The shuttle Columbia was already spinning out of control, its left wing and left maneuvering engine damaged or destroyed, in the last two seconds of data transmission, two officials close to the investigation said, citing a new analysis by NASA.
The analysis, scheduled for release this week, paints a far more dire picture of the shuttle's re-entry than had been previously described.
Investigators had said that the Columbia's last two-second burst of data, sent about 9 a.m. Feb. 1, showed the shuttle on course and in the proper orientation, although it had lost hydraulic pressure, which would have been crucial for landing.
They also reported at that time that the shuttle's rockets had been trying to counter a pull to the left, which could have been caused by roughness on the left side of the orbiter. The area might have been struck by debris on liftoff, or damaged in some other way.
The new information shows the shuttle pivoting like a skidding automobile on its bulletlike path through the sky, its nose veering left at a rate of 20 degrees per second. That rate of skid, or yaw, was the fastest that the shuttle's instruments could measure, so the actual rate could have been higher. At 20 degrees per second, the shuttle would have been in a slow pirouette, completing a circle every 18 seconds.
In addition, sensors in the left wing and in the left maneuvering rockets had mostly stopped reporting, indicating that those parts, if not already torn off, were probably destroyed, one investigator said, according to the New York Times.
The analysis is in a document called "Rev. 14," the 14th revision of the timeline, completed late last week. The Columbia Accident Independent Board, led by Adm. Harold W. Gehman Jr., was planning to release a subsequent version, Rev. 15, which is supposed to be completed today. As the names imply, neither is considered final.
NASA is working with the data in the 32 seconds since the last spoken transmission from the shuttle crew. In that conversation, a ground controller in Houston tells Col. Rick Husband, the flight commander, who is at the controls, that technicians on the ground see the sensor readings indicating that the tires have lost pressure. Husband replies with the word "Roger" and an additional syllable, which sounds like "uh" or "buh." Data transmission continues for another five seconds, then ceases for 25, then picks up for the last two.
While the new interpretation seems to show that the shuttle was on its way to breakup earlier than previously believed, investigators say the loss of voice and data communications itself was not ominous. Such losses are common on shuttle re-entries, and in this case might have occurred because the Columbia was using a relay satellite behind it, over the Indian Ocean, and its vertical tail might have come between the transmitter and the satellite.
The two-second data burst is the last data that investigators will have to work with unless they recover computing devices from the wreckage that might have stored and preserved sensor readings. The shuttle does not carry a crash-hardened data recorder, as civilian airplanes do.
But there is another potential source of information: the condition and location of pieces of debris. Investigators looking at radar data and analyzing winds at various altitudes at the time of the breakup are still hopeful of finding debris as far west as Nevada, and they say that knowing just what parts those were would be helpful. But they have not reported finding any debris west of Texas.
Another scrap of mysterious information from the shuttle's final seconds concerns its autopilot. The data received by NASA showed that the signal to disengage the autopilot had been given. However, that command had not yet been carried out by the shuttle's flight systems when communication was broken for the last time.
Under the conditions of a normal return to Earth, the shuttle flies on autopilot until it is traveling more slowly than the speed of sound. But pilots train to take the shuttle all the way down in case the autopilot malfunctions, and so it is possible that one of the pilots was trying to take control of the yawing craft in its final moments.
It is relatively easy to turn off the autopilot by accident, and in fact this occurred just minutes before the problems with the Columbia became apparent. In the recovered segment of flight deck video of the waning minutes of the flight released by NASA, Husband is heard to exclaim, "Oh, shoot," and to tell mission control that "we bumped the stick earlier," briefly disengaging the autopilot. He quickly and calmly corrected the error.
A NASA spokeswoman, Eileen Hawley, said the possible attempted override in the final few seconds could also have been unintentional. She added that part of the problem is that the data are intermittent, with a high error rate, "and to draw any conclusions from it would be really wrong."
_ Information from the Associated Press was used in this report.