Thirty years ago, this month—had the hands of fate showed greater kindness—two shuttles might have rocketed into orbit within days of each other to deliver a pair of robotic spacecraft toward Jupiter. The first spacecraft, Ulysses, was destined to pick up a gravitational “slingshot” from the Solar System’s largest planet, allowing it to depart the ecliptic plane and explore the poles of the Sun, whilst the second, Galileo, was slated to become the first machine made by human hands to enter orbit around Jupiter. As outlined in yesterday’s AmericaSpace article, Missions 61F and 61G would have launched their payloads atop powerful Centaur-G Prime boosters, built by General Dynamics, whose enormous impulse was precariously balanced against a structural fragility and which had already left lingering worries in the minds of the eight astronauts who might have flown the missions.
Those astronauts—the 61F/Ulysses crew of Commander Frederick “Rick” Hauck, Pilot Roy Bridges, and Mission Specialists Mike Lounge and Dave Hilmers, and the 61G/Galileo crew of Commander Dave Walker, Pilot Ron Grabe, and Mission Specialists Norm Thagard and James “Ox” van Hoften—had been named by NASA in May 1985. According to the agency’s ambitious shuttle manifest for 1986, the missions would have flown within five days of each other, taking advantage of an exceptionally short “launch window” to reach Jupiter, under the most optimum conditions, by July 1987. Orbiter Challenger would have launched on Mission 61F at 4:10 p.m. EDT on 15 May 1986, from Pad 39B at the Kennedy Space Center (KSC) in Florida. She would have flown for four days, before touching down on the Cape’s Shuttle Landing Facility (SLF) on 19 May. Challenger’s landing would have come within sight of her sister, Atlantis, which was targeted to fly from neighboring Pad 39A on the afternoon of 20 May, to begin her own four-day mission.
Had it happened as planned, this would have marked the shortest launch-to-launch interval (just five days) and the shortest landing-to-launch interval (a mere 23 hours) ever accomplished in the 30-year shuttle program. Of course, it was not to be. At the time of 51L, the launch-to-launch record stood at 16 days—achieved between Columbia’s 61C liftoff on 12 January 1986 and Challenger’s untimely loss on 28 January—whilst the landing-to-launch record of 10 days, two hours, and eight minutes stood between the return of Discovery’s 51D crew on 19 April 1985 and the liftoff of Challenger’s 51B astronauts on 29 April. In the aftermath of 51L, the landing-to-launch record was narrowed slightly to nine days, during the STS-56 and STS-55 flight campaign in April 1993. However, the empirical record for the shuttle program was set in the summer of 1995: following Atlantis’ STS-71 launch at 2:32 p.m. EDT on 27 June, her sister Discovery rocketed into orbit to begin STS-70 at 8:41 a.m. EDT on 13 July, creating a launch-to-launch record of 15 days, 18 hours, and nine minutes. And the landing-to-launch achievement of those missions was even more impressive: for Atlantis landed on 7 July, creating an interval of five days, 22 hours, and 46 minutes between the end of STS-71 and the beginning of STS-70.
However, even the gap between STS-71 and STS-70 would have paled in comparison to the aggressive effort to get Challenger and Atlantis into orbit, within the same week, in May 1986. As outlined yesterday by AmericaSpace, both shuttles had undergone extensive modifications, costing about $5 million apiece, which included additional plumbing to load and drain the Centaur booster’s liquid oxygen and hydrogen propellants, as well as control panels on the aft flight deck to monitor its performance.
Worryingly, however, the Centaur’s pressure regulation hardware was not redundant and, worse, a failure of its internal bulkhead carried the potential to rupture its oxygen and hydrogen tanks. Moreover, it was recognized that the huge amount of propellants—which totaled 37,370 pounds (16,500 kg)—could induce “sloshing” and other controllability problems which might hinder Hauck or Walker if the need arose to perform a Return to Launch Site (RTLS) abort, shortly after liftoff. Notwithstanding these risks, the Centaur’s liquid propellants offered considerably more oomph to push payloads like Ulysses and Galileo out of low-Earth orbit and onto trajectories to other planets than solid-fueled boosters could achieve. It was also known that liquid-fueled rockets produced a “gentler” thrust than the notoriously harsh impulse of solids.
“The shuttle was obligated to launch Ulysses and Galileo,” Hauck told the NASA oral historian. “[NASA] needed the most powerful rockets they could have [and] at some point the decision was made to use Centaur, which was never meant to be involved in human spaceflight. That’s important because rockets that are associated with human spaceflight have certain levels of redundancy and certain design specifications that are supposed to make them more reliable. Centaur did not come from that heritage, so, Number One, that was going to be an issue in itself, but Number Two is if you’ve got a Return to Launch Site abort or transatlantic abort and you’ve got to land—and you’ve got a rocket filled with liquid oxygen [and] liquid hydrogen in the cargo bay—you’ve got to get rid of [it], so that means you’ve got to dump it while you’re flying through this contingency abort.
“To make sure that it can dump safely,” Hauck continued, “you need to have redundant parallel dump valves, helium systems that control the dump valves [and] software that makes sure contingencies can be taken care of. Then, when you land, you’re sitting with the Centaur in the bay that you haven’t been able to dump all of it, so you’re venting gaseous hydrogen out this side [and] gaseous oxygen out that side. This is just not a good idea!”
During pre-launch loading operations, the Centaur’s liquids would have been fed through plumbing “tapped” into the shuttle’s Main Propulsion System (MPS) feedlines. Emergency dumping vents—capable of draining all liquid oxygen and hydrogen within 250 seconds of an abort being declared—were situated on opposite sides of the aft fuselage, just beneath the Orbital Maneuvering System (OMS) pods. As part of her validation trials, Atlantis was to be rolled out to Pad 39B in mid-February 1986, just two weeks after the launch complex had been vacated by Mission 51L, with a real Centaur and a mockup of the Galileo spacecraft. Whilst on the pad, the Centaur would have been fueled and a series of “wet” countdown tests conducted. Atlantis would then have been removed from the pad to allow the real Galileo to be installed, and she would have then been rolled out to Pad 39A. By mid-April, she would have been joined on adjacent Pad 39B, laden with the Ulysses/Centaur combo.
Doubts over the reliability of the Centaur riding aboard the shuttle had earlier led NASA to cancel it and fly Ulysses and Galileo atop less powerful, solid-fueled Inertial Upper Stage (IUS) boosters. However, the exceptionally heavy Galileo had both a Jupiter orbiter and atmospheric entry probe and the Centaur-IUS switch, announced by NASA in the fall of 1981, threatened to almost double its journey time to 4.5 years. The cost of the mission correspondingly increased, peaking at around a billion dollars, and in late 1982 Congress pressed NASA to resume work on the shuttle-borne Centaur.
Scheduled to fly on 15 May (61F/Ulysses) and 20 May (61G/Galileo), the missions had launch windows extending for barely one hour. Everything about them had been stripped to the barebones minimum: just four astronauts, no secondary payloads, and even several pieces of crew equipment, including the middeck galley, had been removed. For 61G astronaut James “Ox” van Hoften, it was a stark contrast to his first two shuttle flights, during which he performed several dramatic spacewalks to retrieve and repair NASA’s Solar Max observatory and the crippled Leasat-3 communications satellite. In van Hoften’s mind, the excitement of launching “this goofy thing to Jupiter” hardly compared. “That mission looked honestly really boring,” he told the NASA oral historian. “I had two of the best flights that NASA ever did and I was just on top of the world. [Mission 61G] was going to be a very short flight and there was going to be nothing to it, other than going up and launching this Centaur.”
In early January 1986, NASA accepted a recommendation to fly 61G with Atlantis’ “Phase II” Space Shuttle Main Engines (SSMEs) running at an as-yet-untried 109-percent thrust level. Launching at the standard 104 percent, it was argued, would have meant that the Centaur and the heavyweight Galileo—which tipped the scales at just over 5,000 pounds (2,270 kg)—could not have supported a full load of propellant. This would have limited the Centaur’s margins for the critical trans-Jupiter “burn.” Ulysses, on the other hand, was considerably lighter, at 800 pounds (370 kg), and Challenger’s SSMEs for 61F remained at the standard 104-percent thrust rating.
After achieving orbit, the 61F and 61G crews would have no more than nine hours to get their payloads deployed. This was due to the fact that the Centaur needed to periodically dump its boiled-off gaseous hydrogen to keep tank pressures within mandated limits. After too much time, it would have “bled” so much hydrogen that the remainder would be insufficient to support the trans-Jupiter burn. Three closely-timed deployment opportunities were scheduled for both missions. In the case of 61F, the first “window” opened at 11:10 p.m. EDT, exactly seven hours into the mission. The Centaur’s twin Pratt & Whitney-built RL-10A-3A engines would have ignited about 45 minutes after deployment, generating a combined impulse of 32,200 pounds (14,600 kg).
In Ulysses’ case, this would have set it on course to rendezvous with Jupiter in July 1987 and subsequently pass over the Sun’s northern and southern polar regions in the 1989-1991 timeframe. Meanwhile, Galileo would have followed a somewhat slower trajectory—with then-NASA Administrator Jim Beggs having endorsed a flyby, en-route, of the tiny asteroid Amphitrite—to reach its target. Depending upon whether the Amphitrite flyby would take place (a decision NASA planned to make after Galileo’s launch), it was anticipated that the spacecraft would reach Jupiter in August or December 1988.
Following the loss of Challenger, some initial hopes remained alive that the shuttle might return to flight in the spring of 1987, allowing Ulysses and Galileo to meet the next Jovian launch window in June of that year. However, modifications to the shuttle and its Solid Rocket Boosters (SRBs) pushed the schedule further to the right. And the safety implications of the Centaur were raising more eyebrows. In early February 1986, the Kennedy Space Center’s safety officer refused to approve the booster’s advanced processing, citing “insufficient verification of hazard controls.” Coupled with cost overruns to the tune of $100 million, on 19 June 1986 newly-appointed NASA Administrator Jim Fletcher canceled the Centaur-G Prime program. A new booster would have to be found for Ulysses and Galileo and, eventually, NASA reverted to the IUS.
Could the five-day goal of Missions 61F and 61G have been achieved? Some experts told the Rogers Commission—the presidential inquiry into Challenger—that a lack of spare parts would have brought the shuttle program to its knees by that point. Others countered that launching two missions within 120 hours of each other was possible. Centaur-G Prime technicians at General Dynamics had been given free 28-month calendars in January 1984 … which ended in May 1986. Clearly, the deadline to be ready for launch was immovable. “I’m convinced to this day we would have made the launch window, but it was a sprint to the finish,” said Marty Winkler of General Dynamics. “It was like the racehorse that overtakes you at the end.”
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 the 30th anniversary of the loss of Gemini IX’s Agena Target Vehicle (ATV), which imposed a new challenge to a mission with an already troubled history.