Nevertheless, it was enough to piece together Columbia's horrifying last minutes. Nor was the recorder the only remarkable find. A week after the disaster, Carl Vita -an engineer with United Space Alliance, the company responsible for Shuttle operations at KSC - and NASA engineer Marty Pontecorvo found a video cassette lying on a road near Palestine in Texas. They put it in a greasy Wal-Mart fried chicken bag and sent it to JSC for analysis. ''It'll probably turn out to be Waylon Jennings or a Merle Haggard tape,'' Vita joked, ''and the lab guys will get a kick out of it.''
Not until the end of February did they realise how significant it was.
It was, in fact, the videotape filmed by Laurel Clark during the first part of Columbia's descent and contained 13 minutes of footage showing herself, Chawla, Husband and McCool in jubilant spirits putting on their gloves, telling jokes and admiring the spectacular light show outside the flight deck windows. Little did they know that even as they chatted and looked forward to their Florida homecoming, ionised atoms from the gradually thickening air were entering a hole in their ship's left wing and soon would begin destroying it from the inside outwards.
''Some might view it as a miracle,'' said Charles Figley of Florida State University's Traumatology Institute. ''Suddenly, here is a postcard of these men and women.'' He added that it might provide additional peace of mind for the crew's families by assuring them that their loved ones were happy, blissfully unaware of the situation that would engulf them in the next few minutes and died doing what they loved. ''Each time I watch it'', Laurel Clark's husband Jon said later, ''I get a sense of joy to know they were just having such a great time.''
The surviving videotape ended at around 1:47:30 pm; Clark obviously continued recording after that time, but what had happened next - stored on the outermost edges of the cassette - had evidently burned away during its fall to Earth. About a minute after the surviving tape finished, a 'strain gauge' attached to an aluminium 'spar' just behind one of the RCC panels which wrap themselves in a 'U' shape around the leading edge of the wings measured an unusual increase of structural stress. It seemed that the aluminium was expanding under the steadily increasing heat and beginning to soften.
Each wing consists of upper and lower surfaces connected by an aluminium framework with a total of 22 individually shaped RCC panels attached to the leading edge. The panels are numbered, with '1' being closest to the fuselage and '22' furthest away. Detailed image analysis of Columbia's ascent suggested a piece of foam from the External Tank probably hit Panel 8, which was not visible to the astronauts from the crew cabin or the Spacehab module while in orbit, their view being blocked by the open payload bay door. Even if they had known the panel was severely damaged, they could have done nothing to fix it.
The strain gauge that started measuring unusual levels of stress was situated directly behind Panel 9, ironically in a region of the leading edge that was subjected to the most extreme temperatures during re-entry. Its data was among that recovered from the OEX recorder, and clearly pointed to trouble brewing within a couple of minutes of entry interface and progressively worsening as each second passed. When the CAIB published its final report in August 2003, its authors were convinced beyond any doubt by the strain gauge data that Columbia began her re-entry with a breached RCC panel.
That conclusion was later refined to a particular panel - Panel 8 - by the strength of the gauge's readings and investigators judged the gauge to have been within 40 cm of the point where hot gas was impinging on the aluminium spar. That ultimately allowed them to zero-in on Panel 8. Twenty seconds after the first indication of trouble from the strain gauge, a MADS sensor in the hollow cavity behind Panels 9 and 10 and just in front of the aluminium spars began measuring unusual temperature increases. Moreover, the sensor was both heavily insulated and some distance from the breach.
This led investigators to conclude that the hole in the lower part of Panel 8 must have been at least 15-25 centimetres in diameter. Anything smaller would probably not generate the observed readings from such a heavily insulated sensor so many centimetres away from the actual breach site. It was around five-and-a-half minutes after entry interface, at about 1:50 pm, that Columbia's computers began the delicate process of actively guiding the spacecraft towards Florida by smoothly swinging the nose 80 degrees to the right. Seconds later, sensors attached to the left-hand OMS pod registered an unusual change in temperature.
Instead of steadily climbing, temperatures around the pod rose very slowly; wind tunnel tests would later confirm some of the hot air entering Panel 8 was blowing metallic vapour from melted insulation through air vents in the top side of the left wing. This interfered with the normal airflow around the Shuttle and slowed anticipated temperature increases on the left side of the fuselage. As the re-entry heating worsened, melted Inconel - a heat-resistant alloy used to seal the RCC panels - started spraying Columbia's metallic skin. Then, seconds after 1:52 pm, the aluminium spar behind RCC Panel 8 finally burned through.
The plume now had access to the interior of the wing and immediately cut through sensor wiring attached to the spar, and started heating the aluminium trusses which supported its upper and lower surfaces. All of this sensor data was recorded and stored only on the OEX recorder; none of it was available to the crew or to Mission Control, who still believed they were following a normal re-entry profile. Judging from the amount of spar and other debris recovered, CAIB investigator Pat Goodman estimated that a hole of at least 20 cm across was burned through the aluminium.
By this stage, temperatures imposed on the RCC exceeded 8,000 Celsius - more than twice as much as they were built to withstand. Under normal re-entry conditions, the Shuttle compresses the thin air to generate two shockwaves. This forms a 'boundary layer', just a few centimetres thick, which resists further compression and provides a natural insulator, keeping temperatures on the RCC panels around 3,000 Celsius. However, a smooth surface is essential for the boundary layer to form. With a ragged hole in Panel 8, the protective properties of Columbia's boundary layer were severely disrupted.
''You have this massive thing ... slamming into the molecules as you get farther into the atmosphere,'' explained the CAIB's Jim Hallock, ''and that collision basically disassociates these molecules into their constituents. That's where all this heating comes from. [The Shuttle] uses aerodynamic braking ... so you've got to get down to where the molecules are to slow it down. That's really what's slowing it down - running into these molecules and dissipating a lot of the energy of speed into heat. Once this boundary layer forms, it really protects you pretty well.'' But the hole disrupted the boundary layer, and let the hot plasma penetrate the wing like a blowtorch.
When the aluminium spar burned through, Columbia was still flying high over the Pacific Ocean, 300 miles from the California coastline, and her crew was oblivious to any danger. However, their sense of security would not last. Within 14 seconds, the hot plume started destroying three bundles of wiring running along the outboard wall of the main landing gear wheel well and gas began flowing into the box-like well itself through vents in its door hinges. It was at this point that the first unusual sensor reading popped up on Jeff Kling's monitor, highlighting increased hydraulic fluid temperatures.
By now, however, Columbia's left wing was literally being destroyed from the inside outwards, and the computers would shortly encounter problems with the ship's handling characteristics as it fell ever deeper into the atmosphere. Husband and McCool may have noticed an almost imperceptible tug as the nose pulled to the left under the effects of increasing aerodynamic drag, but this was corrected automatically by the computers which commanded the elevons at the back of the left wing to balance out the discrepancy to the right. By 1:53:28 pm, however, when Columbia crossed the California coastline, she was in severe distress.
Not only were the trusses inside her left wing softening and melting, but so too was its aluminium skin and - critically - the adhesive needed to hold heat-resistant tiles on the bottom and insulation blankets on the top in place. It was almost certainly fragments of these tiles and blankets that ground-based observers saw falling from the spacecraft as they watched its glowing plasma trail traverse the sky from west to east. Some observers saw a bright flash at around this time, which could have represented the point when the plume finally burned through the upper surface of the left wing.
The CAIB's Roger Tetrault later showed journalists a map pinpointing where each piece of debris was found. Interestingly, the most 'westerly' tiles - those that fell away early in Columbia's breakup - were those associated with RCC Panels 8 and 9. ''What that indicates is there was probably a breach somewhere in the 8 and 9 area,'' Tetrault said. ''The hot gas flowed into that breach, heated up the inside of the wing, the [adhesive] which holds the tiles to the outside skin heated up and basically lost its adhesion capability at 400 Fahrenheit and fell off to the west.''
At about 1:54:11 pm, Columbia's computers responded to further disruption of the normal flight path, although on this occasion the forces trying to pull the nose to the left and roll the spacecraft to the left suddenly reversed and the left wing seemingly gained additional 'lift'. The computers readjusted the elevons to counteract the unwanted motions. Again, Husband and McCool may have noticed the adjustments on their displays, but made no attempt to contact Mission Control. Investigators would later blame the perceived lift on the now-weakened lower surface of the left wing, which was bowing inwards under increasing air pressure.
Seconds earlier, ground-based observers had videotaped a large piece of debris fall away from Columbia and vanish in her plasma trail; subsequent analysis concluded that it may have been a large portion of the left wing's lower surface. Additionally, Pat Goodman speculated that conditions might have been so bad at this stage that the melted aluminium spars and trusses could have been lying ''in pools on the bottom of the wing.'' In all likelihood, the only structural support keeping the upper and lower surfaces of the wing together were the RCC panels and the leading-edge spar behind them.
By 1:56:16 pm, the plume of superheated gas had burned through the left-hand wheel well box and began scorching one of Columbia's landing gear struts; shortly thereafter, as the Shuttle swept over New Mexico, a pair of sensors recorded slowly increasing pressures in the outboard tyre. Meanwhile, on the bottom surface of the wing, the concave depression was growing in size and the ship's computers again commanded the elevons to counteract the unwanted deviations from the flight profile. Columbia was still, it seemed, barely in control. That was about to change abruptly.
Sometime after 1:58 pm, her flying characteristics shifted; possibly the left wing began to collapse. ''The vehicle was in control and responding to commands up to that point and, after that point, something changed,'' said United Space Alliance manager Doug White. ''[The flight system] continued to be in control and respond to commands, but the rates and the amount of muscle it needed to continue flying the vehicle the way it should be flown were continuing to increase. Something definitely happened to cause the flight control system to need more muscle and start having to fight harder.''
It was not until the loss of tyre pressure data triggered an alarm in the cockpit that Husband was prompted to radio Mission Control, but his transmission was cut off in mid-sentence. In fact, STS-1 veteran John Young later remarked that the astronauts inside Columbia's cabin might not notice or 'feel' even major changes in the shape of the left wing, even if it fell off entirely. Presumably, Laurel Clark was still videotaping as the first of what would undoubtedly have been many master alarms over the next few minutes began blaring. We shall never know.
By this time the elevon adjustments were proving hopeless and no longer capable of preventing the ship's nose from yawing to the left. At 1:59 pm, two thruster firings, each lasting about 1.5 seconds, were automatically executed by the left-hand RCS jets in an unsuccessful bid to regain control. A second master alarm sounded in the cabin as the elevons' control circuitry failed and seconds later ground-based observers in western Texas saw a large piece of debris falling away from the fast-moving Columbia. In all likelihood, it was most of the left wing.
This was followed, in the first couple of seconds after 2:00 pm, by two other pieces of debris, which could have been the vertical stabiliser fin and part of one of the OMS pods. Despite the dire predicament in which she and her crew now found themselves - out of control, falling nose-first with her left wing missing and probably several other major structural components also gone - one last, reasonably reliable snippet of data hinted that her hydraulic power units and electrical generators were running, her main engine compartment was intact and her communications and navigation equipment were operating normally.
However, the astronauts knew they were in deep trouble. Master alarms would have sounded as Columbia's computers advised Husband and McCool of fuel leaks from the left-hand OMS pod and a multitude of other problems: the cooling system had failed, extreme temperatures were being measured by sensors on the ship's belly and along the left side of her fuselage and the electrical system was experiencing intermittent shorts. Columbia's orientation was changing rapidly, at more than 20 degrees per second, according to recovered data from the OEX recorder, subjecting its remaining structure to aerodynamic stresses for which it was not designed.
''Based on the debris and the track, I think the [left] wing disintegrated in pieces and we probably ended up with very little of the left wing actually attached before the entire orbiter broke up,'' said the CAIB's Scott Hubbard. The computers ''lost control and the wing started to come apart, the sense from the data is that the orbiter went into a flat spin, possibly even going backwards for some period of time.''
At around this time, Columbia's backup flight system computer recorded that one of the flight deck hand controllers - normally used by the pilots to manually fly the vehicle during its final approach to the runway - was moved beyond its normal position. There was some speculation that Husband or McCool, realising the danger, attempted to take manual control, but CAIB investigators would later conclude that the joystick-like device was probably just inadvertently bumped out of position during those horrifying final moments. In fact, until the loss of telemetry, Columbia's digital autopilot remained in control of the ship's descent.
Veteran astronaut Rick Searfoss had followed an almost identical re-entry profile during his stint in command of STS-90 five years before, conducting the first of several roll reversals to bleed off speed while high over west Texas. He wondered if Husband had had a chance, seconds before the alarms began sounding in the cockpit, to glance out of his window to see the area around the Red River where he grew up. ''We'll never know,'' Searfoss told an interviewer sadly in February 2003, ''at least not in this lifetime. I hope perhaps this was one of his very last thoughts.''
Nonetheless, a shudder runs down the spine when one imagines the situation in the cabin during those last seconds. We do not know how the astronauts reacted upon realising that their ship was disintegrating around them. Perhaps, mercifully, they might have be so intently focused on their flight deck instruments, trying to comprehend and respond to master alarm calls and multiple malfunctions, that the end came blissfully quickly. All that is known with certainty is that the OEX recorder, found that March day in Hemphill by Art Baker, stopped working at 2:00:18 pm.
Sometime within the next minute or so - as Charlie Hobaugh tried to regain radio contact, as Leroy Cain and Jeff Kling pondered a problem with Columbia's tyres, as Laura Hoppe wondered how long intermittent communications would last, as Ed Garske and Bryan Austin saw multiple contrails over their homes in Texas, as Andy Cline woke with a sense of foreboding in the woods of Wyoming and as thousands of spectators in Florida awaited Columbia's triumphant homecoming - seven astronauts perished in the United States' second Shuttle disaster.
Was this article helpful?