STS-107 Columbia Tragedy: Were There Options?

Nine years ago, space shuttle Columbia broke up during re-entry, when hot gasses breached its left wing. Re-entry temperatures reached probably around 3000 degrees fahrenheit (just over 1600 degrees Celsius). Today, with the shuttle program's end, NASA envisions manned flights to the moon and Mars via the Orion crew module, while SpaceX's Dragon plans to replace the shuttle as a ferry to the International Space Station. Orion is expected to experience up to five times the thermal effects felt by shuttle astronauts. What technologies can make re-entry less hazardous?

The farther into space a ship ventures, the higher the re-entry temperatures generally experienced. Stardust, an unmanned comet probe, experienced (and survived) re-entry temperatures of at least 2500 degrees Celsius (nearly 5000 degrees Fahreheit).

Heat Protection

With the shuttle, NASA opted for traditional (aircraft) aluminum alloy, insulating it with silicate coating according to NASA's Dethloff, rather than exploring new heat-resistant metals and distributing re-entry heat throughout the "skin" (without insulation).

NASA and Rockwell used silica thermal protection tiles and blankets for much of the shuttle, with porous silica tiles (sometimes containing alumina-borosilicate fiber) coated with borosilicate glass for high-heat areas, particularly on the underside (the craft was angled upwards during re-entry), and reinforced carbon composite (RCC) panels for the nose cone and wing leading edges. Carbon withstands re-entry temperatures of roughly 1649 degrees Celsius.

Although designed to be reused, tiles and some thermal blankets were damaged on missions, and tiles fell off. RCC panels were damaged during STS-107 re-entry. Of 135 shuttle missions, two, including STS-107, ended in crew death.

Orion's Apollo-Era "Avcoat Ablator"

United Space Alliance's Orion will, like previous craft, use silicate covering. The covering however is not reusable shuttle tile but a burn-off Apollo-era disposable shield of a phenolic (phenol resin) honeycomb reinforced with fiberglass, with silica-fiber and epoxy-novalac resin filler. Orion does not use carbon matrix PICA-X (chosen for SpaceX's Dragon, below), because doing so would require multiple tiles, which the Orion-team has considered risky.

Phenolic Carbon PICA-X

SpaceX's Dragon is equipped with a reusable PICA-X heatshield. PICA-X is SpaceX's improvement of NASA's "PICA," a fibrous carbon matrix "impregnated with" a phenolic resin, developed for the Stardust mission to the comets.

Columbia Left Wing Damage

Analysis of Columbia's recovered left wing (Mayeaux et. al., 2004) suggests that the breach was on the lower side, that the first materials to be exposed and melt were insulators behind the RCC panel. Aluminum was not detected in what is believed to be the initial melt. This suggests that the aluminum wing (support) spar melted late.

There was apparently some "overpressure" on the launch pad, and slight overheating of Columbia's left solid rocket booster, although overheating was not exceptional. There was, according to CAIB, however, significant overheating of the left wing right after launch, perhaps seconds after the time of the debris event observed in reviews of the launch.

Efforts to Save Columbia

The military was ready to take photographs of Columbia in orbit, NASA's Michael Card told supervisors, according to Cabbage and Harwood. NASA simply had to declare an emergency. Senior NASA managers quashed this idea. In any case, no photograph may have been close-up enough to show damage. A photograph of the top of Columbia's left wing during STS-107's orbit appears online. It does not show damage, which was apparently on the wing's underside. Houston did model flat tire landings (supposedly coincidentally).

Because Columbia did not return intact, it's impossible to further analyze other in-flight issues, including a water leak on the left aft side of the brand new Spacehab water separation module, and odd current signatures in the left payload bay door motor and other left-side motors (when current was slugglish; these signatures apparently first manifested seconds before launch).

In-flight Repairs: Apollo 13

During Apollo, Apollo 13's service module had an oxygen tank blow (due to improper modifications to the tank). The module soon lost all power as remaining oxygen leaked from the damaged surviving tank, leaving the mission little for return to Earth but what was available in the fortunately yet unused lunar module.

It's a tribute to NASA's Mission Control, to the astronauts, to safeguards built in after the Apollo 1 fire (which apparently prevented short circuits in the ultimately quite-wet-from-condensation Command Module wires), and to perhaps luck with Command Module batteries that the return was successful.

Salvaged lithium hydroxide canisters (for filtering carbon dioxide) from the service module were not designed for hook-up in the lunar module, but NASA guided the astronauts through a "makeshift" hook-up when carbon dioxide alarms sounded. And a decision to tow the "blast-gutted" service module 300,000 miles back to Earth possibly protected the command module's heat shield from damage in space's cold void.

Because of efforts to conserve power on the return, temperatures in the lunar module dropped to 40 degrees fahrenheit by trip end, with moisture condensing on windows and walls.

Columbia: Makeshift Repair Possible?

In the Columbia Accident Board (CAIB) STS-107 investigation, CAIB described hypothetically stuffing any cavity in Columbia's left wing with tools and other metal, plus bags of water, covering this with a thermal insulation panel removed from the payload bay door, and finally securing all with a "teflon foot loop" to "insure" everything stayed in place before return. Then the repair would have been "cold-soaked" in space to freeze the containers of water stuffed into the cavity. Ideally, said the Board, this might have delayed hot gasses' entry into the wing structure itself, although CAIB surmised that the patches themselves would have melted or evaporated.

Another "fix" CAIB envisioned post-accident involved breaking up tiles in "non-critical" places and using these to seal spaces between panels. A tight fit may have been possible, although many factors would have affected it, and there could have been no guarantee. It's also not known how well the tile pieces would have adhered or what was needed to make them do so.

What More To Do?

Saving money was a theme throughout shuttle development according to NASA's Dethloff. However some of the "fixes" below, including ejectable seats (already installed on STS-1, these just needed to be retained) and perhaps even a repair kit or makeshift repair, could have been done at low cost.

Mission to Save Crew Possible?

Physical inspection by astronauts, particularly without proper suits, posed risks but was perhaps possible according to Shayler (2009). Physical inspection was almost the only way to check Columbia, with no cameras integrated with the shuttle's robotic arm (used to move cargo from the payload bay into space for experiments).

Columbia had until February 15 to complete its mission, says Shayler. This was fixed by the available amount of lithium hydroxide on board. Other consumables (oxygen, water, food, propellant) could have been stretched longer.

Atlantis Rescue?

Atlantis was being prepared for a March 1 launch (STS-114). Shayler believes Atlantis could have been moved to the Vertical Assembly Building by January 29th and needed cargo installed on the pad before February 17th. And Atlantis might have been rolled out as early as January 26th, with some things skipped, with perhaps a February 10th launch, if a decision could have been made by January 20th, and preparations begun January 21rst.

Shayler thinks a rescue crew (a minimum crew of four) might have been formed quickly from the astronauts available. Recently returned or currently training Atlantis crewmembers might have been used. There were launch windows on February 9th, not quite so good windows on the 11th and 12th, with a possibility for rendezvous (assuming all was go) on February 13th.

There would of course have been no time to correct for the kinds of issues that apparently caused the Columbia break up. Shayler does wonder what sort of flack NASA would have received had anything gone wrong with a rescue mission.

NASA Changes

Changes made since STS-107, besides introduction of a repair system for tiles (but not for carbon composite fabric/panels on the nose cone and leading edges) which flew on the next shuttle flight in 2005, include more attention to corrosion, pre-launch and during-mission inspection of the orbiter, better seat restraints, and a system for tracking debris at take off.

SpaceX Safety

One concern about the shuttle was its lack of escape system. SpaceX's Dragon (with its low quoted costs) command module claims a "built-in escape mechanism" with eight Draco engines (should one fail the craft should fly). Some however have expressed concerns about SpaceX safety too, including information architecture firm Valador's head Joseph Fragola, who emailed NASA with his concerns. SpaceX countered and sued but settled out of court when Fragola was proven wrong.