STS-1 Columbia

Young took manual control and put it down on Runway 23, on the dry lakebed at Edwards, two and a quarter days and thirty six orbits after Florida.

He came out an hour and eight minutes later and walked the whole way round the outside of it, looking at the tiles.

This is the world's greatest flying machine, I'll tell you that. It worked super.

A spacecraft coming home was not new. Gagarin came home; Apollo came home. What had never happened was the next part.

On 12 November 1981 the same vehicle stood on the same pad and went up again. Not a sister ship, not a fresh build to the same drawings. The same one.

The first flight proved Columbia could come back. The second proved it could go again.

Even then, not all of it came back. The boosters fell into the Atlantic two minutes up, where ships were waiting. The tanks were never collected.

Columbia was the heaviest orbiter ever built, about three and a half tonnes more than the ones that followed: it was the prototype, and carried test instrumentation no other craft needed.

That weight cost it the space station. Too heavy for the climb into that orbit, it was never given a docking system, and the ship that could not go to the station became the fleet's laboratory instead. It flew eleven of the fifteen Spacelab missions.

Foam came off the tank. It had been coming off the tank for years.

This is the part that usually gets told wrong. Nobody ignored it. It was filmed, measured, argued about and written up, and the conclusion was always the same: foam had struck the orbiter before, and the orbiter had come home.

Which was true, and was the crux of the problem. Every strike that had no consequences became evidence that a strike destroys nothing. Twenty two years of coming back was read as proof it would keep coming back.

16 January 2003. Eighty one point seven seconds after liftoff, a piece of foam the size of a briefcase came off the left bipod ramp, where the tank braced against the orbiter's nose.

Two tenths of a second later it struck the leading edge of the left wing at something over four hundred miles an hour.

It was spotted the next day on the launch film, and assessed. Columbia stayed up another two weeks doing science, and everyone believed the vehicle was fine.

By 2003 the entry blackout was gone. The orbiter could reach a relay satellite through the gap in the plasma behind it, and Houston stayed on the loop the whole way down.

So there was nothing to hide in. At 13:59:32 UTC, over Texas, Charlie Hobaugh called up about pressure readings on the left main gear. Rick Husband answered.

Roger, uh, buh. The words stopped in the middle of one, and the data stopped with them.

Rick Husband. William McCool. Michael Anderson. David Brown. Kalpana Chawla. Laurel Clark. Ilan Ramon.

The vehicle was gone. The tape survived.

In the tail there was an old recorder, running since 1981, writing sensor readings onto magnetic tape from places the telemetry never reached. It flew all twenty eight missions. It was not built to survive anything.

It kept writing for another forty six seconds after Husband's last word.

Seven weeks later a search team found it in a field near Hemphill, Texas, one edge three inches into the soil. Slightly damaged. Entirely readable.

Columbia was built to come back, and for twenty two years it did just that.