Inadequate methods of applying and repairing foam on the space shuttle's fuel tank probably contributed to the dangerous loss of a chunk of the insulation during Discovery's (search) launch 21/2 months ago, a NASA (search) investigation team concluded Friday.

So much work is needed to understand the problem and correct the deficiencies that shuttle flights are on hold until at least May, and possibly even next summer.

NASA's Richard Gilbrech, head of the investigation team, said no one may ever know exactly what caused a 1-pound, 3-foot section of foam to break off Discovery's external fuel tank in late July. It was a scary repeat of what happened during Columbia's (search) doomed flight in 2003.

Gilbrech's team suspects workers inadvertently crushed the foam while conducting repairs in that area, or handled it in such a way that resulted in damage. The tank was worked on considerably more than previous ones because of all the post-Columbia modifications.

In addition, thin lines may have been introduced into the foam when it was sprayed onto the tank, weakening the material.

"We don't think in and of itself crushed foam alone could have been the cause of the foam coming off," Gilbrech said. "We believe it was potentially a combination" of things.

The investigation team found no evidence of negligence, said Gilbrech, who is deputy director of NASA's Langley Research Center (search) in Virginia. Workers followed procedures, "it's just we didn't really have an appreciation for the significance that this handling damage could have."

Numerous tests are planned in coming months to establish just how sensitive the foam is and whether it can be easily crushed by workers. Foam is about as well understood right now as steel was during the Industrial Revolution, Gilbrech said.

Another possibility is that the one-inch layer of underlying foam may have cracked because of thermal stresses, causing the layer on top to pop off, Gilbrech said.

"We're attacking all fronts and trying to learn as much as we can, but I don't know that we'll ever pinpoint one of those potentials as the root cause," he said.

NASA already has introduced new techniques for applying foam and is limiting workers' access to vulnerable areas.

Some redesign work will be required in the spot where foam came off Columbia and resulted in a fatal blow to the wing. During Discovery's liftoff, an 8-inch piece of foam broke off that same area. In all, worrisomely large foam chunks flew off in five spots.

Shuttle program manager Wayne Hale said it is too soon to set a launch date for the next shuttle mission but cited May as a possibility. It depends on progress made at the fuel tank assembly plant in New Orleans, which was in the path of Hurricane Katrina.

It will be December before the factory's entire 2,000-person staff is back on the job, Hale said. Only 500 are now working.

Space operations chief Bill Gerstenmaier said 18 more shuttle flights are planned to the international space station and possibly one to the Hubble Space Telescope before the fleet is retired in September 2010.