The Columbia Accident Investigation Board today showed video of Columbia's launching that indicates foam debris falling away from the ship's external fuel tank slammed into the lower leading edge of the orbiter's left wing within a few feet of where it merged with the fuselage.

The white flash is the moment of impact as the external tank material strikes Columbia's wing. Photo: NASA video

The board also revealed that 20 seconds before the foam fell away from the so-called "bipod ramp" area of the tank just under the shuttle's nose, Columbia rocketed through unusually high wind shear. The steering system in the shuttle's left-side solid fuel booster swiveled the rocket's nozzle slightly to counteract the effects of the shear, putting some additional stress on the left side of the vehicle.

This event occurred three seconds or so after Columbia endured "max Q," or maximum dynamic pressure. This number varies from flight to flight, but for Columbia, it was around 741 pounds per square foot. Whether that had anything to do with any subsequent damage to the wing is not yet known, but investigators are looking into the matter.

"At 62 seconds on launch, we saw one of the larger transients we've seen on the solid rocket motor," said Maj. Gen. John Barry. "It was well within parameters, but interestingly enough, the two largest ones we've seen on ascent both happen to be Columbia, both happen to be going on 39-degree inclinations (trajectories), both have lightweight tanks. So we're trying to identify if there's any commonality there as an additional stress load on the left-hand side of the orbiter, because it was with the left solid rocket motor that had this input."

Investigators also are looking into pre-launch repair work in the bipod area of the external tank where foam insulation can be seen breaking away 81 seconds after launch. The shuttle's nose is attached to the forward part of the tank by a two-strut bipod. Just in front of each strut, where they attach to the tank, is an incline made of shaved foam that serves to reduce stress and minimize turbulence as air flows over the area during ascent.

The bipod area of Columbia's external fuel tank. Photo: NASA

Because of schedule changes prompted by the grounding of the shuttle fleet last year due to engine propellant line cracks, Columbia's tank was removed from the boosters that originally were stacked for the mission. Those boosters were used for a different flight. Columbia's tank subsequently was attached to a different set of boosters. During this process, engineers wrote up a "problem report" on damage seen near the left bipod ramp area, but the matter was closed out after a visual inspection. The nature of the damage, and what role, if any, it might have played in the subsequent separation of foam from that area is not yet known.

Regardless of what factor or combination of factors led to the initial damage in Columbia's left wing, investigators increasingly believe a breach at or near the leading edge of the wing, near where it joined the fuselage, led to Columbia's destruction. A jet of super heated plasma appears to have burned its way into the interior of the wing and ultimately into the left landing gear wheel well.

The effects of such a plume are extreme. Board member Sheila Widnall, an aerodynamicist at the Massachusetts Institute of Technology, said she discussed such heating with colleagues at MIT over the weekend and "one of the issues we got into was to talk about the thermal environment seen by the vehicle and by its structure and the properties of aluminium at very high temperatures."

"Without it's oxide coating, aluminum is a very reactive material," she said. "And when it's exposed to very high temperatures and high-speed gas flow, the possibility exists of vaporization of aluminium, which could be followed then by a very rapid burning of aluminum vapor. That's obviously a a scenario that we'll want to follow up."

In any case, damage seen in recovered debris suggests the plume exited the wheel well through seals around the landing gear door. At the same time, a vortex of hot air probably formed along the upper surface of the wing, ripping insulation tiles and blankets away as the disaster unfolded.

Increasing aerodynamic drag on the left wing prompted Columbia's flight control system to compensate, first by adjusting the ship's roll trim and ultimately by firing right-side yaw jets in a futile bid to keep the ship's nose pointed in the right direction. Columbia eventually yawed out of control in a sort of tilted flat spin before breaking up high above Texas.

The leading edge of a shuttle's wing is protected from intense re-entry heating by 22 panels of reinforced carbon carbon composite. The panels are numbered from 1 to 22, beginning near the fuselage. Based on an examination of enhanced video shot during Columbia's launch, the foam debris in question slammed into the left wing at or near RCC panels 6, 7 and 8.

"That piece of debris appears to hit on the leading edge, but on the lower half of the RCC," said board Chairman Harold Gehman. "In other words, the debris doesn't appear to go over the top and under the wing. All the debris appears to go under the wing. So it appears to hit the leading edge, but kind of on the underside of the leading edge and then all the debris follows the wind pattern. It appears to hit somewhere (at or near) RCC panel 6, 7 or 8, no closer than 6, no more than 8.

That section of the wing could not be seen by Columbia's astronauts in orbit. It would have been invisible behind the ship's left-side payload bay door.

Columbia in space during STS-107. Photo: NASA

"As you see after the debris goes by (during ascent), there's no missing tile," Gehman said. "Nothing's missing, so once again we have these tantalizing little pieces of evidence that don't fit together."

A wind tunnel analysis shows the loss of a single leading edge panel of reinforced carbon carbon is not nearly enough to explain the unusual and rapidly intensifying aerodynamic forces acting on the vehicle during re-entry. Three to four panels would have had to be missing to explain the re-entry data and no such damage can be seen in the launch footage.

A single blurry image taken by Air Force researchers in New Mexico using a small telescope just a minute or so before Columbia's breakup is undergoing analysis by photo enhancement experts to find out what additional data might be gleaned. The photo appears to show major damage to the left leading edge area, but the details are literally unclear.

This image of Columbia was taken by the Air Force before the shuttle broke apart. Photo: NASA

"Those photo experts are just now beginning their work in liaison with aerodynamics experts and plasma experts to try to see what the picture will really yield," said Maj. Gen. Kenneth Hess, a member of the CAIB. "But the important part right now is to enhance the photo as much as can be done without adjusting the facts behind the photo."

Another factor that has been discussed in recent weeks is a phenomenon known as asymmetric boundary layer transition - the sudden onset of turbulence during hypersonic flight that could have affected Columbia's left wing more than its right.

Widnall said today calculations show an asymmetric boundary layer transition, acting in the absence of any other problems, would have been too small by a factor of four to explain the unbalanced forces that led to Columbia's destruction.

Based on a detailed timeline of Columbia's re-entry, engineers now know problems began long before the shuttle's breakup. Gehman said the question that is most interesting to him is what came first, "in other words, did a thermal event cause loss of tiles and thereby these aerodynamic forces or did we lose an aerodynamic surface first which then allowed the heat to start unzippering this aircraft?"

"Every time I think I've got it figured out, I get a piece of data that changes my mind," he said. "If the aerodynamic surface were disturbed long before re-entry, that then leads you to believe there was some physical misarrangement of the wing that had nothing to do with entry. On the other hand, all of our photographic evidence indicates if there was a physical misarrangement on the wing it was so tiny it's not visible to anybody else, which would lead you (to believe) that it was some kind of a thermal event which was the triggering event which then started the unzippering process.

"And so, the fact that the orbiter was fighting an aerodynamic misarrangement earlier than we previously had thought makes my riddle very intriguing."

For her part, Widnall sees "a heating event followed by a very rapidly developing aerodynamic event."

"The aerodynamic event, when it occurs, is pretty dramatic and I mulled over that for a while and I looked at the flight profile and I realized that the aerodynamic event occurs at the time when the dynamic pressure is undergoing a rapid increase of roughly 30 percent," she said. "Now that makes sense to me, that the vehicle was kind of coasting along with heat damage and then there was a sudden increase in dynamic pressure that came from the (normal flight profile) and that was the onset of the aerodynamic event. That makes sense."

While she declined to speculate on what caused the thermal event, she said "there is a point in time when the guidance and control analysis indicates the onset of a rapidly increasing, I hate to use the word - yaw moment - because it's so technical, but it's the twisting moment on the vehicle."

Said Gehman: "As the vehicle re-enters the Earth's atmosphere, there is a time when the aero forces start going up more rapidly than a straight line. And lo and behold, at that point the orbiter starts to exhibit non-aerodynamic tendencies, remarkably more."

Investigators and board members plan to visit a Lockheed Martin facility next week to learn more about how RCC panels are repaired, serviced and mounted on the shuttle. Investigators are particularly interested in how pockets, or voids, form in the layers making up the composite material due to oxidation. If Columbia had any RCC panels with similar voids, and if the panels were then hit by falling foam impacting at high speeds, a failure could result.

Putting it all together - bipod foam repairs, wind shear, debris impacts, thermal and aerodynamic events - is a complex job. But it may well be a combination of factors doomed Columbia, not any one specific failure.

"What we're really looking at is a complex failure of a complex system," Gehman said. "It's possible, one of the scenarios we're looking at, it's possible the foam striking a healthy orbiter would not have done enough damage to cause the loss of this orbiter.

"But it's possible foam striking an unhealthy orbiter that had problems in it either due to stresses on launch - we talked about the wind shear, too much heating in transition of years before, aging of the orbiter, like the RCC faults we see, or a whole number of other complex issues - it's possible you could do some damage to this orbiter that ... she could have survived, maybe, at age 10, maybe she couldn't survive it at age 21."

At the Kennedy Space Center, meanwhile, 28,286 pieces of shuttle debris have been catalogued, amounting to about 39,300 pounds of the orbiter's dry weight. Of that total, 25,404 pieces of debris have been positively identified. More than 4,000 men and women continue to search for debris on a daily basis, assisted by a dozen aircraft and Navy dive teams. Sonar scans have located more than 200 submerged targets in two Texas lakes.

Video coverage for subscribers only:
   VIDEO: TUESDAY'S INVESTIGATION BOARD NEWS CONFERENCE QT
   SUBSCRIBE NOW