Share this story
Close X
Switch to Desktop Site

Computer, other problems delay Columbia landing

The National Aeronautics and Space Adminstration's conservative approach to safety was amply demonstrated yesterday when two general-purpose computers on board Columbia failed.

At about 4:45 a.m. central standard time on Thursday, shuttle commander John Young fired a thruster located in the orbiter's nose and reported that the firing was very strong.

About these ads

So strong, in fact, that the firing knocked out one of the shuttle's five general-purpose computers (GPCs). About five minutes later, when another thruster was fired, a second computer went down.

During critical parts of a flight, these computers perform guidance, navigation, and shuttle control tasks. These four units are redundant, running with the same software and processing identical data. The fifth unit is as a backup for the others.

NASA flight controllers then ordered the crew to place the orbiter in free drift. One computer has since been brought back on line.

At this writing, engineers were trying to find out whether the failure is hardware- or software-related. In any case, it appears that GPC No. 1 would be down for the duration of the mission.

Spokesmen for NASA and IBM emphasize that in a pinch, the shuttle can safely land with only one of these computers operating.

But because the opportunity existed to study the problem before landing, planners decided to take advantage of that time to try to troubleshoot the balky unit.

A spokesman for IBM, the manufacturer of the GPCs, says that something similar to this happened on STS-6. During the mission, one of two GPCs that run during the orbiting period failed. In that case, the crew transferred the unit's tasks to another computer, restarted it and gave it less work to do.

About these ads

In addition, one of Columbia's three redundant Inertial Measurement Units, which give the on-board computers attitude and velocity data, appeared to have failed. At this writing, NASA engineers were troubleshooting the problem.

Two other opportunities existed for landing yesterday: one at about 2:22 Pacific standard time. The other occurs at about 3:55 PST.

Follow Stories Like This
Get the Monitor stories you care about delivered to your inbox.