EDITOR'S NOTE: From "Comm Check ... The Final Flight of Shuttle Columbia," by Michael Cabbage and William Harwood, which is being published Tuesday by Free Press, a division of Simon & Schuster. Cabbage is the space editor of the Orlando (Fla.) Sentinel; Harwood is a veteran space reporter for CBS News. Printed by permission.


"The most complicated machine ever built got knocked out of the sky by a pound and a half of foam. I don't know how any of us could have seen that coming. The message that sends me is, we are walking the razor's edge. This is a dangerous business and it does not take much to knock you off."
-- Flight director Paul Hill

******************

Shuttle wings are made of aluminum, the upper and lower surfaces separated by spars and trusses that form a boxlike internal framework. The main landing gear wheel well boxes are located toward the front of each wing, nestled up against the side of the orbiter's fuselage just behind the leading edge.

Behind its protective insulation, the front of a shuttle wing is flat, made up of a panel of aluminum honeycomb material known as the leading edge spar. To give the wing its aerodynamic shape, and to protect it from the most extreme temperatures of re-entry, 22 reinforced-carbon carbon panels are bolted side by side on that flat front surface, creating a smoothly curving leading edge. So-called spanner beams, made out of a heat-resistant alloy called Inconel, provide rigidity. To seal the gaps between RCC panels, thin carbon-composite strips called T-seals are bolted in place to provide a smooth surface along the entire leading edge.

During re-entry, the shuttle's nose is pitched up 40 degrees, which subjects the lower halves of the RCC panels to the most extreme heating. The fittings used to attach the RCC panels to the main spar are protected by heat-resistant insulation that melts at 3,200 degrees.

Whatever happened to Columbia had utterly destroyed this complex system.

Twenty-seven truckloads of wreckage were hauled to Kennedy Space Center between Feb. 5 and May 6. More than 25,000 searchers, who scoured a debris "footprint" that was 645 miles long, found 84,900 individual pieces, about 38 percent of the space shuttle. Each piece or component was cleaned, decontaminated, bar-coded, photographed and entered into a computer database. Wreckage from Columbia's wings, fuselage, and nose section was laid out on a grid in the Reusable Launch Vehicle Hangar near Kennedy's shuttle runway. The most critical RCC panels and attachment fittings -- those numbered 1 through 13 and nearest the fuselage -- were mounted on a full-scale clear plastic mockup of the rounded leading edge that allowed investigators to see each piece in relationship to its neighbors. It also allowed them to map out exactly where the heat went after it entered the leading edge.

The work at KSC was buttressed by analysis by Johnson Space Center engineers of data from the orbiter's Modular Auxiliary Data System, or MADS, recorder and amateur video images of Columbia's disintegration. The inch-wide MADS tape contained information from 570 sensors; it was found by searchers in Hemphill, Texas, on March 19, six weeks after Columbia disintegrated. Ultimately, the Columbia Accident Investigation Board was able to conclude, without qualification, that the foam impact was the root cause of the accident; that the impact had knocked a 6- to 10-inch hole in the lower half of RCC panel 8 on the shuttle's left wing; and that a plume of super-heated plasma entering through that breach had destroyed the wing and triggered the destruction of the orbiter.

The team concluded the foam broke away from the left bipod ramp 81.7 seconds after liftoff and hit the underside of Columbia's left wing two-tenths of a second later. The foam measured 21 to 27 inches long by 12 to 18 inches wide. It was tumbling at 18 revolutions per second. Before the foam separated, the shuttle -- and the foam -- had a velocity of 1,568 mph, about twice the speed of sound. Because of its low density, the foam rapidly decelerated once in the airstream, slowing by 550 mph in that two-tenths of a second. The foam didn't fall on to the leading edge of the left wing as much as the shuttle ran into it from below. The relative speed of the collision was more than 500 mph, delivering more than a ton of force.

On July 7, investigators using a nitrogen-powered cannon fired a 1,200-cubic-inch block of foam weighing 1.67 pounds at RCC panel 8, taken from the shuttle Atlantis. Traveling at 530 mph, the foam blew a ragged 16-inch hole in the RCC panel, vividly demonstrating how much damage foam could do.

***

With the dramatic foam shot at RCC panel 8, all the pieces of the puzzle were finally in place. There was little doubt about what had doomed Columbia and its crew. A second-by-second time line of the final working scenario provided a gripping account of the shuttle's final minutes.

At 8:44:09 a.m. Eastern time on Feb. 1, 2003, Columbia was a half-hour from home. The shuttle had just dropped below an altitude of 76 miles, slipping into the discernible atmosphere 900 miles northwest of Honolulu.

During re-entry, the shuttle compresses the thin air in front of it, creating two shock waves. Those shock waves intersect around RCC panel 9, subjecting panels in that area to the most extreme heating. But the compression of the air in front of the shuttle forms a so-called boundary layer, a region just a few inches thick that resists further compression and acts as a natural insulator. A few inches away from the leading edge, just beyond the boundary layer, molecules are torn apart and temperatures can exceed 10,000 degrees. But the boundary layer keeps temperatures on the leading edge RCC panels at around 3,000 degrees.

A smooth surface is essential for the boundary layer to form and is crucial to a shuttle's survival during the plunge to Earth. If the boundary layer is disturbed for any reason, its insulating effect can be compromised by high-temperature turbulence, subjecting the shuttle's tiles and RCC panels to much more heat than they were designed to handle.

But even as the Columbia astronauts chatted about the light show outside, the hole in Columbia's left wing was disrupting that boundary layer. Ever more air molecules were shooting into the inside of the wing at RCC panel 8 and slamming into the insulation protecting the panel attachment fittings, swirling through the cavity and spreading out to either side. At that altitude, the effect was small. But the shuttle was descending, and the air was getting thicker with each passing second. With Columbia in a 40-degree nose-up orientation, the plume entering the breach in RCC panel 8 was aimed at the upper attachment fittings and insulation. The insulation began melting, and the front face of the left wing's aluminum honeycomb forward spar -- the only barrier between the plume and the interior of the wing -- began heating up.

At 8:48:39 a.m., just four minutes and 30 seconds after Columbia had dipped into the atmosphere, a sensor mounted behind the forward spar, near the point where RCC panel 9 was bolted to the other side, measured an unusual increase in stress. The spar was softening.

About a minute later -- five and a half minutes after entry interface -- the shuttle's flight computers ordered a turn to the right. Up until this point, the shuttle had simply been falling into the atmosphere, wings level, nose up and pointed straight ahead. Now, the ship's flight computers began actively guiding the shuttle toward Kennedy's runway. The shuttle's nose smoothly swung 80 degrees to the right.

Less than 20 seconds after the maneuver, sensors mounted on Columbia's left rear rocket pod measured an unusual change in temperature. Wind tunnel testing would later show some of the hot air blasting into the RCC cavity was exiting through the vents on the upper surface of the wing, carrying thin clouds of metallic vapor from melted insulation.

The firestorm inside the RCC cavity was rapidly increasing in intensity. The boundary layer around the leading edge breach was severely disrupted, and the flow of super-heated air over the lower surface of the wing exposed the protective tiles there to much higher temperatures than they were designed to withstand. Insulation and RCC panel support fittings behind the breach continued to burn away.

Within a few seconds of 8:52:16 a.m. -- the exact time is unknown -- the deadly plume burned its way through the forward wing spar and into the interior of the wing.

The shuttle was still 300 miles from the coast of California. The crew still had no idea anything was wrong.

But with the boundary layer disrupted, the temperature of the atoms and molecules blasting into the wing probably exceeded 8,000 degrees near the leading edge breach itself. Hot gas began flowing into the wheel well through vents around landing gear door hinges. At 8:52:17 a.m., the first unusual sensor reading flashed on a computer screen in mission control: a slight increase in temperature in the hydraulic fluid running through a brake line leading to the left main landing gear.

Columbia's left wing was burning up from the inside out. Twelve seconds after the brake line temperature reading showed up in mission control, the shuttle's flight computers noticed the effects of the damage for the first time as a force, or drag, began pulling the shuttle's nose to the left. After assessing the data for a few seconds, the computers sent commands to the wing flaps, or elevons, on both wings to push the shuttle's nose slightly to the right to balance it out.

On the flight deck, shuttle commander Rick Husband and rookie pilot William "Willie" McCool remained oblivious to their ship's ongoing destruction. They might have noticed the elevon movement on their forward computer displays, but the adjustments were small and would not have caused concern.

Columbia finally crossed the coast of California north of San Francisco at 8:53:28 a.m. at an altitude of 45 miles and a velocity of 15,800 mph. By then, the orbiter was in severe distress.

Scores of amateur shuttle watchers in California and Nevada had gotten up before dawn to watch Columbia's fiery descent. Even first-time observers were struck by the appearance of the shuttle's plasma trail. The super-heated air left in the shuttle's wake glowed in the dark sky like a phosphorescent contrail.

The plume shooting into the wing from the front spar breach may have burned a hole through the upper skin of the wing during this period, perhaps at the same time that many observers on the ground saw a bright flash.

By 8:54 a.m., just 32 seconds after Columbia had crossed the coast -- and just a minute and 44 seconds after the forward spar had been breached -- the outboard wall of the left main landing gear wheel well began melting. A scant 11 seconds after that, the shuttle's flight computers detected another change in the way Columbia's flight path was being affected.

It was as if the left wing had suddenly gained additional lift. The flight computers instantly responded, adjusting Columbia's elevons yet again to exactly counteract the two unwanted motions.

The shuttle stayed on course. Husband and McCool may have noticed the elevon movements as the autopilot responded, but again, they made no attempt to contact mission control for an explanation. In all likelihood, they still believed the entry was proceeding normally.

The increased lift initially puzzled investigators until they pieced together the plume's path through the wing's interior. The melting of the support spars and trusses just behind the forward spar caused the upper and lower wing surfaces to lose their rigidity. The lower wing, which was directly affected by the increasing pressure of the air, bowed inward, forming a depression. It started out small, but as the seconds ticked by and the wing's interior got even hotter, it grew alarmingly. Over the next five minutes, the depression probably grew to some 20 feet in length and 4 feet in width, a concave area more than 5 inches deep. Wind-tunnel testing and computer simulations later showed such a depression could explain the reaction of Columbia's flight computers.

In mission control, the first clear sign of a problem aboard Columbia was the loss of data from sensors in the left wing's hydraulic system. The wires leading to those sensors had been part of a cable bundle attached to the outboard wall of the left landing gear wheel well.

As Columbia was crossing the border between California and Nevada, the shuttle's attitude was down to 43.1 miles. But its velocity was still a blistering 22.5 times the speed of sound. It was 8:54:25 a.m.

Observers on the ground saw or photographed more than 10 debris-shedding events in the next few moments.

At 8:58:03 a.m., Columbia's flight computers detected a sharp change in the aerodynamic forces acting on the shuttle as the depression in the lower surface of the left wing presumably increased in size. At the same time, the drag acting to pull the nose farther to the left continued to increase. Approaching the Texas border, the flight computers again ordered the elevons to counteract the unwanted forces. Several debris-shedding events, indicating the wing was losing additional insulation and structure, were noticed by ground observers.

Months later, Air Force Lt. Col. Pat Goodman, a CAIB investigator, speculated the sudden change in the shuttle's flying characteristics was caused by a major change in the wing's shape. "I believe you can make a case ... that the wing begins to collapse," Goodman said. But the crew still would not have noticed any dramatic change.

They did, however, notice the loss of tire pressure data. The computers triggered an alarm in the cockpit and displayed a message to alert Husband to possible problems with the landing gear. This was the crew's first notification of potential trouble. Husband called mission control, presumably to report the message -- "And, uh, Hou ... " but his transmission was cut off.

Astronaut Charles Hobaugh, sitting to Cain's immediate right, radioed Columbia to let Husband know the flight control team was aware of the alarm and the lost tire data. He added, "And we did not copy your last" to let Husband know he needed to repeat whatever he had been trying to say earlier.

By now, the drag and roll forces acting on Columbia were beginning to reach the point where the elevons could no longer keep the shuttle properly oriented. In seconds, they would reach the limit of their motion.

Husband, perhaps beginning to realize major problems were developing, heard Hobaugh's call and tried to respond.

"Roger, uh, buh ... " It was 8:59:32 a.m. and Columbia was approaching Dallas. Seconds earlier, data from the shuttle suddenly froze on the computer screens in mission control. Down arrows or the letter S, for "static," had appeared on the screens, indicating the numbers were no longer being updated. As it turned out, data were, in fact, still flowing down from Columbia. The signals were garbled, however, and the computers in mission control were programmed not to display potentially corrupted information. Investigators later would be able to extract some of the data. That information, combined with readings stored in the MADS recorder, and analysis of recovered wreckage, eventually allowed investigators to develop a rough time line of events stretching another one minute and 50 seconds beyond Husband's final transmission.

For the astronauts, the final sequence was mercifully brief, but no doubt terrifying.

The left wing had suffered so much damage by now that nothing could be done to keep the nose pointed in the right direction. First two and then four right-side rocket thrusters were automatically commanded to fire in a futile bid to offset the forces pulling the nose to the left. A master alarm sounded in the cockpit as the elevon control circuitry failed. Columbia's nose yawed farther to the left, toward Earth, as the spacecraft began rolling to its right.

In all likelihood, all or part of the presumably collapsed wing suddenly folded over and broke off. At 8:59:46 a.m., a large piece of debris was seen separating from the shuttle. Columbia's backup flight system computer began generating a string of fault messages. Two more large pieces of debris fell away from the shuttle within two seconds of each other starting at 9:00:01 a.m. One of these may have been the shuttle's vertical tail fin ripping off in the hypersonic airstream. The other could have been a large piece of the left-side rocket pod. No one knows.

"Everything just wants to fall over at that point," Cain said. "Because again, this is just like a barn door in wind. If that wing came off as we were falling -- pitching down and falling over ... it is likely that the vehicle then probably broke apart in mid-body area." But not immediately.

At 9:00:02 a.m., two seconds of relatively clean data reached the ground, providing a snapshot of Columbia's condition at that moment.

Columbia's three hydraulic power units were still running, along with the ship's three electrical generators. The main engine compartment was intact, and the communications and navigation equipment in the crew module were functioning normally. The shuttle's life support systems were operational. Air pressure was stable, and the temperature was a comfortable 71.6 degrees.

But all three hydraulic power units had lost pressure, and the ship's reservoirs of hydraulic fluid were empty. The shuttle's cooling system had shut down. Multiple alarm messages intended to alert the crew to problems were being generated by the computer system. Extreme temperatures were being recorded by sensors on the belly of the orbiter and along the left side of the fuselage. The electrical system was showing signs of multiple shorts.

As of 9:00:04 a.m., when the final two seconds of telemetry ended, the fuselage was still intact, along with the right wing and the right rear rocket pod. All or part of the left wing was gone. The condition of the vertical tail fin was unknown.

Just before telemetry stopped, data from the backup flight system computer indicated one of the two cockpit "joysticks," used to manually fly the spacecraft on final approach to the runway, was moved beyond its normal position. That's one way for a pilot to deactivate the autopilot. But investigators do not believe Husband or McCool was attempting to take over manual control. More likely, one of the pilots inadvertently bumped his hand controller during those horrifying final few seconds. The shuttle's digital autopilot remained engaged through the final loss of signal.

Finally, at 9:00:19 a.m., the fuselage began breaking apart. The shuttle was 37 miles up and still traveling 18 times the speed of sound.

A study done for the CAIB concluded the shuttle's heavily reinforced crew module and nose section broke away from the fuselage relatively intact, separating at the bulkhead that marks the dividing line between the cargo bay and the forward fuselage.

Challenger's crew module had also broken away in one piece when the shuttle disintegrated during launch 17 years earlier. As with Challenger, the forces acting on Columbia's crew during this period were not violent enough to cause injury, and investigators believe the astronauts probably survived the initial breakup of the orbiter.

Like Challenger's crew, the Columbia astronauts met their fates alone and the details will never be known. Clark presumably was still videotaping on the flight deck when the alarms began blaring and the shuttle yawed out of control. But the outer portions of the tape -- the portions that might have shown at least the initial moments of the shuttle's destruction -- were burned away.

Investigators concluded the module fell intact for 38 seconds after main vehicle breakup, plunging 60,000 feet to an altitude of 26 miles before it began to disintegrate from the combined effects of aerodynamic stress and extreme temperatures. From the debris analysis, investigators believe the module was probably destroyed over a 24-second period beginning at 9:00:58 a.m. During that period, the module fell another 35,000 feet, to an altitude of 19 miles or so.

Investigators believe the module began breaking up at the beginning of that window. If any of the astronauts were still alive at that point, death would have been instantaneous, the result of blunt force trauma, including hypersonic wind blast, and lack of oxygen. About 45 percent of the crew module was recovered near Hemphill, Texas, including pieces of the forward and aft main bulkheads, the frames from the forward cockpit windows, the crew airlock, and all of the hatches. About three-quarters of the flight deck instrument panels were found, along with 80 percent of the mid-deck floor panels and numerous parts from the crew's seats and attached safety equipment. From an analysis of pressure suit components and helmets, investigators concluded three astronauts had not yet donned their gloves when breakup began and one was not wearing his or her helmet. In the end, however, having sealed pressure suits would have made no difference.

But investigators were struck by the way the crew modules of both Challenger and Columbia broke away relatively intact. The survivability study concluded relatively modest design changes might enable future crews to survive long enough to bail out.

But Columbia's crew had no chance. The astronauts fell to Earth amid a cloud of wreckage and debris.

One of the crew members came to rest beside a country road near Hemphill. The remains were found by a 59-year-old chemical engineer and Vietnam veteran named Roger Coday, who called the sheriff and then watched from the porch of his mobile home as a funeral director drove by to collect them.

"The astronauts have always been my heroes," said Coday, who that afternoon fashioned a cross out of two cedar logs he had cut earlier and erected it at the place where the astronaut had fallen to Earth.

"It's there and we still maintain it," he said eight months after the disaster, still wondering who the astronaut was. "I am a very devout Christian, and I prayed for that person's soul."

COPYRIGHT © 2004 BY MICHAEL CABBAGE AND WILLIAM HARWOOD

Copyright © 2004, Newsday, Inc.









This page revised: February 4th 2004