• • SSC98-X-3 Lessons Learned from the • Miniature Sensor Technology Integration (MSTI) -3 Controlled Reentry • Lesley M. Rahman AlliedSignal, Inc. • 7515 Mission Drive Lanham, MD 20706 • 301-805-3638 • [email protected] Preston S. Diamond • ANSER 1215 Jefferson Davis Hwy, Ste 800 • Arlington, VA 22202 703-416-3434 • [email protected] • Todd C. Probert AlliedSignal, Inc. • One Bendix Road Columbia, MD 21045 • 410-964-7249 • [email protected] Abstract. This paper is presented as an overview of the lessons learned from the controlled • reentry of the Air Force Miniature Sensor Technology Integration program's third (MSTI-3). Since the launch of Sputnik in 1957, the amount of space debris in orbit has • progressively increased to potentially hazardous levels. In light of these facts, the National Space Policy directs the United States to minimize the creation of space debris. NASA has • already adopted a policy to limit the generation of orbital debris, and DoD policy has supported debris minimization for well over a decade. The MSTI-3 was neither designed, nor • intended, to perform a controlled reentry following the completion of its primary mission objectives. In spite of this fact, after receiving direction to prepare for reentry on 10 November • 1997, the MSTI-3 spacecraft was successfully de-orbited and all debris concentrated into a 100 x 10 km square in the Pacific Ocean on 11 December 1997. This report outlines the procedures • and processes developed by the various participants during the reentry of the MSTI-3 spacecraft, discusses obstacles which were encountered, and provides a framework for future satellite • disposal events. • Introduction studying the evolution, mitigation, and characterization of space debris. Basic debris • Since the launch of the first Sputnik in 1957, model projections call for the demand of the number of space debris in orbit has immediate debris mitigation measures such as • significantly increased. Early on, scientists explosion suppression and de-orbit of rocket • became aware of this phenomena and began bodies and payloads after mission completion. • Lesley M. Rahman 12th AIAAfUSU Conference on Small • • • In light of these facts, the National Space the cause of loss of contact to that of a failure Policy directs the United States to minimize in either the transponder or power distribution • the creation of space debris. I NASA has unit. On 15 September 1997, MSTI-2 had a already adopted a policy to limit the well publicized near miss with the Russian • generation of orbital debris, and DoD policy space station . This event influenced the has supported debris minimization for well decision to de-orbit MSTI-3. • over a decade. Congressional direction at the end of FY94 • The decision to execute a controlled de-orbit transferred program management of the MSTI of the MSTI-3 spacecraft following mission program from BMDO to the Air Force Space • completion was the first implementation of and Missile Center, SMCIMTAX. MSTI-3's this policy. As such, no guidelines or primary year-long mission was to collect • procedures were available which addressed short-wave infrared (SWIR) and medium­ how to deliberately reenter payloads after wave infrared (MWIR) background clutter • mission completion. This report outlines the data and scenes measuring temporal, procedures and processes developed by the nocturnal, and seasonal variations. • various participants during the reentry of the MSTI-3 spacecraft, discuss obstacles which Due to a combination of changes in • were encountered, and provide a framework management, funding, and launch vehicle for future satellite disposal events. problems, the satellite development cycle • extended longer than originally anticipated. MSTI Program MSTI-3 was finally launched on 17 May 1996 • In December of 1991, the Ballistic Missile GMT from a Pegasus launch vehicle staged Defense Organization (BMDO) began the out of Vandenberg AFB. After the initial on­ • MSTI program with the objective of orbit checkout, the spacecraft was operated out developing a means of providing rapid access of the MSTI Operations Center (BA TCA VE) • to space to test new sensor technologies. located in Alexandria, VA. MSTI-l was launched less than a year later on • a Scout launch vehicle from Vandenberg MSTI-3 Specifications AFB. The primary objective of MSTI -1' s • week-long mission was to validate the basic The MSTI-3 reflected only spacecraft bus design and the "faster, better, minimal changes from that of the first satellite • cheaper" program philosophy. These, as well in the series. The basic spacecraft design as other mission objectives, were successfully included a fixed, three-faceted GaAs solar • accomplished. array attached to an aluminum structure built around the propulsion system. Spacecraft • MSTI-2 was launched on the last Scout launch electronics and subsystems were either housed vehicle on 05 May 1994. Observations in one of two electronic bays or affixed to the • included fixed ground firings, air-based outside of the aluminum structure. The targets, and actual boosters. MSTI-2 also MSTI-3 spacecraft, benefiting from lessons • supported civilian remote sensing through learned during the two prior missions, dual-use demonstrations. MSTI-2 operations , included improvements in the attitude control, • ended prematurely on 05 September 1994 . power, and data handling subsystems (see when contact with the spacecraft was lost. Table 1 and Figure 1 on the next page). • Further analysis of vehicle behavior narrowed

Lesley M. Rahman 2 12th AIAAlUSU Conference on Small Satellites • • • • • SPACECRAFT PARAMETER CHARACTERISTIC • DESIGN LIFE I Year 85 % Mission Effectiveness (EOL) SPACECRAFT DIMENSIONS 32" Diameter, 56" Height • SPACECRAFT MASS 466 LB (includes P/L and Prop) PAYLOAD MASS 115 LB PROPELLANT MASS 47LB • PROPULSION Hydrazine (for orbit adjust and reaction wheel desaturation) ATTITUDE DETERMINATION AND CONTROL 3 - Axis Reaction Wheels GPS/Star Tracker • POWER 225 Watt Solar Array (EOL) 10 A-Hr NiH Battery Cruise/Observation - 140 W1320 W • STRUCTUREITHERMAL Payload thermally isolated from bus COMMAND AND DATA HANDLING 1750 A - Spacecraft • RISC 3081 - Payload DATA STORAGE (EDMM) 8.64 Gbits @ 25 Mbits/s TT&C SGLS - 2 kbitls Commanding (with PRN range capability) • - 32 kbitls SOH - I Mbitls Mission Data • Table 1. MSTI-3 System SpeedicatioDs • wheels to maintain its nominal pointing attitude. The payload consisted of three • cameras: a short wave infrared (SWIR) imager, a mid-wave infrared (MWIR) imager, • Star Tracker and a visible imaging spectrometer. In addition, the MSTI-3 satellite had a star • tracker which was rigidly attached to its payload supports, replacing the horizon sensor • flown on previous missions. A GPS system was also added to the MSTI-3 satellite to • enable enhanced position and velocity • determination. Eclipse Operations • A-Bay (Reaction Wheels, IRUs) Each October, MSTI-3's orbit subjected the Figure 1. MSTI-3 Line Drawing • spacecraft to increasing periods of time in the The MSTI series of satellites were among the Earth's shadow (see Figure 2). Although the • first of the small satellites to provide precise spacecraft was designed to operate through its attitude control and line-of-sight stabilization eclipse season, two spacecraft anomalies • required for optical background clutter impeded payload operations. characterization and remote sensing. Both the • bus and the payload were designed to provide • pointing control, low jitter, and a high field­ of-regard. MSTI-3, similar to the previous two MSTI satellites, was a three-axis • Figure 2. 1997 - 1998 Eclipse Times stabilized spacecraft which used reaction

• Lesley M. Rahman 3 12th AIAAlUSU Conference on Small Satellites • • • • The star tracker's generation of invalid to be 10 Amp Hours. At this capacity, it was attitude solutions was the first spacecraft possible to execute a minimum of two payload • anomaly experienced during the first eclipse operations a day--even during the height of season. During the 1996 eclipse season, the the eclipse season. Once on orbit, however, • operations team had learned that when the star the usable battery capacity was observed to be tracker's background clutter exceeds a certain less than 3 Amp Hours. At 3 Amp Hours, • threshold, the system inadvertently uses the payload operations could only be executed clutter to generate star field matches. during direct lighting conditions and were • Although the star tracker hardware sets a eventually terminated until the end of the "High Background" flag when the background eclipse season. • clutter exceeds a certain threshold, documentation implied that the hardware The figure on the next page illustrates an • would recognize the background clutter as bad actual orbit during the longest eclipse period data and not pass it forward for attitude (21 December 1996). Towards the end of the • generation. During actual operations, eclipse, the battery voltage comes dangerously however, the hardware did in fact pass the bad close to the first level of the spacecraft's • data forward. Due to this error, an invalid undervoltage (UV) protection. MSTI-3 UV attitude solution was generated since the protection provides three levels of • spacecraft software was programmed to ignore safeguarding: UV -1 switches off the payload; the "High Background" flag and process the UV -2 switches off the storage device, thermal bad data. control devices, and transmitter; and UV-3 • switches off the computer, attitude • To prevent an invalid attitude solution, any determination, and control devices. These source of background clutter (including the levels are tripped in sequence in an effort to • Moon and the Earth's limb) was excluded conserve power as battery voltage gets low. from a 30 0 arc extending from the star tracker's boresight. During the summer, In spite of the power saving measures adopted • precautions only had to be taken when there by the operations team, the vehicle • was a full Moon. During the fall and winter, experienced one UV -3 trip and a number of however, a 100 roll and -40 0 pitch orientation UV -1 and UV -2 trips during the 1996 eclipse • in the southern hemisphere (due to season. The UV -3 trip occurred due to the interference by the Earth's limb) was loss of a sun facing attitude just prior to necessary to secure a valid attitude solution. entering an eclipse period. The UV -1 trips • Unfortunately, this same orientation placed the regularly occurred during the deepest portion • payload in a position where the IR cryo-cooler of the eclipse season when even extreme fell below its minimum operating temperature power conservation modes were barely able to 0 C. • of -40 As a result, the IR operations were keep the power in balance. Any additional halted until an acceptable thermal/star tracker power draw (such as thermostatically­ attitude could be realized after the eclipse controlled heater tum-on or poor placement of • season. a contact) pulled the battery voltage below the • UV -1 trip point. The second spacecraft anomaly had a more • significant operational impact. Before launch, the spacecraft's battery capacity was purported •

Lesley M. Rahman 4 12th AIAA/uSU Conference on Small Satellites • • • • • • 21 December 1996 ~~------~ 12 • 33~~4r------~~==~~ 10 32r---1r------~~~~~~~~~----_t~ 8 • 31~~~~------~~~~_r~~~~~~~~~ 6 & ~~~=====+~~----~~~J~~[I:~!~~~·.;~.~'-----'~'~~~ !L -Essential Bus l 30 +____ 4 Voltage • ~ --Battery Voltage ~ 29t---~~~~--_i~------~~ 2 ---"Solar Array Input • 28r----T----~~~------~~--~------~~~ o

• (D (D (D (D (D (D (D (D (D L() (Y') L() (Y') L() (Y') b .,- N N &'i ti)

• Lesley M. Rahman 7 12th AIAA/USU Conference on Small Satellites • • •

MSTI vs. STS Conjunction Analysis 1st Burn operators suspended all non-essential • Conjunctions computed for 3 days operations and began to tum off non-critical 80 subsystems. All S WIR and MWIR operations • ~ 80 70 ~ were halted by 29 October 1997. The GPS C receiver was powered off on 03 November • § 60 n

ww 1997 and all visible camera operations ended ~ ~~w wi-· ~-~*~-1----- ..0' • on 22 November 1997. Even with these 50 ,.=- • ~. ~. ~ .:- - - ~ - - _. -i --- -- ... power saving measures, MSTI-3 occasionally 40 0 t: • :::r experienced a UV trip. On 25 November 30 1 1997, a negative acquisition occurred due a UV -2 trip. The UV -2 trip automatically • °O~~~~~~~~~~~~~~~~~1~~ powered off attitude determination and control TIme From Beginning of Bum {sec) • devices which in tum caused the spacecraft to lose attitude. The satellite was successfully MSTI VS. MIR Conjunction Analysis2nd Burn reacquired on 26 November 1997 and • Conjunctions computed tor 2 days or until reentry subsequently regained attitude on 29 45 • :t November 1997. Unfortunately, the UV-2 trip :I 40 .. resulted in the delay of the first bum until 02 80 i !!! • December 1997. At this point, MSTI-3 was in 35 n ..0' a 410.2 x 435.2 km orbit with an inclination of 30 I!I. ,. 97.1 0 and a propellant load of25.1 lb. • ... 25 0 t: :::r ~ First Burn • ~ J: .!. o~~--~--~--~--~--~~. o ~ ~ ~ The first bum occurred at 04:29:56 on 02 • TIme From Beginning of Bum (sec) December 1997 GMT over the Hawaii tracking station on the descending node of • Pre-burn Conditions orbit 8721. The bum lasted 1356 seconds, used 12.2 lb. of fuel, and placed MSTI-3 into a • On 25 October 1997, MSTI-3 entered full 224.1 x 420.4 km orbit. The figure on the eclipse for the first time since February 1997. next page shows the ground trace of the first • The duration of umbra (the period of time bum. Hawaii tracking station was in contact when the spacecraft is in total darkness) was with the vehicle for 5.9 minutes prior to • 73 seconds. As the season continued, umbra thruster ignition and remained in contact for duration quickly increased and was projected 2.7 minutes after thruster ignition. Actual • to reach a maximum duration of spacecraft perigee differed slightly from the approximately 24.5 minutes on 21 December predicted target perigee due to off-pulsing • 1997 . To conserve power during eclipse, effects and actual versus modeled thruster spacecraft operators halted all contacts at the • performance (6.5% error). Thule Tracking Station (TIS) on 25 October 1997 and the Oakhanger Telemetry and The first post-bum attempt to contact MSTI-3 • Commanding Station (TCS) on 26 November occurred one orbit later and failed. Even 1997. These stations are located in the • though a signal from the spacecraft was northern hemisphere where the effects of the received by the ground station, the station was eclipse were strongest. In addition, spacecraft •

Lesley M. Rahman 8 12th AIAAIUSU Conference on Small Satellites • • • • • Deorbit Burn 1 • Burn start on 12/02/97 at 04:29 :56 GMT • 90N 90N • 60N ~ 60N • 30N 30N CD 'tJ r • ::::I ON ON I» ...... cr. :.i:I C ~ g, ..J • 308 11) • 308 • 608 608 SJ()S I I I t I I tIl I I I J I I I tit I I 90S • 180W 160W 120W 90W 60W 30W OE 30E 60E 90E 120E 160E 180E Longitude • Figure 4. Ground trace of first burn unable to lock on to it. This brief contact, personnel commanded MSTI-3 to tum on its • however, indicated a relatively healthy Space Ground Link System (SGLS) spacecraft and a nominal bum situation. transmitter and contact with the spacecraft was • Initially, analysts concluded that the station reestablished. Telemetry from the satellite was unable to lock onto MSTI-3 due to showed that a sun presence time out had • inaccuracies in the station's antenna tracking occurred at some point after the bum. A sun parameters (the tracking parameters were presence time out occurs when the solar • based on the predicted orbit). Two additional panels point away from the sun for more than attempts to regain contact were made over the 30 minutes. The result is similar to that of an • Guam tracking station at 07:29 and 08:58 on eclipse in that the battery becomes the sole 02 December 1997 GMT and both failed. At source of power. Once contact with the • this point, Cheyenne Mountain Operations vehicle was reestablished, efforts began to Center had tracked MSTI-3 on several restore MSTI-3 to nominal operations. Table • occasions and had computed more precise 2 compares the pre-bum and post-bum (both orbital parameters. Analysts agreed that the actual and predicted) parameters. • tracking stations now possessed valid antenna tracking parameters and if the station did not Parameters Pre-burn Post-burn Post-burn (Predicted) (Actual) • lock onto the satellite during the next pass Orbit (km) 41 0.2x435.2 209 perigee 224.lx420.4 opportunity, the failed attempt would be Bum Duration N/A 1356 unknown • treated as a spacecraft problem. The next (sec) Fuel Pressure 150.3 N/A 120.82 pass opportunity occurred over the Diego (psia) • Garcia tracking station at 13:35 on 02 Fuel Load (lb.) 25.1 12.23 12.9 December 1997 GMT. When the station Table 2. ComparIson of pre-burn and post-burn parameters • failed to acquire the spacecraft, BATCAVE • • Lesley M. Rahman 9 12th AIAA/USU Conference on Small Satellites • •

At 20:29 on 02 December 1997 GMT, MSTI- perturbed the delicate balance of power, which • 3 suffered another sun presence time out resulted in an increased chance of a UV trip. which resulted in a UV -3 trip. The sun • presence time out occurred over the Guam Orbit Degradation tracking station and was due to over-saturation • of the reaction control wheels. MSTI-3's Orbit degradation was a second consequence attitude control was maintained by a number of increased atmospheric drag. Table 3 (on • of reaction control wheels which spun at the next page) shows the satellite's perigee varying rates to counteract external forces. and apogee heights from 01 December to 10 • The wheels needed to be 'de-saturated' once a December 1997. Due to increased day by switching attitude control to the atmospheric drag at perigee, apogee height • reaction control jets thereby allowing the decreased 2-3 km per day after the first burn. reaction wheels to spin down. Due to the Without a second burn, analysts estimated that • increase in atmospheric drag caused by a MSTI-3 's orbit would have degraded to the lower perigee, the reaction wheels over­ point of reentry by mid to late January 1998. • saturated much faster than anticipated and Rapid orbit degradation also posed an attitude control was lost. To compensate, increased challenge to SMe/TEO personnel as • spacecraft operators increased the number of they performed orbit determination analysis de-saturations performed to four per day. All and created daily ephemeris data products. • MSTI-3 subsystems, with the exception of the star tracker, returned to nominal operations on Date Perigee height Apogee height • 03 December 1997. (1997) (km) (km) 01 DEC 410.22 435.22 • Post-burn Conditions 02 DEC 410.22 435.22 (pre-bum) • 224.10 435.22 Increased Propellant Depletion 02 DEC (post-bum) • 03 DEC 224.09 420.40 After the first burn, satellite operators 04 DEC 224.09 418.55 415.77 • experienced numerous problems associated 05 DEC 225.02 with increased atmospheric drag. As stated 06 DEC 225.02 412.99 08 DEC 225.94 407.44 • earlier, de-saturation events were necessarily increased from one to at least four daily. 09 DEC 225.94 404.66 Increased de-saturation began to erode 10 DEC 226.87 401.88 • Table 3. Orbit Degradation propellant reserves-almost to a level below the minimum propellant required for the • Spacecraft Attitude second burn maneuver. To illustrate, from the last orbit raise on 24 May 1996 until the • Between the first and second burn, all first burn on 02 December 1997, MSTI-3 used operations were focused towards reacquiring a • fuel at an average rate of 0.52 lb. per month. valid stellar solution with the star tracker. Between 02 December and 11 December, Without a valid solution, the vehicle had only • however, MSTI-3's average fuel consumption one axis control, its sun-pointing safe mode, rate increased to 2.4 lb. per month. In addition and therefore, could not be properly oriented • to reducing fuel reserves, each de-saturation in preparation for the second burn. At every available opportunity, stellar acquisition • Lesley M. Rahman 10 12th AIAAIUSU Conference on Small Satellites • • • • • commands were sent to the spacecraft. chart, appears sinusoidal because the vehicle • Unfortunately, after the first bum, stellar is pitching at 0.3 degrees/second while acqulSltlOn opportunities could only be looking for a valid star pattern. The shaded • attempted when the following three conditions areas on the graph represent times when the were all met: angle between the star tracker boresight and • . the Earth's limb is above 43° and the • Not in a period of eclipse spacecraft is not in an eclipse. In this • • Angle between star tracker boresight and example, these shaded areas were the only the Earth's limb> 43 0 times that a stellar acquisition could be • • Angle between star tracker boresight and initiated. the Moon> 12 0 • Finally, the third chart shows the angle The following discussion provides an example between the star tracker boresight and the • of the various orbital conditions which Moon. Again, in order to avoid excess glare, precluded achieving a valid stellar solution this angle had to be greater than 12°. For this • during an eight hour period on 10 December period of time, the 12° condition was 1997.+ continuously met. Note, however, the inverse • relationship between this angle and the cycles The first chart maps the change in altitude of of the Moon (shown in last chart). As the • the satellite. MSTI-3 was in a period of fullness of the Moon increases, the angle eclipse during all shaded areas on the graph. between the star tracker boresight and the • During a normal stellar acquisition, the Moon decreases. Given a few more days, this spacecraft first acquires the axis of the sun and angle would have fallen (and remained) below • then pitches about that axis to identify and the 12° condition until a few days after the full match star patterns. During a period of Moon. • eclipse, however, the sun is not visible to the spacecraft. For this reason, no stellar Stellar acquisitions were initiated only when • acquisitions could occur during an eclipse. In all three of the conditions listed previously this example, periods of eclipse limited stellar were met. During all other times, the • acquisition opportunities by approximately spacecraft was placed in sun pointing safe one-third. mode to prevent the loss of the sun facing • attitude. Unfortunately, since only one axis In addition to the limitations experienced due (the sun pointing axis) was known at the • to the eclipse season, high background initiation of each acquisition, the initial conditions also limited stellar acquisition orientation of the star tracker was uncertain • opportunities. In the southern hemisphere, the and random from one acquisition to the next. Earth's albedo tended to corrupt star position • information, precluding a successful When early commands failed to regain reacquisition. For the star tracker to function attitude, changes were made to the • properly, the angle between the star tracker reacquisition process. Analysts believed that boresight and the Earth's limb had to be at during the eclipse safe hold mode the rate • least 43°. This angle, as shown in the second sensor assembly (RSA) software controller dead band was too large to control the • :: The charts referred to in the text are included as spacecraft at the rate required to properly Attachment 1 at the end of the paper.

• Lesley M. Rahman 11 12th AIAAIUSU Conference on Small Satellites • • • acquire stars. This fact further complicated 11 December 1997 GMT during a normal • the process since stellar acquisitions usually stellar acquisition attempt. Post mission occurred immediately after an eclipse period analysis has not shown why this particular • when spacecraft rates were in transition from attempt succeeded. the RSA and the two-axis rate assembly • (TARA) controllers. On 09 December, the Second Burn spacecraft was commanded to accept rate • information from the TARA instead of the In anticipation of reacqumng spacecraft RSA. • attitude following the first bum, second bum command sequences were uploaded to In a last-ditch effort to reacquire attitude, • spacecraft memory each day. Consequently, mission directors also decided to perform a the second burn command sequence had 'lost in space' maneuver. This option was • already been loaded into spacecraft memory included within the original spacecraft design on 11 December when attitude was reacquired. but had never successfully worked when • At this point, the spacecraft was in a 226.9 x performing on-orbit testing during launch and 401.9 km orbit, had an inclination of 97.1 0, early orbit checkout. In contrast to a normal • and a fuel load of 11.71 lb. stellar acquisition, a 'lost in space' maneuver • will not attempt to acquire the sun but will try Figure 5 (on the next page) shows the to identify and match a star pattern for each predicted ground track for the second bum. reference axis. Due to the numerical intensity • The bum was planned to execute over the of the maneuver, the central processing unit Indian Ocean on the descending node of orbit (CPU) would have taken over an hour to • 8870. A nominal bum starting at 14:11:30 complete just one cycle. In contrast, the state and programmed to last 6000 seconds (until estimation processor (SEP) could perform the • propellant depletion) would place reentry maneuver in less than 15 minutes. For this debris in the eastern Pacific Ocean just north reason, the SEP, rather than the CPU, was • of the equator. Figure 5 also highlights chosen to execute the 'lost in space' potential areas of impact due to lower than maneuver. • expected levels of thruster performance (i.e., premature thruster shutdown, etc). According The maneuver was designed to be executed • to analyses performed by The Aerospace from spacecraft memory subsequent to Corporation, spacecraft debris would have entering an eclipse and just prior to aRTS • posed a hazard to a populated area only if contact. The first 'lost in space' attempt thruster performance was less than 88% of occurred on 09 December but did not succeed. • nominaL During the contact following the maneuver, the spacecraft analyst commanded a normal • While in contact with the Vandenberg tracking stellar acquisition. Although the star tracker station, at 13:37 on 11 December 1997 GMT, identified and matched stars, the analyst did • MSTI-3 was directed to commence the second not have time to lock on before the pass ended bum. The second bum occurred at 14: 11 :3 0 and was not able to sustain attitude. • on the descending node of revolution 8870. Additional 'lost in space' maneuvers were The bum was projected to use the remaining • performed later that day and during the 11.71 lb. of fuel, and to place MSTI-3 into a following two days but to no avail. Spacecraft 224.1 x 420.4 km orbit. • attitude was eventually reacquired at 09:38 on

Lesley M. Rahman 12 12th AIAAlUSU Conference on Small Satellites • • • • • Deorbit Debris Range • Burn Start 12/11197 14:11:30 GMT • • • • ~ , c.e: • .;; (1) ... .. • 30S ~ 30S ~, • ...::.. 60S

90S ... -,-,-,_\.. L.L ...., J_,_L .... J...to...... JJ_,_ ...... J..t. j ..J_1_1_,- ... 1 ... ..1 ""' .. 5_ .... l ...... J JJJ_ ...... "'.l..l J_,_\..\...... J..,. -'-'-'- .. l", ... ..I..I-,-l 90S • 180W 150W 120W 90W 60W 30W OE 30E 60E 90E 120E 150E 180E • Longitude • Figure 5. Ground trace of second burn The vehicle's final pass was over the Diego All subsequent radar observation opportunities • Garcia tracking station. Contact was confirmed that no portion of the satellite established 4.6 minutes after thruster ignition remained in orbit. • and lasted for approximately 5.3 minutes. At this time, the second burn was observed to be Lessons Learned • executing nominally. Table 5 on the next page lists the pre-burn and predicted post-bum Presently, few spacecraft are designed or • orbital parameters. intended to perform a controlled reentry once their primary mission objectives have been • Parameter Post-burn met. Furthermore, documentation governing (predicted) reentry processes or procedures is sparse. In Orbit(km) spite of this, after receiving authorization to • Bum Time (sec) 1353 Fuel Pressure (psi a) not available prepare for reentry on 10 November 1997, the Fuel Load (lb.) 2.5 LB (unusable­ MSTI-3 spacecraft was successfully reentered • trapped in fuel lines) and all debris concentrated into a 100 x 10 km • Table 5. Comparison of pre-burn 2 and predicted square in the Pacific Ocean on II December post-burn 2 orbital parameters 1997. Mission success can be attributed to the • dedication and teamwork of the various Analyses predicted that approximately 45 participants. Comments from the various • minutes after thruster ignition, the remnants of players and organizations concerning lessons the MSTI-3 spacecraft would splash down in learned throughout the disposal process are • the Pacific Ocean, just north of the equator. summarized as follows. • • Lesley M. Rahman 13 12th AIAAIVSU Conference on Small Satellites • •

Advanced Planning participants to obtain current orbital ephemeris • data (i.e., vectors, and element sets). The data All participants agreed that disposal planning should be stored in a standard format. • should commence as early in the spacecraft Availability of standardized, current orbital life cycle as possible. Ideally, the decision to ephemeris data was especially important after • .reenter a satellite at the end of its mission the execution of MSTI-3's first burn since should be made prior to satellite construction increased atmospheric drag rapidly degraded • so that reentry considerations can be the satellite's orbit. incorporated in the design phase. In addition, • funding for satellite disposal, as well as all Effective Communication political ramifications, should be worked out • prior to launch. Absent this advanced Frequent and effective communication planning, participants estimated that. (under between all key players was essential to • normal circumstances), a reentry scenario mission success. Participants recommend that would likely take at least two months to one or more planning meetings be held with • develop. all parties in attendance for future satellite disposal missions. Such meetings should • All participants agreed that the selected area of ideally provide the following: a full briefing debris impact should be prioritized from the on the spacecraft, including orbital • following: non-populated area, wide ocean parameters, propulsion system characteristics, area along ground trace to account for propellant levels and consumption history, • underlover burning and breakup of the power status, full disclosure of specifications satellite, an area which provides optimal radar and characteristics-including manufacturer's • and infrared coverage to confirm nominal burn documentation; a definition of each and reentry location, and an area which participant's role and responsibility; a time line • provides ground station coverage. of proposed events; and a contact sheet listing Furthermore, if independent verification of the address, phonelfax number, and email • satellite reentry is desired, a request to Inter­ address of each party. In order to keep all Range Operations (IRO) must be made well in participants informed and to provide a forum • advance of the date of reentry to insure proper for problem resolution, teleconferences should sensor coverage. be held on a periodic basis. Additionally, a • teleconference should occur prior to major Information Exchange/Standardization events to reiterate each participants role in the • event, outline expected results, and summarize A secondary suggestion for future disposal different anomaly resolution scenarios. • efforts relates to the exchange of data and information. Currently, USSPACECOM does Collision Avoidance (COLA) Analysis • not have an electronic data transfer mechanism. This fact precluded rapid transfer The most current (no more than a couple of • of orbital element sets. In addition, the hours old) ephemeris data must be used for ephemeris data was often sent in an any collision avoidance (COLA) planning as • unreadable format to users-including NASA day-old element sets off of the 151 Command and the Air Force. Participants stressed the and Control Squadron (l CACS) bulletin board • importance of the creation of a central are insufficient. Furthermore, COLA analyses distribution system to be used by all • should account for a certain degree of nominal

Lesley M. Rahman 14 12th AIAA/uSU Conference on Small Satellites • • • • • burning and should evaluate the entire reentry atmospheric drag. Additionally, following the • trajectory. Originally, The Aerospace initial bum, AFSCN RTS schedulers and Corporation evaluated only a few discrete planners had to account for an orbit that was • points at different intervals along the bum. changing significantly every day. This approach resulted in an overstatement of • the risk MSTI-3 posed to the Mir and Shuttle. Detailed Docunientation Further analysis at one-second intervals • showed that the bum would not (and did not) When the decision is made that a satellite is a result in a near miss. candidate for reentry, ensure that all the • necessary and relevant information is well­ In the future, it is unlikely that NASA will documented. Specifically, a safety analysis, • actively participate in reentry events for non­ debris study, and fuel quantity analysis should NASA missions as this activity is under the be completed prior to launch in order to • jurisdiction of USSP ACECOM. NASA needs eliminate any uncertainty which may be to work with USSPACECOM, however, to associated with lack of information later on in • ensure criteria to protect manned assets (i.e., a the process. In addition, a list of all the miss by how much in what direction is too relevant parameters for each attempt should be • close?) are well understood. maintained by one party to ensure consistency between all parallel analyses. The availability • Developed Procedures/Guidelines of any and all spacecraft documentation such as the On-Orbit Handbook (OOH), vehicle • The movement to mitigate orbital debris will schematics, propulsion system no doubt cause satellite disposal activity to characterization, and build papers that include • increase in the future. As such, participants component materials, sizes, and shapes will feel that a centralized agency should be help to improve the accuracy of the reentry • created to regulate the disposal process. Set analysis. guidelines which outline minimal "safe" • operating limits with which to attempt reentry Finally, a Lessons Learned document should need to be established and communicated to be maintained throughout the reentry process • the operational community as soon as in order to capture processes, procedures, and possible. Budgetary constraints often cause information which may prove helpful either • agencies to stretch operations in order to later on in the reentry process, or for future maximize mission objectives. Unfortunately, disposal attempts. • pushing the operational limits of the vehicle could cause the permanent loss of the Acknowledgments • capability to safely dispose of the satellite-as • was almost the case with MSTJ-3. The Aerospace Corporation Reentry operations may require the The Aerospace Corporation's Flight • development of new contingency and on-orbit Mechanics and Astrodynamics Department operation plans. Due to a lower perigee performed calculations for designing the • between the first and second bum, new reentry burns, performed collision avoidance operating procedures were necessarily (COLA) studies, and supplied survivable • developed for MSTI-3 that addressed debris hazard analyses. problems encountered due to increased

• Lesley M. Rahman 15 12th AIAA/uSU Conference on Small Satellites • • •

Inter-Range Operations (IRO) SMCITEO •

The IRO team served as the focal point for SMCITEO, at the RSC satellite control • coordination between SMCITEO and complex, did all vehicle commanding from the USSPACECOM for services such as Early Onizuka TSC prior to June 1997. From July • Orbit Determination (EODET) and collision 1997 until mission end, 'the RSC team served avoidance (COLA) studies. as the AFSCN mission focal point, performed • orbit determination, RTS scheduling, and MSTI Operations Center (BATCA VEl handled all coordination with the AFSCN. •

The BATCAVE performed all bus and payload • mission planning, analysis, and anomaly USSPACECOM resolution. In July 1997, all MSTI • commanding transitioned to the BATCAVE USSPACECOM's Space Safety Office from the Onizuka TSC. Commanding and (Cheyenne Mountain) performed collision • telemetry was routed between the BATCAVE avoidance (COLA) studies between the and the spacecraft via a T-1 link "bent-pipe" spacecraft, Columbia, and Mir, coordinated • through the Onizuka AFSCN node to the worldwide radar observations to locate the Remote Tracking Stations (RTS). vehicle after each burn, and made appropriate • worldwide notifications and warnings of the NASA spacecraft's reentry. •

NASA worked with The Aerospace References • Corporation and USSPACECOM by performing collision avoidance (COLA) 1. National Space Policy, 14 September • studies between the satellite, Columbia, and 1996. Mir. • 2. Satellite Disposal Procedures. Draft SMCIMT USSPACECOM Directive UPD10-39. 03 • November 1997. The spacecraft was under the management of • SMCIMT. Final authorization for reentry of the satellite came from SMCIMT. • • • • • • • Lesley M. Rahman 16 12th AIAA/uSU Conference on Small Satellites • • •