STS-51-L: what went wrong

About eight hours after Challenger and its crew were lost, President Reagan addressed the nation on television; by a quirk of circumstance, a State of the Union address had been scheduled for January 28 well before the shuttle launch was delayed to this date. Reagan’s emoting about the fallen astronauts, for once, didn’t look glib or forced, yet I still thought with a trace of guilt that he nevertheless looked like something of a bumbling ass, prating on about American glory when this was theoretically the furthest thing from the minds of citizens of all of Planet Earth. 20 years later I would come to understand that this sort of mien, an apple-shining grin underlain by a trace of wanting to just get back to the damned ballgame, is perhaps obligatory in order to maintain a proper presidential bearing.

In the weeks following the disaster, when the extended Concord High School community became as morbidly and righteously — if not quite as dispassionately — drawn to the whys and the hows of the disaster as everyone else, it became obvious that the decision to launch the Shuttle in freezing conditions had been a lethal miscalculation.

Rapidly materializing in the relentless scope of public accusation and blame was Morton-Thiokol International (now ATK Launch Systems), the manufacturer of the infamous O-ring that became virtually synonymous with the corporate name. Despite hearing the term too many times to count in the aftermath of the explosion, few people I knew bothered to learn what an “O-Ring” actually was or what it was for, despite its being as simple a thing as it sounds.

In those days, the Concord Monitor was an afternoon newspaper, and the headline on January 28, 1986, succinct, aghast and in a Times-style font, stretched entirely across the width of the page: “Shuttle Explodes.” Although no one was rushing to point such things out at the time, this is technically incorrect. From a physical standpoint, the rapid chain of events — the rupture of a seal in Challenger’s right solid rocket booster (SRB), severe damage to its external fuel tank as a result of exposure to flames from the booster, and the consequent disintegration of the orbiter itself — are well understood.

In the hours before the launch, engineers from both NASA and Morton-Thiokol expressed concern about the cold-weather performance of the booster O-rings, arguing that it might not do its job below roughly 53 degrees Fahrenheit and could therefore allow fuel to escape. The low temperature at Cape Canaveral on the early morning of the 28th was about 30 degrees. When the O-ring failed, effectively shredding the spaceship and killing seven astronauts, it was exactly what engineers had warned could happen. In addition, engineers at Rockwell International, the civilian contractor responsible for most of the Shuttle’s construction, warned that chunks of ice that had accumulated on the launch “pad” itself could cause problems during lift-off, either structurally or as a result of being sucked into the engines (as in a classic airplane bird strike). In the end, these various admonitions were disregarded by both NASA and Morton Thiokol managers, and the launch was initiated in late morning.

Here, diagrammatically, is what the shuttle system consists of: orbiter, two SRBs and an external fuel tank.

The shuttle is powered by a triad of engines which generate approximately 400,000 pounds of thrust at sea level. Each solid rocket booster, meanwhile, produces 3.3 million pounds of thrust on the launch pad, making it the most powerful rocket in the history of human aerospace engineering; its chief fuel is aluminum. These are shed at about two minutes into a shuttle flight (partially explaining why I and others confused the vehicle breakup with a normal event). The external tank contains a mixture of liquid hydrogen and liquid oxygen that weighs almost twenty tons at launch. The boosters are re-usable, but the external tank — dropped from the shuttle moments after the engines are shut down — is not.

Hydrogen and oxygen are simple, lightweight elements, unlike the mixtures of long, unbranched hydrocarbons most commonly used to move civilians around in workaday motorcraft. If the shuttle used a conventional fuel, an age-old problem in rocket propulsion would arise: The weight of the fuel alone would be prohibitive in creating sufficient thrust to allow the orbiter, its payload, and the fuel and fuel tanks themselves to escape Earth’s gravitational clutches. But hydrogen, though it must be cooled to about -250 C to be converted to a liquid, has only about 1/14 the density of water. Its use in rocketry is indispensable.

Overnight, at least one O-ring had not re-expanded completely after contracting in the cold and becoming brittle — this is basic materials science, folks. Although it was not evident at the time, the actual failure of this ring occurred at ignition, toward the back of the right SRB. At this point, oxides produced from the combustion of propellant acted to temporarily seal a gap that otherwise would have allowed the unchecked escape of the extremely volatile booster contents. But as the shuttle rose and picked up speed, the stopgap plug could no longer withstand the shear forces exerted by atmospheric components, even five, six and finally eight miles above the earth.

About 58 seconds into the flight, a burst of flame shot out of the SRB (image). The shuttle continued on for another 10-12 seconds, during which time the flame burned a hole in the external tank. At T + 73 seconds, the tank failed and began to disintegrate; in short order, the beleaguered orbiter, tossed into an unanticipated flight pattern, was blown apart by forces far in excess of those it was built to withstand. The crew members probably lived until the crew cabin, designed generally in the manner of a NASACR driver cage in that it was in some ways a sturdier kernel than the shell of orbiter around it, smashed into the ocean at over 200 miles per hour. They may have been unconscious when the cabin hit the water owing to depressurization in the moments after the structure came apart, but given that three of the four “emergency response packs” in the cabin were found to have been deployed, it is virtually certain that — apocalyptic plumes of smoke notwithstanding — the sudden loss of integrity did not kill the astronauts outright.

The Challenger disaster put the shuttle program on ice for almost three years, and thanks mainly to human factors, a mishap of its magnitude may have been inevitable. The shuttle was never built to allow for a reasonable chance of escape by astronauts in the event of a catastrophe, with the apparent, overconfident aim being to avoid such unsavory things in the first place; yet NASA, as I’ll discuss in the next and final post in this burst of shuttle-related logorrhea, has never taken every step to minimize the chances of such a catastrophe. Taken together, and in the wake of the less-celebrated but equally tragic Columbia explosion of 2003, these factors have spelled the end of the Space Shuttle program, which saw its final launch in July 2011 (Atlantis, STS-135) and will, if all goes as planned, be superseded by the Orion Space Vehicle.

Advertisements
  1. #1 by Anonymous on January 30, 2007 - 1:20 am

    Overnight, at least one O-ring had not re-expanded completely after contracting in the cold and becoming brittle — this is basic materials science, folks.

    I think it has been pointed out that these O-rings have demands not usually put on O-rings – AFAIK they aren’t supposed to seal without the pressure difference, but creeps and deform until sealing. (At least, it would explain the numerous resealings I had to do on vacuum chambers. Most of them without any visible obstructions in the O-ring track.)
    So IIRC combustion material likely always help seal initially. Again – it isn’t rocket science. Or perhaps it is.

    The shuttle was never built to allow for a reasonable chance of escape

    The shuttle is a kludge resulting from conflicting demands. (Yet it is still a beautiful machine.) If the military applications orbit and load requirements hadn’t been thrown in to support its economy, it would have likely resulted in a safer vehicle.
    Another thing that is striking is that it hasn’t adopted best of breed practices. The Buran was superior in safety in many ways (tile thermal load resistance and durability, return flight properties and safety), yet that didn’t seem to warrant trying to purchase available materials and refining the designs towards better safety. The not-invented-here syndrome or merely the usual politics-is-more-important-than-men view?

  2. #2 by Anonymous on January 30, 2007 - 1:20 am

    Overnight, at least one O-ring had not re-expanded completely after contracting in the cold and becoming brittle — this is basic materials science, folks.

    I think it has been pointed out that these O-rings have demands not usually put on O-rings – AFAIK they aren’t supposed to seal without the pressure difference, but creeps and deform until sealing. (At least, it would explain the numerous resealings I had to do on vacuum chambers. Most of them without any visible obstructions in the O-ring track.)
    So IIRC combustion material likely always help seal initially. Again – it isn’t rocket science. Or perhaps it is.

    The shuttle was never built to allow for a reasonable chance of escape

    The shuttle is a kludge resulting from conflicting demands. (Yet it is still a beautiful machine.) If the military applications orbit and load requirements hadn’t been thrown in to support its economy, it would have likely resulted in a safer vehicle.
    Another thing that is striking is that it hasn’t adopted best of breed practices. The Buran was superior in safety in many ways (tile thermal load resistance and durability, return flight properties and safety), yet that didn’t seem to warrant trying to purchase available materials and refining the designs towards better safety. The not-invented-here syndrome or merely the usual politics-is-more-important-than-men view?

  3. #3 by Anonymous on January 30, 2007 - 1:20 am

    Overnight, at least one O-ring had not re-expanded completely after contracting in the cold and becoming brittle — this is basic materials science, folks.

    I think it has been pointed out that these O-rings have demands not usually put on O-rings – AFAIK they aren’t supposed to seal without the pressure difference, but creeps and deform until sealing. (At least, it would explain the numerous resealings I had to do on vacuum chambers. Most of them without any visible obstructions in the O-ring track.)
    So IIRC combustion material likely always help seal initially. Again – it isn’t rocket science. Or perhaps it is.

    The shuttle was never built to allow for a reasonable chance of escape

    The shuttle is a kludge resulting from conflicting demands. (Yet it is still a beautiful machine.) If the military applications orbit and load requirements hadn’t been thrown in to support its economy, it would have likely resulted in a safer vehicle.
    Another thing that is striking is that it hasn’t adopted best of breed practices. The Buran was superior in safety in many ways (tile thermal load resistance and durability, return flight properties and safety), yet that didn’t seem to warrant trying to purchase available materials and refining the designs towards better safety. The not-invented-here syndrome or merely the usual politics-is-more-important-than-men view?

  4. #4 by Anonymous on January 30, 2007 - 1:49 am

    Umm, actually, since the Buran apparently did its first flight in “snow flurries with 20 m/s winds” (while capable of lifting a Buran loading 5 ton more than the shuttle), I think the Energia lifting system may have been best of breed too, at least concerning adverse weather conditions.

  5. #5 by Anonymous on January 30, 2007 - 1:49 am

    Umm, actually, since the Buran apparently did its first flight in “snow flurries with 20 m/s winds” (while capable of lifting a Buran loading 5 ton more than the shuttle), I think the Energia lifting system may have been best of breed too, at least concerning adverse weather conditions.

  6. #6 by Anonymous on January 30, 2007 - 1:49 am

    Umm, actually, since the Buran apparently did its first flight in “snow flurries with 20 m/s winds” (while capable of lifting a Buran loading 5 ton more than the shuttle), I think the Energia lifting system may have been best of breed too, at least concerning adverse weather conditions.

%d bloggers like this: