NASA engineers have conducted computer simulations for years to detect the amount of damage launch debris, such as insulating foam and ice, would do to the space shuttle's delicate heat-shielding skin.

The development of the program, known as the Ascent Debris Trajectory Simulation, reflects engineers' long-standing concern about the kind of debris that fell on Columbia's left wing during its last launch. The simulator was in place as early as 1988.

"It allows you to figure out which tiles were hit, how many were hit and whether they were in critical areas," said Paul Fischbeck, a professor at Carnegie Mellon University and co-author of a NASA-funded report on launch debris and critical mishaps, which used data from the simulator. "Debris has been an issue for a long time."

NASA has not said whether the debris simulator was used last month after NASA engineers reviewing videotape noticed that debris had hit Columbia's left wing. But NASA officials have said that engineering analysis of the debris impact conducted during the fatal mission indicated the spacecraft was safe.

Moreover, Ron Dittemore, the space shuttle program manager, said at a news conference yesterday that investigators had all but dismissed the possibility that a piece of insulating foam or ice from the external tank had caused the fatal damage.

"We don't believe it was the foam," he said. "I don't think it's ice. . . . I don't think this was a chunk of foam solidified with ice."

He indicated that the piece of insulating debris, estimated at 2.67 pounds, was too light, and the air speeds that carried it too low, to cause significant loss of the tiles that form the skin of the orbiter and protect it from the intense heat that occurs during reentry into Earth's atmosphere. As for the possibility that the debris was ice, he said that ice teams had conducted an inspection before the launch and found no cause for concern, and that the insulation was impervious to the rains that had fallen on it as it sat on the launch pad during a particularly rainy December.

Still, the insulating foam has been a source of intense interest for investigators, and a longtime source of aggravation for launch engineers.

The foam, which is manufactured by North Carolina Foam Industries, is applied at a Lockheed Martin facility in east New Orleans, the same place where the massive fuel tanks are made.

The inch-thick layer of foam must meet multiple objectives: It is supposed to keep the 537,000 gallons of liquid fuel inside the tank freezing cold while preventing a buildup of ice on the outside. It must withstand the intense heat of takeoff. At the same time, the foam, which is lathered on like whipped cream but dries hard, must cling to the tank even as it contracts by several inches as the fuel is piped into the tank about six hours before launch.

About 20 NASA investigators have been stationed near the tank plant as part of the investigation, according to a source close to the plant. Investigators have secured "a handful of sites," including parts of the production plant, said Lockheed Martin spokesman Marion LaNasa.

The areas secured include those involved in "installation and application" of foam, he said.

Officials of the foam maker did not return phone calls yesterday. But a message on its Web site said employees were "saddened" to learn of the tragedy.

"We presently have no reason to believe that our product is associated with the Space Shuttle Columbia tragedy. However, we openly pledge our full support to any investigation by NASA," the message read.

NASA engineers have repeatedly made adjustments to the foam layer, hoping for improvements.

After they discovered that gases inside the foam were expanding during the heat of launch, causing the insulation to fall off, or "popcorn," they began poking it with pins to allow the gases to escape without doing damage.

And in recent years, engineers installed a camera on top of the external tank, hoping in part to see how the insulation would perform. It was not successful because debris from the solid rocket booster quickly obliterated the view.

"The pictures lasted just a short time before debris from the SRB hit," said Bill Jeffs, a NASA spokesman. "We got a short, short view, not much."

So even as NASA appeared to be turning its attention away from the possibility that debris from the external tank -- ice or insulating foam -- had caused the tragedy, doubts linger about the launch debris issues that have long worried NASA engineers.

Fischbeck said that once he heard that some debris had hit the wing, he made it a point to listen to the radio on the day of the expected landing.

"I was worried," he said. "I had to listen to it live."

Staff writer Renae Merle in New Orleans contributed to this report.