Background
Spacecraft launches in the 1970’s all utilized government funded, industry purchased, expendable launch vehicles such as Delta, Atlas, Titan, etc. Costs were very high since only a single spacecraft could be launched at a time and none of the rocket parts could be reused. Therefore, in 1972 Congress approved and development started to create the Space Transportation System (STS) of the future. NASA’s vision was that the STS would replace the government’s family of expendable launchers and become America’s exclusive system for access to space. NASA’s case for the Shuttle projected a dramatic step increase in launcher cost-effectiveness through the recovery and reuse of most of this manned launcher system’s components combined with tens of launches per year. Costs would be reduced from roughly $100,000/pound to $100/pound assuming 24 launches per year.
To attract spacecraft customers, NASA established a pricing formula for launch to low earth orbit (LEO). This formula was based on the customer’s utilization of the Shuttles 60 foot long by 15 foot diameter payload bay or the fraction of the available Shuttle payload mass capability, about 55,000 pounds, whichever was the greater. Based on NASA’s total price for a Shuttle launch, adjusted by their pricing formula, projected launch costs for a typical mid-70’s Hughes spacecraft was a fraction (~1/3?) of that charged by the government for an expendable launcher. NASA even indicated to Hughes that their target launch costs were $20M shared among all passengers.
Although Hughes (and others) objected to the shuttle being the only launch vehicle available, Hughes immediately began an internally funded design development activity. The initial concept for a Shuttle-optimized spacecraft was a wide-body spin stabilized spacecraft occupying the full Shuttle payload bay diameter, approximately 1/4 of the payload bay length and about 1/3 of its payload mass capacity. This new spacecraft was called “Syncom IV” and required development of its own propulsion to transfer from the Shuttle orbit (LEO) to Geosynchronous orbit. Syncom IV was to be secured to the Shuttle’s payload bay using a reusable “cradle” adapter encircling the lower half of the spacecraft’s cylindrical drum with 5 contact points. Deployment in orbit was implemented through a “Frisbee” ejection, clearing the payload bay with a small residual velocity and low spacecraft spin. Spacecraft on-board timers would autonomously command deployment of the S/C omni antenna to enable command and telemetry capability, spin up the spacecraft to about 30 rpm and fire the solid rocket motor to achieve the first transfer orbit (approx 9,600 nmi apogee). When Hughes first presented the “Frisbee” ejection concept to NASA, the initial reaction was “You’re going to do what?”.
Undaunted, Hughes continued the development and eventually NASA “saw the light” and supported the effort, though they did not provide any funding to Hughes. This effort included NASA and Hughes jointly developing payload safety requirement standards that became the basis for future payload launches and the creation of a “Demonstration Payload” illustrating the concepts.
The Hughes approach was to keep interfaces between the shuttle and spacecraft as simple as possible. The spacecraft was to be unpowered at launch (except for slight trickle charge on batteries and some heaters on the cradle). The only electrical interfaces were via the cradle to the shuttle Standard Switch Panel (SSP) for deployment initiation and Multiplexer-Demultiplexer (MDM) to monitor cradle release pin position and S/C battery temperatures. Mechanical interfaces were also straight forward, just the five cradle point attachments to the shuttle–-two on each of the shuttle bay longerons and one at the keel. The goal was to optimize the spacecraft for the shuttle, recognizing that it would not be capable of launching on an expendable launcher. The design was named Syncom IV and was to be a new HS-381 product line at Hughes, similar to HS-376, HS-601, etc.
Syncom IV was the brainchild of Alois Wittman and Harold Rosen who handpicked a small-dedicated team to run the program led by Ron Swanson as program manager and Jerry Dutcher as system engineering manager. The system engineering team consisted of Chuck Rubin mechanical and Paul Sengstock for spacecraft electrical interfaces. The tightly integrated program office system engineering and technology division engineering teams minimized the development cost and allowed the program to streamline the “normal” formalized paperwork and meetings that were typical of compartmentalized military and NASA programs.
Hughes conducted a vigorous marketing campaign to demonstrate to potential Syncom IV customers the unmatched cost-effectiveness of the communication performance that could be delivered on orbit for a very attractive cost. Understandably, commercial customers were wary of the risk for a Shuttle that had never flown and a spacecraft designed only for a Shuttle launch. Nevertheless, the state of California expressed interest in using Syncom IV to launch an emergency communications payload that could be used to provide continuous communication services during natural disasters such as earthquakes and major fires. This idea never went beyond the discussion stage since it was overtaken by the opportunity described in the next section.
Navy Procurement
The 1976 Congressional review of the DoD budget directed “increased use of leased commercial facilities”. FleetSatCom, built by TRW and managed by the Air Force for the Navy was having significant technical, cost and schedule problems. The Navy saw a lease arrangement as a way to bypass the DOD dictate that all Space procurements would be managed by the Air Force. In 1978 the Navy held a competition to provide the next generation of UHF satellites and ground systems with all being procured on a fully leased arrangement. The new system was appropriately given the name Leasat. All previous government satellite systems had been procured on some type of cost plus contractual arrangement, which usually grew in cost as the system was developed. The leasing approach was attractive because it provided a contractual way to avoid cost growth to the government by transferring the cost risk to the contractor. Hughes proposed a system that leveraged its work on Syncom IV for the spacecraft bus to provide a low launch cost solution that gave us an edge over the competition. This approach led to an award of a $335M lease contract to Hughes for 5 years of communication service at each of 4 orbital locations. Options were identified for 2 years service extension and Navy could purchase satellites after option exercised.
The Leasat contract provided Hughes an opportunity to expand its business from a spacecraft manufacturer to developing the system architecture and providing hardware for an entire worldwide communications system, including being a service provider. Hughes responsibilities included financing, launching, insuring, building the ground control network and operating the satellites for their lifetime in addition to spacecraft manufacture. Hughes Communications Inc (HCI) was formed as a subsidiary to Hughes Aircraft to provide ground stations and become the service provider to the Navy.
Hughes took on significant financial risk with a lease arrangement as the shuttle had not yet flown and this would be a first “total system” responsibility for Hughes and if the shuttle were to fail (or be significantly delayed), the financial consequences would be severe. However, on the positive side, since the government was mandating that all future spacecraft launches were to be via shuttle, Hughes would have an advantage of several years against other S/C manufacturers that were not willing to take this kind of financial risk but were waiting for the shuttle to prove itself.
Roles of HCI and HSC
The proposal team that won the Leasat contract, led by Sam Silverberg, consisted of engineers from spacecraft systems engineering with support from technology division design and manufacturing areas of Hughes Space and Communications (HSC), and as a result they were spacecraft centric. Many on the proposal team were from the classified area. Hughes Communications International (HCI) was formed as a subsidiary to Hughes Aircraft but it was unclear initially how the new HCI organization should be staffed, as their responsibilities were primarily ground hardware – a very different technology and culture than space hardware. The HSC space segment program was initially run out of the Defense Systems Division but moved to the Commercial Division later when it became clear that the program was more like a commercial program.
Leasat System Elements
A satellite control facility (Operations Control Center, OCC) was constructed at HCI headquarters near HSC in El Segundo. HCI also led the design and construction of four permanent satellite communications stations (SCS) to be located on Navy property (Norfolk, Guam, Stockton, Hawaii) that were remotely controlled from the OCC. These SCSs directly interfaced with the payload and bus to control and monitor the on-orbit satellites. In addition, the OCC remotely commanded and monitored the two Moveable Ground Stations (MGS) located in Norfolk and Guam to provide Leasat launch operations and subsequent on-orbit control until turnover to the Navy permanent facility. The MGS’s would be taken down upon completion of the launch phase and put in storage until the next launch.
Clay T.(Tom) Whitehead was named President and CEO of HCI. Tom did not have spacecraft background; Steve Dorfman was assigned to work directly for Tom as executive Vice President in fall of 1982. Steve replaced him in 1983 when Tom left the company. The spacecraft program manager initially was Dave Braverman. System engineering continued to be led by Jerry Dutcher and the Syncom IV development team previously described. Until Tom Whitehead’s establishment as HCI leader, both Dave Braverman and Sam Silverberg acted in that capacity.
Hughes at that time had a software organization that had expertise in ground hardware and software for classified space related projects (this group was sold to Raytheon in 1997). HCI recruited from that organization. Jack Donahue and Jeff Outwater who played critical roles in software and hardware development for the Leasat TT&C stayed with HSC and played a very critical role for both HCI and HSC. The Leasat program was shut down in 1980 for two years due to Shuttle development issues and delays and when it resumed in 1982 Marv Mixon became the Space Segment Manager. Starting the program up again was very difficult since many key individuals were assigned to other programs and were key to those programs so did not return to Leasat.
The 1979 HSC space segment was organized as a commercial program within the commercial business unit. The initial program manager was Dave Braverman and he built up the Syncom IV systems engineering staff led by Jerry Dutcher. HSC realized they needed additional manpower to augment the Syncom IV team so the following individuals were added to the program office Systems Engineering team: Andy Ott (Launch & Mission Operations and S/C external interfaces), Greg Jarvis (S/C Bus), Larry Nowak (Propulsion and Attitude Controls), Larry Watson (initially TT&C but later Communications payload). I mention these four individuals because they and the Syncom IV team were key to many issues that developed in the subsequent years and supported the program through its completion, even though they too were assigned to other programs intermittently during the several stops caused by shuttle delays.
The Space segment design proceeded relatively smoothly. The HSC systems engineering team worked with the different technology organizations relatively free of oversight. The Navy treated the project essentially “hands off” since it was a lease arrangement and Hughes was taking a huge risk financially as well as technically. HCI acted as a pseudo customer, attending many meetings and design reviews as the program evolved. Their main interest was to make sure they understood the interfaces that affected ground operations as opposed to typical customers that get heavily involved in technical details. HCI could not give direction to HSC and vice versa but we worked together quite effectively. Neither the Navy nor HCI participated in most of the technical meetings with NASA. Launch vehicle integration activities were clearly the responsibility of HSC.
Space Segment Technical Description
Leasat is a cylindrical spacecraft with a diameter of 14 feet and height also 14 feet with its antennas stowed for launch and 20 feet high with antennas deployed. The spacecraft and cradle weigh 17,000 pounds in the shuttle’s payload bay. After separation, Leasat weighs 15,200 pounds and on-orbit beginning of life 3,100 pounds. Three 25 amp-hour NiCad batteries supply power through eclipse periods and the solar array power output after 7 years is 1,200 watts.
The Leasat spacecraft is typical of other Hughes spin stabilized communications satellites in that it has a spun section that provides the necessary electrical power and control of the spacecraft and a despun section for earth pointing that contains the payload electronics. The larger diameter solar arrays provide more power than non-shuttle optimized spacecraft; thus the design permits incorporation of a solid fuel perigee kick motor (PKM) within the spacecraft bus. The PKM is the upper stage of a Minuteman III missile weighing about 7,300 pounds. Surrounding the PKM is a liquid propellant system that augments the initial geo-synchronous transfer orbit resulting from PKM firing and circularizes the orbit at geo-synchronous altitude. Four large tanks containing 4,100 pounds of oxidizer in two and fuel in the other two feed two liquid propellant 100-pound thrusters. Smaller tanks contain helium for tank pressurization and hydrazine (350 pounds) for orbit and attitude control with 5-pound thrusters.
The solid PKM is essentially a second stage integrated into and carried by the spacecraft. The liquid system used for both perigee boost and apogee circularization is the first liquid system ever used for this function. The combination is referred to as “Integral Propulsion”.
The despun communications payload includes two 12-foot long helical UHF antennas that provide both receive and transmit capability, respectively, for 12 UHF repeaters. A Navy Fleet Broadcast function consists of an SHF uplink and both SHF and UHF downlinks. The SHF uplink and downlink earth coverage horns support the uplink and acquisition/timing function. The UHF downlink for Fleet Broadcast is multiplexed together with the 12 transmitters. In addition to providing world wide communications service for Navy ships, Leasat also enabled communication from small remote military facilities of the army, marine, etc and soldiers on the ground using backpacks containing the electronics and a small UHF antenna.
A significant advantage of the spacecraft’s large diameter is that once on orbit the spacecraft’s roll-to-pitch inertia remains greater than one, making it inherently stable throughout life. This simplifies the attitude control system since “flat spin” is precluded and added damping on the despun side is not required.
Leasat is carried on its side in the shuttle payload bay attached to a u-shaped aluminum cradle. Five steel trunnions attach the cradle to the shuttle (4 longeron and 1 keel). Five internal contact points attach the spacecraft to the cradle. Four of these points have motor driven pin release mechanisms. The fifth point uses an explosive mechanism to release a spring that ejects the spacecraft. The cradle remains with the shuttle and is reusable. Two electronic units are mounted on the cradle and distribute 2,000 watts of shuttle power distributed via 2 umbilical connectors to heaters on the cradle and spacecraft.
The shuttle orbit is approximately 160 n. mi. circular. About one day after liftoff, the shuttle is positioned with its payload bay facing the earth and is oriented so that Leasat’s spin axis points in the direction its PKM must fire. The four locking pins are retracted and an explosive device at the fifth contact point releases a spring that ejects the spacecraft in a Frisbee motion. While the spring pushes one side of the spacecraft the other side pivots resulting in a separation velocity about 1.5 feet per second and a stabilizing spin of about 1.5 rpm. In addition to stabilizing the spacecraft, the spin helps settle the liquid fuel, and prevents overexposure of any side to the sun.
Frisbee Ejection From Shuttle
Leasat in Low Earth Orbit
The spacecraft is completely unpowered while in the shuttle payload bay. There are two redundant timers – Post Ejection Sequencers (PES) – that start upon spacecraft separation from the cradle. This separation of the spacecraft from the shuttle is referred to by NASA as “ejection” in order to avoid confusion with the term deployment often used for other activities. Internal spacecraft power is applied to the PESs via two redundant switches that actuate upon separation from the cradle. This starts a time controlled sequence of commands from both PES’s that deploy the omni-directional Telemetry and Command antenna 80 seconds after ejection. About 6 minutes after ejection, two 5 lbf hydrazene thrusters fire to increase the spin rate to 30 rpm. At 45 minutes after ejection the PESs command the PKM to fire, which is the final PES commanded activity. The PKM fires for about 60 seconds and boosts Leasat into its first geosynchronous transfer orbit with an apogee of about 9,600 nmi and perigee at 160 nmi. The spent PKM casing is jettisoned by ground command. Jettisoning the PKM actuates two switches that remove power from the PESs as they are no longer required. Transfer to geosynchronous orbit is accomplished by ground commanding the two liquid 100 lbf thrusters as the spacecraft passes through three perigees(“LAM Burns”). When Leasat reaches geosynchronous altitude, the 100 lbf thrusters are fired again at apogee to circularize the orbit. Leasat is reoriented, its antenna and payload electronics platform despun and UHF antennas deployed. Spacecraft bus and payload in-orbit testing are performed before beginning of service.
Leasat Launch Sequence
Significant Events: Program Start Through System Test
The following attempts to capture the most significant events from Leasat program start through launch preparations for the first launch, with a special emphasis on the effort required for shuttle optimization and Launch and Mission Operations. The biggest challenge prior to launch was the multiple stops and restarts to the program as a result of shuttle development delays, the most significant being a complete shutdown from November of 1980 to January of 1982. During these shutdowns multiple personnel, including many key individuals as well as several program managers were re-assigned with only a few maintaining continuity of the program. In addition, Hughes absorbed significant costs due to the shutdowns.
Fortunately, the prior Syncom IV development activities provided a good start to the program. The program was transferred into the Commercial Business Unit and bureaucracy continued to be minimized. An interesting example of this was the first Payload Integration Plan (PIP) kickoff meeting held at Johnson Space Center in Houston. The PIP is a comprehensive document published by NASA for each payload but with much of the material (including actual text authorship) written by Hughes. The PIP and its many annexes capture electrical, mechanical, operational, launch criteria, etc interfaces with the Space Transportation System, part of which is the shuttle itself. Four Hughes personnel (Marv Mixon, Chuck Rubin, Paul Sengstock and Andy Ott) attended this kickoff meeting and were shocked to find a large conference room overflowing with NASA personnel and the first two rows of 10 seats each on one side reserved for Hughes. NASA was shocked that only 4 people showed up and likely felt “slighted” by these Hughes employees who could not possibly understand the scope of the effort ahead of them. In due time, over the many meetings we had, the NASA responsible engineers for each activity began to appreciate the strong emphasis Hughes had to keep the interfaces straightforward and simple.
The astronauts especially appreciated working directly with Hughes and vice versa. Hughes developed special relationships with the astronauts, including the Leasat program secretary Eileen Keegan marrying one of the astronauts that deployed Leasat (Steve Hawley) and they are still happily married. A special occasion was a visit to the high bay of all 10 of the STS 14 and STS 16 astronaut crews followed by a wonderful reception hosted by HCI. There were many trips to Hughes by the astronauts where they could get familiar with the actual spacecraft they would be launching and there were many trips to Johnson Space Center by Hughes to coordinate the activities required by the PIP and assist in shuttle testing of Leasat interfaces, including the model that was used for astronaut training purposes.
Spacecraft development and hardware integration proceeded relatively smoothly in spite of the many program stops/restarts described. A major challenge to overcome was that the existing Hughes facilities were not large enough to handle a completely integrated spacecraft. We had to transport the spacecraft to TRW in Redondo Beach to perform acoustic testing since Hughes did not have an acoustic test facility and the vibration facilities could not handle Leasat either. We had to design a special shipping container and remove many of the overhead power lines/traffic signals to transport the large spacecraft. The thermal vacuum chambers were too small to accommodate a fully built up spacecraft so we could not perform full RF power testing under vacuum conditions. We could not demonstrate actual shuttle ejection of Leasat since having a fully propellant loaded spacecraft and cradle-combined weight of 17,000 pounds was not practical for testing in a 1-g environment. However, we were able to run a series of cradle ejection tests with a mass model representing the spacecraft.
During the 1970’s, spacecraft command, telemetry and orbital analysis software was undergoing rapid change. Typically, the customer (NASA, commercial, or military) would use their own software to control their on-orbit satellite but Hughes would use its own software for System Test of their spacecraft. The two systems were incompatible so much effort went into making sure the mission software was adequately checked out, usually on software simulators. Since Hughes had total responsibility from hardware development through on-orbit operations, the opportunity existed for a common software system. Unfortunately, the Hughes system test organization was reluctant to use the mission software for commanding and telemetry monitoring during system test; they claimed the Hughes system test software was superior because it allowed direct interface with their hardware, such as spectrum analyzers, power supplies, etc and they had experience with the System Test software. A meeting was called where Fred Schneider presented arguments for using system test software and Jack Donahue presented arguments for using mission software. The meeting became quite contentious and finally Fred said he could not agree to use the on-orbit software being developed by Jack Donahue’s group. The final conclusion for the meeting was clearly stated by Jerry Dutcher to Fred: “You or your replacement will use the on-orbit developed software for system test!”
Having common T&C software for system test and mission operations was a significant benefit. We were able to run many system test procedures with automated procedures (called PROCs) that commanded the spacecraft and verified proper response that were later used in the actual mission. Paul Sengstock wrote the majority of the procedures and those he did not personally author he reviewed prior to use. We were able to run end-to-end system test by remotely connecting the actual mission control hardware and software. In addition, we could use spacecraft on the ground to debug new on-orbit PROCs that needed to be developed for flight spacecraft in orbit. There was a very smooth transition from system test to mission operations as a result of this decision.
Although spacecraft F-1 was originally slated to be the first Leasat launched, due to the many shuttle delays, Hughes decided to launch F-2 first for several reasons, including that the environmental testing extremes were more severe than F2 (protoflight vs. acceptance levels).
Launch and Mission Operations
Hughes HSC personnel supported launch countdown activities and launch at both Kennedy Space Center and Johnson Space Center with at least one technical manager and one senior manager at each site. The Operations Control Center (OCC) at HCI is configured for launch activities and referred to as the Hughes Mission Control Center (HMCC). It was fully staffed, including Leasat system engineering 24/7. Experts from the technology divisions and orbital dynamics provide real time mission support as required. For example, Jerry Salvatore, Loren Slafer and Pete Goshgarian supported every Leasat launch and orbit raising operation, including the many novel, first of a kind, orbital analysis and maneuver planning required for Leasat to reach geosynchronous altitude. A senior Flight Director leads a small team at each Movable Ground Station to assure timely acquisition of the spacecraft upon ejection from the shuttle. On-orbit test is commanded from the HMCC and performance verified at the remote on-orbit test site (Hawaii for F-4).
One of the key HSC systems engineers, Paul Sengstock, author of most of the PROCs used for orbit raising was on medical leave for the first mission. HSC made arrangements for an emergency installation of a second phone line to his home so he could support the mission in real time (another Hughes first? -telecommuting).
The spacecraft is unpowered at launch and ejection is planned for one day after launch so the focus initially is to make sure the shuttle orbital parameters are forwarded to Hughes in a timely manner to enable acquisition as soon as possible.
F-2 Launch (8/30/84, STS 41-D, Discovery)
The F-2 launch was scheduled for 6/26/1984. On 6/25, workers digging a storm drain in El Segundo accidentally severed underground phone lines such that there was no communication in or out of the OCC or HMCC. Hughes established an emergency backup MCC in the Building S41 Orbital Dynamics Lab by moving spare equipment, including the spare PDP 1134 Command and Telemetry processing computer from the actual HMCC. The Orbital Dynamics Lab already had a PDP 1170 with Leasat unique software as it was used for mission planning. The System Test PDP 1134 could also have been used as it also had the Leasat unique software installed. Upon pressure from Hughes, the telephone company dispatched several trucks to the site, worked overnight, and was able to repair the lines without delaying the countdown. Launch, however, was aborted during the final countdown due to Shuttle problems so the emergency backup was not needed. Question for the reader – Do you think Hughes would have called a halt to the launch countdown, considering this was the first Leasat to be launched and it was the first shuttle optimized spacecraft if the communication lines had not been reestablished prior to launch?
During the rescheduled launch attempt 8/30/1984, as luck would have it, a hurricane was passing near the Norfolk Navy station where Hughes had set up the Movable Ground Station (MGS) and its large (30 ft diameter) dish antenna to acquire and track Leasat upon Shuttle ejection. Chris Schram, the lead flight director for Leasat and Jeff Outwater, the HCI MGS expert were at the MGS and though the winds were high (you could hear the walls of the MGS trailer shuttering over the communications line to the HMCC), they felt they could still support the launch. The Leasat system design included ability to launch with only one MGS so decision was made to proceed, as the Guam MGS was available. A lot of frayed nerves were relieved when the winds died down somewhat during final countdown.
The shuttle was launched on time and the following day Leasat was ejected from the shuttle payload bay. All subsequent activities including orbit raising using the solid and liquid motors were nominal. On-orbit test of both spacecraft bus and payload was successful and spacecraft control was turned over to HCI at the El Segundo Operations Control Center (OCC) to provide service to the Navy. The MGSs were taken down and put in storage for the next launch.
F-2 was quickly put into service to support the Iraq war effort. Among the tasks it performed was providing US Special Forces hiding in Iraq an encrypted link back to headquarters in Tampa and Washington DC as well as commanders in the field. The Leasat team felt proud to learn of their role supporting the armed forces.
F-1 Launch (11/8/84, STS 51-A, Discovery)
This shuttle mission not only launched Leasat F1 but also retrieved Westar and Palapa satellites also built by Hughes. All operations for shuttle ejection, orbit raising and in-orbit test of Leasat were nominal and spacecraft control was turned over to the Hughes OCC to provide service to the Navy. The only concern was that a typhoon was threatening the Guam MGS site near the time of liftoff. C’est la vie.
F3 Launch (4/12/85, STS 51-D, Discovery)
Senator Jake Garn of New Mexico was on this flight. Greg Jarvis from Hughes was originally assigned to this flight as a Payload Specialist but politics trumped technology and Greg was re-assigned to Challenger. (Please see separate write-up on the blog of the process followed in selecting Greg to be the Hughes Payload Specialist and how he wound up on Challenger).
The Shuttle launch was nominal and Leasat was successfully ejected from the payload bay the following day. It is interesting to note that Leasat ejection was on 4/13/85 at 13:13 GMT and was documented in real time in the HSC System Engineering Mission Log Book on page 13. Also, 4=1+3 and 8+5=13. Superstitious anyone?
There are two redundant cradle separation switches that actuate upon physical separation of the spacecraft from its cradle. When actuated, these switches provide power to two redundant Post Ejection Sequencers (PES) which then send timed commands to the spacecraft to power up, deploy the omni antenna for command and telemetry capability 80 seconds after ejection, spin up the spacecraft to about 30 rpm and fire the solid rocket (PKM) to raise the spacecraft orbit, all at a predetermined timed sequence. The first indication that something was wrong was at ejection + 80 seconds when the astronaut Rhea Seddon reported the omni antenna did not deploy. The shuttle was repositioned so its windows were not facing the spacecraft since the spacecraft hydrazine spin-up jets were timed to fire at 6 minutes, 35 seconds after ejection and the solid PKM 45 minutes after ejection. When neither of these events occurred, it became clear that commands from neither of the redundant PESs were being issued, but the reason was not apparent.
After much deliberation at Hughes and JSC, it was concluded the only single point failure that could possibly explain this was that both cradle separation micro switches did not activate upon ejection as their plungers were both simultaneously mechanically depressed by a single actuator arm that extended out of the spacecraft through a slot in the cylindrical solar panel. This failure mode was considered remote but the decision was made to rendezvous with the stranded Leasat and attempt to “jiggle” this mechanical arm. The astronauts cut up plastic book covers they had on-board into strips that could be used to “snag” the arm and if it was stuck, perhaps to release it. This contraption looked somewhat like a “fly swatter”. An identical fly swatter was constructed on the ground and used on another Leasat to illustrate its practicality. The mission was extended 2 days and astronauts David Griggs and Jeff Hoffman performed an EVA to attach the homemade “Flyswatter” to the shuttle’s Remote Manipulator Arm. Astronaut Rhea Seddon operated the arm from inside the cockpit, snagging the switch several times. Each time there was clear mechanical reaction to the snag but the spacecraft did not electrically respond. There was nothing more the astronauts could do so Leasat F3 was left stranded in its 160 nmi circular orbit and the shuttle returned to earth.
Fly Swatter After First “Swat”
F3 Stranded in Low Earth Orbit Mission Operations (4/13/85 to 9/85)
Leasat was now stranded in a useless 160 nmi circular orbit without command or telemetry capability from the ground. Since the thermal subsystem was unpowered, predictions indicated that all the liquid propellants would freeze as well as the solid PKM. Liquid freeze/thaw cycles rupturing the propellant lines and the solid fuel cracking were major concerns. In addition, the NiCad batteries would freeze and all spacecraft electronics would exceed the temperature ranges they were designed and tested to.
Within hours of the failure, in the NASA Cafeteria, Joe Engle of the astronaut office at JSC met with Harold Rosen and Chuck Rubin and surfaced the idea of a subsequent rendezvous mission to salvage Leasat. The astronaut core was anxious to display “the right stuff” at that time and NASA wanted to demonstrate the flexibility of their shuttle.
Meanwhile, a very detailed failure review led by Andy Ott concluded that there were no credible single point failures that could have resulted in the failure but the most likely cause was that two redundant separation switches between the spacecraft body and the solid PKM adapter structure were both somehow in the wrong state. The plunger of each switch needed to be depressed into its overtravel region for liftoff and released after the PES fired the PKM. During the first geo-synchronous transfer orbit the PKM would be jettisoned by ground command to remove power from the PESs that no longer had a function to perform. Detailed dimensional and tolerance analysis showed the margin for the plungers being depressed into their overtravel region was low, considering the hysteresis that occurs during shuttle liftoff. It is likely that the “g forces” on the PKM during liftoff caused the switches to open but they did not restore to their pre-launch condition. Measurements on F-4 and F-5 supported this conclusion and a paper search showed that F-1 had to be readjusted during system integration.
One of the major challenges was to convince NASA and the insurance underwriters that approaching Leasat and the subsequent EVA (extra-vehicular activity) to “fix it” would be safe to the astronauts and the shuttle itself. This would be the first unscheduled/unrehearsed EVA of the shuttle era. In addition, Hughes would have to design and build the necessary equipment and electronics to provide power to F3, establish a command and telemetry link, thaw the propellants and batteries and reestablish proper thermal boundary conditions.
Leasat F-3 was considered a total loss. Hughes had paid $15 Million for $200 Million of coverage of the launch but this would not cover total costs. Hughes made a deal with the insurance underwriters and NASA that Hughes would invest $10 Million more of its own money to fund a salvage operation with a subsequent launch. If the salvage were successful, the insurance underwriters would pay Hughes the $10 Million and Hughes would not collect any insurance payments. In addition, Hughes would share Navy lease revenue, which would return most, but not all, of the money insurers paid out. NASA was anxious at that time to prove the flexibility of their new launch system so would not charge Hughes for the extra rendezvous and EVA required tacked onto the Leasat F-4 launch mission. Integration and test of Leasat F-5 (the spare) started since the probability of success for F-3 salvage and subsequent on-orbit service to the Navy was considered low, estimated at 10% by some. Increasing the financial risk to Hughes even more, there was not a signed agreement with the insurers until most of the salvage preparations were complete.
The Digital Electronics Lab at Hughes was assigned the task of building a PES Bypass Box that could be attached to the outside of F3 during a rendezvous/EVA that would enable internal power to the Spacecraft Bus. In order to give the astronauts enough time to return to the shuttle, get back into the cockpit, and maneuver the shuttle to a safe separation distance, this box needed to have a timer that would not start until adequate time had passed. This timer was set to start 13 hours after initiation. Harvey Quan and Rick Hollis of the Digital Electronics Lab were shocked when the program office dubbed this the “12.99 hour timer”.
A small, battery powered hand-held unit had to be fabricated which would connect to the spacecraft despun test access connector to enable the astronauts to manually deploy the omni antenna. In addition, a PKM nozzle heater had to be fabricated and attached to the PKM to help thaw it out. The outer surface of the plate was a “dark mirror” to simultaneously maximize solar absorption and minimize thermal radiation. A portion of the plate was populated with solar cells to power temperature sensors and a small C-band transmitter and whip antenna to transmit temperature data to the ground. Spacecraft capture arm and tools for the installations had to be fabricated by NASA.
Obviously, these efforts required much coordination between NASA and Hughes but the “can do attitude” by both was remarkable. All of this hardware had to be designed, built, tested and delivered within 6 months. Hughes assisted astronaut training at JSC and the astronauts made several trips to Hughes to “practice” on F-4 or F-5.
Modify F-4
Although the evidence strongly pointed to the PKM separation switch installation and alignment as the cause of the F-3 failure, Hughes decided to make significant modifications to F-4, especially since if F-3 salvage was not successful, there was no spare spacecraft to protect against a launch (or spacecraft) failure and the Navy contract called for four operational satellites in orbit. These modifications included separate cradle separation switch levers, umbilical bypass of cradle and PKM separation switches, and spacecraft ground command and telemetry capability upon ejection from the shuttle if the PES sequence failed.
The backup command and telemetry capability required new automated Mission Procedures (PROCs) that were fast since station passes in low earth orbit were only a minute or two. Upgrades to the orbital analysis software were also required. Fortunately, spacecraft system test software and mission software were the same so debugging the PROCs could be done against F4 or F5.
F-4 Launch and F-3 Salvage (8/27/85, STS 51-I, Discovery)
F-4 launch and F-3 salvage were during the same shuttle mission, less than 6 months after the F-3 failure. The launch proceeded smoothly and F-4 was ejected from the shuttle the following day. The PESs turned on as expected and performed all their actions nominally, the backup ground command and telemetry capability was not needed. Orbit raising and subsequent spacecraft bus testing was nominal. The payload was turned on for on-orbit testing and initially performed flawlessly. After 40 hours of full power operation, the UHF payload downlink disappeared.
Meanwhile, F-3 salvage operations were in full swing. The shuttle rendezvoused with F-3 and “Ox” Van Hoften and Bill Fisher performed the longest EVAs in history. First they installed the capture arm to the slowly spinning spacecraft; “Ox” grabbed the 15,200 lb spacecraft multiple times to slow and stop the spin. You could see the Remote Manipulator Arm flex with each capture. The PES Bypass Box and PKM nozzle heater cover were attached to the spacecraft. Astronaut Bill Fisher moved to a protected position behind the spacecraft and fired the omni deployment squib using the Hughes provided hand held unit. The 12.99 hour timer was started. “Ox” manually spun up the spacecraft to about 5 rpm by multiple grabs of the capture arm as the last step before the astronauts returned to enter the shuttle.
Leasat F-3: Approach for Capture
Leasat F-3: Astronaut Bill Fisher holding Leasat in preparation for salvage operations.
For the next 6 months, thermal conditioning of the PKM was monitored from the ground during the approximately two-minute station passes. Ground reception of the PKM temperature data was accomplished by mounting a C-Band dish to the edge of the main MGS tracking antenna at Norfolk and Guam. The baseband signal was recorded on a tape deck snatched from one of the worker’s home audio system and played back post-pass for detailed analysis. Prior to PKM fire, the spacecraft was spun up to 30 rpm using the now thawed RCS Hydrazine jets over many station passes as all burns had to be completed while still in view. The window for firing the PKM by ground command from Guam was 54 seconds. The PKM ignition sequence was initiated by sending a short series of ground commands to the PES Bypass Box affixed to the outside of the spacecraft, that, in turn, applied power to the PESs via test access connectors. The PKM performed nominally and subsequent orbit raising with the liquid bipropellant system also performed flawlessly. Spacecraft bus and payload testing demonstrated that all on-orbit performance met or exceeded requirements. In fact, the F3 battery life was better than any other Leasat spacecraft, even though they had been frozen for many months.
While F-3 thermal conditioning was taking place, the payload in-orbit test team in Hawaii led by Larry Watson from the HMCC was busy trying to figure out why the F-4 UHF downlink disappeared. Since the UHF transmitter temperatures started to increase at the same time as the downlink disappeared (a signature matching that of full reflected RF power), the likely failure was hypothesized to be an open circuit between the UHF multiplexer and the transmit antenna. The fact that every downlink frequency was attenuated to below detectability exonerated the UHF multiplexer itself. An independent senior review team was formed that included consultants and outside experts as well as Hughes to analyze the design and try to figure out what caused the failure. Dozens of high power tests were performed in the laboratory vacuum chamber in an attempt to duplicate the failure. The design proved to be robust and no definitive “smoking gun” was found. The independent team conclusion (not unanimous) many months later was that powdered Teflon at the junction of the output cable to the transmit antenna impeded outgassing through vents designed into the junction boxes. This resulted in pressure in that area going into the corona region and developing into high temperature and high power plasma. The connection vaporized and there was no hope for salvaging the mission. This put extra pressure on the F-3 Mission team (and F-5 entire team) since the program no longer had a spare spacecraft in case of a launch or another spacecraft failure. This also illustrated that Hughes was wise to take a chance on salvaging the F3 spacecraft. (Note: It should be pointed out that many months after the formal failure report, as discussed under the section describing Leasat F-5 modifications, another possible cause was identified.)
As head of HCI, Steve Dorfman was responsible for briefing potential buyers of Hughes on the high visibility Leasat in-orbit problems. Don Atwood, Vice Chairman of GM and ultimately the overseer of Hughes after the acquisition, responded in a big meeting with GM management who were evaluating buying Hughes: “After the Leasat failure most companies would deploy their lawyers to minimize the legal damages from the problem. Hughes was the sort of company that put their engineers to work to solve the problem”. Which we did. And we earned the respect of GM and our Navy customer.
A very special event took place on 11/4/85 when HCI hosted all the astronauts and Hughes team and showed the Salvage Mission Highlights on a very large screen. I am sure the insurance underwriters were ecstatic also.
Challenger Disaster (1/28/86, STS 51-L) Changed Everything
Greg Jarvis, Leasat Spacecraft Bus Systems Engineer, was the Payload Specialist on this mission (see upcoming section in the blog describing the Hughes process for selecting the Payload Specialist and how Greg wound up on this particular mission that had nothing to do with Leasat; launch included TDRSS Spacecraft deployment and Christa McAuliffe teacher in space).
Art Jones, who was the Kennedy Space Center launch interface engineer for Leasat burst into the building S1 conference room where the Leasat F5 Integration and Test Team was conducting their daily system integration coordination meeting with the news that Challenger had blown up. After the initial shock, the meeting dispersed and participants went to different conference rooms to see what happened on television. Emotions ran extremely high; many broke down in tears, including several executives. The conference rooms were filled again when NASA broadcast the memorial from Johnson Space Center and President Reagan spoke; once again many tears, especially when President Reagan hugged Greg’s wife Marcia. Hughes also had a memorial service for all employees that was very much appreciated. Greg had finished the course work required for a Masters Degree in Business Management at West Coast University. Greg mailed a handwritten copy of his thesis to the university the day before the launch. The university had planned to award the degree while Challenger was in orbit, making Greg the first person to have his degree conferred while in space. His thesis was titled “In Search of Excellence” and described Hughes Space and Communications Group character, culture and management style. The manuscript was postdated 1/29/86 and Greg was posthumously awarded the Masters Degree at the Spring 1986 commencement at West Coast.
The Challenger Astronauts: back row left to right El Onizuka, Christa McAuliffe, Greg Jarvis, Judy Resnik; front row left to right Mike Smith, Dick Scobee, and Ron McNair.
The shuttle fleet was grounded. NASA issued a new space policy that precluded launching any unmanned commercial satellites, thus the Leasat (HS-381) product line was now obsolete. However, Leasat F-5 was granted an exception. There was a planned launch for December 1989 by NASA to recover and bring back to earth the Long Duration Exposure Facility (LDEF) that had been in orbit since 1984. The LDEF recovery mission would have launched with the payload bay empty. NASA was embarrassed to launch an empty shuttle so agreed to launch F-5 on this mission.
NASA significantly increased their focus on safety and required reviews (internal and external) of all previously approved documentation and a repeat of the Safety review process. In addition, NASA increased load margin requirements which required significant spacecraft modifications as described in the next section. Since it was not known how long the shuttle fleet would be grounded, the Leasat team was once again disbanded except for a few individuals focusing on NASA interfaces and modifications required to F5 as a result of the new NASA requirements.
F-5 Modifications
F-5 was the “hanger queen” of Leasat until the F-4 total failure since it was originally designated to be a ground spare. Andy Ott was assigned to be the Leasat F-5 Program Manager by Tony Iorillo, the Space & Communications Group CEO. Tony announced “When F-3 failed to activate upon Shuttle ejection, that was a disaster for Hughes. When F-4 failed on-orbit that was a catastrophe for Hughes. I do not know of words in the English language that could describe the effect on Hughes if F-5 is a failure”. Although some felt that this statement put too much pressure on Andy, it was actually a blessing. All the employees at Hughes knew what was at stake and they gave their all to support whatever Leasat needed; a true “can do” attitude.
A complete review of all anomalies and failures for any F-5 units was performed and rework/retest/rebuild performed as necessary. In-orbit history of F-1 through F-4 and any issues/concerns were addressed if applicable to F-5. For example, one of the F-5 wideband receivers was reworked and a brand new one manufactured as a result of this activity. All modifications made to F-4 as a result of the F-3 failure to activate were also made to F-5.
During the detailed review of the process of fabricating the semi-rigid cable between the UHF Multiplexer and Antenna, the process of etching the serpentine spring elements in preparation for soldering was reviewed. These serpentine springs were about 0.5 cm wide and 0.02 cm (?) thick and installed into each junction box between the UHF Multiplexer and antenna to allow for thermal expansion of the center conductor. It was noticed the callout for preparing for solder, an acidic mixture of “Type II” concentration and time under temperature was required. The drawing was old and it was difficult to ascertain whether it called out Type I or Type II. So even though technicians were adamant that they were using Type II, a mixture of Type I was created per the procedure just for a special test and several serpentine elements were immersed in preparation for soldering. After the elements were removed and dried, they were placed on a bench and appeared normal. It was only when held up to the light that it was clear the elements had many small holes, as the acid concentration was too strong. Even one expansion element undergoing this etching process installed on F-4 could explain the total failure. There were many thermal stress cycles during launch and on-orbit operations that could have caused the weakened element to break. The junction box could be in the corona region slowly letting out the air through small vents designed for this purpose. New expansion serpentine elements, double the thickness were manufactured, and the process was carefully monitored. In addition, the entire semi-rigid cable from the UHF Multiplexer to the antenna was replaced with a new one. A full RF powered ambient and thermal vacuum test was performed on the flight spacecraft.
Leasat F-5 in Anechoic Chamber for Full RF Power Testing
A Critical Clearance Checklist was created and verified. Cradle separation switch and PKM Separation switch plungers were verified to be significantly in the over travel region. Separate levers were used for each cradle separation switch. All other clearances were adequate except there was a potential for buckling of the spun barrier splice strips used to cover the seam between sections of the spun barrier. During launch and orbit raising, the spun barrier splice strips are exposed to multiple thermal cycles each day. The strips were reworked and the clearance between spun and despun barriers increased. Verification of this buckling required construction of a facility (at Rodeo Road) capable of spinning the entire spun barrier under on-orbit thermal conditions. Buckling was visually able to be confirmed as each splice strip “potato chipped” and decreased the clearance enough to make contact with the despun barrier had it been there. About a year after nominal on-orbit operations, F-1 began to show a slow increase in friction between the spun and despun sections during seasons the sun was not shining on the barriers. This likely was caused by the buckling phenomena and the solution was to invert the spacecraft twice a year to keep the despun section in the sun. This buckling could also have been the reason spun and despun barriers locked while trying to shake F-4 following the total failure of the payload with the hope that there might be an intermittent connection.
The most significant F-5 modifications were due to new load margins established by NASA as a result of them becoming more conservative following Challenger. This required many F-5 structural elements and joints to be replaced. Since the spacecraft was mechanically fully integrated at this time, an appropriate analogy is “ surgically replacing bones and joints of the patient without disturbing him”. Dr. Sam Bassily was the hands on structural mechanical engineer resolving both the spun barrier buckling issues and leading the surgical team replacing “body parts”.
F5 Launch (1/9/90, STS 32, Columbia)
Once again, the Shuttle delay had resulted in many key individuals assigned to other programs and it was difficult to get them back and they had forgotten much. In addition, several very key individuals had retired, including Larry Nowak and Paul Sengstock. Both Larry and Paul came back to support the F5 mission, illustrating the camaraderie and dedication of the Leasat team and the Hughes culture at the time.
The launch proceeded smoothly and F-5 was ejected from the shuttle the following day. The PESs turned on as expected and performed all their actions nominally; the backup ground command and telemetry capability was not needed. Orbit raising and subsequent spacecraft bus testing was nominal. The payload was turned on for on-orbit testing and performed flawlessly. All on-orbit performance requirements have been met or exceeded, including the 7-year design life. F-5 was eventually replaced, as scheduled, by a UHF follow-on spacecraft and since it was still performing flawlessly leased to the Australian Navy. It is currently providing service to the Australian government, 22 years after launch.
Concluding Remarks
There were many firsts accomplished by Hughes on the Leasat program. Hughes was fortunate the Leasat program was mostly during the late 70’s and 80’s when an entrepreneurial “can do” approach was feasible. Regrettably, today’s environment would never have allowed for that.
Leasat was a challenge in many ways. Although the Challenger disaster resulted in the government canceling all commercial launches on Shuttle, thus making the Leasat design obsolete, it nevertheless was the first stepping stone to grow Hughes Space and Communications Group from a spacecraft manufacturer to a complete communications system architect and service provider. It led the way for programs such as Galaxy, DirecTV, Thuraya and many more to come. In addition, in spite of all its problems, Hughes made a significant amount of money from the program and the Navy is ecstatic with its technical performance and financial protection (the original $335 Million contract for 5 years of operation never increased by even one penny). Most of all, the Leasat team is proud of its contribution to the U.S. Military.
Although this article is written from a Space Segment Systems Engineering/Program Office perspective, I encourage additions to the Blog from your (the reader) perspective. I would like to thank the following who helped me to “remember” the 34 year story; without their inputs and comments the story would have remained untold: Dave Braverman, Jerry Dutcher, Chuck Rubin, Larry Nowak, Larry Watson, Jerry Salvatore and Loren Slafer.
Our Space Heritage 1960-2000 