HOUSTON (AP) _ Columbia accident investigators said Tuesday they are close to zeroing in on where a hole opened up in the spaceship's left wing and strongly suspect the fatal blow was caused by a chunk of flyaway foam at liftoff.

``We're closing in,'' said Roger Tetrault, a member of the Columbia Accident Investigation Board, who added that panel is within 30 inches of knowing exactly where the breach occurred.

A fragment of a seal or panel along the vulnerable leading edge of Columbia's left wing is almost certainly what was missing when Columbia descended through the atmosphere three months ago this week, the board said. This missing mystery object floated away two days into the doomed flight, unnoticed by the crew or ground controllers.

Tetrault said he and colleagues have narrowed down the location of the deadly breach to either part of a reinforced-carbon panel or part of one of the carbon seals on either side, just a little more inboard than previously thought. All the data and debris seem to support this position on the underside of the wing, and further analysis should pinpoint it further, he said.

The top portion of the panel in question _ and possibly part of the two seals as well _ have been found and show extreme heat damage. The sequence of failing sensors and burning wires during Columbia's descent also points to this area.

The gap created by the broken seal or panel let in the scorching gases of atmospheric re-entry and led to the shuttle's breakup over Texas on Feb. 1. All seven astronauts were killed.

``My hope is that we'll be able to get to the bottom of this and be able to say, with some precision, where the hole is and what size the hole is,'' said Tetrault, a retired corporate executive who worked with nuclear submarines.

Tetrault said the hole was at least 100 square inches, and ``the weight of the evidence right now is fairly solid'' it was created by a suitcase-size piece of foam insulation that ripped off Columbia's fuel tank shortly after liftoff in January. It's ``highly likely'' that the foam, the mystery object on flight day two and the catastrophe are all related, he said.

``If things keep progressing the way they are, I think you'll be able to say that it would be highly unlikely not to be'' the foam, he added. ``But I don't think there's anybody on the board who's ready to jump and say at this particular point that it is certain that the foam created the breach.''

The chairman of the board, retired Navy Adm. Harold Gehman Jr., said he hopes to have a working hypothesis in the next one or two weeks, based on ``mountains of data and mountains of debris.''

Nearly 40 percent of Columbia has been found _ 78,000 pieces in all _ and the intensive search in Texas is wrapping up.

While much of the focus in recent weeks and months has focused on temperature, pressure and strain measurements collected during Columbia's final minutes, the board has just initiated an inquiry into similar data gathered during launch. Many of these readings come from an onboard recorder that was found on a muddy Texas hillside 1 1/2 months after the disaster.

Gehman said none of the launch data ``leaps out at you as being startling'' but noted that some readings show ``little squiggles'' in pressure, strain and temperature variations that require more scrutiny.

``It's really premature to speculate if there's anything there,'' Gehman said. ``We really are doing this from the point of view of not missing anything, rather than we smell a rat.''

Also still hanging is a series of foam impact tests at Southwest Research Institute in San Antonio. The tests will begin this week with shuttle thermal tiles and lead up to the most critical targets _ real shuttle wing panels and seals _ in June. Chunks of foam will be hurled at these objects around 450 mph, the speed at which the launch debris is believed to have struck the edge of Columbia's left wing.

___

On the Net:

Columbia Accident Investigation Board: www.caib.us