A fragment of videotape shot by one of the astronauts on Columbia's flight deck during the early stages of re-entry Feb. 1 has been recovered by NASA. But sources say the heat-damaged tape ends before the onset of problems in the left wing that ultimately led to the orbiter's destruction and the deaths of the ship's crew. As such, the tape provides no insight into the mishap.

But sources familiar with the tape say the astronauts showed no signs of any concern as they prepared for return to Earth after a 16-day science mission. The tape has been shown to astronaut family members, the sources say, and will be shown to lawmakers in Washington on Wednesday before its eventual release to the media and public.

The digital video tape, presumably shot by astronaut Laurel Clark, begins around 8:35 a.m., some nine minutes before Columbia fell into the discernible atmosphere 400,000 feet above the Pacific Ocean northwest of Hawaii.

The tape continues for four minutes past "entry interface" and then abruptly ends around 8:48 a.m., four minutes before the first telemetry was received indicating problems in the shuttle's left wing. Sources say no other tape remained on the heavily damaged cassette.

Commander Rick Husband, pilot William "Willie" McCool, flight engineer Kalpana Chawla and Clark were seated on Columbia's flight deck for entry. Crewmates Michael Anderson, David Brown and Israeli astronaut Ilan Ramon were seated out of sight on the shuttle's lower deck.

Unlike the pilots and the flight engineer, Clark had no specific re-entry duties and presumably shot the video recovered by NASA. But that is not yet known.

At a weekly news conference by members of the Columbia Accident Investigation Board, board member Scott Hubbard said the final two seconds of telemetry from Columbia showed all of the orbiter's major systems were running and that its fuselage was still intact.

As CBS News reported late last week, the final two seconds of data, which followed a 25-second period in which no data were received, also showed Columbia's triply redundant hydraulic system had lost all pressure and that its reservoirs of hydraulic fluid were empty.

"The conclusion seems to be that somewhere in that 25 seconds of unrecovered data there was some sort of failure that affected all three hydraulic units," Hubbard said.

Board chairman Hal Gehman provided a bit of additional insight by adding later that no data from the left wing showed up in the final two seconds of telemetry. This is consistent with severe wing damage - or even its complete removal just prior to the final burst of telemetry - that severed all of the hydraulic lines leading to the wing's elevons and landing gear wheel well.

Gehman said engineers studying debris from the shuttle still do not know what caused the catastrophe. But without being specific, he said the ongoing recovery of debris is beginning to pay off.

"Debris continues to be important to us," he said. "We are now beginning to learn some things - that's probably too strong - but we're beginning to see some interesting trends and evidence in the debris."

He then showed pictures of a heat shield tile frrom Columbia's left wing that was found near the town of Powell, Texas. The tile shows severe damage from heating effects, "probably by hot gases."

"I am told this is not typical of a re-entry tile," he said. "This is very unusual. One of the riddles we have to sort out is whether or not this damage was done while the tile was still attached to the orbiter or whether this damage was done after the breakup and that's what it looks like when you try to re-enter the atmosphere in a non-aerodynamic state."

Another fragment of tile found farther west, near Littlefield, Texas, appears to have come from the upper surface of the left wing near the so-called "glove" area where the wing joined the fuselage.

Most investigators believe the catastrophe began with a breach in the left wing that allowed superheated gas, or plasma, to work its way inside. The hot gas then burned through sensor wiring and caused temperatures in the left landing gear wheel well to climb.

The breach may have been the result of damage from external tank debris that hit the wing during launch, impact by space debris during the mission or some other factor. Exactly where the breach occurred - at or near the left main landing gear wheel well or closer to the leading edge of the wing - is not yet known.

Earlier today, NASA has posted visible and infrared images of Columbia while it was still in orbit that were shot by a powerful Air Force telescope in Hawaii. But Columbia's open cargo bay is facing the camera in all of the images and no details of the underside of the ship can be seen.

In any case, engineers have not yet ruled out any "root cause" of the Columbia disaster. Investigators are running computer models to create "virtual" breaches at various points on the wing to come up with locations that at least explain the series of sensor failures and rising temperatures that were radioed to Earth before Columbia broke up.

But Hubbard cautioned that a simple answer might not be forthcoming.

"It's well known that accidents in complex systems often involve a chain of events," he said. "It's not often one single thing caused the whole accident. So this complex series of events needs to be evaluated together. For example, eventually we would like to see if you can couple the external tank shedding event with TPS sensitivities with the calculations that are being done by the aerodynamics folks.

"So what we're doing right now is trying to bound the events in the orbiter, we're trying to go from the town to the ballpark to the seat."

He then provided three examples to illustrate the difficulty of "bounding the event."

"We did a hypothetical study of a 20-square-inch breach in the wing, 4 by 5 (inches), near the main landing gear door or the seal, that seems to account for the temperature rise," he said. "But this is a first cut with an initial set of assumptions. Figuring out what might happen to a plume going inside a wing is a very complex task. It's a good first step, but we don't have an answer there yet.

"Secondly, the sensor data that we've all looked at, the rise in temperature, the sensors going off line, seems to be consistent with wires going through the wheel well being severed. But there is a sensor near the front of the spacecraft that also had an anomaly. How do you account for that? Again, it's a little early to conclude we have a complete story.

"And finally, some initial aerodynamic analysis seems to indicate there was some disturbance going on in the vicinity of the left main landing gear wheel well, but eight different calculations all gave somewhat different answers.

"So the story I'm trying to communicate to you is it's early yet to draw conclusions that we have boxed this story in, that we really have bounded the event. We're making good progress, we need to validate the data interpretations with analysis of debris, experimental tests and in selected cases with independent calculations and perhaps independent tests. We need to look at the system."

As of today, Gehman said, 8,110 pieces of confirmed shuttle debris have been found, representing a bit more than 10 percent of the orbiter by weight.

In a final bit of news from today's briefing, it was reported that investigators studying Air Force radar tracking tapes made during Columbia's mission clearly show a small object measuring about .3 by .4 meters separated from the shuttle on the second day of the flight. The object later re-entered the atmosphere on its own. CBS News reported this "event" on Feb. 7, but it is still not known what it might have been or whether it played any role in the ensuing catastrophe.

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