Officials at the Kennedy Space Center this week will be working out the details of a plan to fix seals deep inside the nozzles of space shuttle Endeavour's solid-rocket boosters (SRBs). The effort represents the first time boosters have undergone internal repairs on the launch pad.
Endeavour's launch, originally scheduled for Saturday, was postponed indefinitely after inspectors found that o-ring seals in the booster nozzles had been slightly scorched on the two latest shuttle flights.
Although the National Aeronautics and Space Administration maintains that the astronauts on the two flights - Atlantis in June and Discovery in July - were never in danger, the components involved are considered "criticality one." This means that their failure could result in losing the orbiter and its crew. The failure of o-ring seals between two booster sections destroyed the shuttle Challenger and killed its crew during a 1986 launch.
Each shuttle uses two solid rocket boosters to put it in orbit, in addition to the liquid fuel engines on board. The boosters each generate 3.3 million pounds of thrust and burn for a little more than 2 minutes before they are jettisoned, fall back to earth, and are retrieved to be used again.
At a briefing Friday evening, shuttle program director Brewster Shaw said the repairs will be tested on a full-scale nozzle assembly at the Thiokol Corporation in Ogden, Utah. Thiokol designed and built the reusable boosters. After the repair technique is perfected, technicians at the Kennedy Space Center will climb into the throat of the SRB nozzles to make the actual repairs.
Like the boosters themselves, the nozzles are built in sections. Two o-rings and a rubber-like compound seal each of four joints to prevent hot exhaust gases from escaping. After examining the scorched o-rings, engineers concluded that the hot gas had found its way out through air pockets in the sealer. On 11 other occasions, gas had burned through some of the sealer, but it never reached an o-ring.
Just prior to Discovery's flight, engineers changed the procedure for refurbishing the nozzles. At Friday's briefing, Mr. Shaw said that ultimately, the problem most likely lay in the procedure, not in the nozzle's design.
"We think, although we haven't proved it, that we are overshooting" the starting point as technicians apply the sealer around the joint's circumference. The two- to four-inch overlap in sealer compresses any air bubbles at that point as the nozzle sections are joined and bolted together. The compressed air weakens the sealer and "the gas found the weakest spot and tried to get out," he said.
NASA officials have extended their investigation into scorching on joint No. 3 to other nozzle joints. To repair the problem, technicians will have to climb into the throat of the nozzle, rout out the nozzle's liner along the joint, then remove and reapply the sealer so no air is trapped.
"Once we can demonstrate that the process does that on offline full scale hardware" at Thiokol, he said, "we'll have the confidence to apply it to the flight hardware," although no tests will be conducted at the Kennedy Space Center to gauge the success of the repair prior to launch. The existing inspection tools, such as ultrasound devices, have failed to give reliable readings on the joints, he added. They identified naturally occurring flaws but did not spot the flaws intentionally introduced as a test.