With Discovery safely back on Earth, NASA's engineers will turn their full attention to figuring out why a menacing hunk of foam insulation broke away from the shuttle's external fuel tank despite 2 1/2 years of work to fix the problem -- the same one that doomed the shuttle Columbia.
NASA Administrator Michael D. Griffin, who has grounded the shuttle fleet until the problem is solved, says he hopes engineers can fix it without postponing the planned launch of the shuttle Atlantis, now six weeks away.
"I think we are going to fix it in short order, and we are going to get back flying," he said late last week. "We don't expect this to be a long, drawn-out affair, to be honest with you."
If Atlantis does miss the September launch window for reaching the International Space Station, the next one opens in November. Griffin said the space agency can still meet one of those windows "by being smart and working hard." While the delay might stretch into next year, he said, "We don't start out by assuming that we can't succeed."
But John Logsdon, director of the Space Policy Institute at George Washington University, discounted that optimism as "rallying the troops."
No one can say how long the delay might be until engineers begin to report on their findings, perhaps later this week, he said.
While engineers wrestled with the foam problem, shuttle Commander Eileen Collins flew Discovery to an on-time landing at 8:11 a.m. EDT, yesterday. The 14-day mission ended at Edwards Air Force Base in California after poor weather in Florida forced NASA to delay and finally cancel landing attempts there Monday and yesterday.
NASA quickly prepared the shuttle for the ride back to Florida atop a modified Boeing 747.
In orbit, Discovery's seven-member crew resupplied, repaired and replaced hardware on the station, whose operations had been curtailed by the grounding of the shuttle fleet.
Two spacewalking astronauts also successfully tested technologies for repairing the shuttle's thermal tiles and panels in case of damage during launch.
And, in an unprecedented and unrehearsed final spacewalk, astronauts Stephen Robinson and Soichi Noguchi plucked dangling cloth "gap fillers" from between thermal tiles under Discovery's nose, eliminating a potential re-entry hazard.
NASA plans to fly its three remaining shuttles until 2010 in order to complete the space station. By 2014, the work will begin to shift operations to an as-yet-undesigned Crew Exploration Vehicle, along with a mix of unmanned spacecraft.
Griffin has rejected as "overreaction" the view of critics that the shuttle is a white elephant that should be retired early.
Despite at least three foam failures on Discovery that exceeded pre-flight limits, he insisted that the mission as a whole was remarkably clean.
Detailed in-flight surveys revealed only about 25 "dings" in the spacecraft's heat shield surfaces, one-sixth as many as the previous average of 145 per flight.
"In the world of engineering," he said, "we did pretty well."
Although he has led NASA for less than four months, Griffin took responsibility for failing to solve the foam problem. But he added, "All we can do at this point is move forward."
At his direction, the space agency has formed several engineering teams to investigate the insulation problem, with help from the tanks' builder, Lockheed Martin. They will report to space station program manager Bill Gerstenmaier, who was to get his first briefing yesterday.
A news conference to discuss their progress could come as early as tomorrow.
Thanks to increased surveillance, engineers have far more data about foam shedding from Discovery than they had from the incident that punctured Columbia's left wing during launch. That damage led to the loss of the shuttle and its crew during re-entry on Feb. 1, 2003.
Because of the Columbia disaster, more than 100 cameras documented Discovery's July 26 liftoff. One of them captured video of the largest slab of insulating foam as it peeled off the tank and blew away. It didn't hit the shuttle, and experts said it was moving too slowly to do serious damage even if it had.
Gerstenmaier told reporters on Friday that engineers now have "a tremendous chance to learn from this exercise."
The insulation on the 153-foot tanks helps keep the craft's 526,000 gallons of super-cooled liquid hydrogen and oxygen fuel from warming up and expanding in the Florida sunshine. But the foam must withstand enormous physical forces.
Gerstenmaier said tank surface temperatures rise sharply as the super-cold fuel is used up, while outside temperatures plummet as the shuttle climbs toward space.
The exhaust plumes of the shuttle's solid fuel boosters scorch the foam. A shock wave created by supersonic winds batters it as the shuttle accelerates. Temperatures on the launch pad vary by season, and wind conditions aloft vary daily, so no two flights are identical.
After the Columbia accident, engineers tested tank foam in a wind tunnel, in flight on F-15s and in the lab. But the forces are too extreme, too variable and complicated to be tested simultaneously, Gerstenmaier said. So each shuttle launch becomes a unique test flight.
Griffin reminded reporters that after more than 40 years and 145 manned space flights, the program is in its infancy.
"This manned spaceflight stuff is hard," he said. "We are still in the tiptoe-in-the-water stages. The equipment and the technology is not what we would have it be. We are learning our way."
The modifications made since the Columbia accident were not expected to end all insulation shedding. But they were supposed to prevent shedding of anything big enough to threaten the shuttle.
They didn't, Griffin said. "The team almost succeeded, but there are four or five pieces where they didn't, and we need to understand why before we can go again."
Gerstenmaeir said NASA's "tiger teams" will try to identify from photos and video which areas of foam on the tank failed and which didn't.
Although the fuel tanks are standardized and much of their construction is automated, Griffin pointed out that some of the work on each one is done by hand. And each tank, as it's being assembled, sustains minor damage and repairs.
It's possible that the sections that failed on Discovery's tank had been repaired in Lockheed Martin's factory in Michuide, La.
"We will never, as long as we manufacture these tanks or much of anything else ... we will never be able to eliminate the fact that there is some custom work for each tank," Griffin said.
But from engineering studies, he said, NASA hopes to learn which kinds of custom hand work provide good results and which don't. It's also possible that some structures are just more vulnerable to shedding than others.
The largest chunk of foam insulation to fail came from the so-called PAL ramp -- a structure that provides a wind shelter for a bundle of hoses and cables that runs down the tank's surface.
Engineers had considered whether the PAL ramp insulation needed to be modified. After what Griffin called "extensive" analysis and discussion, NASA concluded -- incorrectly, as Discovery proved -- that it was not a significant threat.
"Clearly we wish we had made a different decision," Griffin said.
George Washington University's John Logsdon said the scale of the problem will dictate both how long the fix takes, and the future of the space station.
"I think six months [delay] they could withstand," he said. "More than six months would raise some real questions of what can be done [to complete the station] before the shuttles are retired."
The 2010 deadline for that, he said, "is a real hard deadline."
Lockheed Martin spokesman Harry Wadsworth said the company has been "retrofitting" 11 shuttle tanks to meet post-Columbia, pre-Discovery specifications.
One of those tanks flew with Discovery; two others have been delivered to Cape Canaveral. The other eight are still at the Louisiana factory in various stages of completion.