Twenty months after the Columbia tragedy, NASA is still struggling to overcome serious engineering hurdles to resume flying its space shuttles in March, a task that has become even tougher because of hurricane-caused work stoppages and damage to critical facilities.

NASA spokesman Allard Beutel said the Kennedy Space Center stopped work at 6 a.m. Friday and would be closed today while a 200-member damage assessment team inspects the facility.

Beutel said the team remained hunkered down at Cape Canaveral yesterday while residual wind gusts from Hurricane Jeanne whipped through the complex.

Beutel said a preliminary look around suggested that the three remaining shuttles, closed up and protected with sand bags in anticipation of the storm, "are all okay."

But the center's famous Vehicle Assembly Building, which lost 850 side panels during Hurricane Frances earlier in the month, had more panels torn away by Jeanne late Saturday.

Top NASA officials will meet in about two weeks to assess the progress on crucial shuttle modifications in light of the hurricane-caused delays. They will then decide whether space shuttle Discovery's projected March 16 launch can go forward. The next launch window will not open until early May.

Before flights resume, NASA Administrator Sean O'Keefe has insisted that the agency fully comply with 15 recommendations made by the Columbia Accident Investigation Board last year. As of mid-September, NASA had "conditionally closed" five, but it hopes to finish all of them by year's end.

Whether this will happen remains uncertain, and neither NASA nor the agency's Return to Flight Task Group, a panel of experts charged with monitoring progress on compliance, is radiating optimism -- perhaps in keeping with O'Keefe's often-expressed assertion that the shuttles' return to flight will be "milestone-driven" rather than "schedule-driven."

"We believe that NASA has made significant progress in many areas," retired Lt. Gen. Thomas P. Stafford, co-chairman of the task group, said at a Sept. 8 Senate hearing. "At the same time, we believe that the agency continues to face significant challenges and has considerable work ahead of it in some areas before it will be ready to return the shuttle to flight."

Most of the critical return-to-flight recommendations derive from the need to cope with the tendency of the external fuel tank, filled with liquid oxygen and hydrogen, to shed pieces of foam insulation during launch. Investigators concluded that a foam-caused gouge in the "reinforced carbon-carbon" heat-shielding on the leading edge of Columbia's left wing led to the orbiter's disintegration during reentry on Feb. 1, 2003.

NASA has adopted a series of fallbacks to deal with the problem, based on the reality that engineers will never be able to eliminate foam-shedding entirely. The agency focused first on reducing the size and number of debris fragments, then on developing inspection and onboard repair capabilities and, as a last resort, on using the international space station as a "safe haven" where a shuttle crew could wait for a rescue flight if the shuttle was disabled.

Joseph Cuzzupoli, leader of the task group's technical panel, responding to questions by e-mail, said that he did not see any "technical show-stoppers" that would permanently ground the shuttle but that "there is still a significant amount of work ahead for NASA" in areas relating to the external tank.

At NASA's Marshall Space Flight Center in Huntsville, Ala., engineers have removed large foam "ramps" from the V-shaped strut that links the external tank to the orbiter and installed a heater to keep ice from forming. It was a piece of bipod foam that caused the critical damage to Columbia.

Marshall engineers have also developed new techniques for applying foam manually to other parts of the tank. In all cases the wish is to minimize small "voids" in the foam shielding where ice can form and subsequently pop out the insulation during a high-stress launch.

NASA still has no way to detect voids without damaging the foam. "We're developing techniques, with very promising results," said Marshall's Neil Otte, NASA's chief engineer for the external tank, but these will not be ready for a March return to flight.

"This is not an easy problem," Otte said in a telephone interview. "We're taking a system that was designed 25 years ago and we're defining and applying new requirements. What you would like to do is set out your requirements, and then engineer the system."

Because there is no fail-safe method to eliminate debris, the investigation board recommended that NASA develop onboard inspection and repair capabilities for both the ceramic tiles that insulate the belly of the orbiter and the leading edge material on the wings.

Delays in developing a boom to inspect the orbiter's underside caused O'Keefe earlier this year to move the projected return to flight from its original date this fall to the current March time slot. The boom "is right on schedule now" for a November delivery date, Space Shuttle Program Manager William W. Parsons said in a telephone interview. "We will be able to inspect all the areas of the orbiter that we will need to look at."

The task group's Cuzzupoli agreed that the boom no longer presents "significant technical challenges."

Prospects are not as bright, however, for onboard repair. O'Keefe cited the shuttle's inability to make its own in-flight repairs as the main reason for canceling a mission to do maintenance on the Hubble Space Telescope.

Instead, the shuttle will fly only to the international space station, where it can use the station's robotic arm to help with any needed repairs. Engineers at the Johnson Space Center have developed an applicator for the caulking that astronauts would use to patch the ceramic tile, and engineers at the Langley Research Center, in Hampton, Va., are testing the caulk.

Repairing the reinforced carbon-carbon, however, is another matter. Developing onboard capability to repair any size damage is "years out," Parsons said. But "when it comes to smaller holes and cracks, we have a lot of optimism."

The hope is to develop a gun that astronauts would use to spread a special caulking material across a crack or small hole in the shielding's carbide coating, so when the orbiter's leading edge heats up during reentry, the caulk would melt into the crack. "It's still in the development phase, and I couldn't tell you how far we'll get," Parsons said. "I can't tell you if we'll have a certified repair capability."

Failure to plug a large hole in reinforced carbon-carbon would risk a repeat of the Columbia disaster, so NASA, apart from the investigation board recommendations, is exploring the possibility of allowing a shuttle crew to seek "safe haven" aboard the space station for some as yet unspecified time period until a second shuttle could pick them up.

In June, a pair of NASA reports painted a grim picture of the station's ability to support a seven-member shuttle crew and the two-member station team long enough for a shuttle to reach them.

The picture deteriorates further if return to flight slips past the current March date, because the space station's dwindling supplies and support capabilities will be further depleted as time passes.

Parsons said NASA has been deliberately "erring on the pessimistic side" and will seek to "maximize our capability" in coming months by shipping equipment and supplies to the station aboard Russia's relatively diminutive Progress cargo spacecraft.

"We have people that are looking at this, and as we get closer to launch they'll sharpen their pencils more and more," Parsons said. "We'll take a look, and the reality will be what it will be."

A bulldozer sits atop debris from the Kennedy Space Center's Vehicle Assembly Building at Cape Canaveral, where hurricanes Frances and Jeanne tore panels from the building.