NASA engineers at the Langley Research Center in Virginia expressed frustration during and shortly after the space shuttle Columbia's disastrous flight that the space agency failed to investigate thoroughly whether it had been seriously damaged by debris during launch, according to a series of internal e-mails disclosed yesterday.

In one e-mail sent three days before the shuttle disintegrated on Feb. 1 as it hit the atmosphere, a Langley engineer complained that those managing Columbia's flight had chosen not to do simple studies to clarify what risks it would face on landing, and had treated such information "like the plague."

The engineer, Robert Daugherty, also said he understood that NASA engineers at the Johnson Space Center in Houston who were estimating the chances that the shuttle could land safely had "have used words like they think things are 'survivable' but 'marginal.' "

The e-mails made clear that during the flight some NASA engineers were more worried about the shuttle's safety than previously known, and that Daugherty in particular thought his concerns were not being heard.

"We can't imagine why getting information is being treated like the plague," he wrote to his boss, Mark J. Shuart, on Jan. 29.

"I am advised that the fact that this incident occurred is not being widely discussed," Shuart then wrote as he forwarded Daugherty's e-mail to the Langley center's acting director.

Daugherty, a 20-year NASA engineering veteran who specializes in landing dynamics, described engineers in Texas as being particularly worried that foam debris that broke off the shuttle's huge external fuel tank during the Columbia's launch had gouged the heat shielding on the door covering the main landing gear in the left wing.

This is the area where the Columbia's sensors later registered an unprecedented temperature increase during reentry as, investigators now believe, superheated air penetrated the wing and destroyed some of its wiring, causing the sensors to fail altogether. The shuttle broke up within 15 minutes.

After the disaster, officials at the Johnson Space Center said they believed the heat shield had experienced at most minor damage and that they had no reason to worry about the shuttle's integrity or the safety of its seven-person crew. The agency has stated it still does not know why the shuttle broke up, killing everyone on board. The disaster is under investigation by a NASA-appointed commission that has already interviewed Daugherty and plans to do so again, according to a source close to the commission.

"Do we think there's something to that [information]? I think there is," the source said on condition he not be named. "We're looking at that seriously."

The source said the allegations that the potential damage had been inadequately studied were important enough for the board to initiate its own analysis of the debris strike and to investigate how the engineers in Texas handled the information they received from others -- including those at Langley. The probe will also look at the validity of the engineering models and simulations on which the Houston engineers' conclusions were based, the source said.

NASA administrator Sean O'Keefe, who ordered the e-mails released yesterday in response to a request by The Washington Post, made no direct comment on their content. During an interview Wednesday, he reiterated that the agency had no warning of imminent disaster during the Columbia's flight, either from the shuttle sensors or from Boeing's studies of the debris.

"What you are seeing is an inside discussion of 'what ifs' regarding the accident," said NASA spokesman Robert Jacobs. He said it would be up to the commission to decide if the concerns expressed by the authors -- all in messages dispatched from one Langley official to another -- were warranted.

NASA engineers in Texas had consulted Daugherty during the flight because they were worried -- for reasons not fully explained -- that one or more of the shuttle's tires might be deflated during the landing. Daugherty gave them a telephone briefing and then sent a lengthy e-mail spelling out his concerns about excessive heating of the shuttle's wing, which NASA released last week.

NASA officials said yesterday they are unsure if this was the only e-mail communication between the Langley and Johnson centers during the flight. They also said they were not prepared to release copies of e-mails written at Johnson in response to Daugherty's warnings. Nor were they willing to make available any of the officials who wrote or received the e-mails.

The messages released yesterday represent only a small portion of the documents in NASA's hands that relate to the debris strike analysis, according to several officials.

NASA has said that none of these e-mails was seen at senior levels within the agency before the reentry. Flight controllers in Houston discussed the debris issue three times during the 16-day flight and dismissed all related safety concerns during a five-minute discussion on the 12th day, well before the Langley-Johnson dialogue had concluded, according to participants.

Other e-mails and documents released yesterday revealed that after the flight, two other Langley engineers privately expressed alarm that the NASA officials improperly dismissed any possibility that debris known to have struck the shuttle's left wing might have included ice, a heavy material that could conceivably have breached the shuttle's outer skin of glass tiles and reinforced carbon fibers and allowed superheated air to penetrate the left wing.

The board has said it does not know why the hot gas penetrated the wing.

In the studies analyzing the debris impact during the flight, all performed for NASA by the Boeing Corp., the company's engineers assumed that all debris that struck the wing consisted only of foam, a light material that would likely cause less damage to the wing than ice. They made this assumption despite being unable to determine the origin of the debris because of a malfunctioning camera at the launch pad.

NASA had said since the disaster that only one piece of foam struck the shuttle. The agency acknowledged this was incorrect yesterday as it released a Jan. 24 Boeing study that cited strikes by three pieces of foam.

The Johnson center's spokesman, James Hartsfield, said, "it was an advertent error on our part, an unfortunate one." He said NASA's studies during the flight had used the largest piece to determine the "worst-case" impact, and that in new studies it is reexamining whether this was sound.

The question of whether the debris consisted of foam or something else, such as ice or metal, is important because of their significantly different densities. Boeing's Jan. 21 report concluded that if the debris were foam, then the expected degree of damage to the shuttle's heat-protective tiles was low enough to predict a "safe return." But the report also noted that "minor variations in total [impact] energy . . . can cause significant tile damage."

One Langley engineer, Daniel Mazanek, told a manager in an e-mail on Feb. 7 that if the debris was mostly ice -- from water that could have accumulated during the shuttle's five-week waiting period on its Florida launch pad and then shaken loose during launch -- it would weigh 25 times what NASA had assumed, and its energy "would be the equivalent of a 500 lb. safe hitting the wing at 365 miles per hour."

Mazanek's e-mail asked whether "this type of impact [could] carve out a significant channel in the protective tiles?" The engineer, who could not be reached yesterday, said in his e-mail that his analysis was based on the same video footage used by Boeing and by others at the Johnson Space Center.

Boeing Corp. spokesman Ed Memi said it was NASA's decision to land the shuttle after hearing its engineering analysis. "We did our job," he said. A week ago, Boeing Vice President and General Manager Michael Mott said, "In some of the reports I've read, there's an implication that this is done in some sort of a cavalier manner. There's nothing further from the truth. We treat this with the utmost diligence and respect. . . . It's a big, big deal."

The Jan. 24 Boeing study suggests that as the company's engineers refined their analysis over a period of several days, they concluded the debris could have struck closer to the wing's leading edge than they initially believed. Any impact there could be more damaging than farther back on the wing, because the angle of impact would be steeper.

Another Langley engineer, Dennis Bushnell, told senior officials at the research center in an e-mail dated Feb. 5, after the disaster, that "we should have done more analysis of the whole situation . . . [and] taken it more seriously." Bushnell also said the agency should have done more to "minimize ice impingement" during launch, because the resulting damage to the shuttle's protective skin could conceivably alter the flow of air over the wing during reentry and cause grievous harm.

Bushnell wondered in particular why NASA never redesigned a "dump line" that runs down the side of the external fuel tank. The pipe had a history of accumulating ice before lift-off and runs down the left side of the tank, he noted -- the side closest to the orbiter. On several previous flights, Bushnell wrote, ice damage to the wing "traced to . . . the external tank dump line" had been "quite severe."

"Why this dump line was not repositioned to the other side of the tank away from the orbiter, I do not understand," he wrote. Asked about the dump-line issue yesterday, NASA spokesman Hartsfield said: "I'm not aware of any previous focus on the dump line."

Some analysts have speculated that the debris' impact might have been aggravated by unusual roughness on Columbia's left wing. A former astronaut who flew five shuttle missions, Robert L. Gibson, told reporters this week that he had warned NASA about the wing's roughness, but his warnings were ignored.

Staff writers Rick Weiss, Don Phillips, Kathy Sawyer and Eric Pianin contributed to this report.

Technicians at Kennedy Space Center in Cape Canaveral work with wreckage from the shuttle Columbia, collected after the Feb. 1 disaster.