A data recorder recovered in the wreckage of the shuttle Columbia shows hot gas entered the leading edge of the spacecraft's left wing within 16 seconds of the point when the orbiter entered the region of maximum aerodynamic heating during re-entry Feb. 1. Temperature sensors located behind two leading edge panels, both just outboard of the point where engineers believe a deadly breach occurred, showed a sudden rapid spike, or increase, before the sensors failed and dropped off line. The readings, which occurred a minute and eight seconds earlier than previous signs of trouble - suggest Columbia almost certainly began its descent with a pre-existing problem and that it might have suddenly worsened as the ship plunged back into the atmosphere.

"They go way up and then they go to zero," said a representative of the Columbia Accident Investigation Board. Along with showing unusual heating effects earlier than previous telemetry that was downlinked in realtime, "I think what it also indicates is the mechanism of failure was already in place."

Illustrations show where the foam impact is believed to have occurred during launch. Photo: CAIB

Columbia was destroyed by a breach in the ship's left wing. Up until now, NASA engineers relied on telemetry that was downlinked from the orbiter in realtime to reconstruct what might have gone wrong. That data indicated the shuttle began responding to unusual aerodynamic forces around 8:52 a.m., more than a minute after the orbiter entered the region of maximum heating at 8:50:53 a.m. The downlinked telemetry shows the first signs of unusual heating occurred at 08:52:17 a.m. when a brake line in the left main landing gear wheel well began showing an unusual rise.

From that point on, a series of sensors failed as a plume of super heated air burned its way into the left wing, eating through sensor cables and causing temperatures to climb throughout the wheel well. The presumed entry point of the plume - at or near the underside of leading edge panel No. 6 - is believed to have worsened as entry proceeded, causing increasing aerodynamic drag. The shuttle eventually yawed out of control and broke up over Texas about 20 seconds after 9 a.m.

But re-entry readings from hundreds of other temperature, vibration and pressure sensors were stored on a data recorder and not downlinked to the ground. The recorder was recovered by search crews March 19 and after work to clean and stabilized the tape, engineers began work this weekend to examine what data might be present.

The board representative said the tape holds valid data and shows unusual heating in the left wing, behind leading edge panels 9 and 10. One of those sensors is located on the front surface of the wing spar, underneath insulation but directly behind the leading edge panels, while the other is mounted on the interior side of the spar.

Illustration of orbiter wing marks some of the RCC panels. Measurements from the OEX recorder shows temperature spikes behind panels 9 and 10. Credit: NASA

The sudden temperature spikes, which began at 8:51:09 a.m. - 16 seconds after the shuttle entered the region of maximum heating off the coast of California - would be consistent with the entry of hot air through a pre-existing breach in the leading edge. The hot air would have shot down a cavity in the U-shaped leading edge panels, triggering the observed temperature spikes, before working its way into the wing's interior.

The timing of how the leading edge sensors responded to the increased temperature is unclear. One of them failed and went off line 20 seconds or so after its counterpart. But they showed sharp temperature increases that were clearly above what the sensors would detect during a normal entry, the board's representative said.

While it is too soon to draw any concrete conclusions, the OEX recorder promises to provide a gold mine of new aerodynamic and thermal data to engineers trying to re-construct the devastating chain of events that led to Columbia's destruction.

"The board is very, very happy that the data could be retrieved," the board representative said. "They're anxious to see what more we can learn about how the heating was moving along the left wing."

Engineers plan to meet Monday to discuss the status of the OEX data reconstruction. Over the next several days, they will begin the process of weaving the new information into NASA's entry timeline. They also plan to continue work to recover OEX data recorded during Columbia's ascent to determine if any readings might be indicative of external tank foam striking the left wing 82 seconds or so after liftoff.

A preliminary look at the ascent data shows a potentially interesting signature, but officials say it may well be the result of a wiring defect. Additional analysis should resolve the matter one way or the other in the days ahead.