Researchers have begun initial test runs firing external fuel tank foam insulation at a shuttle landing gear door in a bid to calibrate damage prediction software and to assess how much damage high-speed impacts might actually do to a shuttle's heat-shield tiles. Engineers ultimately plan to fire foam debris at a mockup of the shuttle's wing leading edge system, the location of the breach that doomed Columbia.

The large compressed-gas gun used for foam-impact test calibration at Southwest Research Institute in San Antonio, TX. Credit: CAIB Photo by Rick Stiles 2003

In the initial runs, firing foam from a nitrogen gas cannon at the Southwest Research Institute in San Antonio, Texas, only minor damage was observed in the landing gear door tiles, sources said.

Columbia's left wing was struck by falling external tank insulation 82 seconds after launch Jan. 16. Investigators suspect the impact damaged the leading edge of the wing in some fashion, directly causing, or at least contributing to, the creation of a breach that destroyed the ship during re-entry on Feb. 1.

The Columbia Accident Investigation Board is overseeing a complex series of tests at the Southwest Research Institute that will culminate early next month with firings aimed at a highly-instrumented mockup of a wing leading edge.

In three of four initial runs this past week, tiles on a landing gear door taken from the prototype shuttle Enterprise withstood foam impacts at angles of 5 degrees with little or no damage, sources said. The foam samples were not as large as the piece of debris that hit Columbia, but engineers can achieve the desired results by properly varying mass and velocity to achieve the range of kinetic energy imparted in the actual launch impact.

The fourth run in the series, in which the canon fired a foam sample at an 8-degree impact angle and a slightly higher-than-planned velocity, caused visible but relatively minor gouging. The foam that hit Columbia's wing was moving between 416 and 573 mph, engineers believe.

CAIB members have stressed that the landing gear door target was selected early on, before engineers knew the breach occurred in the leading edge, and that these results are not directly applicable to the board's current failure scenario. The real test will come next month, when investigators begin firing foam at the leading edge mockup.