PROJECT LINUS In Greek mythology, Linus is the son of Apollo. As an integral member of the Apollo family, Linus would provide a simple and reliable communication link to the Lunar far-side. Linus used 'by permission of United Artists A Student Design Project Department of Aerospace Engineering University of Michigan Fall Term 1969

ABSTRACT Project Linus is a preliminary design study for a farside lunar communications satellite. The purpose of the spacecraft is to support the Apollo mission while it is on the far side of the moon with both voice communications and high bit rate telemetry. Besides being able to provide a communications link during the forty minutes while the Apollo is out of direct earth communications, the satellite has the capability to support a farside landing. The satellite is placed in a 4000 km halo orbit about the second libration point located 64, 500 km behind the moon on the earth-moon line. Since the moon has a diameter of 3500 km, the satellite will always appear in view of the earth and thus be able to provide the continual communications link with the lunar farside. The orbit however, is slightly unstable so it will have to be periodically corrected during the course of the mission. The 872 pound spacecraft is to be launchedfrom the Eastern Test Range with a three stage thrust augmented Thor Delta rocket in late 1972 and have a lifetime of two years. During the main course of the mission power will be supplied by solar panels. Batteries will provide the power during launch and during the brief periods of occultation. The satellite itself is a cylindrically shaped body with two solar panels attached and 3. 6 meter mesh antenna to provide the needed communication links. Thermal control is a semi-passive system. It consists of radiating areas and heaters that provid.e heat when the communication equipment is not being used. Stability of the spacecraft is supplied by three momentum wheels aligned along the principle axes of the satellite. The communication system is completely compatible with the Apollo communication equipment and the ground based facilities. A 3. 6 meter parabolic antenna with dual feeds will be used to provide both voice and high bit rate communication. The main feed will be along the axis of the antenna and will complete the communication link with the moon. The secondary feed is offset from the main axis a small amount and rotated 360 degrees per halo orbit to establish the earth link for the communications. The communication package itself is a 267 watt transponder.

TABLE OF CONTENTS 1. Introduction 1. 1 The need for a Farside Lunar Communications Satellite 1 1.2 Project Definition 1 1. 3 General Description of the Mission 2 2. Trajectory 2., Introduction 5 2. 2 Equations of Motion about the Second Libration Point 5 2. 3 Orbit Possibilities 7 2.4 Trajectory 12 2. 5 Launch Windows 21 2. 6 Shadowing Effects 23 2. 7 References 26 3. Attitude Control System 3. 1 Introduction 27 3. 1. i Axis Definition 27 3.2 Control System 27 3. 2. 1 Problem Description 27 3. 3 Trans-L2 Coasting and Midcourse Maneuver s 30 3.4 Yo=Yo Despin 32 3.4. 1 Design Decision 32 3.4. 2 System Details 32 3.4. 3 System Parameters 34 3.5 Acquisition Maneuvers 34 3.5. 1 Sun Acquisition 34 3.5.2 Canopus Acquisition 34 3.6 Z-Axis Control 35 3. 7 Attitude Sensors 35 3. 7. 1 Moon Horizon Scanner 36 3.7.2 Sun Sensors 36 3. 7.3 Canopus Star Trackers 38 3. 8 Propulsion System 38 3.9 Momentum Wheels 40 3. 10 Onboard Computer 40 3. 11 Attitude Control Electronics 40 3. 12 References 42 4. Communi cations 4. 1 Introduction 43 4. 2 Antenna Considera tions 46 4. 2. 1 System Alternatives 46 4.2.2 Pointing 47

4. 2. 3 Omni Antennas 47 4.2.4 Development 47 4. 3 System Performance 47 4.4 System Operation 47 4.4. 1 Earth to Apollo Data Relay 50 4,4.2 Apollo to Earth Data Relay 50 4.4.e3 Linus Subsystem 51 4.5 Signal Flow 51 4.6 Transponder Reliability 51 4.7 References 53 5. Power 51. Introduction 54 5.2 Power Profile 54 5.3 Solar Panels 54 5.4 Ag-Zn Battery 58 5.5 Power Control 62 5.6 References 64 6. The rmal 6. 1 Introduction 65 6.2 The Thermal System 65 6. 3 The Thermal Components 69 6. 3. 1 Heat Pipes 69 6. 3. 2 Heaters 69 6. 3. 3 Insulation 71 6.3.4 Thermal Coatings 71 6.4 Thermal Component Weights 71 6.5 References 71 7. Structures 7, 1 Introduction 72 7. 2 Main Structural Components 72 7.2.1 Support Tube 72 7. 2. 2 Equipment Platforms and Top Plate 76 7.2.3 Skin 76 7.2.4 Solar Panel Supports 76 7.2.5 Stiffened Planels 76 7.2.6 Adaptor Plate 76 7.2. 7 Support Rings 76 7.2.8 Struts 77 7.2.9 Solar Drive Mechanism 77 7. 3 Parts Specification 77 7,4 Weight Distribution 82 7 5 References 84

8. Launch Vehicle and Subsystem 8. 1 Introduction 85 8, 2 Description 85 8. 3 Launch Facilities 88 8,4 Launch Vehicle Systems 89 804, 1 Guidance and Control 89 8.4. 2 Electrical Power Systems 89 8,4. 3 Telemetry Systems 89 8.4.4 Range Safety System 89 8.5 Spacecraft Considerations 89 8.5. e Spin-up 89 8.5,2 Spa e raft Attach Fitting 90 8.5, 3 Fairing 90 8.6 Sequencing 90 8, 7 Ascent to Lunar Injection 90 8. 8 Alternative Piggyback Launch 95 8. 9 References 95 9. Implementation 9, 1 Introduction 96 9. 2 Cost 96 9. 3 Work Scheduling 96 9,4 Testing 99 9,o5 Reliability 102 9.6 References 102 Appendix A - Trajectory 103 Appendix B - Attitude Control 133 Appendix C - Communications 142 Appendix D - Power 148 Appendix E - Thermal Control 150 Appendix F - Structures 156 Acknowledgments 161

1 INTRODUCTION 1.1 THE NEED FOR A FARSIDE LUNAR COMMUNICATIONS SATELLITE Despite the remarkable success of the Apollo missions that have flown to date, there is still room for improvement in the total system. This report is directly concerned with such an improvement, i. e. farside lunar communications. The forty minute lack of communications during each lunar orbit probably is responsible for more ulcers and heart failures than anything else on the project. Because the command and service module and the lunar module are out of direct earth communication, they are completely on their own during such critical maneuvers as lunar orbit insertion, LM and CSM separation, and trans-earth injection. At present there is no way for the earth facilities to check out the systems immediately prior to these very critical operations. To alleviate these problems a lunar communication satellite is proposed. It will be able to provide the following service s. First, continual voice communication with the lunar module and the command and service module can be maintained. Also, the onboard systems of both the LM and CSM can be continually monitored. Thus, in case of a malfunction on the far side of the mo-on, the source of trouble can be immediately determined and hopefully corrected. The CSM and the LM onboard computers can be updated just prior to the important farside maneuvers mentioned above, thus eliminating errors and allowing very careful earth monitoring of the farside operations. With such a system, backside operations, both manned and unmanned, become feasible. Thus there-will be-no restriction to nearside exploration because of a lack of adequate communication facilities. Finally, with the tracking data obtained on the lunar farside, accurate mappings of the farside gravitational fields can be made. This new data will not only provide increased scientific knowledge, but will also assure more accurate lunar landings. 1. 2 PROJECT DEFINITION This project is a preliminary design study for a farside lunar communications satellite to support the Apollo missions. It is intended that this satellite be both an experimental and a fully operational satellite. In order to provide some coverage of the Apollo missions, as early a launch date as possible should be used. This allows time for minor developments, but for the most part, present~state of the art will be used.

ERRATA SHEET 1. Page 23 line 4:...motion with a magnitude +26~ (1972) with respect to the equator and a'period of 27. 3 days, and a westerly motion of 14. 5 per hour..... 2. The following change in.the block diagram: (foldout) Encoder Encoder ubcarrier l l Change to\ Subcarrier O scilats~or} 1 - iOscilator Mod Mod ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~I Ir h j 1

The following constraints have been placed on the systems. 1. Very limited technical developments 2. Low cost 3. Simple 4. Reliable 5. Two year lifetime 6. Voice communication 7. High bit rate communication 8. Proper interfacing with CSM and LM communication 9. Compatible with the eighty five foot dishes 10. Capable of backside operations 1. 3 GENERAL DESCRIPTION OF THE MISSION In 1887 Lagrange proved the theoretical existence of five libration points (Appendix A. 1). They have the unique property that at these points the forces due to gravity and centrifugal forces balance each other. Thus a mass placed at one of these points will theoreteically remainthere at rest relative to the earth moon system. The collinear points L1, L2 and L3 however are unstable. Thus a mass placed at these points will have a tendency to drift away from the libration point. Therefore, a small station keeping expense is necessary to keep the mass at these points. This project is interested solely im the L2 point located 64, 500 km behind the moon on the earth-moon line. Our communication satellite is to be in a circular halo orbit of 4000 km around this libration point. Since the moon has a diameter of 3, 500 km, the satellite remains in constant view from the earth and thus can provide the needed continual communication with the farside of the moon-(see Figure 1. 1). Besides the instability of the L2 point itself, the halo orbit also is unstable. It has a tendency to dissociate into a lissajous orbit, which becomes occulated by the moon during the course of the orbit. To correct for this an impulse perpendicular to the earth-moon plane must be applied twice per orbit, This assures a nearly circular orbit. The body of the spacecraft is a c-ylinder forty two inches in diameter and thirty inches high. Inside, fuel tanks, momentum wheels, and all the electronics are located. The thermal control is a semi-passive system with two radiating a reaso During the periods when the communications equipment is turned off, additional heat must be added to the satellite. This is provided by heaters, The main power source is two solar panels 6. 7 feet by 4. 5 feet which supply a total of 410 watts. During launch and during the times when the satellite 's in an occulted region power is provided by a 170 amp-hour capacity battery.

Linus in Halo Orbit Earth w 4 -:v — - - +~ r L2 Z Moon Line of Sight Figure 1. 1 Linus in the Earth Moon System

The communication equipment consists of two omni antennas, one 3. 6 meter mesh antenna, and a transponder. The omni antennas can be used when necessary for telemetry, ranging, and commands, The 3. 6 meter parabolic antenna has a gain of 36. 2 db and a beamwidth of 2. 65 degrees. This is sufficient to handle voice- communications and high bit rate telemetry. A dual feed system will be used with one feed fixed and lying on the main axis of the antenna and pointed towards the moon. The second feed is located slightly off axis and rotated 360 degrees per halo orbit to provide a communication link with the earth. The transponder requires 267 watts of power and operates in the RF Unified S-Band Spectrum. After the 872 lb satellite is put in trans-lunar trajectory by a three stage thrust augmented Thor Delta rocket it will be guided by a combination of gyro package, Canopus sensor, sun sensor, and moon sensor. Stability of the spacecraft will be provided by three momentum wheels aligned along three mutually perpendicular axes. These wheels will also provide the needed torques to keep the main feed of the antenna pointed toward the moon within the required one tenth degree accuracy. The axis of the solar panels must be perpendicular to the ecliptic to provide the most efficient use of the panels. To keep the panels oriented toward the sun, they will have to be rotated approximately once per lunar orbit.

TRAJECTORY 2. 1 INTRODUCTION Project Linus makes use of the second libration point which is one of the collinear libration pqints of the earth-moon system. The L2 point lies on the earth-moon line, 64, 500 km beyond the moon (see Appendix A. 1). The purpose of the trajectory analysis is to obtain reasonable estimates of the AV costs of putting a satellite at the second libration point and maintaining it in an acceptable orbit about this point. The approach taken results in AV costs that are reasonable but not optimal. A fully computerized analysis could be performed to optimize the results. However, due to the lack of time this could not be done, To obtain representative values, an analysis is carried out for a launch date of November 23, 1972. Areas included in trajectory are equations of motion about L2, orbit possibilities, transfer trajectories, launch windows, and shadowing. The reader is cautioned about the inconsistency in units, (e. g. distances in kilometers and velocities in feet per second). This convention follows the reference literature. 2. 2 EQUATIONS OF MOTION ABOUT THE SECOND LIBRATION POINT The coordinate system for the analysis is centered at the L2 point. X and Y lie in the plane of the moon's -motion about the earth. X points away from the moon along the earth-moon line, and Y is perpendicular to X and in the direction of moon' s motion about the earth. Z is normal to the X, Y plane and completes the right-handed system (see Figure 2, 1). The linearized equations of motion about the L2 point are x - 2y (1 + 2k)x = fx y + 2x - ( - k) y = fcy (2.1) z + kz cz k - 3. 19044 The homogeneous solutions for these equations contain sine, cosine, and hyperbolic terms. However, by using certain initial conditions, the hyperbolic terms can be eliminated, The resulting solutions are:

tu:4~sAS a4iue paooD I 'Z osnuTzI c/ /0

x.3433 Ay Siln yt y - A cos y t (2. 2) z = Az cos ( Zt + ao) = 1. 86265 y = 1. 78618 As seen above, the X and Y motions are coupled while the Z motion remains independent. To facilitate analytical computations, all quantities are normalized using the following: Distance Unit 384, 000 km Time Unit 4. 348 38 days Velocity Unit 3356.8 fps Angular Velocity Unit 2. 6617 x 106 rad/sec (Reference 2) In the case at hand, initial conditions are chosen such that Az = Ay ao = -90~ and the solutions become: x =.3433 Ay sin (wyt y Ay cos Cyt (2. 3) z = Ay sin COzt = Ay sin (wyt + a) a = t - t t z y 2. 3 ORBIT POSSIBILITIES The second libration point is naturally unstable and requires translational control to keep a satellite in its vicinity. Therefore, a cost common to all orbital schemes is stability control. Before discussing possible orbits, the occultlation problem must be examined. The occulted zone caused by the moon at the L2 point is 3100 km in radius (see Figure 2O 2). Therefore, the satellite must be offset at least 3100 km from the L2 point to obtain line of sight (LOS) with the earth. Four separate schemes are studied to obtain LOS with the earth. They are hummingbird, free-lissajous, halo, and combination lissajous-halo. In the hummingbird scheme, the satellite is displaced at least 3100 km in a direction perpendicular to the earth-moon line, The satellite is stationary with respect to the earth-moon system which means easy antenna pointing and simplified attitude control. However, high station-keeping costs (about 5000 fps/yr) to maintain this offset and incomplete coverage of the lunar farside make this scheme unfeasible. 7

Occulted Zone 1738 km 6378 km ' | 384, 000 km 64,500 km3100 km Occulted Zone at L2 Figure 2. 2 Occulted Zone

Unlike the stationary hummingbird, the free-lissajous scheme places the satellite in bounded motion about the L2 point. Both the X and Y motions have a period of 14. 67 days while the Z motion has a period of 15. 30 days. This period difference shows up in the phase angle, a (see equations 2. 3). Notice that a is function of time and, since w6z is less than cy, it is always negative. The changing phase angle causes the path, as seen from earth, to appear as a lissajous figure (see Figure 2. 3). The initial circular shape becomes elliptic with symmetry about the 450 line, Becoming more eccentric, the ellipse encounters the occulted zone and the spacecraft loses LOS with the earth. The ellipse continues to become more eccentric until the spacecraft reverses orbital direction and the eccentricity begins to decrease. Eventually the orbit-returns to a circular shape, only the direction is opposite to the initial circular orbit. Now the process begins all over again, only about the other 450 line. The period of this motion is about 358 days. The advantage of the free-lissajous motion is that only the inherent instability of the motion about the L2 point need be controlled. Thus, very low station-keeping costs (on the order of 10 fps/yr) are possible (Reference 1). However, occultation periods of up to 4 days and complicated antenna pointing make this'scheme unfeasible (see Table 2. 1 andaAppendix A. 5). Table 2. 1 Lissajous Occultation Times Orbit Amplitude (km) Per cent Total Time Maximum Single Occultabw~,~, Occulted (%) (days) 3690 40 4.66 4000 32 4.13 The occultation difficulties encountered in the free-lissajous motion can be averted by using a halo orbit. However, maintaining a halo orbit requires a frequency control technique in addition to the orbit stability control. Periodic thrusting, in the Z direction can reduce the Z motion period to equal the X-Y motion period and an almost circular orbit can be maintained. Both posigrade and retrograde halo orbits are possible. As seen from the earth the halo orbit appears circular. It is tipped 180 57' either forward or backward depending on whether the orbit is posigrade or retrograde (see Figure 2.4). Two modes of Z-axis period control are examined here. The first. involves a single burn per orbit, once every 14, 67 days, where the path crosses the Z-axis. The second mode has two burns per orbit, one every 7. 33 days, where the path crosses the positive and negative Z-axis. Both schemes have about the same AV cost (see Table 2, 2). Even though the

I:I z III IV Figure 2. 3 Lissajous Motion (as seen from earth) 10

double Z-axis control means twice as much cycling of the engine (50 times per year versus 25 tmrnes per year) it is selected because of its flexibility. Errors in any one burn can be corrected during a scheduled burn only 7. 33 days later. Should communication be in progress during a scheduled burn, that burn can be delayed or skipped so transmission will not be interrupted. For a delay time of over 7.4 hours, it becomes cheaper to skip the burn and implement a single burn control for that particular orbit. Table 2. 2 Orbit Maintenance AV AV (fps)Control Reason 402/yr single axis z adjust z period 401/yr double axis z adjust z period 25/yr radial X stability To maintain the orbit's stability, occasional thrust are needed in the X direction, The halo model used so far neglects eccentricity of the moon's orbit, solar and nonlinear gravitational effects. In reality, the satellite follows a perturbed path about the halo orbit. The eccentricity of the moon's orbit causes a maximum variation of about 10% of the halo orbit amplitude. Solar perturbations can be approximately 1%o of the amplitude of the halo orbit, and the non-linear gravitational terms shift the center of the halo orbit about 117 km closer to the moon (Reference 2). The stability control scheme consists of radial thrusts to prevent divergence and maintain this nominal perturbed path. With DSN range and range rate measurements available, a reasonable estimate for stability control is 25 fps/yr. The radius of the halo orbit must be large enough to prevent the nominal path from entering the occulted zone. Although a radius of 3690 km (Reference 1) just insures this condition, a radius of 4000 km is chosen to allow a safety factor and to reduce the chance of multipath interference by the moon. At this amplitude the cost for double asxis period control is 401 fps/yr. Therefore, the total station-keeping costs for a 4000 km radius halo orbit is about 426 fps/yr. Because of the reasonable station-keeping costs and continuous LOS communication with earth, the halo orbit scheme is very attractive. The last scheme examined is a combination of the lissajous and halo, In this so-called "power-up, power-down mode", the satellite drifts in the free-lissajous orbit, only requiring low cost stability co9ntrosl until there is a mission for the satellite. Prior to the mission, the spacecraft is injected into the halos orbit so that continuous LOS communication with the earth can be maintained. Now both low cost stability control and high cost 11

Z-period control must be maintained for the duration of the Ymission. After the mission, the period control is discontinued and the satellite drifts back into the free-lissajous, again requiring only low cost stability control. There are two times per orbit when a reaso6nably econormic transfer can be made between lissajous and halo orbits (see Figure 2. 6). At these points, the X and Y velocities of the lissajous motion and halo motion are identical, Therefore, the spacecraft need only be given the Z-velocity of the halo and it will be injected into the halo orbit. Transfers tb both retrograde and posigrade halo orbits are possible, but there is always a trade-off in cost. The magnitude of the velocity change depends upon the phase angle of the lissajous at injection (see Appendix A. 4). If four Apollo missions per year are carried out, each about two weeks in duration, this scheme could save on the order of 50 pounds of fuel over a two year period. With fewer missions per year, the savings would be greater. However, this scheme places added burden upon the tracking system and is more complex than continuous halo. Because of the simplicity and continuous LOS with the earth, the halo orbit is selected as the best scheme for Project Linus. This scheme costs about 25 fps/yr in stability control and about 401 fps/yr in Z-period control. 2. 4 TRAJECTORY Direct trajectories —to the L2 point are discarded because the AV costs at the L2 point are very-high, on the order of 4, 000 fps (Reference 2). Close lunar fly-by trajectories with braking burns at perilune cut the AV requirements considerably. For a given perilune radius, two possible lunar fly-by trajectories give the minimum AV requirements. One is a fast transfer with time fromn earth parking o~rbit to L2 of about 8. 5 days. The other one is a slow transfer with a total time of about 18 days. The critical portion of the trajectory is from perilune to the L2 point. The fast transfer covers this leg in about 3. 2 days while the slow transfer takes about 9. 5 days.- More reasonable results can be obtained from the two-body equations used in the vicinity of the moon if the fast transfer is chosen, Therefore, the fast transfer with total time of about 8.5 days is selected for the trajectory to L2. The trajectory can be divided into three segments. The ifirst one is the earth to moon leg which employs a pat hed conic technique that assurnes a sphere of influence for the moon of 66, 000 km (Reference 7). From perilunle to terminal phase initiation (TPI) is the second segment. Here two body equations about the moon are used. The last segment covers from 12

18~ 57' x Posigrade z z 180 57' y x Retrograde Figure 2.4 Posigrade and Retrograde Halo Orbit 13

Lissajous Surface Posigrade Halo Orbit x ~~~~14 ~ 14

Halo Implementation Halo Implementation Figure 2. 6 Lissajous to Halo Transfer 15

TPI to point of injection (POI) into the halo orbit., This segment employs relative motion equationi which take into accouritf both the earth IS and moon'ts sinfluence. The earth to moon leg for the fast transfer takes about 5. 3 days. The trajectory utilizes a 200 km parking orbit. A characteristic quantity for trajectories is the C3 energy which is twice the energy per mass (C3 = V2 _ 2 p/r). This transfer has a C3 energy of -2, 02 km2/sec2 and requires an injection velocity of 35,815 fps (Reference 7) (see Figures 2.7 and 2.8, and Table 2. 3). Table 2. 3 Earth to Moon Trajectory Time after Distance From Velocity Y Distance From TLI (days) Earth (km) (fPs) (degrees) Moon (km) 1 200,843 4584 75.2 263,557 2 291,432 2780 73.1 219,542 3 346,784 1741 67.0 165,095 4 377,857 998 51.4 101, 357 5 388,508 296 2 -8 0. 7' 31,109 5.33 390,656 7757" 0 " 1,925 (pe rilune) Ace~~ '-y = flight path angle with respect to moon Translunar injection (TLI) occurs 7 hours, 38 minutes and 14 seconds, Greenwich Mean Time, November 23, 1972. One day into the transfer, a midcourse correction (MCC) of up to 50 fps is planned to correct any injection errors. A second MCC of about 5 fps is planned for 10 to 12 hours before perilune. This time period is needed so that ground tracking has ample time before perilune to study the motion and weigh the accuracy of the correction. This information is incorporated into the perilune, or vernier burn, which serves three purposes. It nulls out any small errors picked up by tracking, corrects the direction of the velocity, and injects the satellite into an elliptic transfer to the L2 point (see Table 2.4). 16

y (x 105 km) 2 MCC#l I _ 2 1 0 x (x 10 kinm) Figure 2. 7 Earth-Moon Trajectory (x-y plane)

MCC#1 (x 103 km) noo E MCC#Z! 3 2 1 x (x 10 kinm) Figure 2.8 Earth-Moon Trajectory (x-z plane)

Table 2.4 AV Sequence Reason Approximate Time AV (fps ) MCC #1 (1 day) 50 MCC #2 10-12 hours before perilune 5 Perilune (Vernier) 5.3 days 454 MCC #3 3/4 day after perilune 5 TPI 1 day after perilune 172 MCC #4 10-12 hours before POI 5 POI 3. 23 days after perilune 678 z-axis control 401/yr 802 orbit stability 25/yr 50 2221 fps for 2 yr 154 safety margin 2375 fps Elapsed time between translunar injection and perilune is 5 days, 7 hours, 59 minutes and 47 seconds. Nominal perilune conditions are an altitude of 185 km,- a right ascension angle of 470 30' (measured from the earth-moon line), and a declination angle of -48' (measured from the moon' s orbital plane). As can be seen from the last value, the trajectory is out of the moon' s orbital plane (MOP). A retro burn of 454 fps at perilune reduces the velocity from 7, 757 fps (relative to the moon) to 7,305 fps -and corrects the velocity's direction. This perilune burn injects the spacecraft into an elliptic transfer to the L2 point. The plane of the trajectory is tipped 48' upward from the MOP (see Figure 2. 9). The terminal phase initiation (TPI) burn is applied approximately one day after the vernier burn at perilune. Its magnitude is dependent on orbit injection. For a posigrade injection the TPI burn is 172 fps, but for injection into a retrograde orbit the magnitude of the burn is 169 fps. TPI is the beginning of the crucial part of the trajectory and any sizeable burn induced into the system at this time must be checked by tracking well in advance of L2 so that corrections can be applied if necessary. This is one of the reasons for performing the burn over two days prior to reaching the orbit (see Table 2.5). 19

Trajectory from Perilune to POI as seen from Earth MCC #4 POI T ~PI ~z TPI L T MCC #3 Perilune Burn POI MCC #4 _ Q z z MToon # _ i ~~~Figure 2.9 Moon-POI (Retrograde) (3 — "' Trajectory Perilune Burn 20

Table 2. 5 TPI z (fps) x (fps) y (fps) Retrograde Posigrade Initial V 620 190 9 9 Desired V 769 133 40 -54 AV 149 -57 31 -63 The fourth mid. course correction, like the first, is instituted after a major burn, in this case the TPI burn. Its magnitude is about 5 fps and it is applied 10 to 12 hours prior to POI to correct any error picked up by tracking. The final and most crucial trajectory burn is the point of injection (POI) burn. It is used as a braking maneuver to negate the satellite' s relative velocity and inject the satellite into the halo orbit. Similar to TPI, the magnitude of the POI burn is dependent on injection. For a posigrade injection the burn is 678 fps while a retrograde injection results in a POI burn of 676 fps. At injection into the halo orbit, the satellite has only a Y component of velocity of -65 fps (see Figure 2. 10 and Table 2. 6). Table 2. 6 POI z (fps) X (fps) y (fps) Retrograde Posigrade AV2 -552 450 -66 90 A V2 -552 385 -66 90 AV2 = velocity required to bring to rest at rendezvous point 0 AV2 velocity required to inject vehicle into halo orbit at rendezvous point The remaining velocity corrections needed to keep the vehicle in the desired halo orbit can be placed in the category of station-keeping and this problem is taken care of by attitude control (see Section 3. 5). 2.5 LAUNCH WINDOWS The basic principle in determining launch windows is to find the times in which a translunar injection burn will place the vehicle in an orbit which will include the moon and the L2 point. 21

Retrograde Rendezvous Point/ 8057jks \ / \Ii \ /!1/ \I / I-8~5\' Posigrade I 1 Rendezvous Point Figure 2. 10 Posigrade and Retrograde Rendezvous Points 22

To understand these windows, the antipode must be defined. This is a vector extending from the earth' s center and opposite in direction to the moon. Its motion on a stationary sphere surrounding the earth can be shown as the sum of a sinusoidal latitudinal motion of 14.50 per hour (the earth's rotational angular velocity minus the moon' s angular velocity about the earth) (see Figure 2. 11). The constraints on the mission are: (1) 800 to 1100 azimuth as explained in Section 8.3 (2) A Cape Kennedy launch site to utilize the eatrth's rotational velocity The launch window for each day is the sum of all the times when a launch trajectory from Cape Kennedy passes through the antipode (see Figure -2. 12) For each azimuth from 800 to 1100 there are two times during the day that the path would intersect with the antipode. Thus, the launch windows occur twice each day of the month for lengths ranging from approximately one to four hours. The sum of both windows will be approximately five hours, The inclination of the plane of the transfer orbit to the earth-moon plane (itr) is determined by the angle of the parking orbital plane to the earth-moon plane at the antipode (see Figure 2. 13). This itr varies from approximately + 630 to -63~ over a month, the minimum being +2.50. Larger magnitudes of ir increase the AV2 required, thus the most advantageous times for launch will occur at (itr)mina This takes place when the moon is at the ascending or descending node of the moon's orbital plane with respect to the earth's equatorial plane. Thus, twice each month there is a minimum launch window, but since the change in AV2 with varying itr is small, all previously mentioned windows are acceptable (lteferences 2 and 6). 2. 6 SHADOWING EFFECTS Lunar eclipses occur at the points where the moon, and the satellite will be shadowed by the earth. The solar cells will be able to operate partially in the penumbra, but will not operate at all in the umbra (see Figure 2. 14). Since these are not extremely long periods of darkness, the effects are not crucial as the batteries can easily operate for the necessary time (see Table 2.7). 23

Moon Earth Antipode. Equator Path of Antipode Figure 2. 11 Motion of Antipode Inertial Sphere 8 0 Azimuth Path of Cape 110 Azimuth T1 T Equatorial Plane Moon's Orbital Plane Antipode at T + A T Initial Revolution in Parking Orbit Not Shown Figure 2. 12 Launch Trajectories for Atlantic Injection 24

Pacific Injections Earth +t~r t r — Moon Earth- Moon y y~\ \ y Plane \/B At Alantic Injections Trans- Lunar Plane Figure 2. 13 it Inclinations Penumbra Umbra Sun Earth Moon Linus Orbite Figure 2. 14 Eclipse of Linus Due to Earth 25

Table 2o 7 Lu:nar Eclipses Date Complete Phase (mi) Total Phase (mrnn) 1972, Jan. 30 204 42 1972, July 26 160 1973, Dec. 10 72 1974, June 4 186 1974, Nov. 29 212 76 The moon also shadows the satellite at certain timnes and its effects are more important although not crucial. Since this shadowing occurs near full moons, the chance of coinciding with Apollo missions, which up to now have avoided launches during full moons due to lighting considerations, is negligible. The length of this shadowing varies, reaching as high as 5.4 hours and its probability is approximately five times a year (see Appendix A. 7). The effect of this shadowlng on the size of the batteries and power requirements for Linus are discussed in Sections 5. 2 and 5 4. 2. 7 REFERENCES 1. Farquhar, R. W., The Control and Use of Libration-Point Satellites Suddaar No. 350, July 1968. 2. Final Report for Lunar Libration Point Flight Dynamics Study, prepared by General Electric Co., May-Novo 1968 for NASA Goddard Contract No. NAS 55-11551. 3, Kendrick, J. B., TRW Space Data 1967. 4. Jakson, eorge C, "Libration Points", University of Michigan Seminar Paper, June 20, 1969. 5. Moulton, F. R., An Introduction to Celestial Mechanics, Macmillan C o, New York, Dec. 1947. 6. Frank, M. P. III, Project Apollo Detailed Mission Planning Considerations and Restraints, Nov. 10, 1966. 7. Trajectory Study Rum for Project Linus by NASA Goddard, Groves, Re Nov. 10, 1969. 26

ATTITUDE CONTROL SYSTEM 30 1 INTRODUCTION Two types of attitude control systems considered for Linus were spin stabilization with a despun platform and a combination of momentum wheels and reaction jets. Spin stabilization was discarded because this system utilizes body mounted solar cells and for our power requirements the size and weight of the satellite would be prohibitive. Therefore, the three axes inertial guidance system was chosen because of its compactness and low weight. Hydrazine rockets were chosen Ifor — both midcourse corrections and halo orbit maintenance because of severe fuel tank size and weight limitations. In essence, the attitude control system chosen is an optimum combination of reaction Jets and momentum wheels which attain a high degree of antenna pointing accuracy. 3, 1, 1 Axes Definition The body axis for Linus was arbitrarily chosen to have the yaw axis parallel to the solar panels, the roll axis parallel to the main antenna feed and the pitch axis complete the right hand system (see Figure 3. 1). 3,2 CONTROL SYSTEM 3, 2, 1 Problem Description During the translunar-trans-L2 portion of the trajectory, the power required is enough to merit deployment of solar panels. The objective of the attitude control system, in this phase of operation, is to keep the solar panels in full sunlight and, when necessary, orient the vehicle for a midcourse maneuver o The attitude control of Linus in halo orbit is a problem of properly orienting the satellite so that it keeps a stable orbit with the antenna always pointed at the moon (see Figure 3. 2). The problem is complicated by the fact that Linus is not orbiting about the moon, but is orbiting about the L2 point The natural halo orbit, however, is not ideal. If left on its own, Linus would eventually drift off into deep space. It was found that through periodic thrusts to damp out certain diverging terms in the equations of motion, Linus could maintain an orbit indefinitely. This type of orbit maintenance was called Z-axis and X-axis control. 27

Yaw~~~~Yaw Roll Pitch Side Back Figure 3.1 Axis Definition of Satellite 28

0/ 00\" A Halo Orbit / /. \ / Linus I z ( ~~~~~~~Moon ) )-)O~~~ x~3 —.. Figure 3. Coning of Satellite in Halo Orbit 29

3,3 TRANS-L2 COASTING AND MIDCOURSE MANEUVERS During trans-L2 coasting, the three axis control system utilizes attitude sensors, the rate and position gyroscopes, and the propulsion system. The coarse and fine sun sensors are used for error control about the yaw and pitch axes and the Canopus star tracker reads attitude errors about the roll axis, Angular rotation rates are sensed about all three axis by the gyros in the rate mode. As the errors are sensed, the propulsion system makes the corrections for the desired orientation during coasting. When midcourse maneuvers are required, the position control is switched from the sun and Canopus sensors to the gyros. Linus is then slewed, one axis at a time, to the proper orientation for a midcourse burn, The reaction jets are used to correct for any torques due to thrust vector misalignment. For system block diagram see Figure 3.3. Sequence of Midcourse Events After translunar insertion. 1. Deploy Yo- Yo 's and despin to approximately zero rpm 2. Release Yo=Yo's a, Bring any residual rotation to zero b. Activate rate gyros c. Deploy solar panels 3. Acquire sun a. Antenna is now pointing away from the sun 4. Acquire Canopus a, With Canopus acquired, position control now rests with the sun and Canopus sensors 5. Point antenna at Earth a. Yaw Linus until antenna points at the Earth b. Activate solar drive mechanism and point cells back at the sun c. Check out the antenna 6, Fire reaction jets as needed to maintain proper attitude 7. Midcourse correction a. Switch position control to gyros; gyros are now in rateposition mode b. Slew axes so that roll axis is correctly oriented for the midcourse engine thrust c. Fire engine d. Slew axes back to sun and Canopus orientation e. Turn position control over to sun and Canopus sensors f. Repeat as needed for midcourse corrections or insertion 30

Roll, Pitch, and Attitude Control Roll, Pitch, and Yaw Yaw Jets Electronics Momentum Wheels Gimnbal s on Moon Sensor SensorCaou Course Sun Sen- lecroic I ~lectronics Star isor Pine Sun Sensor Canopus Sensor Ant enna Feed Gyro oll, Pitch, Control Control rror and lectronics ate Gyros Central Computer and Sequencer idcourse round Yo-y o oRocket olar Array f dspina Solar Array Motor l ctnuati Moloto r lectronics, itidcourr 3 Engine Figure 3.3 Attitude Control System Blo k Diagram

8. In halo orbit a. Turn roll axis control over to moon sensor and acquire moon b. Acquire Canopus c, Initiate satellite coning i. e, update computer with phase angles of halo orbit initial conditions do Initiate z-axis control as needed e. Initiate x-axis control as needed 3..4 YO-YO DESPIN During the launch phase of this mission, the third stage of the Delta uses spin stabilization for injection accuracy. After third stage burnout, the satellite is separated from the expended motorcase and despun using the Yo-Yo despin technique. This technique utilizes the transfer of initial angular momentum from Linus by the extension of two diametrically opposed weights connected by steel cables to the main body of the satellite. An initial spin rate of 100 rpm is reduced to nearly zero rpm by releasing two, 1 3/4 pound weights that spin out 26 1/2 feet from the vehicle. Since the cables are long enough to wrap around the solar cells 1 1/2 times, it is required that the cells be protected by an aluminum "U' channel ring which can be jettisoned after the weights have been deployed. 3.,4., Design Decision It is desired to have no angular velocity after despin due to the possibility of tumbling -induced by solar panel deployment. This allows the cable length and end weight to be independent of the initial spin rate. Therefore, the rigid Yo-Yo system (consisting of a cable and end weight) is used rather than the more complicated stretch Yo-Yo system. Also cold gas and hot gas retro devices were considered, but discarded due to the complex control system and weight penalty. 34,,2 System Details Due to limited space on the satellite body for end, weight anchoring and release, it was decided to have the Yo-Yo anchoring and release systems coincide. Therefore, the cable wraps around the vehicle one and one-half times to specify the cable length and allow the end weights to be determined. The despin system is located at the satellite's center of gravity. The end weights are released by kicking a hold down clamp free of the anchor point with a pyrotechnic thruster, The despin ring is jettisoned by two guillotiene actuators which sever the thin double wrap of wire holding the spring loaded ring (see Figure 3,4). 32

Guiderails End Weight / -EndWeigh Solar End Weight A Spring Housing Panels Cable S Guiderails Cd +~ iRing Pyrotechnic Actuator 0 Despin Ring Solar Panels Front View Side View Solar Panels Cable Despin Ring Top View Figure 3.4 Yo-Yo Configuration 33

3,4, 3 System Parameter s I = spin moment of inertia = 46 slug-feet L = cable length = 26.4 feet M = end weight - 1. 75 lbs Mc cable weight -. 9821 lbs Ri initial spin rate = 100 rpm Rf = final spin rate = 0 rpm End Weight Dimensions: half cylinder with hollow for release spring length = 3.0 inches radius a 1.14 inch volume = 6. 18 cubic inches density =. 238 lbs/in3 Cable is 1/8 inch steel,.0031 lbs/in total weight per cable -. 982 lbs,5 ACQUISITION MANEUVERS 3!,.5.lJ Sun Acquisition To put as little drain as possible on the batteries, the sun acquisition maneuver should take place as soon as possible after Yo-Yo despin. Initially the coarse sun sensor is turned on and the satellite is slewed by the reaction jets to within: 170 of the sun line. This is possible since the sun sensors have a 3600 field of view. When a coarse alignment is achieved, the fine sun sensors are turned on and the coarse sensors are tlurned off, The craft is then slewed to within ~ 10 of the sun line and the pitch axis is determined, 3..5,.2 Canopus Acquisition When the sun acquisition maneuver is completed, the Canopus acquisition may be initiated. A precise 3600 roll maneuver is commanded, during which time a star map is telemetered to earth. This star map is compared with an overlay of the predicted star map signal. This overlay will then accurately define the roll angle of the satellite relative to Canopus at the end of the 3600 maneuver. A second maneuver is commanded to orient the spacecraflt to Canopus with a small negative roll error relative to Canopus. The third maneuver in the sequence to acquire Canopus is a command which rolls the spacecraft in a positive direction untlT Canopus recognition is received by the control assembly from the tracker, Upon receipt of the recognition signal, the slew maneuver is terminated and the 34

spacecraft acquires the star (the tracker acquires Canopus and generates the recognition signal as soon as Canopus enters the ~ 40 field of view). The roll axis will then be fixed by the attitude error signal and the rate error from the inertial reference unit. 3,6 Z-AXIS CONTROL The period of our halo orbit Is 14. 67 days and we correct twice each orbit or once every 7. 33 days. Our AV requirement per year is 401 fps or 8 fps per half orbit correction (see Section 2. 3). Linus will fire two. 12 lb thrust engine in the same direction to achieve the needed thrust, The maximum weight of our satellite in halo orbit is 724 lbs making the maximum burn time 12. 5 min. As the satellite mass decreases, due to consumption of fuel, so does the time required for each correction. This analysis assumes that the thrust vector from the reaction jet engines are aligned in the Z-direction. However, since the roll axis of Linus is pointed at the moon, the yaw axis is 3 1/20 misaligned to the Z=axiso Therefore, it is necessary to correct this misalignment before Z-axis control is initiated, This will be done by pitching Linus with the pitch reaction jets. A loss of communication for some 15 minutes will be caused by this, If Linus is communicating at the time a Z=axis control is needed, the firing may be skipped and made up 1800 later by a longer firing. If continuous communication for a period longer than 14. 67 days is needed, the point of application of Z-axis control may be shifted from the top and bottom of the orbit to the left and -right sides of the orbit where the yaw axis of the satellite and the Z-axis coincide. This, however, is much more costly fuel wise. Finally, if no communication is needed, the pitching and yawing motion of Linus can be stopped, This reduced the motion of the spacecraft relative to the X, Y, Z, to one of pure translation and insures the coincidence of the yaw and Z-axis at the top and bottom of the orbit, Pitching and yawing of Linus may, at any time, be initiated by a computer update. 3.7 ATTITUDE SENSORS The concept of system operation involves the use of the s$atelliteCanopus line of sight and the sun vector as the primary reference axis in translunar space and switching to satellite-Canopus LOS and the moon vertical while in halo orbit. Canopus and the sun were used for the translunar period because; (1) they are always in the LOS of the spacecraft; (2) orienting the spacecraft with the sun also correctly orients the so lar panels; (3) since the satellite antenna is pointed toward the earth during translunar flight and the moon scanner is mounted on the antenna edge, the moon scanner cannot be used until the halo orbit is achieved, The change from the sun to the moon as a reference is necessary because the main antenna feed must be pointed toward the moon to increase gain. The rt efore, a moon sensor is needed, 35

3. 7. 1 Moon Horizon Scanner Sensor Description Arrays of thermopile detector elements are electronically sampled in a commuter fashion. There are 40 discrete detectors in each array. Four identical scanning heads, each with its own optical system and detector array, define the four sectors shown in Figure 3.5. Each commutated signal, after filtering, shows a profile equivalent to that which would be generated by mechanically scanning across the space/moon horizon. The spectral band picked up by the scanner is 14 to 40j and the field of view (each sector) is 700 x 100. This detector has no moving parts and each cyclic scan period is 60 milliseconds. Horizon crossover is indicated by transition from zero volts (space in view) to -20 volts (planet in view). This system provides useful spacecraft attitude outputs over a very wide range of altitudes and subtense angles. Reliability is high (i. e. probability of success for a three year mission is 90. 6%) and accuracy is approximately ~. 10. 3. 7. 2 Sun Sensor General Description The sun sensor consists of a +X and a -X unit. The +X unit consists of two coarse solar radiation detectors, one fine solar radiation detector, associated circuitry, thermal provisions and structural members. The -X unit is the same as the +X unit except it has no fine detector. The sun sensor is designed as an unsealed assembly containing no mnoving parts and since the coarse and fine detectors are photovoltalic devices, no power is consumed by the sun sensor during operation except for temperature monitoring thermistor excitation. The operating temperature is maintained between 400 and 1400 F by passive techniques. During the normal mode of operation, the sun sensor provides outputs which control solar array shaft rotation. During the translunar period they are used to orient the spacecraft roll axis and point the solar panels toward the sun. During the acquisition mode of operation, the solar array is locked to the spacecraft body and the sun sensor outputs are used to control the satellite in yaw and pitch, thus aligning the roll axis parallel to the sun vector. The accuracy of this sensor system is ~ 10. 36

z axis Horizon Circle I Vehicle Axis (x) Figure 3.5 Moon 37

3.7. 3 Canopus Star Tracker General Description The star tracker is an optical or light sensing device mounted in a gimballed unit that has two degrees of freedom. The star tracker is locked onto a target star by sensing the misalignment between its optical axis and the line of sight to the star. By using this error signal to drive the gimbals, the star tracker controls its optical axis to be coincident with the line of sight to the star, Thus, the tracker angle is electronically measured to give the satellite attitude angle. The field of view of this tracker is an 80 cone with 1800 x 900 of rotation about two axes. Accuracy is well within 13 arcsecso 3.8 PROPULSION SYSTEM The propulsion system has two purposes: (1) getting the vehicle to LZ, and (2) maintaining it in the halo orbit. This involves two sets of engines; one set for midcourse corrections (MCC) and one set for orbit maintenance (station keeping). Usually the MCC set consists of one main engine to supply the impulse needed and a set of reaction jets to align the thrust vector from the main engine in the proper direction. The stationkeeping set usually consist of an ion or cold gas engine to supply the continuous low thrust needed and the same set of reaction jets to keep the vehicle aligned properly (or to unload the momentum wheels). The case at hand is unique in that contisnuous thrust is not required to maintain the halo orbit. Linus can impulsively be kept in the proper orbit by firing an engine once or twice an orbit. In fact, the size of the engine needed to keep Linus in orbit is of the order of the size of the attitude control reaction jets. Therefore, the project will also use the attitude reaction jets for the orbit maintenance. After sizing calculations of the main engine (see Appendix B) it was found that a 20 lb thrust engine was required for the perilune burln, Thrust misalignment calculations (see Appendix B) indicated that two dual engines at 0.12lb thrust and two quad engines at 0. 12 lb thrust were required. Hydrazine was chosen as the propellant because of the simplifications it allowed in the fuel feed system. It is a monopropellant with an Is of 200 seconds. It will be used in both the main and reaction engines. The propulsion system will be nitrogen pressurized and the fuel will be stored in collapsible bladders, The pressure of the nitrogen must be enough to supply the Hydrazine at 300 psi. The pressure of the nitrogen will initially be 3000 psi in a. 6 cubic foot steel tank (see Figure 3.6),

Regulator Collapsible Bladder Nz To Engine Hydrazine Fuel Attitude Control Tanks Signal Figure 3. 6 Attitude" Control Propulsion System 39

3.9 MOMENTUM WHEELS The momentum rwheel develops and stores momentum for the attitude control system. It is essentially a motor with a large inertia rotor, Motor torque, accelerating or decelerating the wheel, is transmitted through the unit mounting pads to the vehicle structure to maintain a constant vehicle attitude or to impart a desired attitude change on command, The wheel can continue to apply torque to the vehicle until a maximum speed somewhat below motor synchronous speed is reached. At this point the wheel is unloaded by the reaction jets, Unloading signals proportional to direction and speed of rotation are obtained from a tachometer located within the momentum wheel assembly itself. The command for reaction wheel operation may come from a variety of sources depending on the mode of operation: sun sensor, star tracker, moon sensor, or gyro platform. The capability of these wheels in the halo orbit must be enough to keep Linus pointed at the moon and off-set solar torque. The pitch and roll momentum wheels will be 12 inches in diameter and have an inertia of 0. 0647 slug-ft. For momentum wheel calculations see Appendix Bo3 3.10 ONBOARD COMP UT ER The heart of the attitude control system is the on-board computer. The computer commands the proper sequence of events needed to control Linus, Commands can be stored in the memory before launch and the computer can be updated from the ground to control any non-scheduled events. When a signal is received from the ground for a particular command, the computer will either activate the command or store it in its memory as dictated. Because of the importance of the computer in handling all commands of attitude control, a backup system consisting of a command distribution unit is used, This unit is separate from the onboard computer to assure complete control of Linus in the event of computer failure, 3. 11 ATTITUDE CONTROL ELEC TRONICS Several servo meehanisms are needed for precise control of Linus during all phases of the mission. All component electronics have inputs in the attitude control servo-mechanisms, Of particular interest is the servo control needed for the momentum wheels. Shown in Figure 3. 7 is the servo control for these wheels, This figure shows coupling of the servos for small perturbations* P... A po'ssible transfer function for a wheel is K1 (s - a )/Cs where s (d/dt) K1 = gain x moment of inertia of wheel a1 pole of transfer function 40

K s z- x, A K X x = momentum wheel angular speed E = error P.. - perturbation cross-coupling of i torque along j axis zx Z = control position C C y (s a ) r E Cs Y + z position P z + -- c 9, + z z K (s a) z z zz~ z E Cs Figure 3. 7 Block Diagram of Momentum Wheel Control System

3. 12 REFERENCES 1. Abzug, M. J., "Satellite Intertia-Wheel Attitude Control on Elliptical Orbits"', Missile & Space Systems Division, Douglas Aircraft Co., Inc. Santa Monica, California, 1965. 2. Aerospace Telemetry Systems, Bulletin B114, Radiation Systems Division, March 1965. 3. Cannon, R. H., "Some Basic Response Relations for Reaction-Wheel Attitude Control", American Rocket Society Journal, Vol. 1, No, 2, January 1962, pg. 61-74. 4. Corliss, W. R.,."Scientific Satellites", NASA SP-133. 5. Eide, D. G.,Mot and Vaughn, C. A., "EquationofMotin andDesign Criteria for the Despin of a Vehicle by the Radial Release of the Weights and Cables of Finite Mass", Langley Research Center, NASA TN D-1012, January, 1962. 6. Fosth, D. C., and Mitchell, W. Io, "Lunar Orbiter Attitude Control System" AIAA Paper 67-533, AIAA Guidance, Control and Flight Dynamics Conference, Huntsville, Alabama, August 14-16, 1967. 7. Hatcher, Norman M., "A Survey of Attitude Sensors for Spacecraft", NASA SP-145, Langley Research Center, Langley Station, Hampton, Va, 1967 8. Koelle, H. H., Handbook of Astronautical Engineering9 McGraw Hill, 1961. 9. Langford, R. D., "Progress in Astronautics and Aeronautics-Guidance and Control- II," AIAA Series, 1964. 10. Loh, W. H. T., Jet, Rocket, Nuclear, Ion and Electric Propulsion; Theory and Desgn, Springer-Verlag Inc., NewY York, 1968. 11. Miller, J. E., Space Navigation Guidance and Control, Technivision Ltd. Maidenhead, England, 1966. 12. OAO Briefing Supplement, Grumman Aircraft Engineering Corp., Bethpage, New York, July 1967. 13. Showman, R. D., "Simple Processers of Star Tracker Commands for Stabilizing an Inertially Oriented Satellite", NASA TN D-4490, Ames Res, Center, Moffett Field, California, March, 1968. 14.. Walsh To Mo, "A Celestial Attitude Measurement Instrument for Project Scanner", NASA TN D-4742, Langley Res. Ctr., Langley Station, Hampon, Va 42

COMMUNICATIONS 4.1 INTRODUCTION Linus must relay signals for the Apollo program. The communications subsystem is therefore constrained to a large extent by established values of otherwise variable quantities in order to avoid the additional cost of altering present Apollo equipment. A simple transponder which first translates and then repeats a signal can handle most services. Mean values for earth, moon, and L2 relationships are shown in Figure 4. 1. dl =84,000 rl63780 L E 0.26r d2 - 64,500 r2 1738 -355 0 163 d4 = 448,000 r3 - 4000 ~ = 0.510 = 3. 09 Figure 4. 1 The up-and down-link RF Unified S-Band spectrum (Reference 1) is shown in Figure 4. 2. These are approximately 4 MHz channels with about 1 MHz guardbands. S/Norequirements for the various Apollo modes are given in Tables 4. 1 and 4. 2 (Reference 2). Down link PM is shown since FM is used for replay of stored data, a need that would be eliminated. To meet these requirements, several parameters can be varied. 43

APOLLO SPACECRAFT TRANSMIT SPECTRUM 2270 2272.5 MHz 2277.5 MHz 2282.5 MHz 2287.5 MHz 2300 MHz CM-FM INSTRUMENTATION LEM-PM CM - PM MHz TV, TM, UNIT(IU)-FM OR TM, VOICE VOICE TM S-I& -PM RANGE, DOPPLER TM, VOICE RANGE, DOPPLER GROUND STATION TRANSMIT SPECTRUM 2090 2101.8 MHz 2106.4 MHz 2120 MHz LEM - PM CM - PM MHz DATA, VOICE DATA, VOICE RANGE, DOPPLER RANGE, DOPPLER Figure 4. 2 Table 4. 1 S/No Required for Services in the Apollo Unified S-Band Up-Link Threshold Mode Service Quality S/No dB 1 Carrier 12 dB loop S/N 53.9 PRN 2 Carrier 12 dB loop S/N 51.3 Voice 90% intelligibility 55.5 Voice 70% intelligibility 52. 0 3 Carrier 12 dB loop S/N 51.3 Updata 6.7 x 10-b SBER 58.4 4 Carrier 12 dB loop S/N 54.4 Voice 90% intelligibility 59.5 Voice 70% intelligibility 54.8 PRN 5 Carrier 12 dB loop S/N 54.4 Updata 6.7 x 10-6 SBER 64.2 PRN 6 Carrier 12 dB loop S/N 45.9 Voice 90% intelligibility 59.5 Voice 70% intell~igibility 54.8 Updata 6.7 x 10- SBER 66.2 7 Carrier 12 dB loop S/N 46 8 Voice 90go% intelligibility 58 7 tVoice 70 intelligibility 54.8 Updata 6.7 x 10- SBER 66.2 8 Carrier 12 dB loop S/N 51.3 Voice 90%0 intelligibility 56.0 Voice 70% intelligibility 51.5 44

Table 4. 2 S/N Required for Services in the Apollo Unified S-Band Down-Link Thre shold Mode Servi ce Quality S/N dB 1 Carrier 12 dB loop S/N 33.6 51.2 Kbps TLM 10-6 BER 65.6 Voice 90% intelligibility 61.8 Voice 70% intelligibility 5 6. 8 2 Carrier 12 dB loop S/N 23. 9 51.2 Kbps TLM 10-6 BER 66. 1 Voice 90% intelligibility 61,8 Voice 70% intelligibility 56.8 PRN 99.6% acq. in 9 sec. 38.3 2B Carrier 51.2 Kbps TLM Voice/biomed 90% intell/acceptable 64.3/? PRN 3 Carrier 12 dB loop S/N 33.9 1.6 Kbps TLM 10-6 BER 56.7 Voice 90% intelligibility 54. 6 PRN 99.6% acq. in 9 sec. 38.3 3B Carrier 1.6 Kbps TLM Voice/biomed 90% intell/acceptable 57.6/? PRN 4 Carrier 12 dB loop S/N 33. 6 1.6 Kbps TLM 10-6 BER 56. 1 Voice 90% intelligibility 54. 7 Voice 70% intelligibility 49. 7 5 Carrier 12 dB loop S/N 35.8 1.6 Kbps TLM 10-6 BER 47.2 6 Carrier 12 dB loop S/N 31.3 Key 99.6% acq. in 9 sec. 30.5 7 Carrier 12 dB loop S/N 29.4 PRN 99.6% acq. in 9 sec. 38.3 8 Carrier 12 dB loop S/N 34. 2 1.6 Kbps TLM 10-6 BER 51.2 Back-up Voice 90% intelligibility 63.3 Back-up Voice 70%~ intelligibility 49. 7 9 Carrier 6 dB loop S/N 30. 2 1.6 Kbps TLM 106 BER 47.8 PRN 99.6% acq. in 9 sec. 38.3 45

4, 2 ANTENNA CONSIDERATIONS 4. 2. 1 System Alternatives Four antenna systems can be considered, all parabolic reflectors because of high gain requirements. The first covers both the earth and the moon with the same beam, so it is 2 m in diameter with a gain of 17 db and a 3 db beamwidth of 5.50. As evident from the system finally chosen, this would be incapable of handling high bit rate telemetry (52. 1 kbps). The second system provides signals identical to those from near side communications (Reference 2). This requires a diameter of 9 m to give a gain of 44. 1 db, and the 3 db beamwidth is 1~. This system is prohibitive considering the complexity and massiveness of the electronics required for such a small beam to track, for example, the CSM. A feasible system, giving the service required, is a compromise between the first two, This uses two fixed feeds, the central one pointed at the moon and the other offset enough to give a 30 displacement for the second earth beam. As the satellite travels through one halo orbit, it maintains its axis (and central feed) pointed at the moon while rotating about that axis at the orbital rate, thus keeping the offset feed pointed at the earth. This has the disadvantage of complex thermal and guidance problems as well as an almost prohibitive solar panel configuration. The fourth system is like the third, but instead of the satellite rolling about the central feed axis, the offset feed is rotated about the central one. This is the system we chose. It has a 3. 6 m mesh antenna with a 3 db beamwidth of 2. 650 and a downlink on axis gain of 36, 2 db, assuming a 55% efficiency. The gain at the edge of the moon can be found using k 3 - (4.1) db 0 where k angular width of main lobe for decay of kdb 0 3 db beamwidth Allowing ~ 0. 10 for attitude control error, 0k varies from 3. 060 to 3. 600 for maximum and minimum ranges respectively (see Appendix Co 3). As shown in the appendix, as Linus moves from maximum to minimum range, a reduction in antenna gain is balanced by a smaller space losso Therefore, mean values are used for the moon-Linus leg in the link calculations. An antenna gain of about 31 db was determined to provide adequate service in light of amplifier limitations for the up-link (see below). The antenna is rigidly fixed to the spacecraft. A mesh antenna reduces weight and solar pressure. 46

4. 2. 2 Pointing The fixed feed will be pointed at the moon's center by control of the satellite. Moon sensors mounted on the edge of the antenna will keep the error within ~ 0. lO, This accuracy is necessary since a relatively small angular deviation gives an unacceptable loss of gain,. The accuracy does not apply to earth pointing, however, because the smaller subtended angle should stay inside our 2. 650 cone through all orbital variations. The rotation of the latter feed will be directed by an on board computer which can be updated from earth. 4. 2. 3 Omni Antennas Two omni antennas will be included, one on the fixed feed, and. one near the rear of the spacecraft. These will be used for (1) telemetry, commands, and ranging before deployment of the main antenna, and (2) reorienting the satellite if (a) pointing of the narrow beam is in error or (b) after a period when the large dish was not used and pointing was not maintained. A switching mechanism will select the omni antenna if an unacceptable signal arrives from the dish. 4. 2.4 Development Some development of feed and drive system will have to be undertaken to find the best way of rotating the beam. Mechanical rotation has proven reliability, but a multiapertured electronically rotated feed would also be possible. However, the electronics involved are still rather massive for the latter case. 4.3 SYSTEM PERFORMANCE As stated above, adequate service can be provided with the antenna described and amplifiers of 5 and 50 watts output for the Linus-earth and Linus-moon links respectively. The latter must be larger because of the smaller receiving antenna gains. The large output value is used since every decibel is needed to meet up-link requirements and still allow use of a simple non-tracking antenna. Link calculations are shown in Tables 40. 3 and 4.4. 4.4 SYSTEM OPERATION A block diagram is shown in Figure 4.3. The communication package has three basic subsystems. 47

Couple r Diplexer Diplexer Auto Antenna Switch ~ x k 2 ttjLimiter ( C ~) a X k L i m i t e r WA X k 4J Z~~t 2 Limiterr, - 1- 5 1o Cd P. C Ua~aPt ax k2 Li mitt; er c o L n Ln e~~~~~~~~tl~~~~~( Up —Link Up - Link Rece eive r I Rece iverr Phasse (A) (B) MVod Clock x k Prime c ~im 0 Co mmutator ncoder CmadSic & Text I-C arrier ubcear rierreoer f~l kllk~k-(s cilato r 00mnq om(a E~~~~u Range 0 0 0 0o ~b Code U U U I U Mod () PMDcdrCm -00 I -Demod n d T)~~~Lin~e Ix T

Table 4. 3 PM Down-Link Linus-to-earth (b (w PT (5W) 7. 0 LT -2.0 GT 30.8 R = 475, 000 km f = 2. 1 GHz 2 SL -212.4 Lp -0.7 GR 52.0 LR 0 S -125.3 Te 1500 (supercooled) N0 -206. 8 S/No 81.5 moon-to-Linus (1) CSM db (W) LM db (W) Surface db (W) PT10.5 13.,0 13.0 LT -7.0 -5.1 8. 9 GT 26.7 20.0 32.0 SL -195.9 -195.9 -195.9 L p -0.5 -0.5 -0. 5 G 31.6 31.6 31.6 LR -1.0 -1.0 -1.0 S -135.6 - 137.9 -.129.7 F = 2.0 db Ta = 2400K To -= 3140K Te =570~K No -201.0 -201.0 -201.0 S/NO 65.4 63. 1 71.3 Overall S/No 64.9 62.6 70.5 48

Table 4. 4 Up- Link Earth-to-Linus (3) db (W) PT (10 kW) 40.0 GT (85' dish) 53.0 LT 0 R = 475, 000 km f = 2.3 GHz SL -213.2 3 L p -0. 7 GR 31.6 LR -3. 0 S -92.3 F = 11 db T = 1120K Ta = 3140K Te = 39120K N -192. 7 o S/No 100.4 Linus-to-moon (4) CSM db (W) LM db (W) Surface db (W) PT (50 W) 17. 0 17. 0 17. 0 LT -1.5 -1.5 -1.5 GT 30.8 30.8 30.8 SL - 195. 1 -195. 1 -195. 1 L p -0.5 -0.5 -0.5 GR 23.3 16.0 31.2 LR -7.0 -5.9 -9.8 S -133.0 -139. 2 -127.9 F= 11. 3 Ta = 1000K T - 2900K Te = 38800K N -192.7 -192,7 -192.7 S/No 59.7 53.5 64.8 Overall S/N 59.7 53.5 64. 8 49

4,4. 1 Earth to Apollo Data Relay Both of the redundant uplink receivers will operate for the duration of the mission, Earth signals from these are monitored by a test switch unit. This can accept a command but has its own logic. Since alterations in equipment are avoided, the transponder should transmit and receive, for example, present CSM frequencies to and from the CSM. Therefore the CSM relay frequencies between earth and Linus are determined by choosing an appropriate coherent translation, as shown in Figure 4.3, such that kl times the up-link relay frequency gives the presently used up-link frequency. This allows range rate determination of far-side vehicles from carrier Doppler shift. The direct down-link frequency can be used for the uplink earth-to-Linus leg; once direct contact with the CSM is lost, this frequency will not be used anyway. Moreover, Ithis provides a logical translation, economical in bandwidth use, and an already familiar value of kl (221/240). pRS' turnaround provides the range data. After being coherently translated, the signal is amplified by one of the redundant 50 watt traveling wave tubes (TWT). These are controlled by a switching unit operated by ground command. To avoid degradation of system performance, only one channel at a time should be used through the TWT. Otherwise, the 50 watts of power is divided between two signals in the Linus-to-moon link, where the majority of the degradation for the overall up-link already occurs. 4.4. 2 Apollo to Earth Data Relay It is critical that the down-link have the capability for extensive -data relay of consistent quality. To ope.rate four of these data bands simultaneously and economically, five watt solid state amplifiers can be used. These will certainly be available for this mission with near 40% (or more) efficiency, and have low weight and volume. Reliability can be increased by using a multiple relay unit which can send any one signal through another amplifier in case of failure. A very slight degradation would occur if two channels used one amplifier simultaneously. This is because the division of power now occurs in the insensitive leg where less degradation results, for the overall down-link, than would occur in the sensitive space-to-space link for a given change in S/No. In any case, signals could be arranged so only three were transmitted. To increase signal quality in the down-link, preamplification is employed in the down-link receiver. This will reduce the receiver noise figure to 2 db for near future amplifiers using solid state pump oscillators for power. Also note the lower circuit loss for the stationary feed. [No preamplification is. used in the up-link receivers since this is a very strong link already. It is also interesting that the high noise figure of 11 db dominates to such an extent over anteanna temperature that noise spectral density is effectively unchanged for a direct view of the sun (Appendix C.4). Therefore there is consistent signal quality, even during the several hours per month that this occurs, If a preamplifier were used in the up-link, this would not be the Pseudco Random Noise 50

case, Consideration of the sun' s effect would probably not be important if Linus were used in missions similar to current ones. In these, the moon is not full, as it would be above, but less than half-lit so landing may be accomplished with great light and dark relief.] A quadriplexer breaks moon signals into the four separate channels, which are coherently translated, amplified, and finally multiplexed along with Linus transmissions. The Linus-to-earth downlink frequency is taken as the direct uplink for the same reasons explained above, so k2 = 240/221, 4.4,3 Linus Subsystem A standard PCM telemetry and command system will be used. The simple, highly reliable command receiver removes the narrow band of command frequencies from the up-link signals. A separate 4 MHz channel for Linus near present Apollo bands allows use of the same earth feed. A PRN ranging code is transmitted to Linus, coherently translated, and returned to the tracking station on a carrier phase locked to the uplink carrier. Two redundant command demodulators lead to the PCM decoder and verifying unit, This unit can control the command demodulators after receiving a special signal through the uplink receiver. The experimental nature of Linus will probably require a great deal of system monitoring. An extensive telemetry system will use four 25 channel subcommutators and a timing clock connected to one 25 channel prime commutator. This yields a total capacity of 120 channels. If more are required, the increase in weight and volume would be small. The prime commutator will handle all the rapidly changing data while the subcommutators will handle the more stable measurements. 4.5 SIGNAL FLOW The three important cases of signal relay are summarized in Figure 4.4 for clarity. Notice that two occulted terminals use the "long-way-around" route. In this way the design is kept simple since no signal switching is required, i. e., the transponder is an in/out affair, shown schematically in Figure 4.5. Al-so, signals can be monitored at earth terminal c, and a situation is avoided where a and c communicate simultaneously with b. That is, c will retransmit one signal to the appropriate point via the proper frequency, directly in case II, and to Linus for relay in case III. 4.6 TRANSPONDER RELIABILITY The probability of success for the communications subsystem is 97, 5% (Appendix C. 6). To obtain this figure with the system described, a series of developmentswill have to be made to meet the failure rate figures given (Appendix CO 5). Careful design and assembly of the components must be given high priority. Only high quality burned-in parts should be used. With this approach, the transponder will be a highly reliable subsystem over the two year period proposed. 51

C C Q~ j)-;a b a, Case I - a to c Case II - a to b simple occultation a occulted c Q,= a b Case III - a to b a and b occulted Figure 4. 4 Transponder Earth Feed Moon Feed Figure 4.5 52

4. 7 REFERENCES 1. Schmid, P. E., "The Feasibility of a Direct Relay of Apollo Spacecraft Data Via a Communication Satellite," NASA TN D-4048, 1967. 2. "Lunar Libration Point Flight Dynamics Study," General Electric Report, Contract No. NAS 5-11551, 1968. 3. Balakrishnan, A.V., Advances in Communications Systems, Academic Press, Vol. 2, 1966. 4. Bauer, L., "Aerospace Communications Systems, " Aerospace Engineering 485 course notes. 5. Corliss, W. R., "Scientific Satellites," NASA SP 133, 1967. 6. Jasik, H., Antenna Engineering Handbook, McGraw Hill, New York, 1961. 7. Koelle, H. H., Handbook of Astronautical Engineering, 1961. 8. Kraus, J., Antennas, McGraw Hill, New York, 1950. 9 Reed, H. R., and Russell, C. M., "Ultra High Frequency Propagation, " Wiley, New York, 1953. 10. Schmid, P. E., "Lunar Far-Side Communication Satellites, " NASA TN D-4509, 1968, 53

POWER 5. 1 INTRODUCTION Primary power for the satellite is supplied by solar cells mounted on two independently orientable panels. Secondary power, required before the panels are deployed, when they are shaded from the sun, or in an emergency, is provided by a modified Surveyor 205 rechargeable silverzinc battery, Alternative systems using radioisotope thermoelectric generators (RTG's), nuclear reactors, and fuel cells were also considered. These systems were rejected on the basis of high weight and cost. Both the primary and secondary systems can be used separately, or they can be used together in parallel. Both systems deliver a nominal 30 volts DC to the main power bus. Most of the loads operate with this voltage. A power control unit supplies DC voltages other than 30v and AC voltages for guidance and control equipment. Although most power-control logic decisions occur automatically, these functions can be controlled from the ground when special needs arise, 5. 2 POWER PROFILE The power system must provide 342 watts when the communications equipment is operating. The requirement drops to 297 watts when this equipment is off, During the periods when the satellite is shaded by the moon, the necessary power is also 297 watts. These periods occur approximately 5 times/year, last up to 504 hours, and may occur as close as 27. 3 days apart, The nine day launch-to-station flight has a constant power requirement of 278. 5 watts with five mid-course maneuver peaks of 300. 8 watts lasting up to 10 minutes each. Power profiles for the launch-tostation and the on-station periods are shown in Figures 5. 1 and 5. 2 respectively. In order to support a backside lunar landing, it was decided that the power system should be capable of providing 342 watts for an indefinite length of time. This constraint prohibits optimizing the system to have both the solar panels and batteries combine to meet the transmission power requirement. Therefore, the solar panels are designed to provide the full, end-of-lifetirre required power. See Table 5. 1 for power breakdown. 5.3 SOLAR PANELS Primary power will be delivered from solar cell arrays on two rotating panels. The solar cell choice was made due to their light weight and proven reliability, The orientation requirements and geometry of the satellite prohibit the use of body cells. The roll-up array is impractical when it is considered that the array must protrude far enough from the body to avoid the shadow of the 3. 6 m dish-antenna.

Table 5. 1 Power Breakdown Launch-to-Station Communications 42 w Guidance and Control 56. 5 Thermal 180 278. 5 w Mid-Course Corrections 22, 3 w 300. 8 w On Station, Communications On Communications 267 w Guidance and Control 75 342 w On Station, Communications Off Case 1 Normal Operation Communications 42 w Guidance and Control 75 The rmal 18 0 297 w Case 2 Z-Axis Control Previous Requirement 29 7 w Additional Guidance and Control 21.9 318,9 w Case 3 Battery Charge Previous Requirement 297 w Battery Charge 20 317 w 55

400 MCC #3 6. 05 days TPI 7. 05 days 35 0 M #4 0. dy MCC#1 1.0 days MGC#4 8.0 days PO3 8.33 days MCC #2 4 8 days 300 278.5 250 Ul) ON 200 P 150.100 50 0 0 4 5 6 7 8 9 Figure 5. 1 Launch-to-Station Power Profile Time (days)

410 W Power Available 400370 W z-axis Control (10 minutes/week) ~~30~ '~T \ \-~ X-~ ~ Power Required Communications Xm~~ ~ On (Indefinite Cd W 2001 Period) ~ Communications A - Off (Indefinite Period) - Shaded Period-Battery Power Only (5.4 hours up to 5 times/year) Time Figure 5. 2 On Station Power Profile

The arrays will be mounted on single-sided panels which will each be driven by a brushless motor to keep the cells facing the sun line in station orbit. The rotating axis of solar panels will remain approximately perpendicular to the ecliptic (see Figure 5.3). A spring mechanism (see Figure 7. 6), will deploy the folded panels immediately after the despinning of the satellite during translunar flight. At this time sun sensors will orient the panels and they will take over primary power from the battery. The panels will consist of 11, 160 cells. They will be 2 x 2 cm N/P 2 ohm-cm silicon cells of.014 inch thickness. A. 006 inch microsheet coverglass with blue-red filter and antireflective coating will cover the cells, The coverglass will protect the cells from the lesser solar flare radiation effects, however, a 10% allowance has been made in power to be supplied for possibility of solar flare occurances. The blue-red filter will restrict transmission of energy in the solar cell sensitive region. The coverglass and filters, however, do account for a factor of.85 in current transmission losses. The cells will be mounted on 6. 7 x 4. 5 foot aluminum honeycomb panels of 1 1/2 inch thickness, A thin sheet of G-10 glass epoxy will bond the cells on the panels and provide electrical insulation between the aluminum and the solar cells. As shown in Appendix D,1, there will be 93 cells in series strings to supply the required voltage. Each panel will contain 15 strings of four rows of cells connected in parallel for current requirements, The total area of each panel, 30. 15 ft2, includes a packing factor of.85 and an allowance of 1. 95 ft for a center hinge in folding. 5.4 Ag-Zn BATTERY A modified Surveyor 205 silver-zinc battery was chosen to provide secondary power for the launch and shaded periods. Silver-zinc cells were selected over silver-cadmium and nickel-cadmium cells because of greater energy density and higher plateau voltage, It has been shown that the relatively low cycle capability of silver-zinc cells can be extended threefold by replacing two of the six cellophane separator membranes with ine-rt — mat-erials '(see Reference 1), The battery consists of 20 series= connected cells grouped in two 4-cell monoblocs and four 3-cell monoblocs, The cells are mounted in a 7"x 18" x 9" magnesium canister resulting in a weight of 6603 pounds. Since the battery must be mounted on end, a 10% weight allowance for structural support of the upper monoblocs is provided, The total weight at the battery then becomes 73 pounds (see Figure 5.4). The battery provides 30 volts and has a nominal capacity of 170 amp-hrs. However, under the worst discharge conditions (600F and 10 amp discharge rate), the capacity of the battery falls to 125 amp-hours (see Figure 5. 5). This unit is therefore suitable for the requirement of 107 arnp-houars (see Appendix D, 2, 1). The worst possible sequence of shadowing would have an 8-month separation between required deep (70%0) discharges. The battery manufacturer has shown that charged cells kept below 90F retain more than 58

DMoon's Orbit Around Earth Figure 5. Solar Panel Orien t naSEart n Figure 5. 3 Solar Panel Orientation as Seen From the Sun

vl- rh I Modified Surveyor Model 205.........__.... IAg-Zn Battery Using 20 cells,. a, for 30 volt mean voltage ~ i I III ~~ II zI i!1 S It - __ ____ 3 elMnbo L: ' Ui M ~ -' ---- r --- IIJ. I II 4 Cell Monobloc Scale: 1/2 size Figure 5.4 Battery Assembly - End View 60

Capacity (amp-hrs) 1 amp 0201540 60 80 10 amp 125Fig / / t WorstOperatingPoint 50 25 0 20 40 60 80 100 Temp (OF) Figure 5.5 Battery Discharge Characteristics 61

90% of their capacity after 12 months (see Reference 2). Also, the battery could be cycled even when not absolutely necessary. The battery is only required to withstand 10 deep discharges (five times/year for two years). The battery has a cycle capability well beyond this requirement. A 168-hour (1 week) battery recharge period of.5 amp was used, with the assumption that a backside landing would not be scheduled around a shaded period. A higher charge rate could. be used, if necessary, but the input capacity would be reduced. The solar panels could provide enough power for charging the batteries while the communications equipment is operating. However, this would be undesirable since a fully-charged battery should be available for emergency use during a mission. If necessary, full power for communications could be provided for approximately eight hours by the battery alone (see Appendix D. 2. 2). 5.5 POWER CONTROL The power control unit (PCU) conditions the input power to match the varying requirements of the loads and provides circuit protection. The solar panels are sized to deliver 35 volts, allowing five volts for losses (two volts through lines and diodes and three volts through regulators). Revised thermal considerations indicate that the solar panels might provide up to 45 volts, rather than 35 volts. This excess voltage can be used by a switching regulator to perform the dual roles of regulating output voltage and lowering portions of it for particular requirements. Use of such a unit (having an efficiency of about 90%'), is thus more desirable than employing a DC-DC converter to reduce the panel voltage to 30 volts before entering the rest of the PCU. When the communications equipment is not operating, power supplied will exceed power required by about 100 watts (80 watts if the battery is charging). A shunt regulator will channel this excess power to the dissipating resistors on the solar panels. A power module to deliver a 26 volt, 400 cps, three phase sine wave to the gyro motors is provided. A dropping resistor reduc es the current used for battery charging from 14 amps to.5 amps. Voltage cutoff devices act as charge controllers to prevent the battery from charging past 1. 95 volts/cell or discharging past 1. 25 volts/cell. Voltage monitoring of the solar panels provides for automatic take-over by the battery should the solar array fail for some reason, such as sudden panel disorientation. Diodes prevent the battery from discharging into darkened -solar panels, however. All control functions are capable of being monitored from the ground and can be implemented or over- ridden by command from the earth. The entire PCU is 9" x 7!D X 6" and weighs 12 pounds (see Figure 5. 6 for power systems block diagram). 62

Regulator DC-DC Converter Loads Charge Discharge ~Solar jl~ ||Control I Bypass Solar Array I a l......, -..............Power AC o[ a l l lttery Module Loads Shunt Regulator DC-DC.Battery..,.......Converter Loads Figure 5.6 Power System Block Diagram

5.6 REFERENC ES 1. Report 66.310 "Surveyor Model 205 Battery Characteristics and Development of Related Batteries for the Bendix Corporation, Bendix Systems Division, Space Power Group*' The Electric Storage Battery Company, Exide Missile and Electronics Division, 1966. 2. Wylie, G. M. "Sealed Zinc-Silver Oxide Batteries-Part 2" Reprinted from "19th Annual Proceedings, Power Sources Conference" May 18, 19, 20, 1965. 3. Abbott Power Supply Catalog, 1970, Abbott Transistor Laboratories, Inc.:4. Bauer, P., 'Batteries for Space Power Systems" NASA SP 172, 1968. 5. Blueprint No. 205-1000 '"Surveyor 205 Battery Assembly," Missile Battery Division, The Electric Storage Battery Company, 1964. 6. Criswell, D. and Golden, R., "Surveyor Battery Tests," Bendix Systems Division Interdepartmental Memo LVP 528, ND-7407, January 17, 1967. 7. Gibson, R., "Design Data for Space Power Systems," The Bendix Corporation, Aerospace Systems Division, October 1, 1969. 8. Gibson, R., "EASEP/PSEP Solar Panel Development Design of the EASE-PSEP Solar Panel Array, " The Bendix Corporation, Aerospace Systems Division, Report No. EATM-15, November 20, 1968. 9. Pessin, L. and Rusta, D., "A Comparison of Solar Cell and Battery Type Power Systems for Spacecraft," IEEE Transactions on Aerospace and Electronic Systems, Vol. AES-3, No. 6, November, 1967. 10. Simpson, W. D., "Failure Analysis and Design Report of Nimbus 1 Solar Array Drive Failure, Final Report, " NASA CR 95482, August 6, 1966. 11. Szego, G. C. and Taylor, J. E., Space Power Systems Engineering, Progress in Astronautics and Aeronautics, Vol. 16, 1966. 64

THERMAL CONTROL 6. 1 INTRODUCTION When Linus is in the halo orbit it would undergo extreme temperature change's without adequate thermal control' as its orientation to the sun changes and as its equipment is turned off and on. Each of Linus' components have their own particular temperature range at which they will operate efficiently and the satellite' s environment must be controlled to keep the temperature ranges within these survival limits. To accomplish this control each component's placement and heat output must be taken into account to determine the best thermal control to be used. The method chosen should be as passive as possible in order to make the system simple and reliable. There are usually one or two components which put the largest constraints on the system because of their heat output or their operable temperature range. The controlling of these components usually determines the control system to be used. The more flexible constraints on the other components can. then usually be met more easily. The main constraints are placed on Linus' system by the high heat output (180 watts) of the transponder package and the small survival temperature range (600F - 95~F) of the battery package. 6.2 THE THERMAL SYSTEM From the satellite's position during translunar coast and in orbit (see Figures 2,0 7, 2. 8 and 5. 3), it can be seen that tvo sides of the cylindrical body never face directly into the sunlight, By using these two sides as the radiating surfaces this advantage can be used in minimizing the influence of the solar input. The overall satellite system is based on the heat balance equation, iO. e. the heat coming in must equal the heat going out (radiating). qin~ = rad (60 1q) This equation is used to find the radiating area and the component placement required to control the particular temperature range dictated by each component (see Appendix E)o The components are placed on the baseplate and are kept from radiating by gold plating the outside of the packages; therefore, most of the heat is transferred by conduction. It was found that the batteries have to be thermally isolated from the rest of the satellite in order to keep their restrictive temperature range; therefore, they will have their own radiating surface, 9" x 20" (see Figure 6. 1). Heaters will be provided to keep the batteries within their survival temperature range. 65

x Scale: 1" = 8" y z Top Plate Radiating Area _"'.....'"'- '_'".. _... '" _... _ I I 1 ON Insulation _ e i i Base Plate Batteries Figure 6. 1 Radiating Surface 2 (see Figure 7. 7)

The transponder package will have to be placed directly on a radiating surface 33" x 30" in order to maintain its temperature at a reasonable value, which reduces noise in the system. Heat pipes will be conductively connected to the package and the radiating area (Figure 6. 2) to distribute the heat and keep a reasonable temperature level. During the time when the communication equipment is being used, the heat output of this package will heat up the rest of the satellite by conduction through the baseplate, but when Linus is not being used for communicating, either the radiating surface will have to be decreased or heaters will have to be used, When the communication equipment is off there will be 290 watts of power available; therefore, heaters can be used without putting any extra requirements on the power supply. Since radiating surface 1 of Linus is always at a temperature above 320F and below 700F and the insulation minimizes heat input or heat output anyplace else, the components on the baseplate are always between these temperature ranges except those components with a heat output of their own. The temper'ature-of these components will be higher due to the AT induced by the therrnal -re-sistance of: the baseplate (see::Table 6. 1). Table 6. 1 Satellite Component Temperatures (See Appendix E for calculations) Component Condition Tempe rature Batteries Discharging 910 F Not discharging, heaters on 67~F Communication Comm. on 106~F Package Comm. off, heater on 58~F Fuel Tanks Comm. on 700F Comm. off 45~F Computer Comm. on 100~F Comm. off 650F Guidance and Comm. on 1200F Control Electronics Comm. off 850F Solar Panels Cells facing Sun 107OF Back facing sun 145~F 67

Scale: 1' = 8" x Y z Top Plate = Radiating Area, ~ I Heat I 1 Pipes 1 Insulation Insulatio 'I-I jI I I I _ I Base Plate Communications Electronics Figure 6. 2 Radiating Surface 1 (see Figure 7. 7)

6.3 THE THERMAL, COMPONENTS 6.3. 1 Heat Pipes Heat pipes are used when the thermal resistance of the structure is too high to prevent adequate heat flow from a component. They have an almost limitless capacity of heat flow and therefore can distribute a large amount of heat evenly over a large area more efficiently than any metal structure. They work on the following principle. A heat source causes a highly volatile liquid to vaporize and the induced pressure causes the vapor to propogate to the cold area where it condenses inducing a pressure drop and then returns to the heat source by capillary action provided by a wick (Figure 6.3). There must be enough liquid present in the pipe to keep the wick soak.ng wet and the wicek is designed so that the vapor flow is equal to the condensate flow so that the heat source end does not dry up. The wicks are fine mesh metal screens which are attached about one (1) millimeter away from the inside diameter of the heat pipe. There should be no prcblems with the heat pipes working reliably for a long period of time even though the wick and the amount of fluid in the heat pipes must be designed to exact specifications in order to work correctly. They can be ground tested (Reference 2) and the radiating surface thickness should protect them from any meteorite punctureso 6.3. 2 Heaters The heaters are basically electrical resistance devices which are controlled by a thermostat, The thermostats are calibrated to turn the heaters on when their temperature approaches the minimum allowable operating temperature and turn them off when the temperature is about 100F higher than this survival temperature. They consist of a fine resistance wire encased in a thin film of plastic which is applied to a surface by epoxy. These heaters give off five (5) watts of heat per square inch of surface area. Heater Placement (see Appendix E for calculations) Location Number Each Fuel Tank 1 sq in Nitrogen Tank 1 sq in Battery Package 7 sq in Communications Package 24 sq in gq

Condensation Cold Area Pipe i Heat Source Vaporization Figure 6.3 Heat Pipe 70

6.3. 3 Insulation All internal surfaces of the satellite which are not being used as radiating surfaces will be covered by insulation to minimize heat input and heat radiation from these surfaces. The insulation will be a blanket of 40 layers of Al/ mil Mylar which is stacked at 60 layers per inch. 6.3.4 Thermal Coatings The backs of the solar panels will be coated with a stable black paint (Cat-a-Lac) in order to keep the temperature at a reasonable level and to keep the star sensor from sensing the reflectance of this area as a star. The radiating surfaces will be coated with a stable white paint, zinc, oxide/ potassium silicate (Z-93). This coating has a low a/E (0. 2/0. 9) is quite stable in space environment, and can take the handling it will receive during assembly (Reference 1). 6.4 THERMAL COMPONENT WEIGHTS Component Weight Heaters 0. 2 lbs Thermostats 1.0 lbs Insulation 2. 9 lbs Heat Pipes 3,5 lbs Total 7. 6 lbs 6.5 REFERENCES 1. Cunnington, G. R., Grammer, J. R., and Smith, FP J,, Emissivit~ Coatings for Low-Temperature Space Radiators, NASA Contractor Report 1420, Lockheed Aircraft Corporation, Sunnyvale, Calif. September 1969. 2. Katzoff, Samuel, Heat Pipe-s and Vapor Chambers for Thermal Control of Spacecraft, AIAA Paper No. 67-310, AIAA Thermophysics Specialist Conference, New Orleans, Louisiana, April 17-20, 1967. 3. Krieth, Frank, Radiation Heat Transfer for Spacecraft and Solar Power Plant Design, Scranton, Pa: International Textbook Company, 1962, 4. API Thermal Control Systems Standardization Study, No. ETM-038, Bendix Aerospace Systems Division, Ann Arbor, Michigan, Dec. 11, 1967, 71

7 STRUCTURES 7. 1 INTRODUCTION The structure of Linus was designed not only to provide adequate support for the satellite systems, but also to remain within the constraints of thermal control, launch capability, and cost. Structural stability of the satellite necessitated strong components to withstand high launch accelerations, yet launch constraints required the satellite to weigh under 940 pounds. To meet these requirements aluminum (sheet, tubing, extrusions, and honeycomb) was used in all structural components: aluminum combines high strength with light weight as well as good thermal conductivity. Each component has a safety factor of at least 1. 75 so as to ensure adequate strength. Vibrational requirements are met by designing all components to have a natural frequency greater than 40 cps (Reference 8. 1). The major structural constraint was to provide structural support for, and rotational movement of, the solar panels which must extend beyond the 3.6 meter dish antenna. Due to shroud size however, these supports had to be folded which necessitated a specially designed solar array driving mechanism from the spacecraft body to the solar panels (see Figure 7.6). The general satellite configuration is shown in Figures 7. 1 and 7. 2. The main dish antenna is mounted on top of the cylindrical body which houses all the satellite systems. The main factor in determining the size of the body was the fuel supply which makes up over 35% of the total spacecraft weight of 872 pounds. All of the launch loads are carried by the support tube which is the strongest structural component of the body. All of the equipment is mounted on one of:four honeycomb plates with the batteries and communications package also attached to the skin for thermal reasons. 7.2 MAIN STRUCTURAL COMPONENTS 7. 2. 1 Support Tube The support tube is designed to receive forces from the rest of the spacecraft and transfer these forces to the attach fitting. The tube is strong enough to withstand all launch accelerations and has a safety factor of 1. 99, It is 30" long and 18" in diameter,

Omni Antenna Stationary Feed Rotating Feed Main Dish Antenna z y Moon Sensor I ii I I i I'll till Sun gerso Supporting Arms Solar Panel Attitude Control Body Thrusters Omni Antenna Midcourse Correction Engine Star Sensor Scale: 1"=2' Figure 7.1 General Configuration

x z r~ --— Figre 7. Geneal Cof iguatio rss ~~~ T Scale: 1"=2' Figure 7.2 General Configuration

i~~~~~~~~~~~~ I" r - Main Dish Antenna / Solar Panel I~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ i Shroud Ir I \\~~~~ iI I; /r~~~~l I 11 I Ill I.i II t I lI Iil;t~ ~~~l;;l III 1 I' I I~ II' I II llii & I.' I.I II'' \\II rr I I liii I I II ~ ~ ~~~~~~~~~~~~''1 Uil IIIIii ru tIlI II ~ II I II ~.II llL-~ II -J1 1 Il Ii', 1 I i Scale: 1"=2' Figure 7.3 Folded Position 75

7. 2. 2 Equipment Platforms and Top Plate All equipment is mounted on one of these four honeycomb platforms. They are thick enough to handle all launch loads, provide good vibration damping characteristics, and still maintain low weight. The facings are sufficiently thick to provide adequate thermal conduction to and from the various electronic packages. The honeycomb top plate provides sufficient mounting for the antenna. 7. 2.3 Skin The skin is actually three separate pieces: two radiating surfaces and one general spacecraft coating. Each radiating area is mounted on the side of the body facing a solar panel; thermal control requires this. The radiating area on the battery side is 9" wide x 20'1 high while on the communications side it is 33" wide x 30" hiigh. 7. 2. 4 Solar Panel Supports These supporting arms are designed such that as the panel is being deployed by spring-loaded hinges, the lower strut will slide along the upper one and lock into place when the upper strut is horizontal. Pinned joints are used in these struts to negate any bending moments; hence only axial forces exist. The spring hinges are made such that when the arms are deployed, the hinges still have stored energy in them. This requirement is specified regarding the hinges as a safety factor: if the locking mechanisms on the supporting arms fails to lock the arms, the spring hinges will be able to hold the solar panels in the deployed position (see Figure 7.1 ) 7. 2. 5 Stiffened Panels They are designed to give overall rigidity to the spacecraft body and to take any forces from the top plate and deliver them to the base plate and support tube. The panels prevent any forces from existing in the skin while providing a mounting surface for the skin (see Figure 7.5 ). 7.2.6 Adaptor Plate This component has three purposes. It connect the base plate to the support tube and joins the support tube (and spacecraft) to the attach fitting. It also is the surface to which the 20 lb thrust midcourse correction engine is attached. 7. 2. 7 Support Rings These three rings connect their respective platforms to the support tube. They also add rigidity to the support tube. 76

7. 2. 8 Struts The struts are for the purpose of transferring the weight of the solar panels to the support tube. Small vibration rods are pin-connected from the strut mid-point to the support tube to supress launch vibrations. 7. 2. 9 Solar Drive Mechanism This driving mechanism is designed to rotate the solar panels from inside the spacecraft. Once the panels are in the deployed position, a command is sent from the ground to release a spring which forces the driveshaft sleeve into the driveshaft socket yet still remaining engaged to the driveshaft, thus completing the drive train to the solar panel (see Figure 7. 6). If the keys on the driveshaft sleeve do not line up with those in the socket, the driveshaft can be rotated until they do. 7.3 PARTS SPECIFICATION (see Figure 7.4) Item No. Total Wt. # Material A. Support Tube 1 8.5 7075-T6 aluminum t O.050" B. Base Plate 1 9.9 Honeycomb sandwich Core: -. 5" t thick 125x2024x. 00 15 alum. Facings:. 020" 7075-T6 aluminum sheet C. Momentum Wheel 3 2.5 Honeycomb Sandwich Platform Core:. 25 thick. 125x2024x, 0015 alum. Facings:. 020" 2024- T4 aluminum sheet D. Fuel Tank 1 1.3 Same as C. Platform E. Top Plate 1 6. 6 Same as C. F. Skin 3 pcs. 10. 2 2024- T4 aluminum sheet Radiating area t =. 050" Remaining area t -. 010 " G. Solar Panel 4 10. 2 Horizontal: 1.5 " OD. 7075 -T6 Supports alum. tubing t =. 10T Angle: i. 0"! O.D. 7075-T6 alum. tubeing t o.10" 77

Scale: 1"=6" G A,r II Ii AII FII n Figure 7\4 Main Structural Components

Scale: 1"=4"... ~ f 0 ~. ~ I ~.o* - Rivets //'7./Dr I "-Angle Stiffener //'// ~Sheet l [ / Channel Stiffener * '62-'- 75" *, t=.0625" ~ 75" Angle Stiffener Cross Section ~* D9S, - ~t=.0625" 7. 1 K.*.......... Channel Stiffener Cross Section Figure 7.5 Stiffened Panel Details 79

No Scale Solar Array Drive Motor Solar Panel Universal Joint Support Arm Dr iveshaf t Release Spring Go Teflon Coated Bear ings Driveshaft Sleeve Catching 1 I I / Mechanism I ji tt~~rc~ - ~1 illP i r~~~~f I I;1-l-(-ADf u ill I ~~~~D riveshaft Guides Driveshaft Socket Spring-Hinge Spring-Hinge Body Figure 7.6 Solar Array Dr Ive Detail

H. Stiffened 6 11. 0 Facings: 2024-T4 alum. sheet t =.03 0 Channel: 2014-T6 alum..51" x.751" x o5' t =.0625" Angle: 2014-T6 alumr.75" x.75" t =.06251" I. Adaptor Plate 1 5. 0 7075-T6 aluminum J. Support Rings 2 4. 1 2014- T6 aluminum for C&D 1. 01 x. 01. t-. 15" K. Support Ring 1 1.5 Same as J. for Top Plate L. Struts and 2 3.5 75" 1 D. 7075 T6 Mounts alum. tubing t -. 10!" M. Vibration Rods 2..5.2014-T6 alum. tubing N. Spring Hinges 4 2.0 Aluminum and spring steel Solar Drive 2 16.0 Me chani sm Fuel Tanks 5 15.5 & Supports Miscellaneous 10. 0 Total Structural Weight 117.3 lbs 81

7.4 WEIGHT DISTRIBUTION Group Component Weight lbs Guidance & Control Control Electronics 20. 0 Gyro Electronics 13. 0 Gyro Package 7. 2 Computer 12,3 Yo-Yo Despin 10. 6 Fuel 320.0 M. C. C. Engine 2.8 Thrusters (12) 3.0 Momentum Wheels (3) 30.0 Nitrogen & Tank 15.0 Solar Drive Motors (2) 16. 0 Sensors & Electronics (9) 34. 0 483.9 Power Solar Panels (2) 80. 0 Electronics 12.0 Battery? 73.0 165.0 Communications Electronics 5 7. 0 Antenna -dish 21. 0 Antenna-omni 5. 0 83.0 Structures See 7. 3 (Fuel Tanks, Base Plate, etc.) 117.3 Thermal Heat Pipes (2) 3.5 Insulation 2. 9 Heaters & Switches (33 sq. in. ) 1. 2 7. 6 All Groups Miscellaneous 15. 0 TOTAL SPACECRAFT WEIGHT 872 pounds GROWTH MARGIN 48 pounds TOTAL CAPABILITY 920 pou-nds The total capacity 920 pounds is less than the launch capability of 940 pounds because the fuel supply limits the size of the spacecraft which can be adequately controlled for two years to 920 pounds. 82

Scale: 1"=1' Fuel Tanks Radiating Area uidance & Control Electronics Communications Gyro Package Computer oon Sensor ~- - - Electronics ower Heat Pipes lectronicsH Batteries Radiating Area Momentum Wheels Solar Array / '/ Motors Fuel Tank N Tarnk Midcourse Correction Engine Section A-A Figure 7.7 Equipment Layout 83

7.5 REFERENCES 1. Eisley, Joe G., and Jenkins, David R., t Aerospace Structures"9 University of Michigan, 1966. 2. Hexcel Products, "Design Curves", January 1965. 3. Osgood, Carl C., "Spacecraft Structures", Prentice-Hall, 1966. 4. Peery, David J., 'Aircraft Structures", McGraw Hill, 1950, 5. Timoshenko, Stephen, and Gere, James, "Theory of Elastic Stability", McGraw Hill, 1961. 84

8 LAUNCH VEHICLE AND SUBSYSTEMS 8.1 INTRODUCTION In choosing a launch vehicle for Project Linus two basic factors were considered; 1. finding a vehicle capable of placing an 872 pound payload in translunar injection and 2. finding a vehicle with cost in reasonable proportion to total mission cost. With this in mind, the larger families of launch vehicles such as the Atlas and Titan configurations capable of earth escape payloads of 2,300 to 12,500 pounds but costing $17-18 million were decided incompatable for use with the relatively light spacecraft proposed. The availability of the smaller Atlas configuration by the 1972 launch date is in doubt. Thus the uprated Delta (AJ10-118F second stage) with nine-castor augmentation developed by McDonnell Douglas Astronautics Company/Western Division was chosen due to its adequate capability and relatively low cost. The capability of the Delta vehicle is based on the C3 energy (see Section 2.3). For this mission C3 = -2. 02 km2/sec2 which gives a maximum payload capability of 940 pounds. See Figure 8. 1 for payload capabilitie s. Relatively low cost combined with a high degree of reliability experienced on previous launches give the Delta launch vehicle the cost effectiveness necessary to provide a reasonable booster to mission cost r elation. 8. 2 DESCRIPTION A long tank Delta will be used to launch the spacecraft into a tra-nslunar injection trajectory. More specifically the Delta configuration (see Figure 3.2) consists of DSV-2L-1B first stage, an AGC AJ10- 118F second stage, and a TE 364-4 third stage. First and second stage performance parameters are shown in Table 8. 1. The first stage uses a Rocketdyne MB-3 Block III liquid bi-propellant main engine augmented by nine TX-33-52 (castor I) Thiokol solid motors. The liquid propellant consists of the hydrocarbon fuel RP- 1 and a liquid oxygen oxidizer. The solid propellant uses an oxidizer of ammonium perchlorate (NH4C104) and a hydrocarbon aluminum fuel. The combined thrust is 656, 000 pounds with 170, 000 pounds from the MB-3 and 486, 000 pounds from the castor I package. The nine castors burn for 40. 2 seconds aftear which the main engine continues to burn for an additional 178 seconds, bringing the total burn time of the first stage to 218 seconds. 85

CGEner 3 2 rg~ 20 (km /sec 15 10 00 5 940 Payload Is 0 -100 200 500 600 700 800 900 1000 1100 -2. 02 -5 Figure 8. 1 C3 Energy vs Payload Capability of TAT-Delta-TE 364-4

C.D.C.S. Fairing t_'o.... TE 364-4 AJ10-118F DSV-2L- lB Castors Figure 8.2 TAT-Delta Launch Vehicle 87

The second stage is the uprated Delta with an Aero-jet General AJ10-118F liquid bi-propellant restartable engine. This is the same engine found in the Titan transtage and will be available for this Delta configuration in 1971. The liquid propellant consists of Aero-50 and uses nitrogen tetraoxide (N202) as an oxidizer. Aero 50 contains 50% unsymmetrical dimethyl hydrazine (UDMH) and 50% hydrazine (N2H2). The second stage produces a thrust of 9,450 pounds and can burn for a maximum of 324 seconds. The third stage is powered by a Thiokol TE 364-4 currently under development. This uprated TE 364 gains additional performance by increasing the propellant loading from 1440 pounds to 2100 pounds. The solid propellant used consists of polybutadine acrylonitrile with aluminum additives (PBAA) and uses NH4C104 as an oxidizer. The TE 364-4 is expected to be available for Delta mission applications in 1971. Table -8. 1 First and Second Stage Performance Parameters Isp Thrust Burn Time Weight Stage (sec. ) (lbf. ) (sec. ) (lbf.) 9-castor I 274 486,000 40. 2 79,803 1st MB-3, Blk. 3 285 170,000 218 153,867 Combined 282 656,000 218 233, 670 2nd AJ10-118F 296.6 9,450 324 11,895 *total launch weight 249,333 *1Total launch weight includes third stage, 2, 272 lb; payload, 872 lb; fairing, 600 lb; and attach fitting, 24 lb. 8.3 LAUNCH FACILITIES The Delta vehicle will be launched from the Eastern Test Range at an 80~0-1100 launch azimuth. This azimuth range falls within the ETR launch safety limits of 440-1100 and produces an earth orbit inclination of 28.5~-30~0. It also assures two launch windows per day (see Section 2.5), The ETR was chosen to provide an eastward launch trajectory which will take advantage of an additional velocity component due to the Earth' s eastward rotation, Combining both the relative velocity of the vehicle and the rotational velocity of the earth yields a higher inertial or total velocity. The Delta vehicle chosen could not supply the C3 energy required if a westward launch is employed. 88

8.4 LAUNCH VEHICLE SYSTEMS 8.4. 1 Guidance and Control An improved inertial guidance system is housed in the second stage and provides guidance signals for both first and second stages during their respective phases of flight. Guidance signals override the autopilot which is set to command the second stage through a series of ground set turn rates. The vehicle-borne guidance unit generates a pulse that is transmitted back to the ground-based tracking radar for use in automatic tracking and generation of position data. First and second stage control systems stabilize the vehicle through a predetermined trajectory. These systems stabilize control attitude, provide stabilization damping, and reorient the vehicle to the proper trajectory. The third stage is spin stabilized. 8.4. 2 Electrical Power Systems An AC and DC system is used to power both first and second stage components and equipment, AC power is supplied from a three-phase four-wire system with a one-point neutral. The DC system has the negative side grounded. 8. 4. 3 Telemetry Systems First and second stage telemetry systems are similar, each combining the use of pulse duration modulation and frequency modulation (PDM/FM/FM and FM/FM). 8.4.4 Range Safety System Command destruct receivers of the first and second stage are tued to the same frequency and respond to the same RF modulated signal from the ground transmitting station. Dual command destruet systems are installed in the second stage. 8.5 SPACECRAFT CONSIDERATIONS 8.5, 1 Spin-up Spin rockets are mounted on the spin table and fired approxirnately 15 seconds before third stage ignition. The spin table is mounted between stages and allows the third stage to be spun up while still attached to the second stage which does not spin. The third stage is then separated fr om the expended second stage by release of the spin table-to-third stage separation clamp. Clamp release occurs simultaneously with the retro of the second stage to assure a positive separation. The third stage is then 89

ignitedo Fintally, after third stage burnout the spacecraft is separated from the expended motor case by release of fitting-to-spacecraft separation clamp which allows the spring loaded attach fitting to eject the spacecraft. 8. 5. 2 Spacecraft Attach Fitting The spacecraft is supported on the third stage motor by an attach fitting. This fitting includes three separation springs, a separatioin clamp, explosive bolt c.lusters, battery, separation timers, pyrotechniicr delay switches9 and associated electrical components necessary for separation. Physically the attach fitting is a cylinder 9 5 inches high with an 18. 4 inch diamneter weighing 24 pounds and capable of supporting a 1200 pound spacecraft. 8.5.3 Fairing An RF transparent fairing protects the spacecraft from aerodynamic heating and thermodynamic effects while in the atmosphere, The currently available Delta fairing will not house the Linus spacecraft when the TE 364-4 third stage with the larger fuel load employed. How ever a new fairing compatible with the three stage Delta vehicle is being developed for the Canadian government for use in the Canadian Domestic Communications Satellite (CDCS). This new fairing is 240 inches long and has a 77 inch internal diameter (see Figure 8.3), The structural characteristics will be similar to the current fairing which includes monocoque fiberglass construction, The new fairing wil weight about 600 pounds and should be available in late 1971. 8.6 SEQUENCING The launch vehicle staging sequences occur at time intervals specified in Table 8. 2o The sequences start with the occurance of lift-off followed in order by solid moteo)r sepa ration; main engine cut-out (MEC0O); first burn of the second stage engine at separation from the booster; release of the spacecraft fairing followved by the initial second stage engine cut-out (SECO-I). Following SECO-I and coast phase sequences, a final short burn of the second stage occurs. The third stage is then separated from the second stage and ignited following SECO-II. After this burn the spacecraft is ejected from the expended stage. 8. 7 ASCENT TO LUNAR INJECTION Linus will be launched from Cape Ke:nnedy with an optimumrn launch azimuth of 90~. This permits Linus to benefit from the full affect of the earthts rotation. Launch will be initiated on November 23, 1972 at 6 hours 27 minutes 10 seconds9 Greenwich Mean Time. Attainment of a 200 km circulat pardking Oa: bi$t will oCcur ap]prosximilaltely %O minutes afigerl ch, This will requlire the fu 1l. bu rn of the first stage and the fiErst of twro s conld stage bulrns, 90

Payload Envelope - - O -"- j- TE 364-4 _____ - - 4-01 /~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ -— 4 ----Dia 1 I ~L~ --- —Y ~ 8~ 651' D i 842 DiaM Figure 8. 3 Proposed C. D. C. S. Pairing.. -c,~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~O-.a..W. 9, _ B 260, Max Fiue8 3Pooe C.C S arn

Table 8. 2 Basic Sequence of Events Event Time First stage ignition t = 0 Castor burnout t = 40. 2 sec Castor separation t = 75 sec MECO t 218 sec Coast (4 sec) Second stage burn I t 222 sec Fairing separation t 237 sec SECO4I t - 534 sec Coast (approximately 60 mran) Second stage burn 2 (12 sec) SECO-II Third stage burn Third stage burn out t = 4198 sec Separation of third stage and spacecraft about 90 sec after third stage burnout The cicuelaar parkoing orbit wail-give Linus the capability of coasting to achieve its injection point over the Pacific Ocean. This orbit will be at an inclination of 29.40 to the equator (see Figure'8.4). The coast phase of the- trajectory will last about 60 minutes. The vehcle will travel seven tenths of one complete parking orbit revolution. This gives a coast angle of 2500 (see Figure 8.5). The inertial velocity of a vehicle in this orbit is 25,581 ft/sec Linus requires a total velocity of 35, 815 ft/sec; therefore, an injection velocity of 10, 234 ft/sec is needed. To supply the AV needed, the second stage will be restarted and burned for 12 seconds followed by the burn of the solid third stage. This will give Linus a' AzV of 10, 300 ft /sec. This esxtra velocity will take care of any plane changes that may have occubred due to errors in the launch phase. 92

\ ~ ' Launch Site 1. — Equator Parking Orbit Ground Track of Parking Orbit i = inclination angle Launch Vehicle Burnout Figure 8.4 93

\ —.- _ Burnout Launch -1* Injection (I = Coast Angle Figure 8.5 94

8.8 ALTERNATIVE PIGGY BACK, LAUNCH An alternative method of launch is to launch Linus with a Saturn V in conjection with an Apollo mission. There are six compartments in the service module, four for fuel, one for electronics, and the last is presently empty. Linus, as currently designed, could easily fit into this empty bay and thus save the expense of the three stage thrust augmented Thor Delta Rocket. However, this method was not considered, in order to maintain Linus as a completely independent' project. 8. 9 REFERENCES 1. "Delta Payload Planners Guide", McDonnell Douglas Company, Western Division, April 1969. 2. "DSV-2L Space Booster, Payload Planners Guide", Douglas Missile and Space Systems Division, December 1966. 3. '"XM 33 Rocket Motor", Thiokol Chemical Corporation, Revised March 1962. 4. Aviation Week and Space Technology, March 7, 1966. 5. Trajectory Study Run for Project Linus by NASA Goddard, Groves, R., November 10, 1969.

9 IMPLEMENTATION 9.1 INTRODUCTION The previous eight sections have dealt solely with the technical aspects of Project Linus. This section will present a brief discussion of implementing the project. Topics to be discussed include cost, work scheduling, testing, and reliability. These subjects are by no means the only ones to be considered under implementation, but they are the most important. Project Linus is to cost 25 million dollars and take 34 months from design until launch in November of 1972. In order to assure a high probability of success, redundancies must be built into the system and reliable components must be used. Also extensive testing must be done to provide a fully qualified spacecraft launch. 9.2 COST ANALYSIS The total cost of Linus is 25 million dollars. This includes the design, fabrication, and testing of a structural model, prototype, and a qualification test article. It also includes the cost to bitild, test, and deliver two flight accepted spacecraft to the Eastern Test Range. The costs are broken down in Tables 9. 1 and 9. 2. Table 9.1 is divided according to material, labor, and other general expenditures. Table 9. 2 splits the cost into recurring and nonrecurring costs. The nonrecurring costs include those to design, develop, and qualify the spacecraft. The recurring cost is 1.5 million dollars and includes the expenses to fabricate, test, and deliver one flight unit. The cost of the three- stage thrust augmented Thor Delta booster is not included in the cost total nor are such expenses as launch operations or maintainence of the spacecraft once it is launched. The amount estimated is the cost to deliver two flight accepted units to the launch pad and the needed testing and development to assure a successful mission, 9,3 WORK SCHEDULING Assuming a contract is awarded for Project Linus in January of 1970, there are 34 months left until the proposed launch date of November 1972. This is sufficient time to properly design, fabricate, test, and deliver the spacecraft and a backup system to the Eastern Test Range (see Figure 9 1 ). 96

TUab1le 9.1 LINUS COST ESTIMATE Direct Labor Eaxgineering -----------------------— 4$ 5,200,000 Other.. --- —--- ------------------— $. 3,280,000 Labor Burden(@ 110%) -------—; --- —---— $ 9,200,000 Direct Material. --- —-------- -—. ---- $ 2,200,000 Travel1 -.-onM.a.c.-.M --...............$ 400 000 Other Direct Costs --- —---------- 200000 Total Direct Cost and BUrden. --- —--— $20,480, 000 General and Administrative --------- $ 2,660,000 Total Estimated Cost ---..................$23,140,000 Fee or Profit ----—... --- —--—..0 --- —- 1,860,000 Total Program Cost ------------—..-. —$25,25 0000 000 Table 9.2 LINUS COST BREAKDOWN Non-recurring Costs Development --- —-------------------— $2 02 00,000 ~Qualification -. --- —-------—.-...-.$ 1,800,000 Recurring Costs Linus I ---------------------------- $ 1,500,000 Linus II --- -------- - - wan --- —---- $ 1,500,000 Total Program Costs$ 5...........000 97

Figure 9. 1 Work Scheduling ~ -= _ iii~ IZ ~ ~3197 19?1 2 1972 i ii M A.i il A. S... M —U i M. J 0 ID. M.... ' Mognths From Contract -Award ' 'o- _ _ J 6 OF..Majm...... L LL1 I I I I. C.omonent. Assembly Development i 1 P ". P- t. 'i -'. '-...,e._ -.esigt Fabiri.i ation...and...... ' L ' t:'np.~!.! 00 Q u.nalif{tion Test Article..... tFs-br igtioan — and sesembly Fmiight Sn~eraf t Desi. n i r |Fabr'catio Pmnd... A'sembly - P teeieaneefl Tfiead. -a1 E Linus I Delivered Linus II Delivered a ~,i0u..~. I.....-... I I II 1 _ 11 1 1.1 1. I F'eab'c atio.I I iL I-1 1 1.sm A Linus I Delivered A Linus II Dellered A Launch

The work schedule is divided into six main units (Cmonponent Assembly Development, Structural Thermal Test Article, etc. ). Each of these in turn is basically made up of a design, faLrication, and test phase. Both the design and fabrication periods are quie flexible. - By adding rnore designers and-technicians to the project, the time for design and fabrication can be shortened to some extent. The test times, however, are approximately fixed in their length of time, so it is essential to sched-iae around the test periods to obtain maximum benefit from them. Therefore, the final design of the qualification test article is not completed util all tests on the components, the structural thermal test article, and the prototype are completed. This allows for any rough spots to be designed out of the system before fabrication of the flight units are finished. The three milestones indicated in Figure 9. i are the deliveries of the spacecraft and its spare to the launch pad and the launch itself. The delivery of Linus I two months prior to launch allows sufficient time to check out the spacecraft before the launch. If something is found unacceptable, the launch could be postponed a month and the spare used. 9.4 TESTING In order to assure a high degree of success, time must be spent in extensive testing. Therefore five units of each component are to be manufactured. The testing s-equence for Linus is briefly described below. The first tests, Component and Assembly Development Tests, will be to check out the components of each subsystem. The hardware and the.. tests to be perfo~rmed on each unit are summarized in Figure 9. 2. MMost of the subsystems have little new development in them so quite extensive testing like thermal vacuum and life testing are not required. However the rotating feed, solar panel deployment, and momentum wheels are somewhat unique to Linus so additional tests as indicated in Figure 9. 2 are needed. Next a structural thermal test article is fabricated and tested. All subsystems will be simulated as black boxes sufficient to duplicate the center of mass, weight and heat output. The test article will consist of the following: 1. Body structure 2. Thermal idontrol 3. Fuel etanks 4. Solar panel structural simulator 5. Subsystem rinmulators 6. Antennas 7. Instrurmentatio

P4 P i -p - H o H IdO i O 10. Communicatios __s_ _ _ CommP unicat ' - a. Electronic sl C d!.~~ enors.......r c0 Power Conditioning~..... Omni Anten d. P owe r Distributions. Uni.t ___ S.trutures and Therma_. b. 1 e.. ___, a Panel C. Bod. Structur ____. Li__ _. Ra P "onJre d.ThermalControl _. _-___ _ _ _e. asFigure 9.2 Component Testing C Bg.d, IStrucure: _ l L:J!~no

The following tests will be performed on the model: 1. Vibration 2. Shock test 3. Thermal vacuum with solar simulation Development testing is completed with the prototype. This test article will be a complete spacecraft with fully operational subsystems. In step nine below, the entire mission will be simudated in the space environment to see hew the components react as a total system in simulated space. The following is a set of tests to be completed on the prototype. A. Ambient functional 2. Mass properties 3. Vibration (stowed position) 4. Ambient functional 5. Shock (stowed position) 6. Ambient functional 7. Acceleration (stowed) 8. Ambient functional 9. Mission simulation 10. Ambient functional The qualification test article must now be tested in the space environment (acceptance test) and also to the limits the spacecraft was designed (Design margin test). The following isa list of these tests: Acceptance Test 1. Ambient functional 2. Mass properties 3. Vib ratio n 4. Ambient functional 5. Acceptance thermal vacuum 6. Ambient functional Design Limit Test 1. Vibration 2. Ambient functional 3. Shock 4 Ambient functional 5. Acceleration 6. Ambient functional 7. Design limit thermal vacuum 8. Ambient functional

Before the flight models are delivered they undergo acceptance tests as described above. When they arrive at the launch facilities they must go through a set of prelaunch tests which include the following: 1. Spacecraft inspection 2. Spacecraft acceptance test 3. Final checkout on the launch pad 9.5 RELIABILITY In order to assure a high probability cof mission success a numnber of precautions have been taken. First, the above mentioned tests will add to the reliability by spotting poor design before it gets into the final spacecr4ft. Also two flight nmodels will be fabrtcated and checked out to mrneet the lau-chX date in November of 1972. For the most part, already proven subsystems are being used such as the battery pack, sun sensor, reaction jets, and solar cells. Where it is not possible to use completely reliable flight tested components, redundancies have been built into the system. An example of this is the redundancies built into the communications package like the two traveling wave tubes and the multiple switching unit. Overall reliability was designed by attempting to use the simplest methods possible. 9.6 REFERENCE 1. W. R. Corliss," Scientific Satellitest", NASA SP-1330 102

APPENDIX A. 1 LIB RATION POINTS The following figure shows the five earth-moon libration points. All five lie in the plane of the moon's orbit around the earth (see Figure A. 1). C3 ~~~~~~M LI ~~ o L2 Moon Earth L5 Figure A. 1 Libration Points Earth-moon distance = 384,500 km = D Earth - L3 distance - 0.993D Earth - L4 distance D= Earth - L5 distance = D Earth - CM distance = D/82.45 103

Moon - Li distance 0o, 151D Moon - L2 distance 0. 163.D Moon - L4 distance - D Moon - L5 distance - D Lagrangets theoretical law - the centrifugal force of a small body revolving around the common mass center of two celestial bodlies is exactly balanced by the combined gravitational forces of the two celestial bodies. Theory Assumptions 1. The gravitational fields of the earth and moon are uniform 2. No perturbative forces 3. Earth is stationary with the moon revolving around it in a circular orbit The period of one body revolving around another is found by equating the centrifugal and gravitational forces. MV 2( 2 D)2X - GR2 m p = T2TD D2 where: M= mass of central body G = surface gravity of central body R - radius of central body D distance to revolving body T - period of revolving body m -mass of revolving body From this equation it is observed that.2 4wr2D3 T = A. 2 2 Commonly known as Kepler' s third law for circular orbits. By placing a body (satellite) at the second libration point it would revolve about the earth with the same period as the moon. The nomenclature for the calculations is as follows: G = surface gravity of the earth g = surface gravity of the moon R - radius of the earth r - radius of the moon D - earth to moon distance d = moon to satellite distance T = period of the rmoon and satellite 104

R y T z D 0 Moon Figure A. 2 105

The resulting force on the satellite due to the combined fields is: mGR2 mgr2 + 2 A. 3 (D+ d) d The centrifucal force due to the satellite' s rotation about the earth is: mV m r 4rD (f4 +d) m P (D+ d) T T2 equatng; OR2 24 D G+ gr r _A. 5 2 2 A (D + d) d T 2 GOR ~(D+ d) (from Kepler's third law) To facilitate calculations let gr2 = a and d - kD GR Equation A. 6 becomes: 1 a Di + kD (kD + D )2 (kD)2 D3 2+ 2 (1 + k) (k +1) k k2 + a (k + 1)2 = k (1 + k)3 k + 3k4 +.3k3 - a (1 + k) 0 A. 6 using g 0.16, G R 63 00 km r 1760 km a is found to be approximately 0. 012 By substituting a into equation A. 6 and solving it is found that there is only one positive real root. k 0. 165 106

Thus d = kD (0. 165) (384,500) d 6- 4,000 km which is approximately the distance from the moon to the second libration point. The references for this work were Reference 1 and 4 in Section 2. A. 2 RELATIVE MOTION ABOUT THE SECOND LIBRATION POINT R - xe + ye + ze A. 7 x y z r rT e A. 8 rR rT + R - (x + rT) e + ye + ze A.9 For a passive target; Zd.rT - e - e e A. 10 2 2 r 2 x dt rT T rT For an active rendezvous vehicle; f = f e f e + f e A. 11 x y y z z 2 -d r 2 R d -- -' -- e f - 2 (rT + R) e f dr rR e r e = R A. 12 R R r rR d rR r + f A. 13 dT rR When R is sufficiently close to T (AQ small) we have cos A-i1 107

rT x 1 cos T cos Ac or - r 3 r +x R rR T 43 1 _ cos, a, rR3 (rT + x)3 3 ( = rT 3xrT T~ rR (rT + x) R T use x y. z As sume << rT rT rT r(r + - ~)e 3 R [( 2 3 3 4 x 3 4 r R rT T rT rT rT rT rT Neglecting high order terms; r 3 r e e+ A. 14 3 x 3 y 3 z R L\rT T T T d R d (rR rT) A. 15 dT dT +T + e( + f e 3. x 3 y y T, rrT + ( + fx ez — R dR R V - dt - + co xr A. 16 dt /A 108

where W = angular velocity -3 e z dR k e e e dt x y+Y e + z e w x r e e e x y z 0 0 0 x y z =-oye +co xe y ex w x e V = (:k- wy)e + (j + wx) e +ze A. 17 dV dV - - a - d + ~ x V A. 18 dt dt dA d = ('i.ojr +( +i)e +~ e dt A x y z x V = e e e x y z 0 0, 0o (e -oy) (jT +x) Z -- (j+ ox) e + ( - y) e -~ 2 2 a =(x - ZT -0 x) e + ( + Zi - y) e +' e A. 19 2 a =-d A. 20 dt Ecpqating A. 15 and A. 19 we get: x - 2y - x 2x_ + f A. 21 3 x rT r T 109

z = _ z +f A. 23 3 z rT The right side of the equation consists of the gravitational and control terms ] =(~E + FM 3 where ()E is an earth term rT 'rE rM ()M is a moon term A.24 G ME G M M 3 + 3 rE M D(M + M lk M M rE = rM + D where D is the earth-moon distance ((ME + MM) G D3 ME +MME +MM rM - A. 26 = K A. 27 rT xi -2j - (2K+ 1)x=f A.28 cx 2 y + 2x + (K - )y = 29 + EK = f A. 30 c z 110

By normalizing the calculations become simplified Characteristic length D Characteristic time 1/W Characteristic vel DW Characteristic Acc DW2 2 2 2 I xD ADY - DZ (2K + )x =f D.w x - 2w w~Cx 2 - D. y + 2c D _ -c D (K - ) y = f A. 31 2. 2 ^ Dw z + K0 Dz f cz Dividing through by Dw ) and dropping the cap we arrive at the following equations: f x - 2 - (2K + 1) x f' A. 32 D.o f cyK y = f' A. 33 2 cy czDo 'z + Kz -cz A. 34 2 cz Now the constant K is normalized. As sumptions 1. Mean distance between the earth and moon is one 2. The sum of the masses of the earth and moon are one (the moon' s mass is taken to be pL) Thus: ME = 1 - MM B rE 1 + L 1 2 2 rM - Lz Figure A. 3 111

Plugging these values in Equation A. 26 results in-; 3nt 11 3 K= ( + + --- A. 35 3 3 ( L 2 L2 A. 3 AV COST FOR Z AXIS PERIOD CONTROL z = A sin ( t z z - A o cos X t z z z.2 2 2 2 2 z + o z -A to z z z (see Figure A. 4, I) T = - T = z w period of y X period of z -motion y-motion The control scheme is to shorten the z-motion period so that it matches the y-motion period. 1. Double z-axis Period Control The y-motion takes T /4 to traverse one quadrant in the phase plane. The technique is to alYow the z-motion to proceed normally until t = Ty/4. At this time an impulsive burn causes the rest of the z-motion to be skipped and the time to complete the quadrant is the same for the y and z motion. (see Figure A. 4-, II) The z velocity at t - Ty/4 must be reversed (same magnitude, opposite sign). T 4 2w zo - A sin X~ - 2 A sin 2 (- Z I Z% 7 Z 12

0 A z o X Cs O3 Z 4O ec Z 1O A = - COS - sin!m 1 6093 tan 786214 F2. the etre orbt tSie total z-ax is foPer tiod Control FThe a 4000diagram kfor this scheme is shalo in Figure A 4, bit ra-6 radd ft ~4000 km.66g 10 365 6 AV/yr = 4 (1.78618) 5280 (4000 k) 662 x 10 sec 36 km ml ~..'78618 1 4.667 2. S'i.gte z-axis Period Control The phase diagram for this acheme is shown in Figure A. 4, III. -rad 1 km m tan 1. 78618 (2 (1.8 2 - 1. 8 14,67 1 6()93 km [ ~ (2) (1. 86265 1. 786 18) - a02 fps A. 4 LISSAJOUS TO HALO TRANSFER x =.3433 A sin c t Y Y y = A cos co t Y Y z = A sin (c t + a ) a t t y y z y The x and y velocities for both halo and lissajous orbits are identical~ The only difference is the z velocity. 1. Retrograde Halo (a - 0) V z velocity of halo = A o cos c t H Yz Y V z velcity of lissajous A 0 cos (o t + c) L yy -z y 113

I. Phase Diagram A z z A II. Phase Diagram Double z-axis t Control / A z III. Phase Diagram Single z-axis Control lAl~t2 w Figure A. 4 114

V VH - VL = A c Cos o t - cos (o t +a 1A ocY t Y. 3 1V1 Sy a [c o a )~ yt (1 cs a ) + sin o t (sin a)] A. 36 At the intersection of these two curves the coordinates for each curve are equal. Z = Z zH zL A sin o t = A sin (o t + a) Y Y Y Y sin w t - sin ~ t cos a cocs wc t sin a Y Y Y sin a! - cos a + tan w t y sin a 1 tan w t - y 1 - cos a tan tan 2 tan w t = ctn A. 37 Y 2 Solving for o t for various a values and plugging into equation A. 36 gives a curve for AV as a function of a. 2. Posigrade Halo (a - -180 ) z = A sin w t z = -A w cos w t y y yy y AV2 =VH VL - A (-cos Wc t)(l + cos ) + (sin w t) sin a] 2 H L y z y y A. 38 zH zL -A sin w t- A sin (w t+ a ) y y' y Y sin w t - sin w t cos a + sin a cos w t Y Y Y sin a 41 = cos a + tan w t y tan w t (1 + cos a) sin a sina a - tan c t tar y 1~cosa 2 tan c t - tan A. 39 Y 2 115

Solving equation A. 39 for w t for various a values and plugging into equation A. 38 gives a curve for AVYas a function of a. If AV is divided by A w, a nondimensional velocity, AV, may be plotted versus a (see FigureYAZ5). Two transfers are possible for a given a. One is for posigrade halo and the other for a retrograde halo. However, there is a tradeoff in AV cost. For a given a, the light portion of the curve should be used for halo implementation. A. 5 LISSAJO: US OCCULATION To maintain LOS communication with the earth the satellite in a lissajous orbit must perform large enough oscillations to be visible from any point on the earth (see Figure 2. 3). The occulted zone for the earth at L2 is a circle, 3100 km in radius. Slightly larger oscillations by the satellite will enable it to obtain LOS contact with the earth, but only for limited times since the satellite will still periodically enter the occulted zone (see Figure A. 6). The question then becomes how long will the satellite be in the occulted zone? The equations of motion in the y and z directions about L2 are: y =- A cos o t = A sin (w t + 2) Y Y ' Y A.40 z =A cos (co t + ) A sin (o t + (a + )) z z 0 z z' 0 2 The conversion of cosine to sine just alters the initial conditions and is immaterial in this analysis. Thus, let us assume the following: y A sin~ t %Y Y A.41 z -A sinr c t z z Looking at Figure A. 6 it is seen that the y motion enters the occulted region when y AM using equation A.41 AM- A sin e t M: y Y A Asi n c t A y A -=1 M sin A c t A, 42 A v 1 {

AV Expensive transfer 2.I ~~~~~~~~~~~~~~~~~ -- 0Inexpensive 2.0 transfer V1 (retrograde) 1.0 2 (posigrade) 90 180 270 360 Ir^~~~~~~ \ ~~~~~~~-a in degrees A =V1 cos c t (1-cos a )+sinw t(sina) Period = 358 days 1 y y Period a 358 days 1.0t V2 = -cos wy(l+cos a)+sinw t(sina ) For A in km. 2. 0 AV fps V VA c km 2. 662 x 10 sec y z 1.6093 mi Note: All velocities in z direction. A mirror image of these curves is also possible Figure A. 5 CAV vs a

Now one can look at o as the normalizing factor which makes the time a valid argument for the trigonometric operations. Also, by studying the motion in one quadrant, one can find the fraction of the total time the motion is occulted since the fraction in each quadrant is the same. When c t =we have ymax, and AM = Ay Thus, by dividing c t by one gets t Ye fraction of total time. Notice that this figure is al~o dimensionless. Therefore: A c t -1 AM ___t sin A 2 y A 2 -A1 AM 2 sin = T (%) A.43 vT A y y Using the same procedure for z we get, A 2 M 2 sin M = T (%) A.44 Tr A z z To get the fraction of total time the motion in the y - z plane is occulted, A. 43 and A. 44 are multiplied together. A A 4 -1 M -1 AM T =- sin sin A. 43 z 2 A A 1 by z For our analysis, A = A = 4000 km y z -1 2 T.405 (sin (. 775) ) yZ 32 or 32% An other interesting result which should be noted is the maximum period the satellite is occulted. By dividing the occulted region along the z-axis into two symmetrical planes it is seen that the motion in crossing the occulated zone along the y-axis is exactly the same with respect to time and distance on either side of the z-axis. This is also observed to be the worst case. A 1. -1 AM From A. 42 t = sin for one quadrant to completely y o A Y Y traverse the region the time should be dbubled. A 2.- M t = sin A.45 y o A Y Y For our analysis o - 1.86265 y 118

Satellite' s T ajectory M~A~ y A Figure A. 6 Periods of Occultation 119

t 1.075 sin (.775) -.95 or 95% y Using the characteristic time of 4. 34838 days the maximum occulted period in days is: t = 4. 13 days (see Reference I in Section 2) A. 6 APPLICATION OF RELATIVE MOTION EQUATIONS TO THE TERMINAL PHASE OF THE TRAJECTORY Assuming no control forces, the equations are; - 2y - (2K + 1) x - 0 A. 47 y + 2x + (K -) y -0 A.48 z + Kz = 0 A. 49 The coupled equations A. 47 and A. 48 represent the in-plane motion while equation A. 49 is uncoupled and is the out of plane equation. The in-plane equations can be rewritten using operator notation: s- d/dt 2 (s - (1 + 2K)) x + (-2s) y = 0 A. 50 (2s) x + (s (1 - K)) y -- 0 A. 51 Elimination of x from these equations can be accomplished by multiplying A. 50 by 2s and A. 51 by -(s -(1 + 2K)), and adding them. The resulting equation is: y (1 + ZK) (1 - K) + (K- 2)s - s4) 0 y s (K - 2) s + (1 - K) (1 + ZK) -0 A. 52 Solving the characteristic equation yields two imaginary roots and two real roots. Therefore a guess at the solution would be: O t =c t y Qe + B e + C sin w 2t + D cos 2t A. 53 Similarly for x; o Lt -o't x -= E e + F i 2t + H cosZ 2t A. 54 By plugging the suggested solutsisns into Equation A4 47 and A.48 the constants are found to be related as follows:s 120

E 2 Q A. 55 2c F42 =B A. 56 [2 -(l- K)] 2 H — - - C A. 57 o2 f+ (1- K)] F2 2 t(1 ( K)] The four unknowns (Q, B, C, and D) can be expressed in terms of the initial conditions x, y0, cx, and y. t- give: Q 1 -(1 - K)] B [l2 (I-K) C 2 + (1 - K)] x = 2W + - ZW 0 2c I 2c 1 2 2 A. 59 B 2 K)] [022 1 -(1 - K - 2 + (1-K)]( o0~~ 2 L~ — K K)] L - 2A. 60 y Q + B + D A. 61 o = Q BoL + Co A. 62 Solving for the constants and converting the exponential terms to hyperbolic terms results in the following in-plane solutions: 1 2 o + c(1- K) (- sinh c t -1 2 2 WI +K- 1o o -K+l W1 W2, o + 2 (cos h Olt - cos co2t) ~ A. 63 121

Y =2 I +2K&i2('1 + 2siK)n h sin t x + ((o1 - 2K + 3) cos h lt + (o2 + 2k - 3) cos o2t ) = 2 (cos h o 1t cos o Zt) o ___ __2K_ 02 + ZK + 1 (....2.+ sin h co t + sin co | t c1 1 Z Z 022 A. 64 co and o are the magnitudes of the real and imaginary roots, respectively, of EquatiOn A. 52 For. - 012150 K 3,o 19044 o I 2o 15863 01 Z 2 862635 (Reference 2 of Section 2) The Out of Plane Equation To find the homogeneous solution to Equation A. 49 it is rewritten using operator notation. 2 2 z (s +w ) - 0 A. 65 The roots of the characteristic equation are + c3i; therefore the solution is in the form: z = C1 sin o 3t + C2 os co 3t A. 66 Expressing the constants in terms of the initial conditions z and z the solution becomes: o z - sin o t + z cos o t A. 67 03 3 o 3 where w03 irK The T rar nsition from T PI t o POI For given values of the initial position, x, yo, and z, arbitrary values of the ainitial velocities x o y and z wildnot generally result in a rendezvous writhE P0I Therefore it is necessary to determine the desired 122

value s of thleese veloctites, i X, 9, and z, that will insure rendezvous. d ~d o Since the equations ofmotion, Equations A. i7, A. 48, and A. 49 are functions of time t, it is practical to specify time T when rendezvous will occur. With T chosen, it is observed that xT) = y(T) = z(T) = O and using Equations Ao. 63, A, 64 and A. 67 the desired velocities can be found. od and y can be found by using a matrix analysis: setting x(T) y(T) = 0, Equations A. 63 and A. 64 become 0 + K + 3)cos hw 1T - ( K - 3) cos 2T x A. 68 11 w Z + 2 (I - K) i sin h c1T - sin o T y 2 2 / +K- 1 K + O3 -h +sin h 2cT + sin CT x % 1 02 2 e1 + Z,,Ios hclT - cos 2T) Yd + - K +(L sin h1T - + sin a T x 2 (cos hwT cos w >T) 2 2 _ 22K -_ 2 + 2K+ 1 + i K sin h T + sinco T @1 1 W2 2 1 To facilit ate calculations Equations A. 68 and A. 69 are rewritten as follows: e 1 X,, f 2 Y( D pi 3 ~d p14 Yod Ao 70 5 ~ O 5 Y0 t 7 8YOd Ao 7d 123

Rearrangin-hg terrmns we get3 / 4Y p O Xe - 2yo A. 72 13 ~p~1 x.p y A. 73 17 ko 1 8 (od 5 O 6 o Solving for xod and jed9 0 = x od -1 132Yo) 34 d I-5 x, 6 Yo) p8 3 4 7 8 od 3 1 3 (13 o 2 '7 ( 2(5 xo 6Yo) 3 4 37 3 ther efore (P 5 K 18)xo + (14 6 2 8). _- A. 74 '~d 13338 - [4 p7 (1 1 7 1- [3 3 35) X0 + ( z 37 3- P3 6) Yo -- A. 75 ~d [33 8.4 [37 setting z(T) - 0 Eqaation A. 56 becomes ) -~ sin w T + z cos 3 T c' 3 o 3 3 The refore; 124

= = cos o T A. 76 d Before solving for the desired velocities the initial position and velocity must be determined using a geometric analysis (see Figures A. 7, A. 8, and A. 9). A numerical evaluation gives the following results: p1- 2. 45226 p 2. 10318 p o 60414 3 -.27809 4 p5. 34769 136 74308 p =-. 27809 8. 37872 Using these constants and x = -. 059404 and y =.037821 the planar desired velocities become: x f= o 22920 = 769 fps d y =. 03954 = 133 fps d The out of plane velocity becomes: iz = -54 fps (for a posgrade rendezvous *z =.01196) od 0 z 40 fps (for a retrograde rendezvous z - 0088) d To find the TPI 6V1 needed, the initial velocities are subtracted from the desired velocities. Vp (d = X) ex + (; - =r) e ~ (+ - )e A,77 -., d d y od o z Thus, for a posigrade rendezvous: AV 150 e = 57 e - 63 e 1 x y z for a retrograde rendezvous AV = 150 e 57 e + 31 e 1 x y z 125

L~, position Cente r at iendezvous of Moon a =290 48' 2 a "2~~~~~~~~~~~~~ "1 2 4 TPI Point a, 190 48' 2 h, L~~~~~~~~~~~ 2 22. 279 (1923) sin a 2 Y YO = 384, 000 y (33. 541 - 22. 279 cos a2)l923 +.343 (4, 000) X034 0 -. 0594204 o ~~~~38~4,000 x v velocity relative to moon 0 M V =velocity L2 point VL velocity relative to L2 point Figure A. 7 Conversion of Velocity and Position at TPI from Inertial Moon System to Target System (POI)

y 49 = 649 cos 17~ 3' _ = 6 ZO fps ~70 '0 7 Yo - 649 sin 170 3' Yo = 190 fps 0 y -c = 19 48' 22' -c 2 VM = 976 V1 =563 V = 649 X L2 - 6 4 0 - 90~ =17~ 3' Note: All velocities in fps. Figure A. 8 Conversion of Velocity at TPI from Inertial Moon System to Target System (POI) 127

XII xI z s 3.50 x y y L / 2 0 Moon Center x, y, z system centered at Lg x, y, z' (parallel to x, y, z) system centered at POI x11, y", x" rotated system at POI Since 3. 50 is a small ange the rotation is neglected and the ( )' axis system is used. This allows c to have the same direction as in original axis system centered at L2. The linearized equations can still be used with the only difference being in the initial conditions. 128

POI Maneuver POI maneuver = braking maneuver + injection maneuver The braking maneuver negates the satellite' s relative velocity and can therefore be written: 2V = - (T) ex -(T)e -(T) e A. 78 x2 y z 0 By differentiating Equations A. 63, A. 64, 'and A. 67 the maneuver's components are found to be: 1 W 2 + 2(1 - K) (cos h lT - cos, 2T) y 1 + [(W1 + K - 1) cos h!T + (2 -K~ 1) cos 2T] X + 2(w 1 sin h lT + 2 sin w 2T) Yo) A. 79 y(T)= 21 2 (-2(1 + 2K) (coshh T - osh h IT ac 2T) x + [O1( 1 - ZK + 3) sin h wlT - w2( 2 2K - 3) sin ZT y 2(a 1 sin h w 1T + w 2 sin w 2T) x + [( 1 ZK - 1) cosh OlT + (Z + ZK + 1) sin o 2T y A. 80 z(T) = (-co3 sin w 3T) z + cos w 3T T A. 81 3 3 o~d Thus, for a posigrade rendezvous: &V - 552 e + 45 e + 90 e 2 x y z for a retrograde rendezvous iV =-552e +45e - 66e 2 x y z 7 9o

The injection maneuver orients the satellite in the halo orbit and it has a single component in the y-direction. AV = -65 e A. 82 2. y The refore, A An.~V =V ~VA. 83 2 2 2 'o 2 A o 1 where AV 2 is the actual AV needed at POI. A For a posigrade rendezvous: 6AV =-552 e + 385 e + 90 e 2 x y z A For a retrograde rendezvous: AV = -552 e + 385 e - 66 e 2A x y z A. 7 SHADOWING To approximate the probability of shadowing, a period of one year assuming initial shadowing will be-used and each monthly pass of the shadow of the moon over the orbit of Linus will be investigated. (see Figures A. 10 and A. 11). In the analysis, irregularities and eccentricities in the orbits of the earth, moon, and Linus are neglected. From Figure A. 10: a - sin a (64, 500 km) A. 84 From Figure A. 11: Q = sin 0 (3690 km) A. 85 If IQ+a I < A, then Linus -is not in the moon' s shadow. a varies from -5 to +5 over a period of a year. The period of the orbit of Linus is 14. 67 days while the period of the shadow is 27. 3 days. The initial conditions of the test year are: a -0 0 T - dTays. (see Table A. 1) The length of shadowing varies with respect to Linus' location when occulted. Since the satellite is traveling with the shadow at 8 = 90~ the maximum period of darkness occurs at that point. 130

The velocity of the shadow past Linus is approximately 620 km/hour while that of Linus itself is 65. 8 km/hour. The length of occultation at o = 900 is: D sT = -A. 86 o V -V where D = diameter of shadow V = velocity of shadow s V0 = velocity of Linus The solution is T = 5.4 hours. Although this is a maximum case and the shadowing will not, in general, be this long, it will be used as a constraint for the size of the batteries for Linus. Table A. 1 Shadowing Probability T(days) (a )O ( 0 ) a(km) Q(km) IQ +a Shadow? 0 0 0 0 0 0 Yes 27. 3 1.5 310 2000 -2830 830 Yes 54. 6 3.0 254 3350 -3540 190 Yes 81.9 4. 5 209 5030 -1740 2290 No 109. 2 4.0 164 4510 1020 5530 No 136. 5 2. 5 115 2710 3340 6050 No 163. 8 1. 0 58 1160 3130 4290 No 191. -. 5 18 -580 1140 560 Yes 218. 4 -2. 0 324 -2260 -2170 4430 No 245. 7 -3. 5 268 -3940 - 36 80 7620 No 273.0 -5. 0 227 -5610 -2700 8310 No 300. 3 -3. 5 1 80 - 3940 0 3940 No 327. 6 -2.0 1 26 -2260 2990 7 30 Yes 354.9 -.5 72 -580 3510 2930 No 131

A ctual q OtApproximated z t50 <! < 5 A km Oj| bit of < a A Moon Shadow Figure A. 10 Geometry of Shadowing in x-z Plane z _ -- Linus Orbit of Linus ( Od_ e a 00 - Path of Shadow Figure A. 11 Geometry of Shadowing in y-z Plane 1 27

APPENDIX B B. 1 MOON HORIZON SCANNER The horizon scanner system performs the function of a primary sensor generating yaw and pitch error signals and moson edge position signals. Specifically, the moon scanner assembly generates the following signals for use by the attitude control system: (1) The pitch and yaw error signals. (2) Four tracker position signals, indicating the position of the moon-space horizon relative to the vehicle body axis. (3) Four tracker check signals, indicating whether a specific tracker channel is tracking the moon-space gradient or the sun. (4) Four sun alarm signals, indicating whether a specific tracker channel is tracking the sun. Operational De s criptlon The horizon scanner system operates as follows. Four searchtrack units (usually referred to as trackers) contain arrays of infrared detectors (thermopiles). There are forty discrete detectors in each array. These detectors are located in the focal plane of a refractive optical system in such a way that an equal number of detectors in each of the four arrays are covered by the planet image when the instrument is pointed toward the moon center (see Figure B. 1). Conversely if a pointing error exists, the number of detectors covered by the image in the arrays will be different. The outputs of corresponding detectors in each of the arrays are sampled and compared within 60 milliseconds. An output difference indicates an image shift and, consequently, a pointing error in that plane. This type of scanner requires a heat source in the detector area to prevent the ambient temperature of the detectors fromr varying significantly; otherwise small differences in temperature responsivity characteristics of the individual detectors could produce output signal differences that exceeded the' expected difference between the outputs of a detector in the planets image and one outside the image. 133

Array 1 Planet Image Array 2 Figure B. 1 B. 2 DESPIN CALCULATIONS A length of cable will wrap around the circumference one and one-half times. Therefore the length will be 316. 8 inches. The equation relating cable length to the end weight is I 2=M B.1 2 2 (L + a) a L = length of the unwound cable in ft a = radius of satellite in feet I = the spin moment of inertia of the satellite in slug/ft M - end mass of yo-yo in slugs L - 26.4 ft I 46 slug-ft2 a - 2. 8 ft 134

46 slug- ft M =2 (29. 2 ft) - (2. 8 ft) M 46 slug =.0544 slug 844. 8 M = 1.75 lbs R. = initial spin = 10. 46 rad/sec Time' to despin 1/2 1/2 -I + m 46. 15 t= jR~2Ri21 14) sec = 2. 81 sec B. 2 mRR deceleration 10.46 rad/sec 2 2. 8- sec 3.72 rad/sec 2, 81 sec components are 21 in from spin axis 1- sec rad 12 in 32.2 ft/sec maximum tension on cable 3 IR 1/2 T B. 3 max 4 2 (I +2 Ma T = 3, 741. 87 (. 041) lbs max T = 153.42 lbs. max B. 3 MOMENTUM WHEEL SIZING I - 296. slug-ft2 xx I - 294. slug-ft2 YY I = 40.8 slug-ft zz 135

Due to the 3 1/2 coning angle of Linus with respect to the moon there will be a small angular velocity (co): 70 125 rad -7 7.33 days 6.33 x 105 sec = 1 98 10 rad/ The angular momentum is Ico HI:40. 8 (1.98 x 10 ) 8. 08 x 10 ft lb sec for yaw axis HI 296 (1.98 x 10 ) 2.9 x 10 ft lb sec for pitch axis Pitch Wheel I Jco ~t aAt sat wheel for a momentum wheel that saturates at 1500 rpm (157 rad/sec) I sat 2. 9 x 10 -5ft lb sec.85x 10 slug-ft2 owheel 157 rad/sec to correct for thruster misalignment for 1" moment arm sat 14.13 ft lb sec 2 J - = =.09 slug-ft wheel 157 rad/sec wheel The momentum wheel chosen is a 12 in outside diameter 0. 0647 slug-ft2 inertia wheel. This means it will have to be unloaded two times per AV co rrection. Yaw Wheel Going away from the sun the momentum required is 2.45 ft lb sec (to overcome solar torque) see B.4 -6 8. 08 x 10 ft lb sec (to overcome Linus inertia) 2.45 ft lb sec for a wheel that saturates at 157 rad/sec sat wheel 136

2.45 ft lb sec.0137 slug2-ft 157 rad/sec Since the momentum to overcome Linus' inertia is small the same inertia appliers going towards the sun. A 6.5 in diameter wheel with 0. 009 slug-ft is chosen. The wheel will be unloaded a maximum of two times per orbit. Since the roll control torques are approximately the same as pitch control torques, it's wheel will be the same size as the pitch wheel even though there are no torques other than perturbations and rotating feed in the halo orbit. This wheel will be used to take care of any coupling that may exist. The wheel will be unloaded a maximum of two times per 6rbit. B. 4 SOLAR TORQUES The only appreciable solar torque will be about the yaw axis. The moment arm for this torque varies sinusoidally as the moon rotates about the sun as: 1.25 ft sin (cm t ao) where m = 2.68 x 10 6 rad/sec. The solar force is:.94 x 10-7 lb/ft x 60.3 ft = 56.7 x 10-7 lbs. Therefore the solar torque is: 7. 09 x 10-6 sin (wm + ao) ft-lbs. ao is zero if initial conditions are such that the sun, earth, moon, and Linus are on the same line. Since Linus' period is 1/2 of the moon's integrate the solar torques for one of Linus' orbits. 126.8 x 104 Tr Hd - / 7.09 x 10-6 sin cmt dt = (l/cwm) 70.9 x 10-6 cos x 0 - -4.9 ft lbs sec However, only half of this momentum is opposing, the other half turns Linus in the desired direction. Orbit Control During z-axis control two.12 lb thrusters will fire to obtain a AV of 8 ft/sec every seven days. Therefore: 682 lbs Impulse'- Ft - mV = 32.2 ft/sec X 8 ft/sec - 169.6 lb sec ft lb sec ft lb sec 1" moment arm Hd = (169.6)(1/12) ft lb s13ec t l AV correction AV correction Pulse-tuning of AV thrusters may reduce the size of the momentum wheels. 137

B. 5 MOMENTS OF INERTIA Panels and Antenna Deployed 2 = 980 lbs ft 1/2 Mr body Ix X 1/2 Mr antenna I = 378 lbs ft2 1/12 Mc solar panel I = 300 lb-ft X X Hi (/2Ml =I It -17 lbs -ft2 2 X X 1 x Parallel Axis Theorem I -I~,~MI I xx x'x' x 2 4' 9. 85'I = 980 + 378 ~ (30 0 (9. 85)) + (17 + 8 (4A) I = 9573 ft -lbs = 296 slug-ft 138

I yy Body Iy,= (1/1Z) (640) (3 (3.06) + 6. Z25) = 821 lbs-ft Antenna Iy, = (1/12) (21) (3 (36) + 0) = 189 lbs-ft Solar Panel I,, = (1/12) (80) (6.7 + 4.5 ) - 434 lbs-ft Struct I, = I = 171 b-ft yy xx 1. 25 4' ~ ~9.859 -I = 821 + (189 + 2 (1.56)) + (434 + 80 ~/9.85 + 1.56) I = 9436 lb-s-ft = 294 slug-ft2 YY 139

lzz 2 I s/p = (1/12) (80) (4.5) = 134 ft -lbs I,,body = (1/12) (640) (3 (3. 06) + 6. 25) = 821 lbs-ft Z I, antenna = 189 lbs-ft I struct 0 Z Z Spin Moment of Inertia Before Yo-Yo Despin M =21 lbs antenna M = 80 lbs Antenna MA thin M = 749 lbs walled circul t cylinder 6. ' Z. 5' I =1 +I + I~ x IMT MA Ms /p / Solid circular cylinder I = 35.63 slug-ft +. 367 slug-ft + 4 (. 963 + 9. 04) slug-ft2 I = 46 slug-ft x 140

B. 6 Perilune Burn must be executed in 600 sec. To size the main engine, MAV 842, 77 F - A - 842.77 445/600 - 19.5# Therefore choose a 20# thrust engine Burn No. i A~iV fps At rmin. |Fuel Weight Burned Burn No. AV IRS At mi_. (Ibs) 1 50.0 1.09 17 2 5.0 0. 10 1 3 445.0 9.70 56 4 212.0 4.31 26 5 5.0 0. 09 1 6 688.0 13.53 77 X and Z axis;.:-contel, wh.eet unloading. and cushion 142 Total Fuel 320 Data for a 860 lb after 10 lb yo-yo's have been jetesoned 1. 25 Thrust rmax 3.5 Torque - thrust x (L)B_ 1/2 = 20.0 (. 0109) -.218 ft # Two,. 12# thrust jets, located 1.75 ft from c. g. exert a torque of.42 ft#, therefore 13 51 minutes max L =.021 in max 141

APPENDIX C COMMUNICATIONS C. 1 PARAMETERS PT = transmitted power GT - transmitting antenna gain R receiving antenna gain LT - transmitting circuit losses LR receiving circuit losses SL = space loss Lp - polarization loss R = range f =frequency S = signal power available F = receiver noise figure T = antenna noise temp a TL - line and hardware temp T = receiver temp T T T 3140K (Appendix E) o L r T = effectivenoise temp N0 = noise power density S/N = signal-to-noise density ratio C.2 EQUATIONS 8 X c/f, c = 3 x 10 /sec 2 2 G= d C.1 where ~ = antenna efficiency (usually 55%o) d = diameter 142

2 2 SL =(4) RC.3 Te = T/LR + TL (1 - 1/LR) + (F - 1) Tr T/LR + To (F - 1/LR) C.4 N = kT o e -23 where k = 1. 38 x 10 (Boltzmann) 14 Z T = (2 x 10 )/f (0.5/0)2 C. 5 s where T = antenna noise contributed by a direct view of the sun (the sun s subtends 0.5 ) aK K K K- C.6 K + K2 + B where K S /N for a compound link 0 a improvement factor ' 1 K = S/N in first leg K2 S/N in second leg B tran s ponder bandwidth C. 3 ANTENNA GAIN AND SPACE LOSS In Figure 4. 1, d actually varies from 356, 000 to 407, 000 km. Since d = 0. 168 dl, dZ varies from 59, 800 to 68, 400 km. The geometries for the two extremes, as well as the resulting subtended angles are shown below. From the geometries giren below,' the effects at the edge of the moon ( 0. 10) can be surrmarized: Down-Link Range Comparison max min mean R (kn) 69, 900 58, 600 64, 500 kdb -4 -5.6 -4.6 Gk 32.2 30.6 31.6 SL -196.6 -195.0 -195. 9 rnet loss - 164.4 -164.4 -164. 3 143

Maximum Range x!: 1400 kmn M 68,400 X E ~~~~~~~~~~18. 950 M ~~~~~'T~ 68, 400 \ 3.28~ L2 69,90 0 \ 4000 69,900 1.43~ Minimum Range 1.70~ 58, 600..~~~~~~~~~; 58,65 - 16 58,400 144

C.4 EFFECTS OF PREAMPLIFICATION FOR A DIRECT VIEW OF THE SUN The following values hold: T = 3140 K o T = 17000K s L 3 db = 2 R r 11 db -12.6 F ( 2 db 1.59 F = 11 db F- 2 db normal +sun normal + sun T 112~ 18120 1120 18120 a (F -1/LR) 12.0 12.1 1.09 1.09 T /L 560 9060 560 9060 a r T (F - 1/LR) 38000 38000 3420 3420 T 38560 47060 3980 12480 A = 0. 8 db A = 5. 0 db The antenna temperature of 11 2 K is an estimate based on the weighted average of solid angles: 40[(2. 65)2 - (1.63)+ 2900 (1.63)2 11 0K (2. 65)2 145

C. 5 TRANS PONDER BREAKDOWN failure rate weight volume power unit # (10-6 ) (lb) (in3) (W) diplexer 3. 2 1.5 18 quadriplexer 1. 2. 8 10 omni coupler 1. 1. 2 3 five channel multiplexer 1. 2. 8 10 - 5 W solid state amplifiers 5 2. 0 36 72. 0 limiters 5.5 1.25 10 coherent translators 5 1. 25 10 - down-link receiver 2. 2 4, 40 10. O up-link receiver 2. 2 3. 8 35 8. 0 command receiver 1.05 3.0 18 2. 0 command demodulator 2.1 1.0 8 0. 5 55 W TWT 2.05 15 6 432 167. 0 range code demodulator 1.1. 25 4 0. 5 carrier recovery loop 1. 1. 25 3 0. 5 phase modulator 1. 05. 25 3 0. 25 sub modulator 1.05. 25 2 0. 25 cables, switches, sensors, etc..05 5.0 15 PCM decoder/storage 1. 2 2.0 28 1.0 command distributor. 1 2. 0 32 1. 5 commutators 5.1 1. 25 220 1. 25 PCM encoder 1. 2 1. 0 24 1. O0 subcarrier oscillator 1. 1. 25 4 0. 5 clock 1.05. 25 6 0. 25 packaging - -- 9.0 52 totals 2. 9 57. 0 1023 267.0 C. 6 TRANSPONDER RELIABILITY CONSIDERATIONS - In the analysis of the transponder, we assumed system failure if any one of the cascaded components failed. The exponential relation of reliability and time of operation is -Xit -X t - t R total (e ) (e ). ) where Xl, 2X,... Xn are the failure rates of the cascaded components. For analysis, assume that the time of operation of each component is one year (876 0 hrs). This is a reasonable approximation since there will be long periods in the two year flight in which the transponder will not be active, Writing the above as 146

-(X1+X +... + X )t Rtota = e a good approximation for values of R -_.90 is Rtota = (1 XTt) Using XT from C. 5 and t from above, R = 1 - (2. 9 x 10 ) (8.76 x 10 ) total or R tota = 97. 5% 147

APPENDIX D POWER D. 1 SOLAR PANEL AREA The number of cells in series is 35 volts 385 volts/ cl = 93 cells in series for voltage requirement.380 volts/cell To determine the number of cells in parallel 41 0 watts 35 11.7 amps for current requirement 35 volts Considering current transmission losses II. 7 amps = 13. 8 amps required 85 13.8 amps 115 amps/cell = 120 cells in parallel for current requirement. 115 amps/cell This is divided into two panels allow 60 cells/panel in parallel. 93 cells x 2 cm/cell = 186 cm = 6.11 ft 60 cells x 2 cm/cell = 120 cm = 3.94 ft The total area per panel is 56. 11aftcxn 3.94 ft + 1.95 ft (hinge area) =' 30. 15 ft2. 85 (packing factor) D. 2 Ag-Zn BATTERY D.2Z,. 1 Battery Capacity Determination Requirement: Supply 297 watts at 30 volts for 5.4 hours. (Assume 300 w for 6 hrs) 300 w 10 amp 30v 148

(10 amp) (6 hrs) =- 60 amp-hrs Allowance for 25% deterioration 15 over 2 years 75 amp-hrs Assume 7 0% maximum desired depth of discharge: 5 amphrs 107 amp-hrs Battery Requirement.70 D. 2. 2 Emergency Battery Capacity Assume battery charged to 125 amp-hrs. 342 w - 11.4 amp 30 v Assuming 75% depth of discharge: (125 amp hrs) (. 75) = 94 amp-hrs 94 am 4ahrs- 8. 24 hrs (Full power capability on batteries alone) 21.4 amp D. 3 POWER SYSTEM WEIGHT SUMMARY Solar Panels (2) at 40 lb 80 lb Battery 66. 3 10% increase for upright mounting 6. 7 73 73; lb Power control unit DC-AC power module 8. 3 DC-DC converter (2) at 1. 1 2,2 Mis cellaneous 1. 5 12 lb Total power system weight 165 lb 149

APPENDIX E THERMAL CONTROL Symbols q = heat from solar radiation q =heat from planet emitted radiation 0 qr = heat from planet reflected radiation = 0 qi heat from internal equipment int q heat from conduction q heat radiated iR= shape factor AR - radiating area E = emittance coefficient.-Io 2 oF4 0- Stefen-Boltzman constant - 5 x 100 w/ft F T = Temperature A =projected area facing- the sun a = absorptance coefficient 2 S S solar constant = 125.7 watts/.ft + 3 w/ft at 1 A.U. Cth = thermal capacity of segment R = resistivity to heat flow L - average length heat flows k - conductance of material A - cross -sectional area ()0 = radiating area 1 ()2 radiating area 2 q heat from heaters - batteries ()o communications package Cp = specific heat of material m = mass t = time The thermal control systems to be used and the radiating area required is found from the following instantaneous heat balance equation. dT qs qe qr int qc qrad th dt E.1 This could be solved by use of a computer for all areas of the satellite, but most of the time the temperature varies slowly and therefore the average temperature is of interest. The time-average heat-balance equation below can therefore be used with fairly good accuracy. qs qint qrad qs zA a S E.3 s p S 150

qR- FA (T4 E. 4 F - 0.99 1 (see Reference 3 for calculations) Although the above equations are good for giving the average temperature, the following equation for cosnduction must be used to be sure that the structure will conduct the heat away from any heat sources. qc -R E.5 c R R - E.6 kA The above equations have to be iterated several times until reasonable radiating areas and reasonable temperatures are found. This iteration was carried out with the following process, A radiating surface area was assumed and then a value of R and AT is found from Equations E. 6 and E. 5. From AT and the operating temperature of the component a temperature of the radiating surface is found. R T - AT R C component This temperature, TR, is then used in Equation (E. 4) to find a new radiating area, A' R A= R FE T r Below are some calculations for finding the solar input and the temperatures at some important points. Solar Input Radiating area 1 (see Figure 6. 2) q - A ca S S1 Pl s When the sun-satellite line is coincident with the y-axis A - 1.28 ft2 pI q - (1.28)(0.20)(125.7)ft w/ft = 32 w s 1 151

When the sun-satellite line is coincident with the x-axis. A =0 P1 q 0 s1 Radiating area 2 (see Figure 6. 1) When the sun-satellite line is coincident with the positive y-axis A O. 44 ft2 P2 q = (0.44) (0. 20) (125 7)ft w/ft = 11 w qs2 When the sun-satellite line is coincident with the negative y-axis or the x-axis A =0 P2 q =0 S2 Temperature of Radiating Areas The following calculations are carried out when the temperatures are the highest. From Equations E. 2 and E. 3 4 qrad T E.. 7 FA E T R Temperature of radiating area 1 when the sun-satellite line is coincident with the y-axis and Linus is communicating. 2 2 4 (240 w) (144 in /ft2) 1 T-10 24 (30 in) (33 in) (0.9) (5 x 10 w/ft R T - 527 R- 67 oF Temperature of radiating area 2 when the sun-satellite line is coincifdent with the positive y.axis and the batteries are discharging. 152

4 (42 w) (144 in2/ft2) (9 in) (20 in) (0.9) (5 x 100 w /ft OR4 T =522 OR = 620F Temperature of Components The temperature of the components when the sun-satellite line is coincident with the positive y-axis and Linus is communicating. Batteries when discharging From Equations E. 6 and E. 5 L2 _ (2. 24 in) (12 in/ft) -2 kA2 (34. 5 w/ft OF) (20 in) (0. 05 in) R2 0.783 PF/w AT - Tb - T2 = qbR2 = (30 w) (0.783 OF/w) = 23.4 ~F T = 85.4 F b Communications package L _1 (3.72 in),(12 in/ft) R1 kA 1 (34.5 w/ft OF) (30 in) (0. 05 in) R = 0.862 PF/w 1 TT = T _ q180 o o T2 1 qR ) (0. 862) = 39F T = 106 F 0 Heaters Required The following calculations are carried out when the temperatures would be the lowest without heaters. The heat required when the sun-satellite line is coincident with the x-axis and Linus is not communicating. Equations E. 4 and E. 5 are solved by an iteration process to find qh' T2, and Tb. To keep Tb greater than 60~F first assume 153

b then 4 (eo w) (144 in- /ft 1 T210 2 o 4' 2 (9 in) (20 in) (0. 9) (5 x 1 w/ft R 4 8o4 T 4 534 10 R T 481 oR = 21 oF -T Tb - T2 qhb R2 therefore, (60 0F - Zl 0F) hb (0. 783 ~F/w) Continued iteration shows that for qh 35 w b that T2 = 40 F and T = 67 F b It was found that it is desirable to keep T2 above 32 F in order to keep Linus' components at a reasonable temperature level, therefore it was found from Equation E. 4 that the heat required to do this is 180 w. q ired (492 F) (30 in) (33 in) (0.90) ( 5 x 10w/ft R required required 180w Since the heat dissipated by the internal components at all times is 60 w, the heat required from heaters is 120 w. These heaters are placed at the communicatiosns package. Equations E. 4 and E. 5 are then used to find the temperatures of each component (see Table 6. 1 for values). 154

.Thermal Time Constants The above calculations and the values in Table 6. 1 give only the maximum and minimum values of the temperatures. To find the length of time it takes for a component to change its temperature,the time constant must be known. It is found from the following equation: dT q=C mp dt This is approximated as q=C mft p At or At C m Component Time Constant Entire Satellite 4 0F/hr Communications o Package 43 F/hr Batteries 6 F /hr 155

APPENDIX F STRUCTURES The following symbols are used in the structural calculations: A cross sectional area d = distance from the bottom of the support tube to the spacecraft center of cg gravity E -modulus of elasticity F - axial load G - maximum acceleration experienced by the satellite in g' s I - area moment of inertia L = member length M = bending moment m = mass per unit length m - mass of spacecraft r = radius ra = average radius t - member wall thickness S safety factor - poissons ratio W - satellite weight wn - natural frequency =A applied stress -b - bending stress cr= = critical buckling stress Ty - yield stress a axial stress F. 1 SUPPORT TUBE In analyzing the support tube we must compare the magnitudes of two terms. The largest of the two terms determines which bucking formula will be used (Reference 7. 5-5). w r 2 r a a 1/Z 2~BPI 2 r a ) If the first is greater L - 30 'n L 30t For the support tube, r 9in33 a - 1/ (9) 77)] 56.05 156

Hence DonnellV s formula is used:.6 t/r - 10 r /t _a a -E...................... F. 1 cr 1 +.004 E/r y where 6 E = 10. 6 x 10 psi - 72,9 000 psi y o6 o0S /9 -o0 9/ 05 - 10. 6 x 106 = 22 000 psi o00r 1+ 4 10. 6 x 10 1~0+4 72, 000 The applied stress is due to both axial and lateral accelerations. Mr WG a eA -a +b A + F. 2 where A= 2. 83 in G = 30 W = 872 lbs M =Wd - 872 (21) - 18, 300 I = 115 in 872 30) I18, 300 (9) A 2.83+ 115 9550 + 1480 = 11,030 psi S cr 22, 000 S -=.99 cA 9,11 030 F. 2 BASE PLATE To analyze the base plate we must approximate the anrnulus -shaped plate by a square shaped plate with sides a and b. We then proceed as fsllows (Reference 7.5:3): 157

M 1 qa F. 3 where p- 0479 depends on ratio b/a a = b 36 in Load _ 544 30g14 1 psi q Are a - 11460ps M ( 0 479) 14. 05 (1 296) 873 in -lb in M A t t whe re t honeycomb core thickness - 1.5" tf honeycomb facing thickness. 02" 873,A (1.5) (. 02) 29, 100 psi 67S 2000 31 TA 29, 100 Fo 3 SOLAR PANEL SUPPORT ARMS F P 270 lbs (assuming 5g accelerations).~ 3L2 / F2 sin $ =F P 2 _ 270 = 510 lbs 2,/~~~ F2- sinc -.529 PF1 = F ctn = - 270 (1. 6) = 432 lbs 158

Pinned joints imply only axial forces exist in members. F 1 432 A = - — 917 psi tension F2 510 -A2 =2 1624 psi compression AZ A. 314 Using Euler~ s buckling formula 2 rE IE =C~~~~~~~ - ~~F.5 cr. G L 2 2 2 whe re 6 E 10 o6 x 10 psi L2 = 56 in 2 A2 = 314 in 4 I2.0392 in 2 6 %r,2 iT (l06 =... )(' 0392) cr.2 2 (56) (.314) 3960 psi Horizontal support is in tension hence buckling formula is not applicable. It can fail in yield stress only. S - Y 7- - 78. 5 not critical 1 Al 917 cro 2 3960 2 A 1624 244 Vibrations set up during the use of the midcourse correction engine are very small and donu t affect the supporting arms. During launch, however, ti is necessary to constrain the supports to the body in order to keep the natural frequency a ove 40 cps. 159

F. 4 STRUTS The stress in this cosmponent is far from the critical buckling stress as calculated from Euler' s formula; hence, the stress analysis of it will not be presented here. However the vibrations which are set up in this strut are critical and will be analyzed. We will assume pinned supports for a conservative answer. The differential equation governing a vibrating rod under a compressive load is EIwIv + PW + mw - 0 F. 6 where P = axial load let w(x, t) - X(x)T(t) which results in two equations: 2 T + w T = 0 (time) F.7 iv 2 EIXiv + PX - mw X = 0 (space) F. 8 Solving the above equations we get for the first natural frequency 2 p 2 2 r w El wI = L- 2 F. 9 where for the struts tE -10.6 x 106 psi m.024 lb /n I.0165 in P 750 lbs 2 T | 7r (210. 6 s 6) (.0165) El.024 (81) 81 107 500 rad /sec Wo = 328 rad/sec = 52. 3 cycles/sec This is in accordance with launch specifications. 160

ACKNOWLEDGMENTS We would like to thank those who have given up their time to us for detailed critique and recommendations. Prof. T. C. Adamson The University of Michigan Prof. J. V. Breakwell Stanford University Prof. Harm Buning The University of Michigan Mr. James Burns Bendix Aerospace Systems Division Mr. Darryl Criswell Bendix Aerospace Systems Division Mr. Donald Crosby The University of Michigan Dr. T. N. Edlebaum Scientific Resources Corp. Mr. Richard Gibson Bendix Aerospace Systems Division Mr. Robert Groves Goddard Spaceflight Center Prof. G. I. Haddad The University of Michigan Mr. S. J. Harris Mc Donnell- Douglas Mr. Daniel Horten Wright Patterson Air Force Base Mr. A. Mullenkopf Bendix Aerospace Systems Division Mr. C. Murtaugh Bendix Aerospace Systems Division Prof. Wilbur C. Nelson The University of Michigan Prof. J. Ao NiLchlls The University of Michigan Mr. Frank Nich olson General Electric

Mr. Rob er Peltzer The University of Michigan Prof. R. L. Phillips The University of Michigan Dr. Pltzer Goddard Spaceflight Center Prof. WO F. Powers The University of Michigan Mr. Presiall Bendix Aerospace Systems Division Prof. L. L. Rauch The University of Michigan Mr. Andrew Rolinski Goddard Spaceflight Center Mr. Paul Eo Schmid Goddard Spaceflight Center Mr. Robert Schultz Bendix Aerospace Systems Division Mr. R. J. Simms Bendix Aerospace Systems Division Dr. J. Shea Ratheoon Mr. Charles Shokkos Bendix Aerospace Systems Division We would like to give special thanks to Prof. Wilbur C. Nelson for his guidance throughout Project Linus and also to Caroline Rehberg for typing this report. In addition we would like to thank Mr. W. James Lane for his assistance,

PROJECT MANAGER G ERALD M. SIVULKA TRAJECTORY 1 LATTITUDE & CONTRO COMMUNICATIONS POWER BRUCE GERDING SCOTT DAVIDSON BOB CLEMENT SAM NAKAMURA DAVID NOSHAY MILAN GROSS CAREY GROSS SHELDON RUNDELL GREG OPDYCKE BILL WOODS MIKE KERN TRAJECTORY CONSULTANT THERMAL STRUCTURES LAUNCH GLEN ALLVORD TOM HEPPINHEIMER JIM SHAW BRUCE KIMBALL JOE ZUKOW