Challenger 51-L: O-Rings, a Cold Launch and the Decision to Go

Seven crew members died on January 28, 1986, when the Space Shuttle Challenger broke apart on mission STS-51-L, 73 seconds into flight according to the launch call of "Challenger, go at throttle up" and the loss of data moments later. A presidential commission concluded in its final report of June 6, 1986 that the disaster was caused by the failure of a rubber o-ring on one of the solid rocket boosters to seal properly, due in large part to freezing temperatures before liftoff. The temperature at the 11:38 a.m. liftoff was 38 degrees, 15 degrees colder than any previous launch. The Commission also identified serious flaws in the launch decision-making process.

Background: a shuttle sold as a space truck

By the early 1970s funds for NASA were drying up, and plans for a space station and crewed missions to Mars were postponed indefinitely. The reusable Space Shuttle, originally meant to be a supporting ferry to the station, was thrust into the spotlight. Congress and the Nixon administration were sold on a system so reliable that it would be a moneymaker. The shuttle was designed to fly up to sixty missions a year. Robert Hotz, editor of Aviation Week and Space Technology and a member of the Presidential Commission, called the idea of a space airliner a cargo truck a "grand illusion".

President Reagan declared the system fully operational after the fourth orbital test flight. Yet only nine missions were flown in 1983 and 1984 combined, and launches were repeatedly delayed.

The joint that was known to be flawed

In 1973 Thiokol won the contract for the reusable boosters, although it had been ranked lowest in actual engineering design. The concern centered on the field joints of the twin solid rocket boosters. Each booster was a stack of four cylinders, each 12 feet wide and over 25 feet tall, assembled at the Kennedy Space Center. The tang of the upper segment slid into the clevis of the lower, and each joint was sealed by two synthetic rubber o-rings about a quarter of an inch thick. Thiokol described the second as a redundant backup.

Recovered boosters showed the joints were not sealing as intended. In a phenomenon called joint rotation, the segments bulged at ignition and the tang and clevis bent away from each other instead of toward each other. NASA's own engineers at the Marshall Space Flight Center had predicted this. In a 1979 memo, senior engineer John Miller called the Thiokol design completely unacceptable, noting that the secondary o-ring could become disengaged from its sealing surface. In late 1982 NASA reclassified the joint as criticality 1, with the failure effect "loss of mission vehicle and crew". Managers nonetheless kept calling the opening an "extrusion gap" and treating the secondary seal as redundant.

Warnings in 1985

• January 24, 1985: Discovery launched on mission 51-C in the coldest launch on record up to then, at 53 degrees. Thiokol's Roger Boisjoly, the company's top expert on the field joints, found massive hot gas blow-by in two joints, over 110 and 80 degrees of arc.
• April 29, 1985: After mission 51-B, an inspection of a nozzle joint in May showed two-thirds of a secondary o-ring eroded away and the primary seal destroyed.
• July 1985: Boisjoly wrote a memo to Thiokol's vice president of engineering, Bob Lund, warning of "a catastrophe of the highest order", loss of human life. It was stamped company private and never reached NASA. A seal task force was formed at Thiokol.
• July 23, 1985: NASA headquarters budget analyst Richard Cook wrote a memo saying failure in flight would certainly be catastrophic. Headquarters engineers had told him the shuttle "could blow up" and that they held their breath every time it launched.
• August 19, 1985: At a headquarters briefing, Thiokol and Marshall proposed a quick fix and recommended that flights continue. Houston's program office, and so the astronaut office, was not invited. A slide stating that cold temperatures aggravate the o-ring problem was reportedly deleted by Marshall before the presentation. The decision was that flights would continue.

The Commission concluded that the briefing was detailed enough to require corrective action before the next flight. Commissioner Richard Feynman described the pattern: a flight with blow-by flew and nothing happened, so the risk was assumed lower, "a kind of Russian roulette". By October the Thiokol seal task force felt it was a paper tiger without real authority, and its chair Bob Ebeling wrote a memo that began with the word "help" and ended "this is a red flag".

The night of January 27

Challenger's mission 51-L carried Christa McAuliffe, chosen from 142 applicants as the first teacher in space. The launch slipped from January 26 to the 27th and then the 28th, the day of President Reagan's State of the Union address. A cold front brought a predicted launch temperature of 29 degrees.

Thiokol engineers met and agreed to try to stop the launch if the cold materialized. In an afternoon call Boisjoly said he did not want to fly; the data suggested 53 degrees was the limit. A telecon began at 8:45 p.m. Thiokol engineers argued that cold, harder o-rings would seal more slowly. Boisjoly could not quantify his concern. Vice president Joe Kilminster declined to recommend launch, and Marshall's George Hardy said he was appalled but would not recommend launching over the contractor's objection. Marshall's Larry Mulloy asked, in effect, when Thiokol wanted him to launch, noting that there was no launch commit criterion for joint temperature.

Thiokol asked for a five-minute offline caucus, which lasted about 30 minutes. Six days earlier NASA had announced it was seeking a second source for the boosters, putting Thiokol's contract, worth over $1 billion, in question. In the caucus, engineers Boisjoly and Arnie Thompson tried to be heard, and general manager Jerry Mason told Lund to take off his engineering hat and put on his management hat. The poll of the managers was four votes to launch, none against. At about 11 p.m., Thiokol recommended launch and, at Reinartz's request, faxed it in writing.

Thiokol's Al McDonald later raised the ice on the pad and the storm in the booster recovery area, and said he thought the concerns would be passed on. Stanley Reinartz testified that he made the decision not to escalate the issue to level 2.

The temperature at the joint was measured at 8 degrees overnight, which was not reported. Early on January 28 an ice inspection team found the pad covered in ice. Rockwell's Rocco Petrone said Rockwell could not assure that it was safe to fly, but NASA proceeded. The Commission concluded that NASA appeared to be requiring Rockwell to prove it was unsafe rather than safe.

Launch and loss

At ignition the joints flexed, and the cold, hardened o-rings separated from their sealing surfaces and were vaporized. A fragile seal formed by the products of combustion stopped the leak. About 60 seconds later it broke, and escaping gases burned at the external fuel tank "like a giant blowtorch". The tank ruptured and the orbiter broke apart. The crew compartment emerged intact, and the seven astronauts survived the breakup but died when the compartment struck the Atlantic nearly three minutes later. Flight control called it "obviously a major malfunction". Boisjoly, who had refused to watch, joined colleagues in the viewing room and whispered when the vehicle cleared the tower that they had dodged a bullet.

The crew named in the tribute were Onizuka, McAuliffe, Jarvis, Resnik, McNair, Scobee and Smith.

Investigation

Boisjoly and Thompson were sent to the failure investigation team at Marshall and found NASA management trying to find something other than low temperature as the cause. On February 3 Cook wrote a memo saying there was a growing consensus that a field joint burn-through caused the explosion, and that it had probably been preventable. When his supervisor took the copies back, Cook took his papers to the New York Times, whose front-page story on February 9 gave the first detailed explanation of how the joints failed. Reporter Philip Boffey shared a Pulitzer Prize that year.

President Reagan announced the Presidential Commission, chaired by William P. Rogers with Neil Armstrong as vice chairman, and gave it 120 days. In the first closed hearing on February 6, Judson Lovingood told the Commission that Thiokol had ultimately recommended launch, without mentioning its objections. At the February 10 closed session, as Rogers was about to adjourn, McDonald revealed that Thiokol had originally recommended not launching below 53 degrees. Commissioner Hotz said the reaction was that this would be worse than they thought.

In a public session on February 11, Feynman put o-ring material in ice water and showed it did not spring back at 32 degrees. The day before Cook testified, Michael Weeks described him to the Commission as a financial person who was not very knowledgeable about the whole program.

The Commission also concluded that NASA and contractor management first failed to recognize the joint as a problem, then failed to fix it, and finally treated it as an acceptable flight risk.