CAPE CANAVERAL, Fla. – Investigators trying to figure out what destroyed space shuttle Columbia focused immediately on the possibility that its thermal tiles were damaged far more seriously than NASA realized by a piece of debris during liftoff.
Just a little over a minute into Columbia's launch on Jan. 16, a chunk of insulating foam peeled away from the external fuel tank and smacked into the left wing, which like the rest of the shuttle is covered with tiles to protect the ship from the extreme heat of re-entry into the atmosphere.
On Saturday, that same wing started exhibiting sensor failures and other problems 23 minutes before Columbia was scheduled to touch down. With just 16 minutes to go before landing, the shuttle disintegrated over Texas.
"As we look at that now in hindsight ... we can't discount that there might be a connection," shuttle manager Ron Dittemore said on Saturday, hours after the tragedy. "But we have to caution you and ourselves that we can't rush to judgment on it because there are a lot of things in this business that look like the smoking gun but turn out not even to be close."
Just a day earlier, NASA had given assurances that the launch-day incident was absolutely no reason for concern. The space agency did an extensive engineering analysis that included a frame-by-frame examination of the launch video, and concluded that any damage to Columbia's thermal tiles would be minor.
If the liftoff damage was to blame, the shuttle and its crew of seven may well have been doomed from the very start of the mission.
Dittemore said there was nothing that the astronauts could have done in orbit to fix damaged thermal tiles and nothing that flight controllers could have done to safely bring home a severely scarred shuttle.
The shuttle has more than 20,000 black, white or gray thermal tiles that are made of a carbon composite or silica-glass fibers and are attached to the shuttle with silicone adhesive.
Loose, damaged or missing tiles can change the aerodynamics of the ship and allow heat to warp or melt the underlying aluminum airframe, causing nearby tiles to peel off in a chain reaction. If the tiles strip off in large numbers or in crucial spots, a spacecraft can overheat, break up and plunge to Earth in a shower of hot metal, much like Russia's Mir space station did in 2001.
In Columbia's case, the shuttle broke apart while being exposed to the maximum re-entry heat of 3,000 degrees on the leading edge of the wings, while traveling at 12,500 mph, or 18 times the speed of sound.
"I would say that the tiles are the No. 1 candidate" for causing the disaster, said Norm Carlson, a retired NASA test chief and former launch controller.
Dittemore said that the disaster could have been caused instead by a structural failure of some sort. He did not elaborate.
As for other possibilities, however, NASA said that until the problems with the wing were noticed, everything else appeared to be performing fine. NASA officials said, for example, that the shuttle was in the proper position when it re-entered the atmosphere on autopilot. Re-entry at too steep an angle can cause a spaceship to burn up.
Law enforcement authorities said there was no indication of terrorism; at an altitude of 39 miles, the shuttle was out of range of any surface-to-air missile, one senior government official said.
A California Institute of Technology astronomer, Anthony Beasley, reported seeing a trail of fiery debris behind the shuttle over California. Dittemore said he was unaware of the sighting and would not speculated on what it meant.
If thermal tiles were being ripped off the wing, that would have created drag and the shuttle would have started tilting from the ideal angle of attack. That could have caused the ship to overheat and disintegrate.
The rust-colored foam that covers the shuttle's 154-foot external fuel tank is just lightweight polyurethane, but it can damage when the shuttle when the spaceship is hurtling into space at high speed, Carlson said.
Dittemore said that even if the astronauts had gone out on an emergency spacewalk, there was no way a spacewalker could have safely checked under the wings, which bear the brunt of re-entry heat.
And even if they did find damage, "there's nothing that we can do about tile damage once we get to orbit," Dittemore said. "We can't minimize the heating to the point that it would somehow not require a tile."
NASA did not request help in trying to observe the damaged area with ground telescopes or satellites, in part because it did not believe the pictures would be useful, Dittemore said.
For example, long-distance pictures did not help flight controllers when they wanted to see the tail of space shuttle Discovery during John Glenn's flight in 1998; the door for the drag-chute compartment had fallen off seconds after liftoff.
The shuttle was not equipped with its 50-foot robot arm because it was not needed during this research mission, and so the astronauts did not have the option of using the arm's cameras to get a look at the damage.
It was the second time in just four months that a piece of fuel-tank foam came off during a shuttle liftoff. In October, Atlantis lost a piece of foam that ended up striking the aft skirt of one of its solid-fuel booster rockets. At the time, the damage was thought to be superficial.
Dittemore said this second occurrence "is certainly a signal to our team that something has changed."
Another NASA retiree, Jose Garcia, who was a technical assistant, did not want to speculate Saturday on a possible cause but said budget cuts throughout the 1990s resulted in the elimination of many safety checks and balances during launch preparations. He went public with his concerns, all the way to President Clinton, in fact, but said nothing changed.
"The managers always say, `It's safer than it's ever been. Safety first.' All those words come easy," Garcia said.