SAN ANTONIO - A foam projectile fired at the leading edge of a space shuttle's wing yesterday all but destroyed a piece of its thermal armor, blasting a 16-inch hole in the material and providing evidence of what may have doomed Columbia.
"We have found the smoking gun," said Scott Hubbard, a Columbia Accident Investigation Board member who oversaw the test.
The test, which fired a 1.67-pound chunk of foam out of a nitrogen-powered cannon at roughly 500 mph, was aimed at the same spot where insulation that broke off Columbia's external fuel tank smashed into its left wing during its Jan. 16 launch.
Columbia disintegrated over Texas on Feb. 1, destroyed - investigators had theorized - when superheated gases burned through the aluminum frame of the breached wing. The crew of seven was killed.
But until yesterday's test, at the Southwest Research Institute in San Antonio, there had been no convincing evidence that an insulation material not much denser than a Styrofoam beer cooler could breach the reinforced carbon carbon (RCC) that protects a shuttle wing's leading edge.
"I felt surprised at how it appeared, such a dramatic punch-through," Hubbard said. "But it is the kind of damage, type of damage, that must have occurred to bring down the orbiter."
The circular hole of roughly 256 square inches punched by the chunk of test foam brought hoots and gasps of "Wow!" from technicians and other onlookers. Six previous tests had produced far less damage, though one had cracked the protective carbon coating and another had produced a dime-sized crack along a seal that connects two of the panels.
The result seemed certain to raise new questions about whether NASA might have been able to spot a breach if the agency had sought ground-based or space-based telescopes to inspect the shuttle, which agency officials had declined to do.
Hubbard declined to say whether spy satellites might have spotted such damage, but he pointed out that it would have appeared as a black hole in a black piece of carbon sheathing.
Paul Czysz, a professor emeritus of aerospace and mechanical engineering at St. Louis University, said the actual damage to Columbia's wing was probably smaller than the hole seen during yesterday's test. Otherwise, he said, the shuttle would have broken apart earlier.
"Once you had a crack through the leading edge, all the way through like the second test, it was going to fail," he said. "All you're arguing over now is, is it going to burn up over Los Angeles, over New Mexico or over Texas."
Ultimately, he said, Columbia's destruction was a foregone conclusion the instant the foam hit the wing.
The RCC - which also forms the distinctive black nose cone of the shuttle - wards off gases as hot as 3,000 degrees that flow over the wing's leading edge at re-entry into Earth's atmosphere.
But it is not designed to withstand debris impacts. And NASA had never tested its resistance even to small pieces of foam, let alone the briefcase-size chunk that peeled off Columbia 82 seconds after liftoff. It was the largest piece of foam ever seen to come off a fuel tank, though smaller pieces have routinely come off the tank for years.
The investigation board studied launch film to determine that the breakaway piece of foam that hit Columbia weighed 1.67 pounds and was traveling at more than 500 mph when it struck the orbiter's left wing at an angle of about 15 degrees to 20 degrees.
Those same factors were matched for the tests. But the cannon previously was aimed so that a corner of the foam chunk hit the wing mockup. The foam glanced off - but still cracked the leading edge panels and a seal.
During yesterday's test, the cannon barrel was turned so that the 11.5-inch edge of the foam struck the RCC panel directly, delivering nearly a ton of energy.
The bull's-eye also was moved from near the relatively stronger edge of the RCC panel to near the center of panel 8, which at 28 inches is the widest of the wing's 22 RCC panels and the one investigators think was struck during Columbia's launch.
Finally, panel 8 was mounted alongside RCC panels 9 and 10. Investigators had wondered if damage from the foam hit might be magnified by the connections between panels.
Panel 8, for which there is only one spare available for the entire shuttle fleet, was taken from Atlantis and had logged 27 flights. Columbia was on its 28th mission.
Panels 9 and 10, for which there are many spares, were taken from Discovery. That shuttle has had 30 flights, though the two test panels may not have been used for all of the missions.
Together, the panels cost $2.4 million and will not be used again on a shuttle.
What happened was unmistakable: The foam punched through the RCC as if it were made of thin glass.
Hubbard said one reason for the extensive damage yesterday could be the variability between RCC panels, which he said can differ as much as 70 percent in strength when new. On top of that, RCC panels weaken with use and time.
Kevin Spear writes for the Orlando Sentinel, a Tribune Publishing newspaper.