They also are looking closely at what may be two key pieces of Columbia debris -- a 2-foot piece of one wing, including an attached chunk of thermal tiles, and a 300-pound cover of a landing gear compartment, possibly the site of a sudden temperature rise moments before the shuttle broke apart.

One day after Columbia's Jan. 16 launch, military radar detected an object moving rapidly away from the shuttle. NASA said it is unknown what the object was, but the possibility that it could have been ice from a waste water vent sent investigators back to a detailed search for evidence that the shuttle may have formed ice throughout its mission.

Adm. Hal Gehman, head of a board investigating the Columbia accident, said Sunday that the object detected near the shuttle could have come from the spacecraft itself and could be ice.

He said the U.S. Space Command of the Air Force, which monitors objects in space, is providing data on the object to the investigators.

"These reports are emerging now right now," Gehman said. "It's too early to say if they mean anything."

The waste water vent, which is under the shuttle cabin, in front of the left wing, is used to expel into space both urine and surplus water generated from the shuttle's fuel cell power system.

Usually the water shoots out into the cold vacuum of space as a spray of crystals, but on at least one shuttle mission, in 1984, the water formed a basketball-sized chunk of ice on the lip of the vent. At the time, NASA engineers were so concerned the ice could damage the shuttle wing during re-entry that they ordered the astronauts aboard Discovery to use the shuttle's robot arm to break off the ice ball.

That heavy robot arm, which wasn't necessary for Columbia's 16-day science mission, was left off so more experiments could be added, and the waste water vent could not be seen from the cabin by the seven astronauts. NASA spokesman Kyle Herring said it's possible ice could have formed and not been detected, even though heaters were installed on the waste water dump valve after the 1984 mission.

When Columbia fired its rockets to drop out of orbit, it could have sent any accumulated ice slamming into the wing where other data suggests there was severe damage to the thermal protection tiles. The theory is unproven and is only one of a number of scenarios being probed by engineers.

Although Gehman and the other members of the Columbia investigation board were appointed by NASA, Gehman said their charter gives them the authority to conduct testing in laboratories not affiliated with the space agency.

He said Sunday that the board will split up into three teams and each will gather data at different NASA centers. This will speed up the investigation, Gehman said. The board has 60 days to complete its investigation. Some critics said the board needs more time, noting that the commission that investigated the 1986 Challenger accident required 120 days to complete its investigation.

NASA administrator Sean O'Keefe said Sunday that no theory has been excluded.

"Nothing is off the table," he said on CNN. "We're going to let the Columbia accident board guide us in terms of their findings about what caused this accident."

More than 12,000 pieces of debris have been located in Texas and Louisiana, including what appears to be a hatch door with a hydraulic opening and closing mechanism that was found Sunday. O'Keefe said the debris will be transported to Kennedy Space Center starting this week where investigators will attempt to reassemble as much of it as possible, though it won't be easy.

"There is certainly no way we are going be able to reconstruct it. The pieces are just absolutely mangled," O'Keefe said. "It's an awful lot of tangled stuff."

The wing segment and landing gear compartment door found in Texas have captured the attention of engineers because they could have been near areas where the shuttle registered a rapid temperature rise during the last minutes of flight Feb. 1.

Gehman declined to say Sunday if the wing was from the left or right side and said he didn't know which side the landing gear door came from.

Mission Control received data from Columbia that showed a sudden rise in temperature in the left landing gear compartment and along the left side of the fuselage. The data also shows that there was increasing wind resistance from the left wing, forcing the autopilot to rapidly move control surfaces and fire jets to maintain stability. The craft seemed to be losing the control battle, engineers said, just before all communications with Columbia stopped.

NASA's shuttle missions are on hold now, but O'Keefe said Sunday that the agency is still preparing to resume flights as soon as the cause of Columbia's breakup is determined and any shuttle flaws are fixed. "We've still got folks aboard the international space station," he said.

(China Daily February 10, 2003)