A dozen or so officials gathered in a windowless room at the Johnson Space Center in Texas on the 12th day of the space shuttle Columbia's final flight for a routine briefing about its progress. The mood in the room was relaxed and upbeat, despite nagging questions about some insulation that apparently had peeled off a fuel tank and struck the left wing 81 seconds after launch, causing undetermined damage.

It was at this 6 a.m. meeting when a flight engineer named Ralph Roe told the assembled officials -- including a representative of the crew and an official charged with policing the flight's overall safety -- that some NASA engineers and contractors had modeled the potential damage to the wing's heat-deflecting tiles in their computers.

Their conclusion was that even if several tiles had been destroyed or deflected off the orbiter, and the shuttle's underlying aluminum skin had been dented, the skin would not burn through in the 3,000-degree temperatures of reentry into Earth's atmosphere. There was, Roe said, no "safety of flight issue."

NASA said yesterday that it still believes this judgment was correct, and that something else must have caused the shuttle's destruction. That may well turn out to be the case, but a review of the initial deliberations, including interviews with informed NASA officials, showed that mission management discussions of the problem -- while the shuttle was still aloft -- were brief and relatively unquestioning.

No one in the room during that early morning meeting dissented, said several NASA officials privy to the deliberations. No one asked detailed questions about the modeling or asked to look at the raw mathematical data. No one at the meeting called for more review, and no one called for independent consultations with tile or insulation experts who did not work for NASA or its contractors. No one from NASA's headquarters in Washington was at the meeting, although someone might have been listening via an audio link, officials said.

About five minutes after the conversation began Jan. 27, "the issue was dispensed with, basically," said a participant who asked not be named. Roe "had worked it to its conclusion, and we felt it was done."

When the meeting ended after about 30 minutes, a one-paragraph summary was widely disseminated. But from that moment until the 16th day of the flight, Feb. 1, when the Columbia disintegrated while reentering Earth's atmosphere, no NASA managers met again to discuss the problem.

Most post-flight speculation about the crash, both inside and outside NASA, has focused on the insulation that struck a glancing blow against the left wing just after launch. NASA is looking at a range of other hypotheses. One that has received fresh attention in the past few days is that the shuttle might have been struck during flight by space junk, something unrelated to the insulation, which also might have helped disable its heat protection during reentry.

But the fervor with which the agency and members of an outside commission have reexamined the tile-insulation hypothesis contrasts with the unhurried pace and more narrow scope of the agency's internal review during the flight.

Michael Kostelnik, a top shuttle official at NASA headquarters, has described the agency's in-flight review as "a very stringent process drawing on the past experience, models and a lot of different activities." On Feb. 2, Ron Dittemore, the Houston-based shuttle program manager, said: "The technical community got together, and, across the country, judged" the impact.

NASA declined yesterday to permit an interview of Roe or of Mark Erminger, a safety official present at many of the team's meetings. But the agency said in response to questions that the management team discussion was brief, and that "no more than three or four people worked on the actual engineering analysis of the foam-tile impact." It said these were all NASA employees and contractors, including engineers from Boeing and United Space Alliance. NASA's top space flight official, William Readdy, later clarified this statement, saying these individuals at Johnson Space Center were supported by many other engineering experts.

Kostelnik said Tuesday that he was "not privy to details" of the analysis during the flight. He also said Dittemore "didn't participate in those particular things," although he was informed of the judgment "of those teams of professionals."

Alex Roland, military historian at Duke University and a frequent critic of NASA, cautioned against leaping to conclusions but said his impression is that the shuttle program is operating "in the same institutional atmosphere" that led to the 1986 Challenger explosion, that of "a program constantly stressed by problems you couldn't afford to fix."

Days before Columbia's liftoff, the craft and its crew passed through a rigorous Flight Readiness Review at the Johnson Space Center in which top officials from the Kennedy and Marshall space flight centers joined in certifying the craft's readiness for launch.

Space shuttle officials take this procedure seriously: NASA rules bar a liftoff unless the heads of these centers or their designees sign the bottom of a five- or six-page document summarizing the resolution of all safety-related issues raised during servicing of the orbiter and final flight preparations.

No mention was made in the Columbia's final readiness review -- or in two previous flight-readiness certifications for planned Columbia flights in August 2000 and March 2002 that were deferred at the last minute -- of any concern about tiles or insulation on the external tank, according to two people who saw it.

The document instead focused on last-minute analysis of a series of cracks found in a ball joint near fuel lines for the main engines of other shuttle orbiters, besides the Columbia; it concluded that even if the same cracks had recurred in the Columbia, they posed no risk to the flight.

So the flight was cleared without any new discussion of a problem that arose on at least two previous space shuttle flights -- the shedding of some insulation from the external tank during launch -- or the loss of tiles on virtually every previous flight without serious consequence.

The readiness review did examine the risk that the orbiter or its crew would be lost to a strike by orbital junk -- the potentially deadly detritus of man's past forays into space, including bits of metal, solder and computer chips from hundreds of exploded boosters and satellites. Since space debris started building up, NASA has worried about it.

Because of the orbiter's extraordinary speed in space -- on average 17,000 mph -- a strike by a grain of sand packs as much energy as a bowling ball traveling 60 mph on the ground. A strike by junk the size of a pea is comparable to being struck on Earth by a 600-pound safe traveling 60 mph. NASA has taken great pains to minimize the risk of a debris strike in orbit, including reversing the craft so that the crew compartment is protected by the tail.

The risk estimates are based on complex modeling of the prevalence of debris in or near the path of the orbiter throughout its flight, and draw on orbital debris data catalogued by NASA and the Air Force. In this case, the Final Readiness Review declared there was 1 chance in 370 that either the crew or the orbiter would be lost because of a strike by space junk -- a figure well within NASA's standard safety limits. NASA officials said yesterday that the analysis is being redone as part of the inquiry into the calamity.

The launch Jan. 16, according to a brief internal NASA report three hours later, was "nominal with no problems identified." But within one day, NASA technicians spotted a puffy white cloud as they reviewed film from the 1960s-era video cameras trained on the orbiter, its external fuel tank and its rocket boosters during launch.

The delay in retrieving and analyzing the film means that even if a massive strip of heat-resistant tiles had somehow been peeled off during launch, the crew would not have known about it quickly enough to turn the orbiter back to Earth before it reached outer space and faced a high-temperature reentry.

"No one looks at tiles and insulation in real time during ascent," said a NASA engineer who asked not to be named. "There is no way to know about any [tile-related] problem before it reaches orbit." As a result, NASA is betting heavily during each launch that the tiles, which are essential for the craft to return to Earth, will remain almost completely intact.

The video of the Columbia clearly showed something, presumably insulation surrounding the tank, striking the leading edge of the orbiter's left wing, which is protected by a special carbon coating. The discovery was first briefed Jan. 18 to the mission management team of flight controllers at Johnson Space Center -- those who direct the crew from the ground and are responsible for every aspect of the shuttle's operations after liftoff.

Each of the team members is responsible for different sub-systems on the orbiter, such as propulsion, guidance, electrical or environmental control equipment, as well as "flight integration," the payload -- in this case a large scientific experimentation capsule -- and the mission's overall safety. Most are senior members of the shuttle programs, and on this particular flight, there were no newcomers in the room.

"This is a group of engineers working in concert with each other," said one official, noting that almost everyone on the Columbia's mission management team had been together for at least 20 missions. "No one from the outside world is involved. This is all the home-grown-operations world," with highly specialized experience, another official said.

Although Columbia's crew was kept busy performing dozens of scientific experiments, the mission was not considered particularly taxing on the ground. And so the overall management team decided not to meet over the three-day weekend between Jan. 19 and 21, despite the uncertainties associated with the insulation strike on the tiles.

They were not briefed again on the issue until they returned, when Roe reported that his team had studied the videos frame by frame, only to conclude that "the resolution . . . is insufficient to see individual tiles." But Roe also said the reflection of light from the tile surface after the dissipation of the puffy cloud provided "no indications of larger-scale damage." Roe and deputy chief flight director Linda Ham agreed not to request a visual inspection of the tiles by Air Force ground cameras because the cameras often provide poor images and the risk was considered low.

More modeling was then performed by Roe's engineers, culminating in the briefing Jan. 27. Any member of the team could have requested a more exhaustive examination of the problem at that meeting, four NASA officials said, but none did.

Staff writer Kathy Sawyer contributed to this report.

A piece of what is believed to be the underside and rear of the left wing of the space shuttle Columbia lies against a fence near Nacogdoches, Tex.