O-Ring Seals Were Below Freezing

By Michael Isikoff
Washington Post Staff Writer
Monday, February 24, 1986

The temperature of the key rocket booster part implicated in the explosion of the space shuttle Challenger was about 29 degrees Fahrenheit at the time of liftoff -- about 9 degrees colder than surrounding air temperatures, according to new National Aeronautics and Space Administration studies.

The findings, not yet publicly released, cast new light on the prelaunch warnings by engineers from rocket maker Morton Thiokol Inc. that the Jan. 28 mission should not proceed because of the unusually cold weather.

NASA and Thiokol engineers had known for nearly a year that the rubbery O-ring seals that join segments of the booster rocket had suffered erosion in 50-degree temperatures recorded during the previous coldest shuttle launch in January 1985.

In effect, the new findings indicate that the actual surface temperature of the O-rings was far colder than their safe operating range, thereby bolstering theories that a failure of the rings allowed flames to escape from the rocket, causing the explosion. At the same time, the findings do not support recent reports of extraordinarily low temperatures -- 7 and 9 degrees -- on another part of the rocket.

The findings of subfreezing surface temperatures on the critical right solid rocket booster are based on complex calculations taking into account a variety of factors -- including wind and sunlight angles.

Richard P. Feynman, a member of the presidential commission investigating the disaster, said in an interview today that the subfreezing temperatures could have been deduced by NASA because air temperatures were in the 20s a few hours before liftoff and about 31 degrees only a half-hour before launch.

Feynman said it was "no miracle" that the O-ring temperatures were as low as 29 degrees because "things don't heat up that fast."

At the same time, the commission member discounted previous reports that there were surface temperatures as low as 7 and 9 degrees on parts of the right rocket. These temperature readings, recorded on hand-held infrared sensors by a NASA ice team, have now been determined to have been unreliable because the instruments were not given enough time to stabilize, he said.

Calculations by NASA and the presidential commission last week indicate that the actual temperatures on those parts of the right rocket were about 16 and 19 degrees. Yet Feynman said that even these temperatures are not "relevant" to the inquiry into what caused the disaster because they were on a different side of the right rocket from the suspected O-ring failure.

Even less relevant, he said, were reports that winds blowing over the shuttle's supercold external fuel tank caused temperatures there to dip to 8 degrees below zero. New calculations indicate the temperature was about 2 degrees, a figure that could be explained by the 420-degree below zero temperatures inside the tank.

Nevertheless, the temperatures are lower than expected and the presidential commission is still concerned as to why they were not reported to NASA officials who gave the final green light for the launch, according to commission sources.

The temperature readings, taken about three hours before launch, were mentioned over an internal radio used by the ice team but never reported to top launch officials in the "firing room" here.

"There's a real question that if they saw something as peculiar as they saw, shouldn't they have reported it higher up," said one source close to the commission. "Apparently, they felt it was their own numbers, that it was not that important."

In a related development today, the Navy reported that it had recovered a 200-pound piece of debris thought to be from the shuttle's external fuel tank. The wreckage was found 100 feet below the surface about 25 miles east of here.

© 1986 The Washington Post Company