Computer analysis and the results of a dramatic test last week indicate a critical experiment on tap Thursday may help investigators determine, to the satisfaction of most, that a foam strike during the shuttle Columbia's launching was the initiating event that led to the ship's destruction. While the root cause will never been known with 100 percent certainty, engineers believe Thursday's experiment could go a long way toward eliminating any lingering doubt about what damaged the leading edge of the shuttle's left wing, creating the breach that ultimately doomed the ship and its crew.

Foam strikes a mockup of the space shuttle wing leading edge system during a test last week. Credit: CAIB

"From my perspective, and I'm not speaking for the board now, I think it moves us a lot closer toward a very likely initiating event," said Scott Hubbard, a member of the Columbia Accident Investigation Board overseeing the tests. "I think it establishes foam can break (wing leading edge panels) in a realistic environment and for my own sake, I think it moves us a significant step toward establishing that is an initiating event. There could have been other initiating events, but I think this moves us a lot closer to saying the foam can do this kind of damage."

Columbia was destroyed Feb. 1 by a breach in the leading edge of the ship's left wing that allowed super-heated air to burn its way into the interior. At issue is what caused the breach in the first place. During launch Jan. 16, a large piece of foam insulation broke off Columbia's external fuel tank and slammed into the underside of the left wing's leading edge at some 530 mph. A NASA analysis during the mission concluded the foam might damage heat-shield tiles, but that it wouldn't cause any safety of flight concerns. The analysis all but dismissed any possibility of significant leading edge damage.

Since the accident, engineers have been trying to pin down exactly what sort of damage might actually have been caused such an impact. They have refined their initial estimates of the mass of the foam, based on enhanced photography and computer analysis, and they now believe it struck the leading edge, while rapidly rotating, at an angle of 15 degrees.

During the test last week at the Southwest Research Institute in San Antonio, Texas, a 1.7-pound piece of external fuel tank foam insulation was fired at a mockup of a shuttle wing leading edge by a nitrogen gas canon. The foam, traveling at some 531 mph, struck a Fiberglass leading edge panel, taken from the prototype shuttle Enterprise, at an angle of 20 degrees. The higher angle was chosen to account for rotational energy in the actual impact that cannot be simulated. The Fiberglass panel, mounted in position No. 6, served as a stand in for the reinforced carbon carbon panels actually used on the space shuttle.

The results of last week's foam impact test shows the T-seal moved out of position. Credit: CAIB

Fiberglass is 2.5 times tougher than RCC carbon composite material and engineers were surprised by the results. The impact lifted and deformed a so-called T-seal between panels 6 and 7 while generating up to seven times the force computer models had predicted. Nothing broke and there was no direct gas path into the interior of the structure, but the higher-than-expected impact force caught investigators off guard.

"I was standing there and I heard all these assertions about well, this only weighs a pound and a half and (the leading edge) is tough stuff," Hubbard said today. "When the gun fires, there's a blast wave that comes out and you can feel it and then I saw the streak and clearly, something happened. I thought to myself, this is more than I had expected. ... I thought oh my God, this is something! This isn't just a light bounce.

"People's intuitive sense of physics is sometimes way off. You don't feel this can do anything. But you fire this at 500 mph and there you saw it, I really did think oh my God, this is really an impact, this is a significant effect. Then when we actually looked at it and saw things spread apart and all that, that's when it really came home to me what one-half MV squared really means."

He was referring to the formula for computing kinetic energy. While the mass of the foam is low, the velocity is very high and that is the term in the equation that is squared. The impact force delivered by the foam is roughly equivalent to catching a basketball moving at 500 mph.

Hubbard and his team originally planned four foam shots at Fiberglass panel No. 6 before shooting at a real RCC panel taken from the shuttle Discovery that is similar in age to the panels making up Columbia's leading edge. But given the higher-than-expected forces observed during last week's test, engineers have decided to forego any additional shots at panel 6 to make sure they don't inadvertently damage any critical support hardware. Instead, they plan to shoot at Discovery's RCC panel Thursday.

"Because the observed effects, the fact that the peak loads were higher than expected in panel 7, the fact that we deformed permanently the T-seal, the fact that the loads in panel 6 were in some cases a factor of seven higher than anticipated, all led us to believe we probably had learned as much as we needed to and that we should proceed to the RCC," Hubbard said.

Thursday's test will use an identical piece of foam as the one fired last week. The foam will hit at the same velocity and impact angle. Computer models, adjusted with a correction factor to account for the different strengths of Fiberglass and RCC carbon composite material, indicate Thursday's impact will deliver more than enough energy to break the RCC panel.

"Those predictions would say that we have by maybe 70 percent more force than we need to break the RCC," Hubbard said. "Now, whether it actually turns out that way or not, that's why we do the experiment. But the analysts are saying it looks like it'll break it."

If it does, or if it causes any remotely significant damage, investigators will have solid evidence the launch day foam strike could have been the root cause of the Columbia disaster. Not absolute proof, but solid evidence.