. No. 70Sd4289, General Electric Space Systems Organization, Philadelphia, Pa, 24 November 1970. Gaetano, A. F., Manager, Astrionics Engineering, Lockheed Missiles and Space Co., Sunnyvale, Calif., (personal correspondence, 25 February 1971.) Koelle, Heinz Hermann, ed., Handbook of Astronautical Engine ering, McGraw Hill, New York, 1961. Osgood, Carl C., Spacecraft Structures, Prentice Hall, Englewood Cliffs, N. J. 1966. 77

Project MISSAC, University of Michigan Aero 483 Design Project, April 1968. Project REMUS, University of Michigan Aero 483 Design Project, December 197 "Proposal for Applications Technology Satellite F and G" Phase D, Book 1, Vol. 1, Technical Proposal, SD Proposal N-21630, General Electric Space Systems Organization, Valley Forge, Pa., 18 July 1969. "Proposal for Applications Technology Satellite F and G', Phase D, Book 1,.Vol'. IH,,..Reflector, SD Proposal.N-'21630, General Electric Space Organizatic Valley Forge, Pa., 18 July 1969. Shames, Irving H., Engineering Mechanics, Statics, Prentice Hall, New Jersey, 1966. "Space Erectable Large Aperature Reflectors", LMSC-A946613, Lockheed Missiles and Space Co., Sunnyvale, Calif, 26 March 1969. "Television Broadcast Satellite (TVBS) Study", Vol. I, II and III, NASA cr-72579, General Electric Co., Space Systems Organization, Philadelphia, Pa. 15 November 1969. TRW Space Data, TRW Systems Group, Redondo Beach, Calif. 1967. University of Michigan Aerospace 414 Class Notes, Fall Term, 1970. 78

LAUNCH VEHICLE 7.1 'INTRODUCTION The basic mission requirement dictates that a launch vehicle capable of placing a 783 pound useful payload into an equatorial, earth-synchronous orbit be used. Selection of such a launch vehicle was based primarily on four considerations: 1) capability, 2) low cost, 3) reliability, and 4) orbit injection accuracy. The required capability of placing a 783 pound useful payload into this particular orbit virtually eliminates the Scout and Delta family launch vehicles because they lack the necessary power. This requirement is satisfied by an Atlas SLV-3A/Agena D/Burner II launch vehicle configuration which has a useful payload capability of 900 pounds. This specific vehicle costs approximately $10. 0 M and is $.0. 5 M less in cost than the next size larger launch vehicle, the Titan III-B/Agena D. The Atlas' long history of successful launches coupled with the equally dependable operation of the second stage Agena D reflect their inherent high reliability characteristics. Because the mission requires a. high degree of accuracy in positioning MEDUSA in its orbit; a vehicle with a highly sophisticated guidance control system was essential. The Atlas SLV-3A/ Agena D/Burner II launch vehicle provides such accuracy because all three stages possess their own individual guidance control systems. Thus, with all the above considerations taken into account, the selection of the launch vehicle is the ATLAS SLV-3A/AGENA D/BURNER II (Figure 7.7. 1), 7~ 2 LAUNCH SITE AND WINDOW MEDUSA will be launched due east from the Eastern Test Range (ETR), thus taking complete advantage of the earth s rotation. Complex 13 at the ETR is equipped to handle the SLV-3A/Agena D launch vehicle configuration; thus, only slight modification will be required to accommoda te the additional Burner II third stage. When MEDUSA first reaches its on station possition, the initial co ditions should be such that the satellite will not go into an orbital eclipse for a maximum period of time. These eclipses occur around March 22 and September 22. In order that this state exist, the satellite will be lawunkched on or shortly after October 16, 1974 at 0500 EST. If this date is missed, April 16, 1975 is the next optimal date of launch. In either case, the satellite would be available for educational purposes for the 1975 school year. 79

Convair Station - 245 Modified 2MLT Delta Fairing l ~]84 in dia. 313 in 68 1 -.. Agena D 206 Boo ster Adaptor 313 347 _._ Adaptor Extension 490 129 ft 1 20 in dia. SLV-3A (Uprated) 1133. 1310 Figure 7.7. 1 SLV-3A/Agena D/B-II MEDUSA Launch Vehicle 80

7.3 LAUNCH VEHICLE WEIGHTS The following is a breakdown of the Launch Vehicle weights for the mis sion:pound s SLV - 3A 3 24, 768 Booster Jettison (7487) Sustainer Jettison (7765) Flight Expendables (309,1 26) incl. 1"000# F. P. R. Booster Adaptor (290) Booster Extension (100) Agena D 14,957 Agena Jettison (1430#) incl. Agena/BII Adaptor Flight Expendables (13,527) incl. 60# F. P. R. Burner II 1,9361 BII Jettison (360) Flight Expendables (1001#) Shroud (Jettisoned at SECO) 1,030 B -II/Payload Adaptor 22 Total Liftoff Weight, less Payload 342, 138 7.4 LAUNCH VEHICLE SYSTEMS The following subsystems are contained in the Lalnch Vehicle: 7.4. 1 Guidance and Control SLV-3A - General Electric radio guidance is employed. The system calculates the required vehicle position and velocity and commands the flight correction necessary for mission objectives. The flight control system provides stabilization and attitude control during the powered phase of the flight. Three axis control is attained through the use of engine displacement. Agena D - an all-inertial guidance system provides error signals whenever the vehicle deviates from its prescribed attitude. The control system uses both pneumatic and hydraulic control forces to correct deviations. A three-awis, body-mounted gyro reference system coupled with two horizon scanners determine the attitude of the vehicle. A reaction control system employing Freon 14/nitrogen is utilized during the parking orbit coast. 81

Burner II - guidance is provided by precision strap-down gyros and a velocity meter. Control during powered flight is provided by a H202 reaction control system. This R. C. S. also provides the vehicle with velocity vernier thrust after main engine burnout. A separate nitrogen jet control system provides the attitude control during coast. 7.4. 2 Electrical Power Systems SLV-3A - the Atlas vehicle is provided with 28 vdc main vehicle battery power, as well as 28 vdc pyrotechnic battery power. Also available is 115 volt, 3 phase, 400 cps power provided through an inverter. Agena D - the Agena is provided with the same type of power system as the SLV - 3A. The type of batteries used are 2- Type VI-A batteries with an average of 2, 640 watt-hours of power available. Burner II - batteries must provide power for a nineteen hour duration, therefore a total of 114 amp-hours of battery power were necessary for the mission. This provides the stage with 28 vdc power. During parking orbit >co'ast, an electrical interface from the Agena provides the B-II with power. 7.4.3 Telemetry Systems The telemetry systems of the vehicles are: SLV-3A - a PAM/FM/FM telemetry system is employed. An 18 -channel telepack is used of which five channels are commutated to supply 26 measurements each. Agena D - a system similar to the SLV-3A is employed. Burner II - the system is similar to the SLV-3A with a 15 watt VHF transmitter and three antennas. 7.5 LOADING FACTORS The total load factors of a spacecraft utilizing the Atlas/Agena ascent vehicles are based on either total spacecraft weight or spacecraft equipment weight. Thus, for a spacecraft weighing 783 pounds, the total load factors are: Loading Condition Longitudinal Lateral Maximum Lateral 3. 0 g 2.7 g Maximum Longitudinal 7. 35 g 2.35 g Load Factor - g's 82

During launch the spacecraft will experieance simultaneous sine and random vibrational accelerations. The min nimum and maximum sinusoidal vibration accelerations over the entire frequency spectrum are: Minimum 0. 3 g(RMS) Maximum 1. 0g (R MS) The minimum and maximum random vibrational accelerations over the entire frequency range are: Minimum 0. 1414 g (at 20 CPS) Maximum 5.0 g (at 1000 CPS) 7.6 FAIRING (Figure 7.6.1) The fairing chosen for this mission was a modified two meter, long tank Delta fairing. A modification was made to the length of the fairing- an additional four feet were added to the mid-body portion of the shroud. The mission required a fairing with an internal diameter of 76 inches and payload envelope length of 247 inches. A nominal development and testing cost has been added for the modification. The fairing weight has been estimated at 1030 lb. 7. 7 SPACECRAFT ATTACH FITTING The satellite will be supported on the Burner II third stage by an attach fitting specifically designed for the spacecraft (see Figure 7. 7. 1). This fitting will be fabricated from 2024- T4 aluminum alloy and will weigh 18. 50 pounds. The attach structure will be more than capable of supporting the 783 pound satellie during the launch and orbital accelerations. Because it will be necessary to keep the Burner II attached to the satellite after final orbit insertion; the attach fitting will be permanently mounted to the satellite and the Burner II. An additional allowance of 3.5 pounds has been made to provide for thermal insulation between the Burner II and spacecraft. 7. 8 BURNER II The guided Burner II stage was chosen as the apogee kick msotor for three reasons: 1. Accuracy - accura~yis improved about 50 percent over a spinstabilized A. K. M. 2. Stability - because of the extrerme length of the feed structure as compared to the satellite diameter, a spin-sstabil zed configuration may have caused extreme tumbling of the vebicle to occur during transfer to synchronous orbit. 83

22 22 / — 22R Ref - If ~Feed Feed Support 313 Atta ch Fitting 54 | A A I All dimensions I 77 54~1~84 If -- Bune I

/f~ ~~, \ ~ X~ ~22*375" R ~/.///l\ /'\ ~ as~~ -— / /~~~ 19., 3.18" R ' // \\X i,. - 1 X, /20~ X'~~ k\ / ~ _..,-.,-.,...\-,..... ---.1. _ Figure 7.?. A~tta~h' Fitting 85

3. Thermal - by placing the A. K. M. outside the spacecraft, -thermal problems were greatly simplified. Had the mostor been incorporated into the satellite, active thermal control may have been required, and then only for a short time. 7.9 USEFUL PAYLOAD Synchronous Transfer 2166 lb Flight Expendables 1001 lb On-Orbit, 2. 1 Inclination 1165 lb B-II Jettison Wgt. incl. adapter 382 lb Useful Payload 783 lb The launch vehicle chosen has the capability for growth of useful payload to 900 lb. This may be accomplished by inccreasing the fuel loading of the B-II to 1168# and by increasing the burn tim es of the liquid stages of the launch vehicle. 7. 10 REFERENCES "Atlas Launch Vehicle Family for SpacecraIt Conatractor Planning ", General Dynamics-Convair Division, Report No. GDC BNZ 69-01!, October 1969. "Atlas Launch Vehicle Family for Spacecraft Contractor Planning," General Dynamics-Convair Division, Report No. GDC BGJ 67-002, April 1967. "Mission Planners Guide to the Burner II", The Boeing Company, Report I J2860a1-5,,:April 1968. "Advanced Atlas Launch Vehicle Digest"9 General Dynamics-Convair Division Issue No. 2, April 1967. 'rLaunch Vehicle Estimating Factors'", NASA, January 1969. "Delta Payload Planner's Guide", McDonn ell Doug.as Astronautics Company, April 1969. Personal correspondence, Convair Division, General Dynamics, E. J. Hujsak, February 16, 1971, 86

8 ORBITAL ANAL YSIS 8. 1 FINAL ORBITAL REQUIREMENTS The mission requirements fora long life, commruwnicatiorns satellite serving Alaska make it necessary to positiOn the satellite in a low inclination, geosata.ionary orbit over the Pacific Ocean, A geosa;:io:na.- orbi" is one with a period of one sidereal day in the equatorial plane (inclination = 0 degrees) and a theoretical radius of 22, 766. 9 nm. However, due to perturbations caused by the sun and moon and earth's non-ideal gravitat-ional f-1ield, th(. orCbit is altered (see Section 8.8). MEDUSA' s orbit will have a radius of 22, 767. 2 nm with a planned initial inclination of 2. 1~. In order to optimize the operating period of the television system without the loss of the continental U. S. uplink or serious degradation of the transmissions 1700 W was chosen as the on station longitude. A series of rendezvous and walking maneuvers will be used to reach the on st atio longitude because the initial geo. ationary longitude will: be to the east. All calculations and data are based upon three- sigma error data. The general configuration of the maneuvers is shown in Figure 8. 1. 1. 8. 2 MISSION TIMETABLE day hour min sec event 0 0 0 0 Launch 2 36 Atlas BECO (Booster Engine Cut-Off)' 5 24 Atlas SECO (Susainer.Engine Cut-Off) 5 2 6 Fairi ng Separation 5 44 Atlas VECO (Vernier Engine Cut-Off)l 5 50 Atlas Jettison 8 42 Agena Ullage Initiation 9 2 Agena Ignition!1 26 Agenas MECO, (Main' Engine -Cut-Off) Parking Orbit Insertion 26 37 1 st Node 5 34 55 Agena Ullage Initiation 5 35 15 Agena Ignition, Transfer Orbit Inj ection, 8th Node 5 36 46 Agena MECO 5 39 33 Agena Jem$ison 5 49 33 Agena Re tc omraneuver 10 50 35 Burmner II Ignicion, Geo st'.onaa. Orbit Circ ula ri at i, 87

r.^ Peru T. O. I. M / 28, 5 Equatorial Plane 26.5~ 5 '% u:5 Parking Orbit T. O, IL vC 2) Parking Q Orbit Geo stationary Orbit \ M ~Transfe r Orbit / G. O. C. Transfer Orbit ~M u~ 2~.13 Geo stationary Orbit Equatorial Plane Figure 8. 1. 1 Orbital Configurations 88

10 51 2 Burner II MECO 10 52 0 Deploy Reflector 10 53 0 Deploy Solar Arrays 10 58 0 Pow~er Up Attitude and Control and Telemetry 11 45 0 Align Satellite for Rendezvous Terminal Phase Initiation 11 50 0 Terminal Phase Initiiation 23 42 0 Align Satellite for Rendezvous B raking Maneuver 23 48 2 Rendezvous Braking Maneuver 23 50 0 Align Satellite on Alaska 1 0 0 0 Television System on with Ala ska Uplink 1 0 1 0 Power Down Bur:ner II Guidance and Control 18 On Station Longitude, Walking Maneuver Terminated, No rmal Operations 8.3 LAUNCH PHASE MEDUSA will be launched from the Eastern Test Range and injected into a nominal 100 nm circular parking orbit with an inclination of 28. 5~. Nodes (equatorial crossings) are as shown in Table 8. 3. 1. Table 8e 3. 1 Parking Orbit Nodes Node Orbits- Longitude Time hour min sec 1 0 2.9 E 26 37 2 1/2 171.80E 1 10 42 3 1 19, 2~W 1 54 48 4 1 1/2 149. 80E 2 38 53 5 2 41.2 W 3 22 59 6 2 1/2 127.~70E 4 7 4 7 3 63. 30 W 4 51 10 8 3 1/2 105.70E 5 35 15 9 4 85.3~W 6 19 21 10 4 1/2 83, 70E 7 3 26 89

Note that all odd nodes are desceending nodes (iL e,, Southward crossing) and all even ones are ascending nodes. 8.4 TRANSFER ORBIT INJECTION AND TRANSFER ORBIT The transfer orbit's function is to take MEDUSA fromn the parking corbit to geostionaiy altitude at a suitable longitude for the circularization maneuver. MEDUSA will orbit at parking altitude for 3 1/2 re7olations before Transfer Orbit Injection (T. 0. I. ). This choice gives a suitable apogee loxngitude and is an ascending node. Since the orbit is to be biased down 2. 10 at geosfaionary, this reduces the required plane change by 20. A descending node at this point would have increased the plane change by 20. See Figures 8. 1L1 and 8.4, 1. Also, in order to minimize the velocity required for plane changes, a reduction of 2. 00 of inclination of the orbit will be made at perigee of the transfer orbit (see Appendix G. 2). Following are the parking and transfer orbit parameters, along with their maximum errors. Parking Orbit Radius 100 + 3. 6 nm Parking Orbit Velocity 25568 4 39 fps Parking Orbit Inclination 28. 50 Parking Orbit Period 88 min 11 + 9 sec Longitude at Perigee 1105 7 East Perigee Burn Increment o062 + 40 fps Transfer Orbit Radius (Apogee) 2Z2767 + 10 rnm Transfer Orbit Velocity 59236 + 77 fps Transfer Orbit Inclination (Apogee) 26.5~ + o 3 Transfer Orbit Period 10 hr 30 min 40 + 740 sec 8,5 GEOSTATlONAIY ORBIT CIRCULARIZATION At the apogee of the transfer orbit the Burner II solid proEspellant stage will be burned to circularize the satellite's orbit and correct the inclination to the desired 2. 10. Note that this node is a descending node (see Figures 8.1. 1 and 8. 5. 1)0 The Burner II guidance and propuElsion-,n systems have accur acies of 0. 80 and 39 fps. These accuracies and errors in transfer orbit injection lead to the following orbit parameters: Apogee Longitude I53.o 1 West Apogee Burn Increment 5,750 + 39 fps Post Burn Velocity as088 + 116 fps Orbital Inclination 2. 1 + 1. 0 Orbital Period (Nomrinal) 23 hrs 56 min 4 se(c 90

Parking Orbit Equatorial Plane 28. 50 Z6. 50 AV = 8, 062 fps Figure 8. 4. 1 Configuation at Transfer Orbit Injection Transfer Orbit Z. 13o Equato rial Plane 26. 5 5V = 5750 fps Figure 8. 5. 1 Configuration at Geostationary Orbit Circularization 91

If desirable, a small mid -course cc0:rrection (appro-ximatelly 20 fps) can be made during the transfer orbit with the Burner II vernier propulsion system. In addition the final burn attitude of the Burner II and the pr-ogrammed use of its vernier propulsion system bewil l used to) minimize the efec-ts of errors in the transfer orbit and in the ci(rcularization maneuver~r 8. 6 RENDEZVOUS MANEUVER In order to minimize the fuel requirements to get on station while starting operations at the earliest time, MEDUSA will rendezvous with gpodsationary altitude at its injection plus one hour longitude. The.aneuver will be initiated one hour after circularization in order for there to be sufficient time to clearly determine the post circuarization trajectory. This first rendezvous maneuver, the terminal phase initiation (T. P. I, ) ill be made by MEDUSA under Burner II guidance and wi11 be as shown in Figu-re 8.4. 1. One half orbit or 11 hrs 58 min later the rendezvous braking maneuver (R. B. M. ) 1will be made at g-ogiionary, altitude to alrmost circularize the orbit at this longitude. During Ro B. M. MEDUSA"s`radial velocity will be elimin.ated but a net 10 fps eastward velocity will be retained. This will put MEDUSA into itfs walking orbit. Based upon calculations whose results are shown in Figure 8.6. 1 (see Appendix G, 3), and the initial requirement for the walking initiation, 137 fps is allocated for rendezvous, 8. 7 WALKING ORBIT At the coxne'lusion of the -rendezvous braking maneuver, MEDUSA will be on a geos'atioanaiy- orbit but 10 fps fast,, As a resulit t will follow a slightly elliptical orbit with perigee at gecotionc;,arya altitude and a period longer than one sidereal day, The perigee will therefore dri~ft westward towv 'ards the final on station longitude. For 10 fps 'eas-tward, the perigee will regress at approximately 10 per day MAfter 16 to 20 days a perigee will occmur. less than 10 from the station longitude and MEDUSA will bur n its pro pulsioir system at that perigee to fix it's longitude It will then be on station, and will be able to assume full operations, An allocation of 1 fps is m1ade for the iwalking.. maneuvers, See Figure 8. 7. 1. 8.8 PERTURBATIONS 8. 8. 1 Introduction Were the earth spherically symnrret i-al and isolated in spa( e, the problem of establishing a geo onanry euato ital orlbit for a satellite would be simple, One could merely equate the period of the orbiting sa2elite (a function o y of the eradius ~for a ic caulr orbit) to one sidereal d:a, and solve for the desired distance from the center of the earth. Ho'i7yver,,the earth is not spherically symnmetrical, and the presenzce of the sun and moon prevents 92

Hi AV = -123,7 V = -108,4 X x AV = -2,6 AV = -2.6 Y Y Fa st. Slow AV = 108.4 AV =123,7 X. x AV = 2.6 AV = 2 6 Y Y LOW Terminal Phase Initiation Maximum AV= 124 fps Hi AV =1.2 AV -1.2 x x AV =2.6 AV 2. 6 Y Y Fast -. Slow AV = -1.2 aV = -1. 2 x x TV = -2. 6 AV - -2.6 y y Low Rendezvous Braking Maneuver Maximum AV = 3 fps Total AV = 124 fps + 3 fps + 10 fps = 137 fps 10 fps is maximum AV to start walking maneuver Figure 8.6.1 Rendezvous Analysis (AV's shown are those for worst cases of injection error as indicated by axes labels" Maximum errors are + 10 nm altitude and_ 116fps.) 93

T. P. I. Actual position and velocity when determined Geo statiomr3 71T*lti TM*, R. B. M. On Station Longitude True Longitude at T. P. I. Earth Powered Maneuvers R.B.M.* R.B.M. + AVl 10 fps/l~/orbit TM' Trim maneuver to terminate longitude drift. Figure 8. 7. 1 Walking Orbit 94

even a local isolation in space. As a result, the idealized equations predicting geosdtationaityY orbital characteristics m-ust be modified, 'They must be modified not only because of solar and lunar gravitational attractio ns and the mass asymmetry of the earth, but Jlso because of atmospheric an.d electromagnetic drag, radiation pressure, and relativistic effects. For MEDUSA, however, the last four forces are negligible. Therefore, only the perturbing effects of the gravitational $attractions of the moon and sun, and the triaxiality of the earth (which includes both oblateness and the ellipticity of its equatorial section) will be considered. 8.8.2 Corrected Geoistanaf r adius The Keplerian radius of the geo.ationary satellite is increased by. 30 nm due to the earth's equatorial bulge (see Appendix G. 4). In addition, the gravitational attraction of the sun and moon is found to alter the steady-state radius by anywvhere from-O.27 rnm to O 34 nm depending on the initial geometry of the sun-earth-moon-satellite system. Both these corrections to the initial steady-state geod;ionary radius can be included with initial orbital injection co rrections. 8.8.3 In-Plane Perturbations The effect of the ellipticity of the earth's equatorial section is more serious.arai edther of those mentioned in the previous section. Because MEDUSA will be stationed at 1700 West longitude, or 470 west of the nearest stable point, it cannot be in truly geokationary orbit. Instead it will tend to drift towards the 1230 West longitude stable poinat. After two months of uncorrected drift, the satellite will have moved about 100 east from its desired on- station position, with a corresponding ra dial displacement of -13. 3 nm (see Appendices G. 5 and G. 6). If this drift went unanswered, the satellite'vould make large angle oscillations about the stable point with a period of about 1. 6 years, However, this drift can be corrected at the cost of 16. 9 ft/sec/year in East-West station-keeping (see Appendix G0 7). Another source of radial deviation fromn the steady- state value is due to the oscillatory characteristics of the sun-earth-moon-sate lite system (e. g. this includes the angular rates of rotation of the earth-moon system around the sun and about it-s center of mass). The maximum possible radial deviation from these effects is less than 1o 7 nm. The accummunnlated amount of radial error will be corr-ected periodically (see Appendix G,, 7)e 8.8.4 Out;-of-Plan Perturbations For satellite lifetimes less than 10 years, the prsincipal effect of the sun and moon is to produce a rotation of the orbiltal plane aoway from the equatorial at the rate of 60$5250/yea.e Resultling then, is a north-ou'h oscillation of the satellite witih a period of 1 sidereal day and with an incliniation amplitude 95

of n(0.8525~) after n years. At geoa~tionanr velocity, it would require 150 ft/see/year to correct this rotation. On the other hand, if the initial orbital plane is biased away from the equatorial by 5/2(-0.85 250) or -2. 130, MEDUSA's orbiting plane will always be within Z. 130 of the equatorial plane (during the 5 year lifetime), an acceptable value. The non-spherical nature of the earth cases no significant out-ofplane perturbations of the geo ~ationay-' satellite. 8.9 REFERENCES Atlas Launch Vehicle Fa~mily for Spa~cera Contractor Planning, General Dynamics Convair Division, April 1967, Frick, R. H., ''Orbital Regression of Synchronous Satellites Due to the Combined Gravitational Effects of the Sun, the Moon, and the Oblate Earth", RAND Report R-454-NASA, August 1967. Frick, R. H. and Garber, T. B, "Perturbations of a Synchronous Satellite", RAND Report R-3990NASA, May 1962. Kendrick, Jo B., TRW Space Data Third Edition, TRW Systems Group, TRW Inc,.f, Redondo Beach, California 1967. Launch Vehicle Estimating Factors, NASA OSSA, Code SV, Washington.-D. C. January 1969. Marks, R. W., The New Dictionary and Handbook of Aerospace, Bantam Books, 1969. MISSAC, University of Michigan, Department of Aerospace Engineering Design Project, April 1968. Mission Planners Guide to the Burner II, The Boeing Company, Aerospace Group Space Division, Seattle, Washington, April 1968. 'vOrbital Flight Handbook", NASA SP-33 Part I, 1963. SCANNAR, University of Michigan Department of Aerospace Engineering. Design Project, April 19.70. Space Planners Guide, United States Air Force, Air Force Systems Command, July 1965. Syncom Orbit Primer, Hughes Aircraft Company, Space Systems Division, July 1964.

9 PROGRAM PLANS AND COST 9. 1 SUMMARY SCHEDULE The NASA Phased Project Planning approach was used as a basis for the scheduling of Project MEDUSA. The schedules are assumed to start at the completion of this project. In general, an 11 -month period is allowed for Phase B (which includes system synthesis, analysis, evaluation, and selection), a 24-month period for Phase C/D through Qualification Tests, and about 12 months allocated for Phase D (flight hardware fabrication, tests and final checkout). With overlapping of flight hardware fabrication with final design and qualification tests, the overall time would be about 41 months. Figure 9. 1. 1 shows the summary schedule for MEDUSA. 9. 2 COST The cost of MEDUSA is based on the fabrication and launch of two satellites from the ETR in October 1974 for the first and April 1975 for the second. The major development cost would include the cost incurred for the design and testing of -'a) TWT in 800 MHz range, b) a 30r reflector to be used in 800 MHz range and c) a 76" ins-ide diameter", extended shroud. The total:saystern- 6cost.is given. in Table 9. 2. 1. Table 9. 2. 1 Cost Estirmte Phase Cost (Mi lions $) Phase A Feasibility Study Phase B 2. 5 System Analysis Preliminary Design Program Definition Phase B/C 16.0 Complete Analysis Engineering Development and,Tests B readboard Phase C/D 36. 0 Final Design Fabrication Qualifi cation Fabrication of back-up vehicle 6. 0 97

1 st Launch Vehicle 10. 0 2nd Launch Vehicle 9. 0 Total Satellite Cost $79. 5 Total Ground Receiver Cost (100, 000 units) 12.5 Five-year Operational Cost 5.0 Total System Cost $97.TT million:' 9.3 REFERENCES Feasibility Study for a iht7.ight TV Broadcast and Communications Satellite, General Electric, November 24, 1 9 7 0.d Hesselbacher, R. W., An Evaluation of Television Broadcast Satellite Systems, General Electric, AIAA Paper No. 68-1061, October 21, 1968. Merz, E. J., Basic Elements of System Engineers, Lecture Notes. Jan. 27, 19 Television Broadcast Satellite (TVBS) Study, Vol. I, General Electric, August 1, 1969. 98

-.L71.. 2- 4 6 84.10 12-14 16 182. 24 26. 30 3234 36 38 40 42 46 48 R Feasibility Study a 5 e Preliminary Report Submitted p Project Definition.a Complete System Analysis - S Final Report Completed &Submitted e. Evaluation A B Proposal Complete A Phase C & D Contract Awarded R Long Lead Component Design & a. Development e Other Components Subsystem Development Testing B/C System Development Testing 'O P h Component Qualification a System Qualification c/D p Fabrication, Assembly &'Checkout h a Prelaunch Operations Launch MEDUSA I A D Launch MEDUSA II g Development r 0 Prototype u n Qualification d Ground Receiver Manufacture 5 Installation in Schools ys Installation in Homes t Table 9. 1. 1 Summary Schedule

APPENDIX A COMMUNICATIONS A. 1 THE EARTH'S GRID AS SEEN FROM THE SATELLITE The instrument used to map out the beamn in Figure 2. 2. 1 as wr-ell as to plot and evaluate other tenat ive beams is the drawing of the earth's grid of longitude and latitude as seen from the satellite. This draw~ing is presented here as Figure A. 1. 1. It is a projection of the earth's grid onto a plane perpendicular to the line joinirng the satellite to tIhe cenaer of the earth. The lines of longitude and latitude appear as ellipses on this drawing. The lines of longitude are marked relative to the satellite, i. e. the satellite longitude is marked 00, and 200 E means 200 east of the satellite location. In this way the same drawing may be used for any satellite location. To map out a beam, one superimposes on the grid of Figure A.1. 1 the beam cross sections of Figure A.i.. 3 These beams are labeled according to their angular diameter. They are plotted for a beam centered on the subsatellite point. For other points the beam sections should be slightly enlarged and their shape slightly elliptical. These effects are negligible. Figure A. 1.2 also shows cointours of equal satellite elevation on the earth. as viewed from the satellite. A. 2 FEED DESIGN The requirement of circular polarization when combined with that of a 105 MHz bandwidth dicitates an axial mode helix as the feed element. However, the half angle of the rim of the reflector as seen from the feed is 590 (Figure A. 2. 1), which is too wide for a helix, For this reason a helix in a horn is chosen as the tentative design. The dimensions of the helix are shown in the figure. They were chosen following reference 1. The diameter of the circular waveguide is chosen so as to allow a cutoff free space wavelength oEf 50 cm ('the longesi in our band is 40 cm). The length of the circular waveguide is chosen to allow over one wavelength (X = 37.5 cm, 800 MHz cerirtral wavelength, corresponds to Xg = 56. 8 cm). Finally, the diameter of the opening o~f the horn is chosen to provide a 10dB half width of 590. The equations used to relate the above quantitites were X =.. 706 x (diameter of vwaveguide). x= (-y~~ ~'q g c 100

70 7~~~~~~~~~~~~~~~~~~~ 60 /~~~~~~~~~~~~~~~~~~~ /-/.~~~~ 50~ ~~~~~~~// / j ~~~~~~~~~~~~~~~30~ ~~~~~~~~/ I~~~~~~~~~~~~~~~~ / I i I ~i i 800W 30W 20W 0W 0E 200E 30 40E 50E. E7 E Figure A. 1. 1 Earth Grid

Figure 1.2 Ground Elevation Angles Figure A 1. 2 Ground Elevation Angles

Figure A. 1.3 Beam Sections 103

Feed i1(r —.5c4 m IEm HegIlol ix n coldctor.,.i. o..t.r: 7. r5i.e I + 47 4 6elr8nl, 1. 5r.2 cmI I7. oan=s56!. SCALE 1:10 foot 52' i i 'SCALtE lcn per foot F' i A2. 1: ee] es ia... 104

J1( sin 4) 2 rD = 0, 1 -k sin where D is the diameter of the aperture, and 4 the 10dB ha.lf angle. The West Coast feed (aimed at San Francisco) must be displaced 4. 20 corresponding to 0. 965' in a direction 51~ west of north as vievwed from the satellite. The 0. 965' displacement is not enough to allow for another horn. For this reason a bare helix is used for the displaced feed, with the tip of the helix in the focal plate. The displaced helix is a scaled down version of the main feed helix. The scale factor is 8/il = 0. 727 corresponding to the change in frequency from 800 MHz to 1100 MHz used on the uplink. The 10dB angle of the bare helix is 490 which covers only (tan 490/tan 590 ) - 0. 48 of the area of the 30' reflector reducing its gain from 38. OdB to 34. 8dB. A. 3 LOSSES Estimates of the propagation losses are given in Table A. 3, 1. They are based on an elevation angle of 70 The ionospheric absorption (line 1) is due to polar cap absorption events. These have been known to reach gdB at 100 MHz. The effect is inversely proportional to frequency and to the sine of the elevation angle, which leads to 0. ldB at 800 MHz, At higher frequencies the effect is negligible. The absorption by oxygen comes from the curves in Figures 7-2. 15, 16 in Reference 2 o Absorption by water vapor (Figure 7-2. 14, there) is negligible. The estimates of refraction loss (line 3) comes from Figure 7-2. 17 in Reference 2. This effect is independent of frequency. Cloud losses (line 4) are neglected at low frequencies. 'The value of 2. 5 GHz is adapted from GE's estimate for the Demonstr+ation Service to the U. S. by adjusting four the lower elevati n. The figure for 80 4 GHz is taken from GE's Community Service to India, unadjusted. This is a rough estimate, and it is hoped that the tropical climate of India increases the loss in the same amount as does the high latitude of Alaska. This is probably an underestimate in the rainy Southern panhandle of Alaska. (In any case, the frequency of 8.4 GHz is not actually used in the projecst. ) Faraday rotation does not affect our circunlarly polarized beams. However, a 0.5 dB loss due to imperfect circular polarizatlion is included (line 5). The effect of fading (line 6) affects only the lowest frequencies. Line 7 shows the total propagation loss. 105

Losses due to pointing errors of the ground antenna are given in Table A. 3. 2. The pointing error is made up of 0. 150 allowved at the ground antenna together with a maximum 2. 1 daily variation in the eievation of the satellite due to orbit inclina-ion. The resulting 2~ 250 error makes the 8.4 GHz system unusable and leads to a high loss also at 2. 5 GHz. For these frequencies the table shows alsoP loss estimates at a pointffig error of O. 5. To limit the error to this amount,, gnorth- south stationkeeping of the satellite becomes necessary, No pointing loss is included for the main satellite reflector. The 0. 10 error in pointing (sec. 4. 5) is taken into account by decreasing the effective diameter of the half power beam (sec. 2 3). No separate estimates were mnade for ohmic and imperfection losses of the antennas. These are lurnped into the antenna efficiency parameter estimated at 55% for all antennas. Table A. 3. 1 Propagation Losses (dB) Frequency GHz 0. 8 1. 1 2, 5 8.4 1. Ionospheric absorption 0. 1 0. 0 0. 0 0. 0 2. Absorption by oxygen 0. O 0. 4 0. 4 0. 4 3. Refraction loss 0. 1 0. 1 0 1 0. 1 4. Cloud loss 0.0 0O 0 0. 7 3. O0 5. Polarization mismatch 0.5 0. 5 0. 5 0. 5 6. Fading 0.9 0.9 0. 0 0. 0 7. Total 1. 6 1. 9 1. 7 4. 0 Table Ao. 3 2 Pointing Losses of 5. 2v Ground Antenna (dB) Frequency GHz 0. 8 1. 1 2.5 8.4 Pointing error 2. 25 0. 2 0.4 2.0 - 0. 5~ 0.! 1.1 106

A. 4 NOISE ESTIMATES Estimates of the receiver antenna noise temperature for the various link calculations in this report are presented in Table A, 4. ii. The estimates are broken into contributions from sky, earth, ohmic loss in the a.nternnaa and man made noise. Sky noise temperatures are composed of galactic and cosmic noise, and atmospheric noise. In the case of uplink calculatsion the atmospheric contribution is absent. For downlink calculations the atmospheric noise tem.perature is taken at an elevation of 70. These sky temperatures are presented in Table A. 4. 2. The data comes from Figure 5 in Reference 3. The sky and earth temperature contributions to the antenn)a noise in Table A. 4. 1 use the sky temperatures of Tabl]e A.. 4. 2 and an earth temperature of 290 K. These are multiplied by the fraction of the antenna field of view (weighted by gain) subtended by sky or earth respectively. In the case of 0. 8 GHz, 30% earth and 70% sky is used to agree with GE data for the same anthenna under the same conditions in their Alaska project. A.t higher frequencies the same 5. 2' antenna has a narrower field of sight and sees onl.y sky. In the uplink calculation the 30' reflector acting as re ceiver sees about 35% of sky over the rim of the earth (sec. 20 2). The telemetry ground antenna is high gain and sees only sky. The telemetry satellite antenna is isotropic; the earth covers about 0. 6% of its field of view. Man made noise is broken iynto side lobe contribution and main lobe contribution. The latter is present because of the low elevation, The figures for 0. 8 GHz are again those used by GE in their Alaska study. They arrive at the main lobe contribution by assumirnng that mann-made noise flls half of the antenna field of view. This assumption is retained also for 1;he higher frequencies, although it may be overconservative. In any case, man made noise decreases at higher frequencies to insignificant levels. The assumption that it goes down in proportio n to f 2.3 (f is the frequency) has een rused to obtain the figures for the other frequiencies4 In addition to the antenLna n.oise Fw must consider the zoise temperature of each receiver. We approximate this by the noise termperature of the first stage of amplification. In the crucial case of the TV dowvi. ink.calculation itis'flt:that thhe 2900I K' postulated by GE in their India project5:i:~ tooolOw-, o.:En the preseiit p erojct, 600 K are allowed, The noise contribution of the telemetry receivers is estimated on the basis of noise figurkes and operating temspe aturel s as showr iL Table A. 4. 3. For these cases one adds the receiver noise temperature to the antentna noise temperature to obtaln thE total noise tnemppe atu -e that appe a-s in the l1i nk calculations~ 107

In the case of the TV uplinlk onje n-iulit also take into ato~a the noise contribution of the multiplexe r The multiplexer involve s a los s cf. 4dB_ corresponding to 27. 5%. One therefor e must combine 27. 5% of the. rultiplexer maximum temperatur e of 3 11 K with 7 2. 5 % of the antelna. te.mperab p ue of 2250K to obtain 2500K. To this one adds the temperature of - the firsst transistor amplifier; with a. noise figure of 6dB and operating temperature of 311 c1K the noise 'temperature is 12370K. Puttying the two contr ibutions together one finds 14870K as the noise temperature to be used in the uplink calcaiulation Table A. 4.1 Receiver An'tenna Noise Temperatures ( ~K For the Link Calculations Frequency GHz 0.8 0. I 2. Z 203 5 8_ 4 Link down up u1 dM dow-r down Sky 45 8 5 25 24 30 Earth 87 188 0O 0 0 Ohmic 29 29 3 4 29 29 29 Man Side lobes 29 0l3 2 made Main lobe 25 2 22 18 1 Total 442 225 3 9 B1 18 _ _ 103 94 Table A. 4. 2 Sky Temperatures (G~K) Frequency GHz 0.8 1. 1 2. 2. 3 z 5 8.4 Link down up p down d own down down m et vtel em et v) Galactic and 48 23 5 5 4 0 Co smic Atmo spheric 6 -20 20 3 Total 64 23 5 25 24 3 0 Table A. 4. 3 Telemetry Reeiv-e Noise '$tempe atures i '- -i; 1. Ground Satellite Noise figure 1. 5 2. 5 Operating tem'perature 31-00K 3 100K Receiver noise temperat~ure 1550K 065~K 108

A. 5 LINK CALCULATIONS AT OTHER FREQUENCIES Table A. 5. i shows dow-i nk callculatLons at the frequencies 800 MHz (UHF), 2. 5 GHz (S-band), and 8. 4 GHz (X-band). All three calciatia. s are based on the FM modulation system of the prese nt project involving a noise bandAwzidth of 25 MHz. All calculations are for a 5. 2' receiver antenna and a 3, 00 half poower beamwidth necessary to cover Alaska, The receiver noise allowed is 608oK in all three cases. The antenna noise comes from Appendix A. 4. The 800 MHz calculation takes into account the pointing loss due to a satellite daily cycle of 2, I~0 At higher frequencies it is assumed that satellite stationkeeping has reduced the total poia.ting error to 0. 5~ (section A. 3). The table is arranged to show that as far as geo netrical faetors are concerned, a variation in frequency changes the diamete of the satellie antenna but leaves the ratios of power received to power transmitted (line 3S) unchanged. On this ideal level the highest frequenc(y is favored as it correspn -ids tO the smallest satellite antenna (2, 35'). When losses are taken into account, the S-band system is favored from the point of view of power conservation,9 The higher frequencies suffeir from pointing errors (even though more rigorous stationkeeping is assumed), This is more than made up for by lower antenna noise. The X-band system, in wever, also suffers from cloud losses. The compelling reason for use of UHF in the present project is cost of ground r eceiving equipme nt as explained in Sections 2. 1 and 2. 7. A. 6 COMMAND OF THE TRANSPONDER The telemetry command system will allow control of the transponder during the mission. This is to incaude the ability to sw7itch the power on and off for the components that need power as weil as the a.bility to switch the RF signal from any component to its backupo In, the case of the TWTs, it should be possible to switch the RF signal from a. qy of the three channels to any of the six TWTs (120 possibilities) and from any thhree TWTs into the mnultiplexer (20 combinations ). In the case of twos way s'v itches one commrand Should:sufi e tlo throw the switch over. In the cases wher t-here are man.y possibilities a mcre complex system has to be employed. This could i.n-volve a numbering systlem -or the different switch stat es and a small -rnumber of comnm3ma.nd s t*hat wold ake it possible to run through a sequence of states to the desired final stbat-e o g. 3 commands: increase by one, increase by ten, incr ease by a, hundred). The actual system to be used is not spe.cified at this time; wkue allo s7 tvwenty commands for the pu pose of swiltching the RF signal thsrough diffe - en TWTs. The estimated number of different commrands needed to control 'the ranspen der is given in Table Ao 6. 1. The RF switc6hes at the output cf thLe TWTs at4 to be electrom chan.-arflcal and designed to avoid lIosses, All other swithes are to be electroinic. 109

Table A. 5. 1 Downlink Calculations at Other Frequencies 1 Frequency GHz 0. 8 2.5 8.4 2 Ground Diameter ft 5. 2 5. 2 5. 2 3 receiver Gain dB 19.5 29.5 40.3 4 antenna Cost $ 22 200 200:5 Inst. cost $ 60 60 60 6 Free space attenuation dB -182. 8 -192. 8 -203. 2 7 3 + 6 -163.3 -163.3 -163.3 8 HP beamwidth 3.| 0 0 3o, 00 3.0~ 9 Trans antenna gain dB 35. 0 35.0 35. 0 10 Trans antenna diameter 29 9. 3' 2. 35' 11 Ideal power ratio received/transmitted dB -1 28. 3 - 1 28. 3 - 28. 3 12 Propagation losses dB -1.6 -1. 7 -4. 0 13 Pointing error dB -0.2 -0.5 -1.1 14 Actual power ratio dB -130. 1 -130. 5 -133.4 15 Noise temperature OK 1050 711 702 16 Noise dBW -124.4 | 26o 1 -126. 1 17 Required C/N dB 9.0 9. 0 9. 0 18 Required carrier dBW beam edge -115.4 -117. -117. 1 19 Required carrier dB beam center -112.4 -114.1 -114.1 20 Required RF/channel dBW 17./'7!6.4 |19. 3 21 Watt 59 44 85 22 Cost of ground adapter electronics $ 12 16 184 23 Total groaund adaptation costi 6 + 5 + 22 $ 95 275 425 All costs are based on GE estimaltes for an audience of 10 and lo'p.wer noise than allowed here. 110

Table A. 6. 1 Transpo:nder Commands Turning components on and off (except TWTs) 16 TWT heaters 6 TWT voltages (Anode, Helix, Collector) 6 Switching RF between components and backups (except TWTs) 12 Switching RF among TWTs 15 TOTAL 55 A. 7 LIST OF FUNCTIONS MONITORED BY TELEMETRY Following is a list of the various satellite functions monitored by the telemetry system. Most of these do need the accuracy afforded by the twenty bit word, and often infosrmation about two different functions is grouped in the same word. The list indicates the. nature of functions monitored, the number of analog readings made, and tentative assignments of the number of words and of the prime commutator or the subcommutators. Number of Words No. of Prime System Function items commutator Sub commutators Battery Pressure 2 1 Temperature 2 I Solar Array Angle 2 1 Tempe rature. 2 i Current 2 1 Attitude Control Yaw error 1 1 Roll error 1 1 Pitch error I 1 Reaction wheels 3 3 Thruster s On off 16 2 Temperature sensors 10 5 Transisto r Voltage 7 amplifiers Current 14 7 RF input i 4 14 RF output 14 14 Os cillato r Vo ltag e 4 2 Current 4 2 Output 2 2 Mixer RF input 2 2 RF output 2 2 TWT Heater voltage 6 6 Heater current. 6 3.111

Anode Voltage 6 6 Anode current 6 3 Helix voltage 6 6 Helix current 6 3 Collector voltage 6 6 Collector current 6 6 RF input 6 6 RF output 6 6 Transponder connector status 2 Telemetry equipment 16 8 8 Totals 188 24 115 Number of words available 24 128 A. 8 LIST OF COM.MANDS Command Number of distinct commands Telemetry system power up 1 Decoder enable 1 Initial deployment 1 Deploy antenna 1 Pulse jets 12 Momentum wheels (on-off) I Transponder control (see Table A. 6. 1 for details) 55 Total 92 Number of commands available 1 28 A. 9 REFERENCES 1. Edward F. Harris, "Helical Antennas" Antenna Engineering Handbook Henry Jasik, Ed., McGraw-Hill, New York 1961 Chapter 7. 2. GE TVBS Section 7. 2. 2. 13 3. Justin R. Hall and al, "Space Science Review" 8, 595 (1968) Figure 5. 4. TRW TVBS Section 2. 8 5. Table 7. 2-1 of Reference 2. 112

APPENDIX B POWER B. 1 SOLAR ARRAY CALC ULATIONS Power required = 725 watts Array Degradation (5 years) 1 - (. 795) radiation ~ (. 89) thermal cycles ~ (.852) solar misalignment (.86) temperature ~ (.90) uncertainties =.53 = 53% 7 25 watts 725 watts (:E.O. L.)= 1545 B. 0. L. 47 Cells in Series Calculations (Figure B. 1. 1) 28 volts (bus) + 2 volts (diode + line losses) = 30 volts 30 volts = 68 cells in series.440 volts/cell Each string is split in half and cells are spaced using.032 spacing. Panel Width = 2. 5 ft. Cells in Parallel Calculations (Figure B. 1. 1) 1545 watts - = 55. 2 amperes 28 volts 55. 2 amperes = 456 cells in parallel.121 amps/cell Using.016 spacing. Panel Length = 30. 5 ft + 2 ft Dummy = 32. 5 ft. B. 2 BATTERY CALCULATIONS Power required = 533. watt hours (maximum eclipse + shadowing at 5 years) 5,3,3 watt-hours = 38.'1 amp -hrs 28 volts (.5) DOD Thus, a 30 ampere hour battery was selected, sacrificing depth of discharge (DOD) for a significant weight savings. B. 3 TRANSMIT1 CE THROUGH REFLECTOR For calculation of solar array power, the following curve is necessary. Figure B. 3. 1 gives the solar energy transmittance characteristics of the 30 ft reflector from which the amount of power received through the reflector by the solar arrays was determined. 1P4

120: PEAK POWER POINT CELL TEMPERATURE 115~ F 2x2 cm, 2 Ohm N/P Cell 100oo + SILICON THICKNESS.008 inc CURRENT (mA) - ACTIVE AREA 3.9 cm2 80 l SUNLIGHT SIMULATOR 140mW/ 60 40 20 0 I. FIGURE B. 1.1 0 e1.2.3.4 5.6.7 VOLTAGE (volts) LOCAL ANGLE of INCIDENCE vs TRANSMITTANCE of SOLAR ENERGY THROUGH the ANTENNA MESH 80 PERCENT 70 TRANSMITTANCE 60 -40 30 - 20 - 10 0aI.. 1, 1 — || ' FIGURE B.3.1 O 10 20 30 40 50 60 70 80 90

APPENDIX C ATTITUDE CONTROL C. 1 CALCULATION OF THE CENTER OF PRESSURE The satellite is symmetric about the z-axis, so the center of pressure will be located somewhere along it. Using the bottom surface of the satellite as a reference the following table results: Satellite Area Distance from (Area) x Section (ft2) Reference (ft) (Distance) Body 12. 00 o 50 18. 00 Feed 4. 20 17. 70 74.60 Reflector Torus 4.60 3, 35 15.40 Solar Panels 18 3.00 0.50 91.50 Feed Support 1.00 9.85 9.85 Reflector"' 61.00 5. 6s 344.00 Burner II 16.40 3.50 -57.40 28Z2 2 ftz 495 95 ft3 Therefore, the distance from the reference point to the center of pressure is, 495,95 495. 95= 1.76 ft 21. 1 in 282.20 Since the reflector is slit transparent, it was calculated to have an effective area of 61. 0 ft C. 2 MOMENTS OF INERTIA Moments of inertia about the principle axis were calculated to be: Izz (1295 slug-ft )max ' (1256 slug-ft 2)in I = 1608 slug-ft2 IXX =454 slug-ft )ma, (2415 slug-ft )min C. 3 DISTURBING TORQUES a) Solar Pressure = (A)(M)(P) where - = torque due to solar pressure A = area of the satellite M = CP-CM offset (0. 075 ft) P = radiation pressure constant for a tota]Lly reflective surface (1. 805 x 10-7 'lb/ftZ) 116

About the pitch axis; Tmax (Ax M)max x P = (282. 2 x 0. 075)(1. 865 x 107) ft-lb -6 = 3. 94 x 10 ft-lb This torque is cyclic each day as a sine curve, with nodes at local noon and midnight. This is a result of the effective CP-CM offset going to zero at these times. Torque Time Noo Midnight Midntght The total momentum absorbed in one 12 hour period is 4 M = T (4. 32 x 10 sec)(2/wr) rmax max which is the area under the sine curve for a 12-hour period. M = 0. 1085 ft-lb-sec. max This momentum buildup will be cancelled during the next 12-hour period. This results from the fact that at the node the effective moment arm shifts to the other side. The net result over a 24-hour period is, therefore, zero. About the yaw and roll axes the torques are effectively zero. b) Gravity Gradient The equations for the gradient torques are: T = 1. 5co (I - I ) sin 20 pitch o0 roll yaw pitch T = 2 OW (I 2 )I sin 20 yaw0 oC pitch roll yaw =20 I 2( - I ) sin 20 Troll = o ( pitch yaw) roll (these equations ignore any second order coupling effects) -4 where 0q = orbital rate (0. 728 x 10 rad/sec) I -= moments of inertia 0 - error from desired orienlsation

_0 0 0 = 10 max 0 6=.- mraxOe =.m li max yaw roll pitch T = 1.5 (.728 x 10 -4)2 (2454 - 1256) (.00349) pitch -6 = (3 32x 10 ft -lb)m -4 2 T = 2(. 728 x 10 ) (1 608 - 24:.15)(0 0349) yaw =(-3.19 x 10 fti-lb)max T1l = 2(. 728 x 10 4) (1608 - 1256) (.00698) = (2.60 x 10 ft-lb )max This gravity gradient torque is also cyclic. Thus, the net effect over a period of time is zero. c) Thruster Misalignment During the east-west station-keeping maneuvers, or on any translational motion or correction, a torque might be exerted on the satellite due to a thruster misalignment. This torque, and the resulting momentum, can be calculated. An assumed misalignment of 0. 1 inches would have the following effect: T= thrust x misalignment = (. 1 )(. 1 in)(1 ft/12 in) = 1. 65 x 10-3 ft-lbs M = r x At F = ra = M AV/At Ms = satellite mass 36. 3 slugs At = MSAV/F AV stationkeeping requirements for one year is at most 16.9 ft/sec F = 2(0. 1 lb) = 0. 2 lb M =TMsAV/F = 5. 03 ft-lb- sec/yr. d) Other Distrubing Torques Included here are: 1. RF Radiation Pressure Torques, which in this satellite is effectively zero, because the RF pressure acts along a line through the center of mass, so there is no moment arm. 2. Gravitational and aerodynamic effects are also effectively zero, because at synchronous altitude they are very small. 3. Meteorite impacts are dis regarded because of the low probability that it will ever happen. 4. There are many other effects that act on the satellitethat cause torques, bLut they also are so small that they can be neglected. 118

C. 4 MOMENTUM WHEEL SIZING The major goal in sizing of the wheels is making sure that the cyclic torques do not exceed the saturation momenrtum of the wheel. The major cyclic torques that occur are solar pressure and gravity gradient torques about the pitch axis. This results in a maximum momentum input of 0. 155 ft-lb-sec. The wheel's momentum should exceed this by a good margin of safety. The wheel chosen has the following specifications: Type: 3 Bendix Type 1 791 290 Advent Reaction Wheel Momentum:. 456 ft-lb-sec at 1000 rpm Inertia: 0, 0!39 slug ft2 Weight: I Q. lbs Dimensions: 7. 5" dia x 3. 5 Power: 4. 6 watts at 26 v Operating Temp: 50~F to 1200F Considering the weight and power requirements, this seems to be a good wheel for the satellite. C. 5 UNLOADING OF THE MOMENTUM WHEELS M = momentum load of wheel = 1.456 ft-lb- sec M = TAt, At = M/T T torque produced by thruster during unoading 0.4 ft-lb for yaw thruster = 0. 45 02 ft-lb for pitch thruster = 0. 350 ft-lb for roll thruster Azt = 3. 64 sec for yaw thruster burn time 3. 24 sec for pitch thruster burn time - 4. 16 sec for roll thruster burn time Also, FAt/Isp MF F = force of thrusters (2 x 0. I lb thrusters) Isp specific impulse of fuel (Isp 150 sec for hydrazine) MF = weight of fuel required for each unloading MF - (0. 2)(At)/150 lb = 0. 00485 lbs for yaw wheel urloading 0. 00432 lbs for pitch wheel urnloading = 0. 00554 lbs for ro ll wheel ualoading

The only envisioned unloading of whee&Ls Nwill be due to the thruster misalignment torques experienced in east.vwest station-.keeping. Due to this, the wheel will be unloaded about 3. 5 times a year, thant is, i7. 5 times over a 5 year period. Fuel required for this unloading of the yaw Niheel over the five year period is: MPF 5 yr = (18)(0. 00485) lb 0. 0875 lbs C. 5 ATTITUDE SENSING EQUIPMENT SPECIFICATIONS Sun Sensors Type: 6 Adcole Model 10042 Aspect Sensor 1 Adcole Model 9378 Aspect Electronics Dimensions: Sensor 3, x 2. 25't x.625" Electronics 4. 5 " x 2. 843" x 7. 75" Accuracy: at a sun angle of 32. 5 accuracy of +. 5 Coverage: Field of view 1280 x 1280 Power Required: Total system: 3 watts at 26 v System Weight: 8 lbs Operating Temp. Sensors: 185F to -161 F Electronics: 1853F~ to -220F Earth Sensor Type: 1 Barnes Synchronous Altitude Horizon Sensor Model 1 3- 163X Dimensions: 8o 75' x 3. 75" x 3. 5" 7. 0" Accuracy: +. 050 pointing iith 200 pitch and roll deviation Coverage: Li-near Range + 20~ in pitch and roll axis Power Required: 3. 5 watts at 26 v System Weight: 9. 0 lbs Operating Temp: 14Z F -to i0 O Rate Integrating Gyro Type: 1 Honeywell G0G87 Miniature Integrating G.yro Dimensions: 3o 56 " x 2 15" dia Input Freedom: + 100 Power Required: ~ 3 watts System Weight: I lb Operating Temp: -65 F to +150~F 1 20

Rate Gyros Type: 3 Honeywell GG440 GNAT Miniature Rate Gyro Dimensions: Mounted in a GG312 Gyro Package 3, 5 x 3,62 x 3. 211 Input Rate: 6 per sec Linearity: + 0. 5% Full Scale Power Required: 10. 5 watts at 3. 6 watts each System Weight: 2. 6 lb s Operating Temp: -65~F to +2000F C. 6 IMPULSE AND FUEL WEIGHT CALCULATIONS The formula for impulse is (AV) (Ws) Ig where Ws = satellite weight =1170 lbs g - 32. 174 ft/sec and the formula for fuel weight is Wf= I sp where Isp = specific impulse (sec) 1. Injection Errors Here the AV = 127 fps, so I =(127)(1170) 1 32. 1 74 Ii = 4618 lb-sec and taking I = 210 sec, sp 4618 fl 210 W = 21.99 lbs. f i 2. Walking Orbit For this maneuver, AV = 20 fps, _ (20j(1170) 2 32 -2 174 12 = 730 lb-sec

Again taking I -210. sec sp 730 Wf2 21 0 Wf = 3. 46 lbs. 3. E-W Station-keeping In this case, the AV is at most 16. 9 fps per year, which for five years is 84. 5 fps-. Therefore, I (84. 5)(1170) 3 32. 174 I3 = 3073 lb-sec Since the small thrusters will be used here and since they are operated in pulsed mode, the specific impulse is I = 150 sec sp W 3073 f3 150 Wf3 = 20.42 lbs 122

4. Large Thruster Misalignment Misalignment torque due to the ten pound thrusters would show up during the large translational maneuvers. A possible misalignment of 0. 3 degree in the thrust vector is chosen for calculations. The torque produced is T = 2(9. 66 lbs) (22 in) (2 in )(sin 0. 30) T = 0. 206 ft-lbs. Since the momentum from this torque is W AV M = TAt and At F g then, TW AV s F g M = (0. 206)(1170)(150) (19. 32)(32. 174) M = 58. 1 ft-lb-sec. The small thrusters would be used to correct the misalignment torque. Using the minimum torque available in order to determine the maximum fuel weight used, T = 0. 175 ft-lb, the fuel weight is FM W T I, where F = 0. 2 lb f5 T I sp _ (0. 2)(58.1L w = fS (Oo 175)(150) Wf5= 0. 443 lb. This is an impulse of I5 = (0.443)(150) 15 = 66.5 lb-sec. 123

APPENDIX D THERMAL CONTROL Symbol s q = heat from solar radiation s qe = earth emitted radiation q = earth reflected radiation qdi-= heat dissipated from equipment on board qR = heat radiated into space F = shape factor AR radiating area A = projected area facing the sun E = emittance coefficient a = absorptance coefficient -6 watt r = Stefan-Boltzman constant - 5. 02 x 10 ft R T = temperature S= solar constant = 130 watts/ft2 C = specific heat p m = mass t= time The instantaneous heat balance equation is dT q 4q + -q- C m- D. qs e qdis *R p dt Since temperature changes will be quite slow except when the satellite enters eclipse, fairly good results can be obtained by setting dT/dt = 0. Also, in earth synchronous orbit, qe and qr can be neglected relative to qdis and qs (except during eclipse when qs = O and di is greatly reduced). Now equation 1 becomes qs qdis qR D. 2 where q A aS s p and 4 qR = FARE oT4 124

Main Body Heat Balance Case 1 (see Figure D. 1) The main body radiators are coated with Z-93 white paint (a = 0. 2, E = 0. 9). Because of the presence of the antenna and the solar array, which restrict the radiaor 's view of space, F was taken to be 0.45. 4 qdis + qs = FARE T qdis= 94. 6 watts ( see Table D. 1) dis q = 13.7 watts s AR 6 ft T =860F Case 2 (see Figure D. 2) AR 6 ft qs = 43 watts q - 94. 6 watts dis F = 0.60 4 qdi + q FA RET T= 80 F Feed Body Heat Balance To simplify the analysis, it is assumed that the entire feed structure is at uniform temperature and that the K-truss attachment points are designed for negligible thermal conduction across them. Case 1, Maximum solar input (see Figure D. 3). A = AI sin O + AZ cos 0 p 1 2 A1 3 - 3.14ft A2= 4-0 f~ dA = 0 = 0 = 38 1 dO 125

A 5. 08 ft max To find T max F = 0. 9 AR 14. 24 ft2 q = aA S 132 watts s P max q s = 447 watts dis 4 qdis + q FAR ET T = 1030 F Case 2, Steady-state feed body temperature when q = 0. F = 0. 9 AR = 14. 24 ft2 q = O s q. = 447 watts dis qdis + = FAR EOT T = 67 F Thermal Time Constants AT qR At C m p Main body q = 94. 6 watt:s C = 0. 23 BTU/lbm-P p m= 440 Ibm AT (94.6)(3,413) 3.. F/hr At (0. 23)(440) 126

Feed body qdis 447 watts q = 81.6 watts (just prior to eclipse) q = qdi - 529 watts C = 0. 23 BTU/lbm- F p m= 130 Ibm A T (5 29) (3 41 3) 6 00F/hr At (0, 23)(130) A Lower Bound on Feed Body Temperature A T T T TT min e At max T - 900F = temperature at the moment of eclipse e T = 1. 2 hrs = maximum time spent in eclipse max Tmin = 90 - (1. 2)(60) = 180 F This is still within the survival limits of all the components on the feed. (see Table D. 1 for the survival limits of all components in the satellite ). 127

Solar Array Drum 1. 5L,, I Solar Radiation Earth 2 Radiator w Solar Radiation.4751' 8.10 N N Earth S 128

Solar Radiation TI___ 4 _ _ _ _ 2' ~.. Radiator Z.4~~ 2.~~5'4~~ ~Solar -Radiation 1. 23. 5 N Figure D. 2 Main Body Heat Balance Case 2 129

.,....... Solar Radiation A = projected area of the end plate = A sin 0 A P = projected area of the end plate - A1 sin 0 A-2 = projected area of the cylinder = A2 cos 0 pZ22 A 3.14ft A2 =4. 0 ft2 Figure D. 3 Feed Heat Balance Case 1 130

Table D. 1 Summary of Heat Dissipated Feed (watts) Main Body (watts) Communications 402. 0 6. 1 Telemetry and Command 0. 0 26. 1 Attitude and Control 4. 5 34. 3 Power 40. 28. 1 Total 446.5 94.6 Table D. 2 Heat Dissipated and Temperature Limits of Components Communications Unit Heat (watts) Temp. Range ( F) min. max. 3 TWT's 11 2/tube -4 +185 Multiplexer 66.0 -4 +158 7 Transistor amp's 0. 5/am -4 +158 Mixer 0. 0 -4 +158 Oscillator 2.6 -4 +158 Filte r 0.0 -4 +158 Total -408. 1 Main body 6. 1 watts Feed 402. 0 watts Total -408. 1 watts Telemetry and Command Unit Heat (watts) Temp. Range (0F) min. max Commutator 1 0 -4 +150 3 Sub-commutators 0. 5 each -4 +150 Encoder 2.0 -4 +150 Decoder 2. 0 -4 +150 Diplexer 0.1 -4 +150 Transmitter 9.0 -4 +150 Receiver 3.5 -4 +150 Command Distributer 3, 0 -4 +150 Clock 2. 0 -4 +150 Oscillator 1.5 -4 +150 Demodulato r 0.5 -4 +150 Total 26. 1 Main body 26. 1 watts Feed 0. 0 watts Total 26. 1 watts 131

Attitude and Control 0 Unit Heat (watts) Temp. Range ( F) min.:max. 6 Sun sensors 0. 0 -160 +185 Sun sensor electronics 3. 0 -22 +185 Earth sensor 3.5 +14 +140 Rate integrating gyro 3. 0 -65 +150 Rate gyro 10. 5 -65 +200 Computer 5. 0 -4 +150 3 Momentum wheels 13. 8 +50 +120 Hydrazinre 0. 0 +35 +230 Total 38. 8 Main body 34. 3 watts Feed 4. 5 watt s Total 38. 8 watts Power Unit Heat (watts) Temp. Range ( F) Min. max. P. C. U. (on feed) 40. 0 -4 +150 P. C. U. (main body) 10. 0 -4 +150 Array drive 8. 1 Battery 10.0 +45 +100 Total 68. 1 Main body 28. 1 watts Feed 40. 0 watt s Total 68. 1 watts 132

APPENDIX E STRUCTURES E. 1 CENTER OF GRAVITY CALCULATIONS The center of gravity calculations are based on the assumption that weight is concentrated at the center of individuals components. These individual weights, times their distance out along the z-axis from the base, are added up, and then divided by the total weight of the spacecraft. Without the Burner II, C. G. = 40 52 in lbs = 51.4 in 790 lbs This high value for the C. G. would have placed to high of a requirement on the power needed by the momentum wheels. But with the Burner II left connected, the C. G. is moved down to a more favorable position. (Information on exact weight allocation in the empty Burner II was not available. Single the exact piositiori of its, C, G. was 'not know, estimates of component weight s were used. ) Location of C. G. with the Burner II still attached, 25,800 in lbs C.G. =- = Z. 0 in i, 171 lbs (this value is very close to the center of pressure (21. 1 in) ). E. 2 SIZING OF SUPPORT STRUCTURE FOR THE SOLAR ARRAY PACKAGE The entire weight of the solar array system must be carried by the support columns that hold the frame. These colums are therefore sized for this condition. To insure the necessary strength, the entire weight was assumed to be supported by the four columns next to the wall, that hold the frame at the bearing mounds. Then all the columns were made that size to insure stiffness of the entire assembly. 391 lbsJ |The column used for support is a channel beam with the following properties: I =.015915 in Area =. 145 in2 12"11 y=.345 -. 4Thu p EI _1,640,000 cr 2 144 124 lbs L - 11,400 bs ~cr.145 '78,500psi 78,500 > 66,000' 0i cr - 66,000 F 133

Considering the launch load factors the four supports must each carry the following loads; horizontal 124 lb s vertical 391 lbs The actual stress on the column is therefore given by -r = + b a b P where a- a A _ My Gb- I 391 (6) (124) (. 345) =.14 -+ 18,800 psi.145 +.015915 The safety factor is given by;,cr - 66,000 S. F. = cr 66- 1,3 51 G' 18,800 E. 3 FUEL TANK MOUNT The design criteria for the fuel tank mount was again considered to be the launch loads. horizontal 62. 45 lbs vertical 1 84. 80 lbs The columns are square channels with the following properties; I =.02865 in4 184.8 191.63 2 891. Area =. 19 in 62.45 L = 15 in 1T2 EI P =2 13, 110 lbs cr. 2 L 13,110 = 69, 000 psi cr.19 191.63 (89. 1) (15) (.5) 19. 02865 81.1 S.F. =66,000 = 2.61 ~2~ ~:4,310ps 134

E. 4 REFLECTOR SUPPORT COLUMN The loads are assumed.to be evenly distributed between the six columns. These are square channels again, which have the following properties: I =.1017 in 58. 7 lbs Area -. 29 in2 l mc TE 11,610 lbs cr 2 184 lbs 30" 11, Gc =2 =40, 000 psi cr 29 The load is concentrated approximately as shown, M = ((184) (8. 25) t (58. 7) (30)) 3,278 in lbs 1840 (3,278) (.75) - +' 30, 500 cr.29.1017 S.F.-30'00 = 1.31 E. 5 REFLECTOR The degradation in gain due to the reflector not being a perfect paraboloid is given by Lockheed as: JAG I = 21.0(N)379 (d/X) 867 where N = number of ribs = 24 d = diameter = 30. 0 ft X= wavelength = 1. 230 ft Thus, IAGF = 0.0484 db. The gain of a perfect paraboloid.is computed from 2d22 Gt= l 2 -135

where n = aperture efficiency - 0. 55 Thus, Gt = 35.08 db. E. 6 REFLECTOR SUPPORT CALCULATIONS To facilitate the requirements of a first design, the possible loading is represented in Figure E. 6, 1. The reflector torus was represented by its effective center of mass (for the purpose of these calculations the c. g. was placed four inches out from the inside radius of the torus). The resultant reaction loads are as shown. Representing the uniform loading on the faces of the lower support by point loads seemed reasonable for the purpose of this study. Note that the loading represented is to be calculated for the maximum possible loading encountered during the launch phase (i. e., 11. 025 g vertical and 3. 5 g lateral). The reaction forces at the assumed support points were calculated as a result of the eccentric loading of the reflector torus. Reflector weight - 100 lbs Maximum load experienced at.-launch; Lateral load = 350 lbs Vertical load = 1102o 5 lbs. Therefore, the maximurmn loading experienced by each support system is 58. 33 lbs lateral and 184 lbs vertical.- Note that the lateral loading may be in the plus or minus direction. For example, the eastsupport experiences a positive acceleration which implies that the west support is experiencing a negative acceleration. - This results in variations of the loading on the upper and lower supports on the opposing sides of the spacecraft. Only the maximum loads were used as design parameters. Therefore, the maximum will be indicated in parenthesis if the calculated value is not the maximum. Figure E. 6. 1 is the effective loading taking into account the six degree tilt of the reflector torus. Summing forces and moments: E F 0 - R 183 z A F 0 R -R -R 57.9 MRef = 0 = 3RA + 4RB + 1.5 RA -4(183): x ~ Y 136

Solving the above for RA, RA, RB results in: x y x RA =183 lbs y RA = 98. 54 lbs A RB = 40. 54 lbs(90. 17 lbs) x For the magnesium alloy used (AZ 31-H24) a yield strength of 29000 psi must not be exceeded. The stresses in the structure were approximated in the following way. For the lower support the loading is represented in Figure E. 6. 1 At Section A. (Figure E. 6. 2) - (1. 80)(0, 1)(3. 0) + (0. 875)(0,1 )(I. 75) (0o. 1)(3. O+ 1. 75) = 1. 46 in I= (0.) (3) + (0 1)(3. 0) (0. 34)t. 75(0 1) + (0.1)(1. 75)(.585) 4 = 0. 1 39 3 in4 - 183 (1.5)(-1.46) (183)(1.5)(1.85 - 1.46) I. 1 39 3. 1 39 3 - -2880 psi, 768 psi And for comparison consider section B (Figure E. 6. 2) y = 2. 25 in 3 2 LI = (I 5)iz3 + (0. 1)(1 5)(2 4)2] (5 3 ( 1) 1, ' 12 2. 971 in4 A = 0.1 (1.5 + 5. 3 + 1.5) 2 = 0.83 in P My 11,2 A I -9 8. 44 + (183)(7.5)(2.5) 1,2 0. 83 2 2.971 137

RB B 41f 183 lb Ref l 57.9 lb 4" A 1. 5' Figure E. 6. 1 Thickness = 0.1 in 7T7 1. 1.75" A B Figure E. 6. 2 138

1,2 = +1035.5 psi, - 1272. 5 psi The upper support calculations were made in a similar manner, with the resulting stresses at the root; "1,2 = 11707. 2 psi, -4172.8 psi All of the above calculations resulted in stresses below the maximum of 29,000 psi. As for the possibility of a 6000 ft-lb torque (72, 000 in-lb) at the time of the reflector deployment, a momrent of 1Z, 000 in-lbs/support is assumed to be supported at the lower support: 2(1.5) (0. ) (0. 1) (5. 3) lowe r 12 12 =.5664 12000 (0o75) lower 0.5664 The upper support is assumed to take little or none of the load. Thermal deflection. The thermal coefficient for the magnesium alloy is 14. 0 x 10 in/in- F. The temperature range experienced by the members is 500F - 1500F. Arbitrarily picking 700F as the original alignment position, the expected deflection is 9. 8 x 10 in/in. For the lower member this means a horizontal extension of 5. 88 x 10 in, assuming six inches from the satellite body to the inside radius of the torus. For the upper member the deflection would be 4. 63 x 10-3 in. The difference is; 5.88 x 10-3 - 4.63 x 10-3 = 25 x 10 -3in. Therefore tan = 0.000125/8 tan ~ = 0. 0001562 = 0. 00894 deg the two structural pieces being eight inches apart at the torus. 139

This would result in a maximum tip deflection of about 0. 0563 in which, in a first approximation, is acceptable for the system. At the.spacecraft/support interface, the maximum difference in thermal expansion would be expected to be; Temperature range: 80 F (inside) - 150 F (outside) Assuming the system was built at 70 F the maximum difference would occur at 150 F. Implying; For the Aluminum alloy: -6 (12.9 x 10 )(80)(5.5) = 0. 0000568 in For the Magnesium alloy: (14 x 10 )(80)(5. 5) = 0. 0000616 in which seems to be even much less than manufacturing tolerances, and thus defining the compatability of the alloys at the interface. E. 7 FEED SUPPORT STRUCTURE The feed support members were sized by examining a simple truss diagram and assuming that each section takes 1/3 of the load. 158 lbs 158 lbs Vertical Strutt 1- g 158 lbs O. D. = 4. 0 in; t =.05 in;L = 184 in fi444 lbs 444 lbs Area -. 62046 in2 I= 1.21 in4 Clamped Enl s 47r E -r = 214,732 lbs p = 2 cr L 14,732 - 23,700 psi cr. 62046 140

444 (7796) (2) 13, 566 psi.62046 1.21 23,700 S.F. = 566 1. 75 13.,566 1002 lbs Upper Diagonal Strut O. D. = 2. 35 in; t = 0.08 in; LL= 73. 158 lbs Area.5454 in2 I I.3215 in4 106 lb| pinned ends, Tr EI \\P~r 2= 6, 230 lbs cr 2 ~0 c.46,230 80 ~ (F|5 r 5454 = 11,400 psi 1002 = 5454 1840 psi *5454 i0,990 S. F. 199 = 5.97 (sized for stiffness, taper, and connections) 1177 lbs Lower Diagonal Strut 108 lbs 0. D. = 2. 5 in; t = 0. 08 in; L = 74.3 A --- I =.445 in4 167 lbs Area =.608 in2 pinned ends, 2 P _ 7 EI = 8, 300 lbs cr 2 81cr 8 = 13, 630 psi cr.608 1177 ar=.608 = 19.20 psi 141

13,630 S. F. 1920 7. 12 The other important criteria in this K-truss design is to have a fairly large natural frequency. The natural frequency of the struts are as follows; Vertical Strut (assuming worst representation for frequency response) Fixed Frequency member (4.91) 1/2 in longitudinal (L) In cps M weight per unit length. 4.91 ( (1. 21)(10.4 x 106) 379 cps..... 1.062046 Fixed pinned member 48. 2t 1/z on = 2 = 20. 17 cps transversal L The diagonal struts are considered to be pinned at both ends (transversal natural frequency is lowest). Upper Diagonal: 30. 9 EI 1/2 453 cps n (73. 1)2 o Lower Diagonal: 30.9 EI 1/ =74 23) - 49.4 cps n (74. 3) 2-M These were considered to be the problem area for frequency response. The other members should have even higher natural frequencies. 142

APPENDIX F LAUNCH VEHICLE PARAMETERS The following are the launch vehicle parameters used in this report: SLV-3A Thrust (lbs) Booster (2) 370,000 Sustainer (1) 60, 000 Vernier (each) 700 Specific Impulse (sec. ) at S. L. /Vac Booster 254. 3/292. 2 Sustainer 217. 0/3 10. 3 Vernier 191/238 Agena D Thrust (lbs) 16,000 Specific Impulse (sec) at Vac 292. 5 Burner II Thrust (lbs) 9600 Specific Impulse (sec) 286 143

APPENDIX G ORBITAL ANALYSIS G. 1 ORBITAL PARAMETERS AND EQUATIONS Geophysical Constants 16 3 2 t = 1.40765392x 10 ft /sec = E = 2rwadians/sidereal day = 7. 2921 x l0-5 radians/sec TE 1 sidereal day = 86164. 09 sec rE = earth radius 3443~ 9367 nm 2. 09257 x 107 ft 1 nautical mile (International) = 6076. 11 ft Circular Parking Orbit Parameters (Nominal) r = 3543.9367 nm pkg Vlcpkg = 25,567.7 ft/sec (circular velocity) Tpkg = 88 min 1t sec pkg Geostationary' Orbit Parameters (NIominal) T TE - 86164.09 sec geo0 E =geo =E 7. 2921 x L0 rad/sec r = 22, 766. 90 nm = 1. 3833421 x 10 ft g eo Vlc = 100870 5 ft/sec (circular velocity) geo 8 h = 19,322. 97 nm = 1. 17408 x I0 ft geo Keplerian Orbit Equations (1) r V =r V a a p p (2) 2a- r + r a p (3) VIV c =-i (circular velocity) (4) V- V 1 / 2- r/a 3/2 (5) T G. 2 CALCULATION OF OPTIMUM PLANE CHANGE AT PERIGEE AND APOGEE OF THE HOHMANN TRANSFER ORBIT If a satellite is able to make a plane change just once, it shouid do so at the highest altitude possible where the velocity vector, whose direction must be altered, has a smaller magnitude. However, MEDUSA will make two burns 144

(the first is to kick from the 100 nm parking orbit into Hohmann 'Transfer Orbit; the second is to kick into geo~oaifonary orbit at apcogee of the Hohmann Transfer Orbit) where a plane change could be made. As a result~, the total characteristic velocity required to reach geo-tciona.ry altitude in the equatorial plane can be minimized by making appropriate plane changes at perigee as well as apogee of the Hohmann Transfer Orbit. Following is t'he calculation of the plane change at perigee (a) and the plane change at apogee (p)e At Perigee At Apogee C I Y1 f 2V 2 2 1x x2 -!1;Y2 C1 is the required perigee velocity for a Hohmann Transfer Orbit with apogee at geos kionary altitude and perigee at 100 nm. C2 is the local circular velocity at 100 nm C3 is the local circular velocity at geo/adioQnary altitude C4 is the velocity the spacecraft has at geosaionary altitude by virtue of that being its apogee of the Hohmann Transfer Orbit from 100 nm. a + p = (Initial inclination of the satellite orbital plane to the equatorial) - (amount that the satellite orbital plane at geostationary altitude will be biased away from the equatorial to neutralize the solar and lunar out-of-plane perturbing forces) a +f = 28.50 - 2. 130~ 26. 370 ~/-. - 2 z 2 2 2 AVJ y= +X ( -C C1 sin a + (C cos a C2) YI 1 2 C2 JI2 2 \l 2 2 2 AV y2~(X C) C3 sin + (C cos- C4) To minimize AV1 + AV2 which is a function of a only, set d/da (AV1 + AV2) 0 2C1 sin a cos a + 2 (C1 cos a - C2) (-C1 sin a) 2 C1 sin a + (C1 cos a C2) 45S

-2 C3 sin, cos 5 + 2 (C3 cos 1 C ) (C3 sin 15) + o - 2 3 sC + (C3 cos p C4) 2 C1 C sin C C sin C1 + C2- 2 C1 C o2 os 23 + 4 2 C3 C4 cos Substituting C1 33,630, C 25 568 C3 0,088 C4 5236 We get: '2 2 sin a sin 2.42 - 2. 33 cos a 46.4 - 37.9 cos which can be solved numerically to give a = 10 57' 20 and P = 24. 370 Substituting these values into the equations for AV1 and AV2, AV, = 8062 ft/sec V-t (AV! + AV ) miimm = 13,812 ft/sec AV = 5750 ft/sec For comparison, had all the plane change been made at apogee, = 00 AV! = 8060 ft/sec 1V (AV + AV)V 0 = 13,940 ft/sec AV, = 5880 ft/sec So by making the best possible plane change at perigee, we benefit by 13940 - 13812 = 128 ft/sec The total characteristic velocity to reach geo sationaly altitude (in a plane biased 2. 130 away from the equatorial) is V = V + (AV + AV ) - 1340 benefit of launching with earth's rotation at 28. 5 at 100 nm latitude latitude V = 38, 040 ft/sec c G. 3 RENDEZVOUS Shown below is a diagram of the coordinate system utilized in calculating the rendezvous laneuver.

R R = Rendezvous Vehicle, MED USA T = Target Position +X | T Figure G. 2 Rendezvous Coordinate System Since it was decided to rendezvous with the point on geostAtionary at the longitude of the terminal phase initiation, contributions to the required velocity changes due to east-west errors are eliminated. It should be noted that MEDUSA may be either high or low and fast or slow with respect to geostationary, after circularization. Also, the desired on station location is many degrees to the west (-X), It is possible to write the equations of motion for a point, R, inan inverse-square gravitational force field in terms of a coordinate system fixed to another point, T, also moving in the field. The only restrictions on the analysis are that the points R and T be very close together compared to their distances from the center of the force field, and that the target, T, be in circular orbit. The rendezvous method chosen for MEDUSA does not violate these conditions. The equations for rendezvous are functions of the initial X and Y positions and the initial X and Y velocities as seen from the coordinate system on the target as well as functions of the specified time, t, between T. P. I. and R. B. M. which occurs at rendezvous. The equations for rendezvous are presented below: Xd(O)- =14y(0)[ 1 - cos ot] - [6y(O) cot - x(O)] sin cot xd(0)= 8 t[3 sin ot - (1 - cos cot)] 147

(O) 0)(3 wt cos wt 4 sin (t) - 2 X(O) (L - cos +t) t [3 sin wct -8 (] - cos ct) Wot where x(O) and y(O) are position values at T. P. I. o is the angular rate of the target, and xd(0) and Yd(O) are the velocities needed at the time of T. P. I. to rendezvous in time t. AV d (0) (0) TPI YXV Yd (0) j (0) YTPI d and NI 2 2 AV T = AV + V TPI XTPI YTPI The rendezvous braking mnaneuver requires the following velocity changes: AxMV = -3d (0) - 6coy(0)] + [ -2yd(0)] sin ot RBM +[4>d(0)+ 6coy(0)] cos wt A= [ 2d(0) + 3y(0)o] sin w t + [ d(O)] cos Wt SO: Av + AV R B M x y RBM XRBM RBM The total Delta V required for the rendezvous maneuver is the sum of the increments at T. P. I. and R. B. Mo: LWV = AV + Total = AV TPI RBM G. 4 CORRECTED GEOSTATIONARY RADIUS Definition of symbols r - geostationa.:y satellite orbital radius r - Keplarian geostationary r'adihUs J - non-dimensional correction factor (oblateness term) = 1. 08219 x 1O R - radius of earth go - gravitational acceleration at earth:s surface OE - angular rate of rotation of earth 148

2 3 go R 2 3 gR _o= (32. 174)[ (3443. 9)(6076. 1 _ 0Eo( 22 2 0E ~86164 r = 22,766.9 nm c0 2 R r =r (1+ 1 J c 0 2 2 2 r 0 (1.08219 x 10-3 )(3443. 9)2 22766. 9 (1 + )- 22, 767. 2 nrn 2(22,766. 9) The effect of the earth's oblateness is to increase the geos:tationai rtadius by.30 nm. G. 5 CALCULATION OF TIME TO DRIFT 100 IN LONGITUDE Definition of Symbols y - satellite angular position relative to the minor axis of the earth's equatorial section (positive if satellite position is east of the closest stable point; negative if satellite position is west of the closest stable point) y - initial value of y Ay - variation from ~o 2) - non - dimensional co rection factor (equatorial ellipticity term) =-5. 35 x 10 0E - angular rate of rotation of earth t - time R (2) o 2 2 o 2Tr rad Ay (9 J - 0 sin 2y ) t -10 x -. 1745 2 2 E o 360 r c 2 -. 1745 t = -6 3443o9 L 2r,2 -9(-5. 35 x 10 ) 3 9 8616 (-.99 756) 22_~_76,~) 86-64 t = 5.47 x 100 seconds = 63. 2 days G. 6 CALCULATION OF RADIAL DRIFT FOR 100 DRIFT IN LONGITUDE Definition of Symbols t - time of drift = 5.47 x 10 sec (see Appendix G. 5) J(2) -non-dimensional correction factor (equatorial ellipticity term) 2~= — 5.35 x.10 149

0E - angular rate of rotation of earth r - geostationa'yr satellite orbital radius Y - see Appendix G. 5 o - initial value of y 2 (2) o Ar = (-12 J2 sin ) 2 E 2 B o r 6 344 3.9 _ 2_ 6 Ar = -12 (-5. 35 x 10- ) ( 7 -67. Z) 86l) (-. 99756)(5. 47 x 10 )(22, 767. 2) Ar= -13.3 nm G. 7 CALCULATION OF EAST-WEST STATION-KEEPING VELOCITY REQUIREMENTS If not corrected, MEDUSA would make large angle oscillations about the 123~ West longitude stable point. The correction can be made as follows: a) The satellite drifts for t seconds at which time - r- + Ar (2)2 o wvhere Ar = (-J2 Ro/rr 0E sin Zy ) t (see Appendix G. 6. y is -47 whr o c E to ' since 1700 West longitude is west of 1230 West longitude. So both J2(2) and sin 2y "are negative. Therefore, Ar is negative and rt is less than r. As 0 a result, the period of the satellite is less than the period of the earth's rotation, and the spacecraft moves eastward as expected. At - t0 the satellite has a =1 0 2 2 tangential velocity relative to the earth of V = 18 J R /r E t sin 2y and negligible radial velocity. Hence it is in circular orbit with inertial velocity V1 AVo + (r + r) = r E + 6J(2) R2/r E2 to: sin 2~y E' c cE 2 o cE 0 b) At t - to, the satellite is injected into a Hohmann Transfer Orbit with perigee at r +F Ar and apogee at rc - Ar. A. AV1 6 J2(2) R 2/r, 0 2 t sin 2y, increase in V1 will put the satellite into the desired transfer orbit in such a way that at apogee, 1/2 a sidereal day later, another AV1 increase will give the satellite a circular velocity (at ra = rc - Ar) of V2 = -AVo + E (r - 4r) rc 0E - 6J2) R /r 2 t0 sin 2yo. Since r is greater than r the satellite will now tmove westward back towards the initial position -, as Ar approaches 0. 50

The complete correction cycle takes time 2to and requies a velocity impulse = 2AV1. So the velocity change per unit operating time is AV1 = 6J (2) R 2/r E 2 sin 2y = 16.97 Isin 2~Yo = 16.9 ft/sec/year for =J2 o c E - 0 ' = -47. 0 The theoretical derivation above does satisfactoilly describe the east-west station-keeping method. However, the final result of AV required/ year is found to be about 7 ft/sec/year instead of 16. 9 ft/sec/year. This new figure results from new information regarding earth gravity field cdharact'eriistiIcs as compiled by the Syncomr 2 satellite. This new figure for station-keeping requirements gives more insurance that attitude and control can adequately provide east-west station-keeping. 151

ACKNOWLEDG EMEN TS Mr. Rodney Ko. Bergman, Applications Engineer Watkin, Johnson Company, Calif. Professor Stuart Bowen University of Michigan, Aerospace Department Professo r Harm Buning University of Michigan, Aerospace Department Mr. John Dysinger General Electric, BSC Systems Manager Mr. Harvey Do Faram Benddix Aerospace Systems Division Mr. Claude Gabriel Fairchild Hiller Corp. Mr. A. F. Gactano, Manager Astrionics Engineering, Lockheed Missiles and Space Company Mr. Charles Hunt General Electric, Chief Engineer ATS F/G Mr. E. J. Hujsak General Dyngamics, Convair Division Mr. Harold Lanning Bendix Aerospace Systems Division Mr. E. J. Merz General Electric Systems Analysis Manager Mr. John E. Miller NASA/GSFC Mr. P. J. O'Leary General Dynamics, Convair Division Professor Richard L. Phillips University of Michigan, Aerospace Department Professor W. F. Powers University of Michigan, Aerospace Department Professor J. E. Rowe University of Michigan, Electrical Engineering Department Mr. H. F. Schulte Unive sity of Michigan, High Atilude Laboratory Professor John E. Taylor Universsty of Michigan, Aerospace Department Mr. Richard J. Simms Bendix Aerosspace Systen s Division 152

Mr. John Vidolich University of Michigan, Electrical Engineering Department Mr. R. L. Woods, ECM Product Sales Manager Huges Electron Dynamics Division, Calif. Special Thanks Go To: Mr. Richard Gibson Bendix Aerospace Systems Division Mr. Paul I. Pressel Bendix Aerospace Systems Division Professor L. L. Rauch University of Michigan, Computer Information and Control Engineering And Also To Professor Wilbur C. Nelson, for his assistance and guidance through this project. 153

Project Manager Roger Lundberg Communications:-~ I Satellite Subsystems Launch and Orbital Axnnon Katz ILarry Brinser Bill Ahrens Paul Atkins Don Johnson Ed Rafalko Dan Braun James Kourt Ross Richardson Gary Haviland Mike Van Guilder Sharon Setter Jim Ladley Frank Turkovich David Lawrence Adolph Lohwasser Reg Modlin Del Patterson john Van Roekel PERSONNEL ASSIGNMENT

UNIVERSITY OF MICHIGAN 31111111111III 1111111111151111111 775 3 9015 03483 7750