Fragments of a Mystery

It came as a big relief early last week when NASA investigators determined that, yes, it was clearly a loose chunk of insulating foam that had damaged the shuttle Columbia's skin and led to its crack-up in the skies over Texas. Of course, they acknowledged, it was also possible that the ship actually started breaking up over California, and it might not have been foam that killed it, but a meteor. Or turbulence. Or an explosion in the wheel well. In fact, it began to seem that the only thing NASA could say with certainty was that nothing seemed to be certain at all.

Even as searchers scoured much of the southern tier of the country for remains of the ship (and a smaller, more sacred swatch for remains of the crew), fresh theories and new clues abounded. Images of Columbia's demise were sent in by amateur videotapers, and reports of evidence were phoned in by freelance debris hunters. Part of the leading edge of one wing turned up near Fort Worth, Texas, while a rear wing section was examined in the eastern part of the state, near Nacogdoches. Researchers dug up old NASA memos warning of just the kind of accident that may have claimed Columbia. Experts sought to reassemble 32 seconds of vital, if patchy, data that sputtered down from Columbia after voice communications were lost. As NASA scrambled to manage events, officials in Washington began taking sides, some sharpening the long knives for the agency, others lining up to defend it. "Space exploration will go on," says Senator Mary Landrieu, Democrat of Louisiana. "[But] there will be intense investigations."

Those investigations got under way even before the shuttle debris was cool. The most notorious piece of evidence was the bit of hardened foam that fell from the external fuel tank during lift-off, striking Columbia's left wing area. Applied like shaving cream, the foam dries to the hardness of a brick, which could conceivably damage the fragile external tiles that protect the shuttle during its fiery re-entry. When it was later disclosed that the spacecraft had spent 39 days idling on the pad before launch--enduring episodes of freezing rain that could have loosened the foam further--the case seemed closed.

But there were problems with the theory. First of all, the foam may seem as hard as a brick, but it isn't nearly as heavy. Even if the debris had been moving at 1,000 m.p.h. when it struck the shuttle's left side--about twice as fast as it was actually going--computer analyses suggested it could have done little damage. "It's difficult for us to believe...that this foam represented a safety issue," said shuttle program manager Ron Dittemore. That, at least, was the agency's position on Wednesday. On Thursday, however, NASA administrator Sean O'Keefe seemed less certain, saying--in what some saw as a mild rebuke of the well-regarded Dittemore--that no possible cause of the accident was being ruled out yet. Dittemore then modified his own public statements accordingly.