Fragments of a Mystery

(2 of 4)

If the foam was not behind the disaster, the wheel well might have been. Some of the flaky temperature readings that came down from the ship in its last few minutes originated in the left well, leading to fears that explosive bolts intended to help lower the wheel if it became stuck might have blown, damaging the ship. But the very purpose of the bolts is to detonate in the wheel space and do so safely. What's more, the well temperatures rose only about 40ºF in the last minutes of the flight, worrisome but not nearly high enough to trigger or confirm a serious explosion. Said Dittemore: "A 30-to-40-degree rise does not constitute cause for concern."

With the foam and the bolts moving down the list of likely causes, a meteor hit moved up. Few people suggest that a cataclysmic collision simply blew the ship out of the sky--not so low in the atmosphere, anyway. But up in orbit, a bad ding by a rogue rock could have done enough damage to cause serious drag as the ship descended through the atmosphere, and Columbia indeed heeled sharply to the left before it disintegrated. Pits and gouges in the protective tiles are common during flight; ships routinely pick up close to 100 of them. But for that very reason, a meteor remains a long shot: with 22 years of experience, NASA knows small hits seem to do shuttles little damage, and the Columbia crew never reported anything big.

The confusion may be cleared up as NASA continues its hunt for debris--particularly if it finds the first bits that fell from the ship. Shuttle tiles carry serial numbers that correspond to a particular part on the spacecraft's underside. The pieces on the ground thus form a sort of bread-crumb trail leading back to the area on the spacecraft where the problems began. Find the westernmost part, and you have pinpointed the trouble spot. "That would be very, very significant," says Dittemore.

Originally, Columbia's wreckage was thought not to have fallen west of Texas. Then an astronomer from the California Institute of Technology reported that he saw what looked like debris trailing the ship as it passed overhead. An amateur videotape shot in Arizona seemed to show the same thing. Most tellingly, NASA released a photograph taken by a high-powered Air Force telescope as Columbia soared over New Mexico. It had been widely reported that damage to the left wing was visible in the picture, but the resolution turned out to be too poor to reveal anything conclusive. The agency was hopeful that the videotapes might yield more, but skeptics cautioned that even during routine re-entries, bursts of plasma can mimic the appearance of debris.

Late in the week, the agency began speculating about a far less obvious cause of the accident: atmospheric turbulence. As the shuttle descends, air moves smoothly over the wings until, at about 150,000 ft., it begins to churn and swirl. As long as the shuttle is moving slowly enough, at about 6,000 m.p.h., it can handle the transition. If the shift happens higher in the atmosphere, when the ship is moving faster, the heat and stress increase dramatically. This can sometimes make a shuttle pull to one side--just as Columbia did. Early turbulence can be caused by sudden shifts in air density, but it can also be caused by pitting on the wings or otherwise ragged tiles.