Software The Biggest Stumbling Block

Meanwhile, despite the setbacks, Columbia was now firmly on schedule to lift off on 10 April 1981, which, as it happened, was two days shy of the 20th anniversary of Yuri Gagarin's pioneering spaceflight. The six-and-a-half-hour launch 'window' -necessitated by a need to have adequate lighting conditions to satisfactorily photograph Columbia's ascent for engineering analysis - opened that day at 11:50 am GMT.* The window also provided for daylight landing opportunities at White Sands Missile Range in New Mexico, should a launch abort require Young and Crippen to perform an emergency return to Earth after one orbit.

Shortly before 9:00 that morning, after a traditional astronauts' breakfast of steak and eggs in the crew quarters, Young and Crippen boarded Columbia for what turned out to be an uneventful countdown - at least, that is, until its final stages. Then, with just nine minutes to go, during a pre-planned hold, a problem cropped up in one of Columbia's five GPCs. It was described as a 'timing skew'; in effect, the backup flight software was unable to synchronise with the primary set.

Unlike previous manned spacecraft, the Shuttle is totally dependent upon its computers to run the main engines, move the elevons, control its heading and operate the thrusters, to name just a few of many thousand different functions. These units are so critical that five GPCs are carried: four primaries, which run the same software and 'vote' before issuing commands, and a backup. If one of the primaries disagrees with the others, it is 'outvoted' and considered faulty. The backup contains its own, different set of flight software, so that if all four primaries became corrupted, it can take over control.

The problem that Columbia experienced on 10 April was essentially that the four primary GPCs were not communicating with each other correctly. Taking advantage of the lengthy launch window, the liftoff was rescheduled for 3:20 pm as computer engineers wrestled with the software, but when a solution could not be found it was decided to stand down until 12 April. A disappointed Young and Crippen clambered out of Columbia and would spend the next couple of days maintaining their proficiency flying the STA.

Meanwhile, the GPC problem was isolated late on the 10th and the countdown resumed next day. ''The software'', remembered Gordon Fullerton, who flew Columbia in March 1982, ''became the biggest stumbling block. The software in these computers not only control where you fly and the flight path, but almost every other subsystem! Getting the software wrung out and simulators writing the checklists ... we didn't really have it nailed down by STS-1. There were a lot of

* All times throughout this book are given as Greenwich Mean Time (GMT). This has been done to avoid possible confusion, as different time zones apply in Florida, California, New Mexico and other locations mentioned in these pages. As a general rule of thumb, Florida is approximately five hours 'behind', New Mexico some seven hours 'behind' and California around eight hours 'behind' GMT.

unknowns [but] you just finally have to set a launch date and say 'We're going to go'. You cannot be 100% sure of everything.''

Young and Crippen again departed the crew quarters in the early hours of 12 April and took their seats on Columbia's flight deck. Both were clad in bulky US Air Force high-altitude pressure suits, which afforded them full-body protection and were destined to be worn by the first four OFT Shuttle crews. Since these were considered 'test flights' and were also equipped with ejection seats, the full-pressure garments were mandatory; on later operational missions, when restrictions were relaxed somewhat, it was intended for astronauts to fly in lighter overall-type flight suits and helmets.

If an emergency had necessitated their use, the rocket-propelled ejection seats would have fired Young and Crippen through two overhead hatches, but they could only be used to an altitude of 30.5 km, meaning they would not realistically work during the ascent phase and only at selected intervals during re-entry. Astronaut Jack Lousma, who commanded Columbia's third test flight in March 1982, would later remark that his Shuttle launch was far riskier than his Apollo ascent a decade earlier and his opportunities to escape in the event of an emergency were much reduced.

Whereas most previous manned spacecraft had taken the form of ballistic capsules attached to the top of expendable boosters - so that, in the event of problems, an escape rocket could lift them several thousand metres into the air and parachute them a couple of kilometres out to sea - the Shuttle did not offer that option. An on-the-pad emergency would have precluded the use of the ejection seats, because the astronauts would have hit the ground before their parachutes had opened; also, ejections during the first couple of minutes of ascent would have sent them straight into the SRBs' roiling exhaust plumes.

Consequently, the seats could only realistically have been used at selected points of re-entry, after the period of maximum atmospheric heating, and even then the astronauts' chances of survival were slim.

Young and Crippen again boarded Columbia on 12 April, lowered their visors and encountered their first problem: neither man could breathe properly. It turned out that a quick-disconnect fitting for the oxygen system, situated beneath the control panel, had been mispositioned. After this was resolved, the countdown proceeded smoothly.

0 0

Post a comment