Wednesday, February 1, 2017

February 1, 2003 -- The Space Shuttle Columbia broke apart over Texas, after the thermal shielding on its left wing failed. American astronauts Rick Husband, William McCool, Mike Anderson, Kalpana Chawla, David Brown, and Laurel Clark, and Israeli astronaut Ilan Ramon, died.

Space shuttles were clad in what was jargonistically called the "Thermal Protection System". There were multiple elements to this system: Not every part of the shuttle was exposed to the same amount of heat or stress during reentry, so some elements of the system were stronger than others. Where the temperatures and stresses were highest (on the shuttle's underbelly), high-temperature reusable surface insulation (HRSI) tiles were used. The chin of the orbiter and the leading edges of its wings received less-intense heat during reentry, but were subject to even greater stresses. Here, reinforced carbon–carbon (RCC) tiles were used. These tiles, developed and manufactured by Ling-Temco-Vought (now Lockheed Martin), were very thin, just one-quarter to one-half an inch thick, and made from graphite rayon cloth impregnated with a phenol resin. There were 22 of these RCC tiles on the leading edge of each wing (usually referred to as "5-left" and "5-right", or "12-right" and "12-left".) For the areas around the cabin windows, the upper edges of the wings, and the tail, low-temperature reusable surface insulation (LRSI) tiles were used. LRSI tiles were the same as HRSI tiles, but were thinner and smaller and colored white. The remainder of the shuttle was covered in felt reusable surface insulation (FRSI), a blanket-like material similar to the fire-resistant material firefighters use.

Space shuttles were powered by an external tank and two solid-fuel rocket boosters. The external tank contained most of the fuel for the shuttle, which consisted of liquid oxygen and liquid hydrogen. Both were extremely cold. This could cause frost and even ice to build up on the external tank, adding an immense amount of weight to the rocket. To prevent this, the external tank was covered in inch-thick foam insulation. This not only helped keep the oxygen cool, but also discouraged frost and ice build-up. Ablative thermal tiles (like those used on the Apollo missions) were attached to the upper part of the tank, where compression of the atmosphere as the shuttle reached supersonic speed created high temperatures. The solid rocket boosters didn't need insulating foam, but they did need heat and stress protection. So their nose-cones were also covered in foam.

A V-shaped strut (the "bipod strut") connected the external tank to the shuttle. Umbilical connections at each wing sent fuel to the shuttle's engines, and allowed the shuttle and tank to share data and power during ascent.

* * * * * * *

Early tests showed that as the shuttle launch system lifted off, the nose of the external tank remained relatively free from compression heating and stress. But just behind the nose, about where the bipod strut connected the tank to the nose, immense shock waves formed that created one of the hottest and most stressful environment around the tank. Moreover, airflow around the struts was complex and could cause further heat and stress.

Initially, NASA believed that the usual foam insulation around the struts would protect them. But additional testing showed this wasn't good enough. After experimenting with several solutions, NASA realized that building up the foam insulation in this area and then shaping it into a ramp-like structure could diver the shockwaves away from the struts -- protecting them.

The foam would be sprayed onto the bipod struts and fittings once the shuttle was mated to the external tank at Cape Canaveral. The foam was sprayed by hand and applied evenly. The two chemicals which made up the foam had to be mixed in an exact ratio, which meant the spraygun had to be in perfect working order and the nozzle completely clean in order for the right ratio to be sprayed. After it dried, the foam would be carved by hand into the shape of a ramp. The foam was visually inspected after carving to ensure there were no holes, cavities, or other problems.

Was visual inspection good enough? Amazingly, NASA never answered that question. They never ran tests to see if minor changes in hand-spraying technique introduced small but critical faults, fractures, or voids into the foam. They never ran tests or drills to ensure that the equipment was cleaned, prepared, and used properly. The agency never cut open any foam to see if it was, in fact, free of problems. Later tests showed that the physical shape of the bipod strut made applying the foam difficult. Both the hand-spraying and the need for the spraygun to emit an exact ratio of elements introduced subsurface defects. Some of these were fracture lines, where the foam could shear into two or more pieces under stress.

More about the tragedy behind this cut....

NASA really needed to answer these questions, because if foam broke loose from the external tank during lift-off, the foam could strike the thermal protection system and damage a tile. NASA established a fundamental design requirement for foam: It could never come loose from the external fuel tank. Working under this assumption, the heat-resistant shielding didn't need to be very strong. After all, it had to be heat-resistant, not impact-resistant. Engineers subsequently designed the thermal protectin tiles to be about to withstand impacts with a kinetic energy of less than 0.006 foot-pounds. (A foot-pound is a unit of energy equal to the amount required to raise one pound a distance of one foot.) This meant that just pressing your fingernail into a tile could damage it.

By every standard of the day, NASA's engineering of the bipod strut and foam ramp was acceptable. At the time, engineers worked on the mechanical aspects of the strut first, then turned this over to the thermal team, which turned it over to the de-icing team. This approach meant that the strut was designed for optimal mechanical performance, but its design hindered the application of the insulating foam. Today, engineers from all three teams would work together at the same time, which would help avoid these problems. By every standard of the day, NASA's testing of the foam ramp was also acceptable. NASA engineers sprayed foam into a solid block, cut it into shape, and then subjected it to wind-tunnel and heat tests. It passed with flying colors. But today, these tests would spray foam sprayed onto a bipod strut, and then subject the foam and strut to stress and heat. Finally, NASA broke a critical rule of engineering: "Test what you fly, and fly what you test." This rule of aeronautices and rocketry stems from the fact that ground-tests, wind-tunnels, flamethrowers, and the like cannot even come close to the speeds and heats and stresses of actual spaceflight. Ground-test results must be extrapolated to real conditions. But this extrapolation isn't always correct. That's why sensors are always applied to the test vehicle in every conceivable way, to see if in-flight experience confirms the ground-test data. NASA never did this, so it never knew how the foam was acting in the real world.

* * * * * * *

The shuttle began shedding foam immediately.

On STS-1, the shuttle's maiden flight in 1981, more than 300 tiles on Columbia were damaged by foam, ice, and launch pad debris and had to be replaced. A whopping 16 tiles came lose and were lost (although none of them were in critical areas). Engineers were shocked; some even said that, had they known that such a large amount of foam could be shed, they would not have cleared Columbia for flight.

During STS-7, Challenger's second mission (it launched on June 18, 1983), bipod ramp foam came loose. NASA declared the event an "In-Flight Anomaly", which meant the problem had to be resolved before the next shuttle launch, or it had to be proven to pose no risk to the crew. There wasn't much time to resolve the issue: Challengerwas due to launch again on August 30, 1983. At Challenger's Flight Readiness Review (the meeting where everything is verified that the shuttle is safe to fly), the anomaly was declared "not a flight or safety issue." Astonishingly, no one had yet figured out the cause of the foam shedding, and no hazard analysis or engineering attention had been given to the problem. The Flight Readiness Review made its declaration without any evidence to back up its claim. The rationale? The shuttle came back safe, so loss of foam must be safe.

This was a blatant breach of NASA's safety norms. And no one said anything.

On STS-27, the third flight of Atlantis (it launched December 2, 1988), foam from the right-hand solid rocket booster nose hit the orbiter's nose 85 seconds into the flight. This debris then struck the orbiter's window, blocking the crew's view as they ascended into space. There was clearly heat shield damage from the impact. But because of the location of the damage, it could not be completely seen using the shuttle's robot arm. The images that did exist were downloaded to NASA. But this was a classified Department of Defense mission, which meant that the shuttle had to encrypt all its communications. This degraded the images, and left NASA convinced that the thermal protection system was intact. Mission Control told the crew not to be worried. Commander Robert L. Gibson, however, was convinced the shuttle was in serious danger, and secretly resolved to blame NASA publicly over an unencrypted channel if his spacecraft began to break up on re-entry. It turned out that Gibson was right: After landing, more than 700 heat-resistant tiles were found to be damaged, and one was missing entirely. A burn-through almost like the one that doomed Columbia 15 years later had nearly occurred. A post-mission damage report warned NASA that attention to foam varied with the amount of damage the foam made. The report warned that foam was a Safety of Flight Issue, and demanded better monitoring of tile damage to identify any trends. No such improvements were made. The report also recommended that a thorough investigation of foam shedding be conducted. No investigation was made.

On STS-32, on January 9, 1990, Columbia suffered the loss of bipod foam.

On STS-35, December 2, 1990, Columbia suffered a signficant foam shower. In the Flight Readiness Review for the next shuttle mission, NASA declared foam a "re-use or turnaround issue" for the first time, downgrading the risk from Safety of Flight. It did so, even though there was no evidence supporting a downgrade.

On STS-42, January 22, 1992, NASA launched Discovery without having resolved the foam-shedding from the previous shuttle flight. This was the first time NASA launched without resolving a problem first. Discovery suffered a signficant debris shower, incurring 159 hits to its thermal protection system. NASA was later described the incident as "unexplained" or "isolated event", and said foam shedding was "not a safety-of-flight" concern.

On STS-50, June 25, 1992, bipod foam was lost on a Columbia mission. A Hazard Report issued by NASA after the event called the loss an "accepted risk". It did so without any evidence to support the downgrade from Safety of Flight.

On STS-52, October 22, 1992, bipod foam was lost on a Columbia mission. The loss of foam was the fourth bipod event, and went undetected. (NASA would only discover the incident in the post-Columbia disaster investigation in 2003.)

On STS-56, April 8, 1993, a large number of tiles on Discovery were damaged. The after-mission report called the damage "within experience" and declared it an "in-family" issue (that is, one that NASA alone would correct and not report to its contractor or sasfety watchdogs). It did so without any evidence to support the downgrade from Safety of Flight.

On STS-64, September 9, 1994, bipod foam was lost on a Discovery mission. The loss of foam was the fifth bipod event, and went undetected.

On STS-87, November 19, 1997, Columbia suffered 90 hits to its thermal protection system. Among them were 12 holes one inch or larger. NASA characterized the damage as "less than average", even though it wasn't. In fact, the agency was so alarmed that it began testing ways to make foam sturdier. NASA knew it had a huge problem, but refused to reclassify foam shedding a a Safety of Flight issue. NASA knew it had a problem, but refused to fix it before the next shuttle flight. Instead, NASA tried one solution after the other over the next nine flights. None worked. Frustrated, NASA simply declared issue resolved: Foam shedding was an "acceptable risk" as of STS-101 (May 19, 2000) when Atlantis flew.

On STS-112, October 7, 2002, bipod foam was lost on an Atlantis mission. The foam gouged a hole in the insulation on one of Atlantis' Solid Rocket Boosters.

* * * * * * *

Following STS-112, a working group analyzing launch imagery recommended that bipod foam loss be reclassified as an In-Flight Anomaly. In a meeting chaired by Shuttle Program Manager Ron Dittemore, Shuttle Program Integration Manager Linda Ham and a number of other managers decided against that reclassification. Instead, the meeting participants set a a new date for resolving the foam shedding problem. The new deadline was after the launch of STS-107 (Columbia), which meant that STS-113 (Endeavour) would -- like other shuttles before it -- fly without a major saftey problem being resolved. The group appears to have been influenced by the fact that foam had been shed so often, and without any danger to the shuttle, that they now felt foam shedding was an acceptable risk. Moreover, Node 2 of the International Space Station was due to be flown by the shuttle, and holding up launches to resolve the foam issue meant a major public set-back. The team knew this, and decided to risk the lives of the astronauts rather than suffer a public relations disaster.

Endeavor's STS-113 Flight Readiness Review was conducted on October 31, 2002. Dittemore and Ham were present. Jerry Smelser, the Program Manager for the External Tank Project, presented two PowerPoint slides about foam shedding at the reivew. On his first slide, Smelser mischaracterized the foam loss as "never...a 'Safety of Flight' issue". In another bullet point, Smelser concluded that because bipod ramp foam shedding had occurred on only three of the 110 previous shuttle flights, there was little risk of it occurring again. This is a classic risk assessment mistake, similar to assuming that because you've rolled six on a dice twice the likelihood of rolling a six a third time is low. Moreover, Smelser made the mistake of assuming complete data -- which he knew he did not have. There had been large amounts of foam loss in other areas, which Smelser ignored. He knew that no data existed for many launches, but assumed (consciously or not) that foam shedding "must not have occurred". That assumption was completely unwarranted; it meant he was unaware that the shuttle had suffered six, not three, bipod ramp foam losses. On his second slide, Smelser laid out seven rationales for why Endeavor was safe to fly. His first, that the bipod ramp design had not changed since STS-54 (January 13, 1993) was factually incorrect. It had changed with STS-76 (March 22, 1996), which meant that data from launches prior to STS-76 might not be the same as data after it. His fourth (foam work was "performed by experienced practitioners"), fifth (foam "application involves craftsmanship in the use of validated application processes"), and sixth ("no change in inspection") rationales do not logically alter (up or down) the risk of future foam loss. His seventh rationale ("probability of loss of ramp Thermal Protection System is no higher/no lower than previous flights") is the classic risk assessment mistake, repeated from the first slide.

The NASA Headquarters Safety Office also presented a report at the same meeting. The report estimated a 99 percent probability of foam remaining intact around the bipod strut area. This analysis was based on the fact that bipod ramp foam loss had occurred in only two of the previous 48 flights. Once more, a classic risk assessment mistake had been made. The report lessened the risk even more by including right bipod foam loss (which had never been seen), relied on available imagery for right bipod foam (which was far less than for left bipod foam), and limited the calculation to missions between STS-64 (the last time bipod foam had been seen shedding) and STS-112 (the most recent incident). In fact, foam loss was extremely common: Post-Columbia disaster investigations showed that foam was shed on 65 of the 79 shuttle missions for which imagery was available. On 14 of the remaining 34 missions for which there was no imagery, foam loss could be inferred from the placement and number of hits on the shuttle. Bipod foam loss occurred on 10 percent of all missions.

Finally, Endeavor's STS-113 Flight Readiness Review ignored the damage done by foam on STS-112.

NASA requires flight managers to sit in on reviews of the two previous missions, as well as the launch to come after. This helps ensure that flight directors are kept up to date on the latest issues, problems, achievements, and changes. Linda Ham and her managers did so. She was unhappy with the STS-112 Flight Readiness Review, which she strongly criticized as "lousy" in a later email to Shuttle Program Manager Ron Dittemore. It was still "lousy" for STS-113. Nor had it changed for STS-107. Nevertheless, Ham refused to ground the shuttle.

* * * * * * *

STS-107 (Columbia) was launched on January 16, 2003. Foam from the left bipod strut separated from the external tank and struck Columbia's left wing 81.9 seconds after liftoff.

To this day, no one knows why the left bipod foam ramp broke free.

Later image analysis showed that the piece of foam was from 21 to 27 inches long and 12 to 18 inches wide, and relatively thin. The foam was traveling with the shuttle at about 1,568 mph. When the foam detached, it began to slow down (in part, because it was no longer powered, but also because its low density and shape shed energy and thus reduced its speed). Essentially, the orbiter ran into the foam at a relative velocity of about 545 mph. The foam was making about 18 revolutions per second as it fell. Its trajectory was parallel to the shuttle's fuselage, at a 5 degree angle away from the external tank, and with a slight outboard movement. The actual orientation of the panel of foam to the wing at impact cannot be determined.

The foam hit RCC panels 6-left through 9-left. Because the shower of post-impact fragments could not be seen passing over the top of the wing, analysts later concluded that the debris had impacted the left wing below the leading edge.

The debris impact was identified on January 17, the second day of the mission. That same day, an object -- probably a piece of damaged RCC tile -- drifted away from the orbiter. Although military radar observed the piece floating off, this was not discovered until after the accident.

A series of organizational snafus, informal analyses and messages, poor communication, and other errors now deeply compromised NASA's ability to determine if the impact had caused any damage. Managers repeatedly assumed "no damage", even as lower-level engineers became almost frantic with concern. Use of informal channels led upper-level managers to assume there was nothing to worry about. Managers also relied on informal advice from one engineer who was a non-expert in the RCC tiles -- and this engineer gave plenty of advice about how there was "no damage" (feeding their own ego while speaking about issues they were unfamiliar with). Top shuttle program managers focused on sticking to schedule rather than worrying about safety, and repeatedly dismissed attempts to assess damage by saying "there's nothing we can do about it now".

In fact, NASA had plenty of opportunity to determine if Columbia was in danger. Although no launch pad footage of the bipod ramp existed, a Columbia crew member had filmed the bipod ramp from inside the shuttle. NASA could have asked for the complete footage to be downloaded, and did not. NASA could also have asked the Department of Defense to image Columbia using ground-based telescopes or satellites. It never did.

Could anything have been done to save Columbia? Yes. NASA could have ordered a spacewalk, during which the crew would have attempted to pack the damaged area with insulation, metal, and other items. Although this stuff would be burned away during re-entry, Columbia's own internal structures would be saved and the shuttle might -- might -- land safetly. NASA later judged this to be an "extremely high-risk" solution, however. Much more likely, NASA would have sought to send Atlantis into space to rescue Columbia's crew. Atlantis was almost ready to move to the launchpad. NASA would have to had to rush its refurbishment and safety checks, and had a flawless countdown. Columbia carried enough carbon dioxide filters (and more than enough food, water, and oxygen) to allow Atlantis to reach it, with even a few days to spare.

But since NASA never knew about the damage to Columbia, no rescue mission was attempted.

* * * * * * *

At 8:10 AM on Feburary 1, Mission Control notified Columbia's crew they were "go" for de-orbit. Columbia was flying upside down and tail-first. At 8:15:30 AM, Commander Rick Husband and Pilot William McCool executed the de-orbit burn, which lasted 2 minutes and 38 seconds. Husband then turned Columbia on its belly, turned the nose to face foward, and slightly pitched the shuttle nose-up.

Entry Interface (EI), the point at which the shuttle entered discernible atmosphere (about 400,000 feet) occurred at 8:44:09 AM.

At 8:48:39 AM (EI+270 seconds), a sensor inside the leading edge of the left wing showed that the wing was under strain higher than normal. This data was not intended to be sent to Ground Control or to the crew aboard the shuttle. One of Columbia's internal data back-up units recorded the data (and, luckily, the unit survived the crash).

At 8:49:32 AM (EI+323), Columbia excecuted a planned roll to the right. This banked the spacecraft and decreased the orbiter's rate of descent. It also reduced heating on the spaceship's surfaces.

At 8:51:16 AM (EI+427), the damaged RCC tiles on the left wing failed and the wing was breached. Superhot plasma, as hot as 8,000 degrees Fahreheit, now began to flow into the wing. Aluminum melts at 1,200 degrees Fahrenheit, and the interior aluminum strust inside Colubmia's wing began to heat toward the melting point.

At 8:51:46 AM (EI+457), Columbia began to yaw {sideslip) to the left as the damaged front of the left wing created drag.

At 8:52:00 AM (EI+471), temperatures on the leading edge of Columbia's wings reached 2,650 degrees Fahrenheit.

At 8:52:05 AM (EI+476), Columbia's yaw became so pronounced that the shuttle's computers were forced to automatically trim the aileron to compensate. Neither the yaw nor the trim change would have been obvious to Mission Control or the crew.

At 8:52:17 AM (EI+488), the left main landing gear brake line in the wheelwell began to show a temperature rise. This indicated that significant internal damage had occurred to the wing struts.

At 8:53:10 AM (EI+541), four hydraulic sensor cables that ran from the aft part of Columbia's left wing through the wiring bundles outside the wheelwell failed.

At 8:53:38 (EI+569), Columbia's yaw exceeded all previous flight experience.

At 8:53:44 AM (EI+575), the first debris was shed by Columbia. This eight-pound piece became visible from the ground two seconds later.

At 8:53:46 AM (EI+577), people in California watching and filming the shuttle's flight from the ground saw debris being shed from Columbia. The superheated air surrounding the orbiter suddenly brightened, causing a streak in the shuttle's trail. During the next 23 seconds, observers witnessed four distinct events as chunks of debris came off the shuttle. Many of these people, who had seen shuttle landings before, knew something was horribly wrong. In the next 15 seconds, temperatures on the fuselage sidewall and the left Orbital Maneuvering System pod (the housing for one of the main engines) began to rise. Hypersonic wind tunnel tests indicated that the increased heating on the Orbital Maneuvering System pod and the roll and yaw changes were caused by substantial damage around RCC panel 9.

At 8:54:13 AM (EI+601), temperatures in Columbia's wheelwell began to rise rapidly, which indicated that the superheated air coming through the wing leading edge had breached the wheelwell wall. At the same time, observers on the ground noted additional significant shedding of debris. One of these objects probablly weighed about 190 pounds, which means Columbia's physical integrity was now seriously compromised. There was no indication that either Mission Control or the crew knew the shuttle was breaking up.

At 8:54:14 AM (EI+602), as the left wing deformed due to loss of structural integrity, it exhibited increased lift. Columbia began rolling to the right to compensate. Left wing drag was also worsening, causing addiitonal leftward yaw. Observers on the ground noted additional significant shedding of debris ovber the next 30 seconds. There was no indication that either Mission Control or the crew knew the shuttle was breaking up.

At 8:54:24 AM (EI+613), technicians in Mission Control told the Flight Director about the failure of the hydraulic sensors.

At 8:54:25 AM (EI+614), as Columbia crossed from California into Nevada, a bright flash was visible from the ground. Witnesses observed another 18 similar events in the next four minutes as Columbia streaked over Utah, Arizona, New Mexico, and Texas. There was no indication that either Mission Control or the crew knew the shuttle was breaking up.

At 8:54:32 AM (EI+623), several hundred pounds of debris broke free from Columbia. Visible from the ground four seconds later, this is the brightest debris shedding seen by the public. There was no indication that either Mission Control or the crew knew the shuttle was breaking up.

At 8:56:30 AM (EI+741), Columbia initiated a planned roll to the left.

At 8:56:55 AM (EI+766), Columbia completed the planned roll to the left.

At 8:58:03 AM (EI+834), the aileron adjustment exceeded all previous flight experience. This was likely due to additional, severe wing deformation. Ground observers saw more debris break from the shuttle. There was still no indication that either Mission Control or the crew knew the shuttle was breaking up.

At 8:58:20 AM (EI+851), Columbia shed a thermal tile. This became the most westerly piece of debris recovered.

At 8:58:39 AM (EI+870), an alarm light flashed in Columbia's crew cabin and an alarm tone sounded, indicating loss of pressure in the left main landing gear tires. Mission Control received the same alarm. Husband and McCool called up the fault page in the shuttle's computer to review the information.

At 8:58:48 AM (EI+879), the shuttle crew contacted Mission Control. "Roger, uh, Hou --" That was not unexpected, as garbled messages were common during re-entry as antennae adjusted to track the shuttle.

At 8:58:49 AM (EI+880), an alarm flashed and sounded in Columbia's crew cabin, indicating low inboard tire pressure. Mission Control received the same alert.

At 8:58:56 AM (EI+887), an alarm flashed and sounded in Columbia's crew cabin, indicating low outboard tire pressure. Mission Control received the same alert.

At 8:59:06 AM (EI+897), an alarm flashed in Columbia's crew cabin, indicating that the left main landing gear was down. Mission Control received the same alert. This false indicator was due to a failure in the landing gear sensor system. This is the first indication the crew has that something is seriously wrong. Columbia's crew has just 1 minute and 12 seconds left to live.

At 8:59:15 AM (EI+906), pressure was lost on both left main landing gear tires. Mission Control saw the messages and began evaluating the indications. Flight Control radioed Columbia that it did not understand the crew's last transmission.

At 8:59:26 AM (EI+917), an abrupt change in Columbia's aerodynamics occurred as the left wing suddenly deformed significantly. Yaw right and left began to occur, and the left wing's drag worsened even as its lift increased.

At 8:59:29 AM (EI+920), Columbia began to yaw and roll beyond the ability of the aileron to compensate. Two of the shuttle's control jets began to fire continuously to compensate. It was not unusual for the jets to pulse as needed throughout re-entry. But continuous firing indicated a severe problem. Although a light on Commander Husband's control panel lit up, this light was not easily noticed.

At 8:59:32 AM (EI+923), Commander Husband was heard to say "Roger, uh [cut off in mid-word]." It was the last communication from the crew. Telemetry was lost at the same moment by Mission Control. The loss of telemetry was expected, as the shuttle was switching from western antennas to eastern antennas. At this point, sensors inside the shuttle indicated that all but one crew member was in their seat and strapped in. The final crewman was just ingressing their seat. Several crew members did not have their gloves on, and none had their visors closed.

At 8:59:33 AM (EI+924), Columbia's Master Alarm sounded and flashed, indicating the failure of a wire bundle. The crew probably reset the alarm and called up a page in the computer to begin assessing the problem.

At 8:59:36 AM (EI+927), the third control jet began firing continuously.

At 8:59:37 AM (EI+928), the fourth control jet began firing continuously. Ground observers saw a massive brightening of the shuttle. The crew lost all hydraulics and control of the shuttle, and the flaps and aileron began to "float" without power, moving randomly.

At 8:59:46 AM (EI+937), an alarm sounded in Columbia's crew cabin as the nose of the shuttle pitched up uncontrollably. The shuttle's nose was now vertical in the air. As the shuttle began to corkscrew, the crew saw shadows move inside the cabin and saw the horizon disappear. The crew was tossed from side to side, their upper bodies moving forward as they simultaneously sank into their seats. Over the next 34 seconds, the crew experienced ups and downs in G-forces, up as much as 3Gs. A bright piece of debris was also shed at EI+937, followed by a second, dimmer piece two seconds later. More and more debris is shed from the shuttle, visible from the ground as brightening, shining objects separating from the main glow, puffs in the trail, and splitting of the trail.

At 8:59:49 AM (EI+940), pieces of Columbia's left main engine pod began to be shed.

At 9:00:03 AM (EI+954), Columbia's autopilot was turned on. This indicates that either the commander or the pilot was still executing commands.

At 9:00:04 AM (EI+955), the left wing separated from Columbia. The orbiter began to fall out of control at 10,000 mph.

A short time after 9:00:05 (EI+956), the Pilot McCool attempted to restart the auxiliary power units (APU). He must have naturally assumed that the loss of control was due to an APU failure, when in fact the hydraulic system itself had ruptured due to the heat.

At 9:00:18 AM (EI+969), Columbia began disintegrating. People on the ground heard a loud sonic boom. A major brightening occurred, indicating the shuttle was breaking up. The forebody (including the crew cabin and nose) separated from the midbody, causing all power to be lost in the crew cabin. The shuttle's break-up occurred not because of stress, but because heat had significantly damaged the shuttle aft of the crew cabin. The first crack appeared on the right. The crew cabin remained attached to the midbody for a few seconds before finalling pulling free.

At 9:00:25 AM (EI+976), the two major pieces of Columbia became discernible from the ground. The crew would have felt a jolt as the crew cabin pulled free from the rest of the shuttle. But in breaking free, the cabin was also damaged, and a number of small breaches above and below the crew deck occurred. The crew cabin depressurized in two or three seconds. None of the crew had their helmet visors down and locked. So although their spacesuits were feeding them air, the air was being sucked away. At least two crew members did not have their gloves on (which would have exposed them to more decompression effects), and one did not have their helmet on. Autopsy data showed that none of the crew managed to close their visors. All of the crew suffered from depressurization effects, which would have rendered them unconscious within just a few seconds. Their breathing would have stopped as well, although heart and brain function would have continued for another two to four minutes. The crew cabin also began tumbling. Whenever a crack in the cabin faced forward or downward, superhot plasma invaded the crew module at breach-points. Melt near the cracks melted, and globs of molten metal were found on the crew's spacesuits, seats, and restraining straps. (There was no indication that the interior of the crew module was hot enough to cause death.) The crew module's rotation increased to one every two seconds, with internal stresses as low as -1G and as high as 5Gs. Crew harnesses either failed to lock, were blocked by debris, or improperly functioned. If any crew were still alive, they now suffered lethal neck and upper-body injuries as their heads and torsos snapped around in the spinning capsule.

At 9:00:53 AM (EI+1004), a significant brightening event began around Columbia's forebody, indicating that it was breaking up. First, the nose came loose from the crew module, exposing the crew's bodies to air and fire. Within 10 seconds, the crew module itself began to come apart. The upper flight deck stayed intact five seconds longer than the lower deck. The crew capsule completely disintegrated by EI+1019. The terrific speed of the air stripped the spacesuits from the crew's bodies, causing additional injury. The wind and stress then dismembered the crew.

At 9:01:10 AM (EI+1021), Columbia's crew module and midbody disappeared from video, as they had slowed enough for the plasma around them to dissipate.

At 9:35:00 AM (EI+3051), all debris from Columbia reached the ground. It had taken half an hour for the debris to fall the 39 miles to the Earth.

* * * * * * *

More than 25,000 people (nearly all of them volunteers) took part in recovery operations. Searchers covered more than 2.3 million acres, an area the size of Connecticut. More than 84,000 individual pieces were located, representing 38 percent of Columbia. The largest pieces of debris traveled the farthest, landing in western Louisiana. Most debris tended to land in clumps, or "debris fields". More than 2,000 debris fields were located.

Searchers also found human remains. Among the remains were arms, feet, a heart, legs (some intact, some partially intact, some just bone), a torso (ripped in half at mid-chest), and a skull (stripped of all flesh). Some remains showed signs of exposure to high heat, such as scorching. More remains than this were found, but NASA swiftly censored any news about them out of respect for the families. DNA testing showed that remains from all seven astronauts were recovered. These remains were turned over to the respective families.

Based on the post-accident recovery of debris, NASA determined that it was possible for astronauts to have survived -- had they had better spacesuits, better oxygen and internal heat, and an automatic escape system.

* * * * * * *

Commander Rick Husband's remains were buried at Llano Cemetery near Amarillo, Texas. Pilot William McCool's remains were buried at the U.S. Naval Academy in Annapolis, Maryland. The remains of Michael P. Anderson, David M. Brown, and Laurel Clark were buried at Arlington National Cemetery in Arlington, Virginia. Kalpana Chawla's remains were cremated, and her ashes were retained by her family. The remains of Ilan Ramon were buried Moshav Nahalal Cemetery near Nahalal, Israel.

President George W. Bush signed legislation on April 16, 2003, which authorized Arlington National Cemetery to erect a memorial to those lost in the Columbia disaster. Artist Barbara Prey designed two bronze plaques, which were affixed to a grey granite marker. The memorial was dedicated on February 2, 2004.

* * * * * * *

Ron Dittemore resigned from NASA in April 2003. He was appointed president of ATK Launch Systems Group, which later was named the prime contractor for the first stage of NASA's Ares I rocket. He is currently retired.

Linda Ham was reassigned after the disaster, and appointed an assistant to the director of engineering at the Johnson Space Center. She left that position in December 2003 to take a job at NASA's National Renewable Energy Laboratory. She took leave from NASA from 2004 to 2006, then returned to the Johnson Space Center and was named Technology Integration Manager and then Transition Manager for the Constellation rocket program.

No NASA personnel were fired or reprimanded for their actions leading up to the Columbia disaster. Most remained at NASA, and moved up in the hierarchy afterward.

No comments:

Post a Comment