Fifteen years ago, next week, seven men with ancestries in four sovereign nations—the United States, France, Switzerland, and the United Kingdom—embarked on one of the most difficult flights of the entire shuttle era. Astronauts Curt Brown, Scott Kelly, Steve Smith, John Grunsfeld, British-born Mike Foale, Frenchman Jean-Francois Clervoy, and Switzerland’s Claude Nicollier roared into the darkened Florida sky aboard Discovery, in pursuit of the malfunctioning Hubble Space Telescope (HST). During their eight days in orbit, they performed three spacewalks, each lasting in excess of eight hours, and became the only shuttle crew to spend Christmas off the planet. However, as described in yesterday’s AmericaSpace history article, STS-103’s own fortunes had been changed by the need to implement fleetwide wiring inspections of all three orbiters and by problems and failures with the iconic Hubble itself.
The seven astronauts represented the most experienced shuttle crew, in terms of number of previous flights, ever launched in the program’s 30-year history. With Brown on his sixth mission, Foale on his fifth, Nicollier on his fourth, Smith, Grunsfeld, and Clervoy each on their third, and only Kelly as a “rookie,” the men of STS-103 had no fewer than 18 shuttle flights behind them when they rode out to the pad on the evening of 19 December 1999 for launch. Selection to join STS-103 was a particularly complex first flight for Kelly. “Once we get very close to the telescope, the commander will start flying the vehicle manually,” he told a NASA interviewer in the weeks before launch. “Prior to that, we’ve traded off on who performs some of the burns, but as we get closer I jump in the commander’s seat. I’ll perform the last few rendezvous burns. Curt will get in the back of the vehicle, preparing to fly the orbiter manually. We’ll come up from ‘below’ Hubble and slowly decrease our closure rate until we’re traveling at a very slow closure and to close proximity. It’s very critical that we conduct these burns very precisely, because, if we didn’t, we wouldn’t be able to catch up. We could overshoot the vehicle or undershoot it. Whether we would be able to recover from that or not is hard to say, so it’s a very precise task that we do.”
With three critical gyroscopes having failed aboard Hubble during the course of 1999, it was recognized that a further failure would force mission managers to place the telescope into protective safe mode and suspend scientific observations until the shuttle’s arrival. Since it cost NASA about $21 million, per month, to keep the telescope operational, it was not a minor problem. On the morning of 13 November 1999, these dire fears were realized when a fourth gyroscope failed. “The failure mode here is the breakage of flex wires, the little hair-sized wires that carry current into the motors,” explained HST Project Scientist David Leckrone of the cause of the gyroscope problems. “We think the breakage comes about because those things get embrittled because of oxygen that’s trapped in the fluid in which they’re suspended, and that oxygen has been there since the things were built.” Since the gyroscopes were designed to perform with very low levels of vibration, their operating fluid was extremely thick (about equivalent to 10W-30 motor oil) and tended to hold onto its oxygen. “It’s that oxygen that’s causing the embrittlement,” said Leckrone. “That says the failure is not run-time related. It’s age-related.”
By the time that Hubble’s fourth gyroscope failed, the STS-103 shuttle stack was on the pad, but liftoff remained at least three weeks away, forcing controllers to place the telescope into “Zero-Gyro Sun-Point Safe Mode.” Its aperture door was closed and the twin solar arrays were oriented parallel to the body of the spacecraft. Hubble was then positioned such that the arrays were perpendicular to the Sun, rotating every hour. Until the arrival of Discovery and her crew, the telescope would return only engineering data, slowly spinning at a single revolution per hour to maintain stability. “We don’t like having a spacecraft sitting there not working,” said HST Program Manager John Campbell, “but hopefully it’ll only be another three weeks to the servicing mission and we’ll be back online again shortly thereafter.”
It was actually closer to six weeks by the time Discovery reached orbit on 19 December 1999 to begin a slightly truncated eight days of operations, and three EVAs, to bring the telescope back to operational service. On the second day of their mission, they reduced the shuttle’s cabin pressure to prepare the four spacewalkers for operating within their space suits. With Steve Smith and John Grunsfeld scheduled to perform EVA-1 and EVA-3, it was Mike Foale and Claude Nicollier who had drawn the short straw by having just one spacewalk to perform. They were tasked with EVA-2 and the repeated delays to Discovery’s launch had caused their EVA-4 to be scrubbed. They had already lost EVA-6 from the six-spacewalk flight, when Servicing Mission (SM)-3 was broken into two missions in March 1999. In a space-to-ground interview on 20 December, Foale took the news in his stride. “Well, they come, they go, and you have to kind of roll with it,” he said. “I try to look at the big picture. Down the road a bit, there’ll be other flights, other chances. Certainly, I would regret not getting as many EVAs as I could, being greedy as I am, but this is fine and I understand why the program made that decision.”
Two days into the mission, at 7:34 p.m. EST on 21 December, Curt Brown maneuvered Discovery to within range of Hubble and Jean-Francois Clervoy grappled the telescope with the Canadian-built Remote Manipulator System (RMS) mechanical arm. “Houston, Discovery, we have a good capture,” Clervoy reported. “We have Hubble grappled!” Based upon pre-flight estimates, the retrieval occurred seven minutes ahead of schedule. At the time of capture, the two spacecraft were flying over the Gulf of Mexico and Houston, prompting Capcom Steve Robinson to tell the crew that observers had seen them in the sky, a little dimmer than Jupiter and a little brighter than Saturn.
Kicking off an ambitious trio of EVAs, nearly an hour ahead of schedule, Smith and Grunsfeld switched their space suits onto internal battery power and floated into Discovery’s floodlit payload bay at 1:55 p.m. EST on 22 December. Their primary task was the replacement of all six gyroscopes, housed in pairs within three Rate Sensing Units (RSUs), in order to restore the telescope’s stability for science operations. Described by John Campbell as “Job One,” the task was completed with crispness and perfection, as Smith worked inside Hubble to electrically disconnect the RSUs and Grunsfeld—anchored on the end of the RMS—drove out the bolts with a power tool. Smith then installed battery voltage regulator kits to keep the telescope’s batteries properly conditioned. Following this activity, mission controllers conducted “aliveness” tests to verify that all of the RSU electrical connectors and pins were in place.
Whilst these tests were underway, Smith and Grunsfeld moved to open a pair of valves to allow residual nitrogen to vent from the Near-Infrared Camera and Multi-Object Spectrometer (NICMOS), which had been installed by the SM-2 shuttle crew in February 1997, but which suffered a thermal short a few weeks later and had been out of action since January 1999 due to a dwindling coolant supply. It was anticipated that during the second half of the mission, SM-3B, tentatively scheduled for launch in the spring of 2001, a gaseous neon cooling system would be installed to enable NICMOS to continue scientific research. Unfortunately, when the spacewalkers tried to remove the valve covers, they refused to budge. Mission Control advised them to photograph the covers and press on with the voltage regulator work, but were quickly called back to make another attempt with a high-torque wrench. This time, Smith succeeded in removing the covers and loosening the valves. Although it was not considered a critical task, it was part of the much-desired plan to get NICMOS operational.
The valve problem and difficulties closing and latching the aft shroud doors left them behind schedule, and the intended 6.5-hour EVA quickly surpassed its target and ran to eight hours and 15 minutes by the time Smith and Grunsfeld repressurised Discovery’s airlock at 10:10 p.m. EST. This made it the second-longest spacewalk in shuttle history at that time, eclipsed only by the eight hours and 29 minutes which astronauts Pierre Thuot, Rick Hieb, and Tom Akers spent outside Endeavour during STS-49 in May 1992. Congratulated on a job well done, EVA-1 had completed virtually all of its tasks. The only item which had not been ticked off the list was the installation of handrails, ahead of the Fide Guidance Sensor (FGS)-2 replacement by Foale and Nicollier during EVA-2. However, the replacement of the six gyroscopes had already secured completion of 80 percent of STS-103’s critical tasks. “The gyros are in!” exulted John Campbell. “We watched as each of them got its power on, cheered each one of the six and were really pleased to see all six come up. We know from the aliveness tests that the central computer can use all six gyros when [it] needs them.”
Next day, at 2:06 p.m. EST on 23 December, Foale and Nicollier entered the payload bay and pressed on with their first task to install the $7 million DF-224 computer, which comprised a radiation-hardened Intel 486 processor to replace the 1970s-era device aboard HST. In making his first spacewalk, Nicollier became the first European Space Agency (ESA) astronaut selected by NASA to perform an EVA, an accolade to which Jean-Francois Clervoy made reference in a congratulatory message. Built around the Intel 80486 DX2 processor, and running at 25 MHz, the new, chair-sized computer was made up of three identical boards for redundancy and was 20 times faster and carried six times as much memory as its predecessor. Additionally, it could operate on just 30 watts of electrical power, as opposed to 100 watts for the earlier computer, but NASA engineers joked that it represented the most expensive 386-to-486 upgrade in history. “A 486 by most people’s standards is out of date, but this is a pretty special 486,” Foale explained before the mission. “It’s able to withstand all the radiation and there’s very strong radiation up at the high altitude the Hubble flies without causing the program to crash.”
It was also a difficult upgrade, since it was hard for Foale to see critical connectors on the left side of the box. He had to take exceptional care not to impact connectors on a nearby data management unit. “All the connectors are on the side of the box where you can’t see them,” said John Grunsfeld before the mission, “so Mike has to do that basically without the aid of stereoscopic vision. He’ll have one eye as he’s reaching around to do that and on the left of him, on the [aft shroud] doors is the data management unit and huge bundles of delicate cables. The challenge there is how do you jam yourself as close as you can to that without touching it, so you can see the connectors you have to disconnect, without damaging the cables.” Providing clearance cues, Nicollier positioned himself close to the aft shroud doors and Foale had successfully hooked up the new computer by 4:30 p.m. EST. Shortly afterwards, aliveness tests confirmed that it was functioning as expected.
“Most excellent!” Grunsfeld told the spacewalkers from his station on Discovery’s aft flight deck. “The brains of Hubble have been replaced.” Foale and Nicollier’s next task was the installation of FGS-2, which had been removed during SM-2 in February 1997 and returned to Earth for refurbishment. It took two attempts to properly seat the sensor in its alignment rails, but was completed successfully, prompting Steve Smith to radio Mission Control to ask if the spacewalkers could press on with the replacement of the Optical Control Electronics (OCE) package. The OCE replacement was meant for EVA-3, but Steve Robinson replied from Houston that Foale and Nicollier’s spacewalk was already heading toward the seven-hour point and they should call it a day and begin cleaning up their work site. By the time the spacewalkers returned inside the airlock at 10:16 p.m. EST, they had been outside for eight hours and 10 minutes, concluding the third-longest EVA in shuttle history.
STS-103’s final spacewalk got underway at 2:17 p.m. on 24 December, marking the first U.S. EVA ever performed on Christmas Eve. Smith and Grunsfeld replaced one of the Engineering/Science Tape Recorders (ESTRs) with a new Solid State Recorder (SSR) and installed equipment to aid with fine-tuning the performance of the new FGS-2. They also installed a new S-Band Single Access Transmitter (SSAT), to replace one of two identical units which had failed in 1998. The transmitter, which funneled data from Hubble through NASA’s Tracking and Data Relay Satellite (TDRS) network to ground stations, was not designed for orbital replacement by spacewalkers, and including numerous fiddly screws and cable connectors, which quickly placed the astronauts behind schedule. By five hours into EVA-3, they were 45 minutes behind the timeline.
“We have to work with tiny little bolts,” said Grunsfeld. “It’s kind of like if you had to repair a watch wearing winter gloves. I’m trying to deal with little screws that are non-captive and if you drop one, obviously it’ll float away.” The lengthy S-band transmitter replacement procedure required several other tasks, including the installation of stainless steel foil insulation panels over six of Hubble’s equipment bays, to be deferred to the next servicing mission. As circumstances transpired, they managed to install covers over two bays, before they were called to begin cleaning up their work site and closing out the spacewalk. By the time Smith and Grunsfeld returned to Discovery’s airlock at 10:25 p.m. EST, they had secured the fourth-longest EVA in shuttle history, at eight hours and eight minutes. This placed STS-103 in the record books for having the second-longest, third-longest, and fourth-longest EVAs in the shuttle program, and the mission boasted more than 24 hours and 33 minutes of spacewalking time in total.
Although much attention had focused upon the spacewalkers, the task of Jean-Francois Clervoy as RMS operator was a monumental one. His task was incredibly delicate, requiring him to manoeuvre the EVA crewmen and their tools with great precision and stop and start their motions at specific times. However, he could not apply the arm’s brakes. “The arm will be in running mode at all times, so I will have to protect the sticks to ensure that nobody bumps into them,” he said. “I will have to stay concentrated for several hours and Claude Nicollier will be my backup for that job, when Steve and John are outside, and John Grunsfeld will be my backup, when Claude and Mike are outside.” By his own admission, Clervoy believed that Nicollier or Grunsfeld would probably take over about 20 percent of the whole EVA time flying the RMS to allow him to take a breather. As EVA-3 came to a spectacular conclusion, Clervoy could celebrate a job well done.
Pausing before he returned inside Discovery, Smith took a moment to congratulate flight controllers, engineers, and scientists who had made the repair work possible. “In the last three days,” he said, “we put 13 new [components] in the Hubble Space Telescope with 100-percent success and we appreciate greatly the great folks who are working on this holiday, Christmas Eve 1999.” Grunsfeld added his own words of praise for their space suits, which he described as “incredible machines” which enabled the HST repair work. This proved a slightly premature statement, for whilst in the airlock, shortly before repressurization, his suit failed to transition from internal battery power to Discovery’s power supply. At this point, Grunsfeld had only about 30 minutes of battery power remaining in his suit. Ultimately, the suit was left on battery power throughout the repressurization process, and Claude Nicollier later inspected it and noticed a bent pin in the multi-pin electrical connector.
Only hours after the completion of EVA-3, it was Christmas Day and STS-103 became the first (and only) shuttle mission to remain in orbit over the festive period. In fact, counting Apollo 8 in December 1968 and the crew of Skylab 4 in December 1973, this was only the third occasion that American astronauts aboard an American spacecraft had spent Christmas off the planet. The day began with season’s greetings for Curt Brown as the STS-103 crew was awakened to the sound of Bing Crosby’s “I’ll Be Home for Christmas.”
Christmas in space was a fact that Scott Kelly had taken time and pains to discuss before launch with his five-year-old daughter. “I told her that we were going to point the telescope at the North Pole and get a picture of Santa,” he said. “She was all excited and really didn’t mind too much her dad being away for Christmas.” The crew received messages from their families and shared a meal of duck liver and other delicacies from south-western France, supplied by Jean-Francois Clervoy. Speaking to Mission Control, Brown conveyed his own take on the Christmas message. “The familiar Christmas story reminds us that, for millennia, people of many faiths and cultures have looked to the sky and studied the stars and planets in their search for a deeper understanding of life and for greater wisdom,” he said. “We, the Discovery crew, in this mission to the Hubble Space Telescope, are very proud to be part of this ongoing search beyond ourselves. We hope and trust that the lessons the Universe has to teach us will speak to the yearning that we know is in human hearts everywhere; the yearning for peace on Earth and goodwill among all the human family.” John Grunsfeld added that he wished for peace on Earth in the new millennium.
Later in the day, NASA Administrator Dan Goldin sent his congratulations to the crew. “This proves once again you need people to do things,” he told them. “You just can’t send a robot up there to go fix that telescope. It was like watching a ballet. It was just fantastic what you folks have done. I’m very, very pleased.”
“Well, sir, thank you for the words, but we must put that credit on the folks who trained us and prepared us for this,” replied Curt Brown. “They did a great job and I think it shows. During the EVAs, we did exactly as planned and had a number of surprises that Hubble threw at us with bolts that were a little tighter and things that were a little bit different than planned. As you mentioned, humans are able to accomplish the chances without any trouble and robots, obviously, wouldn’t be able to do that.”
On Christmas afternoon, the final steps ahead of redeploying Hubble into space got underway. By 4:19 p.m. EST, the latches holding the telescope to its support structure in the payload bay were released, and at 6:03 p.m. Jean-Francois Clervoy successfully released the telescope back into free flight to continue its voyage of scientific exploration, with the first data anticipated from the rejuvenated observatory about two weeks later. As it drifted away into the inky blackness, ground controllers reported that it was operating normally on its own, holding steady with its new gyroscopes, which prompted a unanimous cheer from the STS-103 crew.
Owing to the year-end rollover problem associated with the much-hyped Y2K computer bug, it was essential that Discovery was back on the ground, and powered-down, well ahead of New Year’s Eve. Having said this, NASA engineers were confident that the date rollover from 1999 into 2000 would not pose any problems for the shuttle’s computers. “We have done everything that we could do to make sure we are ready for the year-end rollover,” said STS-103 Entry Flight Director Wayne Hale. “We have tested, we have done everything we know to do to make it as seamless as possible. As far as we know, everything would work just fine if we worked right through 12 midnight on 31 December.” Two landing opportunities existed at the Kennedy Space Center (KSC) at 5:18 p.m. and 7:01 p.m. EST on 27 December, with further options in both Florida and at Edwards Air Force Base, Calif., on the following two days. Although NASA had announced before launch that KSC and Edwards would both be kept on standby as “joint primary” landing sites, Hale stressed that Florida carried the edge, because of the $1 million cost of ferrying the orbiter across country from the West Coast to the East. Eitherway, it was mandatory that Discovery made landfall before 29 December.
Early on the 27th, the concern at KSC appeared to center on high crosswinds, which forced mission managers to wave off the first landing attempt. At length, Brown and Kelly were instructed to fire Discovery’s Orbital Maneuvering System (OMS) engines at 5:48 p.m. EST to begin the 73-minute descent to Earth on the second landing opportunity of the day. The burn lasted almost five minutes and produced a spectacular fireshow for ground-based observers, as the shuttle’s glowing plasma trail enthralled skywatchers from southern Texas to New Orleans and the coast of the Gulf of Mexico. Discovery’s ghostly touchdown on Runway 33 at 7:01 p.m. ended a mission less than one hour shy of eight full days in orbit. It was a fitting and spectacular end to the second millennium AD, and closed out almost the first four decades of human space exploration. It would have seemed difficult to countenance, on 1 January 1900, that humans would carry the ability to fly space vehicles and deliver and repair space telescopes, far above Earth, and as the 20th century ended the crew of STS-103 offered their own moment of reflection on the KSC runway. “We shouldn’t forget this is almost the end of the century, this is the end of a millennium,” said Mike Foale. “We started this century dreaming about leaving the planet Earth and flying. And human beings have achieved that.”
This is part of a series of history articles, which will appear each weekend, barring any major news stories. Next week’s article will focus on Apollo 8’s “Christmas story” from orbit around the Moon in December 1968.