Despite longstanding concerns about the design of the space shuttle's O-rings, NASA chose not to monitor their performance in flight, even though, according to rocket engineers, it should have done so.
In-flight monitoring could have yielded insights into how well the synthetic rubber O-rings prevented escape of hot gases through joints in the solid-rocket booster.
"You would have learned a hell of a lot," said J. Edmund Fitzgerald, a Georgia Institute of Technology professor who was a consultant for Morton Thiokol Inc., the booster manufacturer, on the design.
That information conceivably could have alerted National Aeronautics and Space Administration and Thiokol engineers to potential problems with the O-rings, prime suspects in the Jan. 28 Challenger accident. Investigators say they believe that one or both of the two O-rings in the right solid-rocket booster failed to seal, allowing gas burning at 5,900 degrees Fahrenheit to spew out the side.
While it would be nearly impossible to attach sensors to the O-rings, Fitzgerald said, their effectiveness could easily have been gauged by monitoring pressure in the booster joints.
Fitzgerald and other solid-rocket engineers argue that a simple pressure sensor plugged into the inspection port in the booster segment joints would yield a millisecond-by-millisecond profile of how well the O-rings were coping with pressures generated by the launch.
"If the sensors indicated full pressure between the O-rings, then you know the primary O-ring never worked and the second one could be working," Fitzgerald said. "If there's 100 to 200 pounds-per-square-inch pressure, you'd know the inner O-ring didn't seal immediately" and if pressures changed later, "you'd know there was blowby," or erosion of the O-rings permitting a leak.
That sort of real-time data would complement NASA's practice of analyzing the O-rings recovered after flight, he said.
NASA and Morton Thiokol check booster joint pressures in their "static testing" facilities in Utah where the O-rings and the solid-rocket booster segments are subjected to flight simulations. However, NASA engineers agree that there can be significant differences between static testing results and actual launch conditions.
Testimony during the presidential commission hearings Friday revealed that the problem with the O-rings was being investigated last July, but no effort was made to obtain flight data of O-ring performance. The study relied exclusively on ground testing and analysis of O-rings recovered after flight.
According to engineers at Kennedy Space Center here, the shuttle's solid-rocket boosters never had a joint pressure monitor, even during initial test flights.
Similarly, documents and discussions with NASA engineers at the Marshall Space Flight Center in Huntsville, Ala., indicate that monitoring O-ring performance in flight was never seriously considered.
"We did not instrument a joint because I didn't think we had a problem," said a NASA engineer who asked not to be identified. ". . . I don't think it would have gotten us useful information."
Lawrence B. Mulloy, manager of NASA's solid-rocket booster program, said he was not part of that program when problems with the booster joints first surfaced, "but as far as I know, no one even thought about" using telemetry.
"From what I have researched, there were tests and analysis done to better understand joint rotation and blow-by but there was nothing done relative to telemetry.
"To my knowledge, it was never considered or proposed by anybody," he said.
However, solid-rocket engineers expressed surprise that NASA did not consider monitoring O-rings after the shuttle flight in November 1981 that revealed signs of O-ring erosion.
"Once you saw the erosion, that should have sent up a big red flag," said Fitzgerald, conceding that he has benefit of hindsight.
"I know this is Monday-morning quarterbacking," said Henry Shuey, a solid-rocket engineer with extensive military background, "but after the 1981 flight , I would have instrumented the joint by putting a pressure gauge on the port."
But Shuey pointed out that joint pressure data may not have been helpful in anticipating the Challenger disaster if the O-ring problem is due to some external factor such as ice, grease or damage during assembly rather than a basic design flaw. On the other hand, he said, data transmitted to Earth during flight "would certainly help you to figure out what went wrong" if there was an accident.
One NASA engineer who asked that he not be identifed said that there are potential drawbacks in putting pressure sensors on the boosters. One, he said, is that the sensor could be a potential tunnel for a gas leak; another is that the sensors and the wires that connect the sensors to the ground transmitter could create turbulence that would affect the booster's flight path.
Both Shuey and Fitzgerald dismiss those arguments, saying that a properly engineered sensor is unlikely to allow leakage and that boosters during the first shuttle flights were laden with sensors.
"There was a lot more instrumentation going on earlier and it didn't create any problems," Fitzgerald said.
Most of those 350-plus sensors came off the solid-rocket boosters at the end of 1984 as the Marshall center deemed the boosters to be "flightworthy as far as the instrumentation was concerned." The boosters now have three pressure sensors that indicate when the boosters ignite and when they should separate from the shuttle.
"That is the absolute minimum configuration," Fitzgerald said. "That's not being done to gain knowledge, that's purely functional."
In the past, however, NASA has usually chosen to place sensors on critical systems if it is not too costly, especially if the reliability of those systems is uncertain.
"As a general rule, that's what's done," said James R. Thompson, vice chairman of NASA's Challenger investigation task force and former NASA associate director for engineering in Huntsville.
"There are always tradeoffs," said John Clark, an RCA official and former director of NASA's Goddard Space Flight Center, "but my experience with NASA has always been 'when in doubt, instrument.' "