An unexplained last-minute problem with a main engine's cooling pump Friday forced a postponement of the launch of the shuttle orbiter Atlantis until Monday at the earliest.
A power surge caused the pump to stop operating and forced an automatic abort of the launch 31 seconds before liftoff. If the pump itself is damaged, said launch director Robert B. Sieck, replacing it would delay the launch until "well into next week."
In reviewing videotapes of the pre-launch sequence after the mission was scrubbed, NASA engineers also discovered a second problem--a pinhole leak in the insulation of a line that carries liquid hydrogen used as a coolant back to the external fuel tank. That insulation will have to be repaired over the weekend as well.
John Dutton, director of engineering for Martin Marietta, emphasized that the problems that caused the delay were simply "two isolated problems . . . neither one serious enough to cast a cloud on the other components" of any of the three main engines.
NASA officials will meet Sunday morning to determine if the launch can proceed on Monday or if it will have to be further delayed.
"It just wasn't our day," said Sieck, who added that getting Atlantis ready in time for a second launch attempt Monday "would be optimistic."
The delays came at the end of a countdown that had proceeded virtually flawlessly throughout the week. Friday dawned sunny with barely a cloud in the sky, and the crowds that gathered on the causeways outside the space center were in a festive mood.
Only one problem had been encountered before Friday.
On Thursday evening, when technicians were withdrawing the rotating service structure that encloses the shuttle for pre-launch maintenance, they discovered a gash in one of the insulating tiles protecting a thruster used for maneuvering the orbiter in space. Replacement of the tile required three hours but didn't cause any delay.
A more serious problem arose about 12 minutes before the scheduled ignition when a balky range safety computer caused a five-minute delay in the countdown. The range safety computer is used in making a decision to destroy the shuttle rocket if it should veer sharply off course during launch and threaten a populated area.
Technicians got the computer operating again, however, with plenty of time left in the 23-minute "window" during which launch could be achieved.
But then about 41 seconds before ignition and 10 seconds before shutdown, engineers noticed a sharp power surge through the recirculation pump on engine one, with the current jumping from its normal level of two to three amps to a high of 21 amps. That surge apparently disabled the pump and stopped it from functioning.
The pump circulates supercold liquid hydrogen--which is also used as the rockets' fuel--through the engine before ignition to keep it cool. If the engine were not cooled, thermal shock at ignition might cause it to crack or overheat and melt. After ignition, the flow of cooling hydrogen is maintained by the powerful pumps that force the fuel into the rocket engines.
Sieck said that the pump shut-off did not endanger the shuttle because interlocks in the shuttle and on the ground would prevent ignition of the shuttle engines if the pump were not operating.
The second problem involved the line by which liquid hydrogen is pumped back into the external fuel tank after it has been used to cool the engines. That fuel line is jacketed by a second line, filled with inert liquid argon that serves as insulation.
About 15 minutes after the mission was scrubbed, technicians reviewing videotapes noticed a thin plume of gas being ejected from one of the flexible joints--called "bellows"--in this line. Then there was a larger release of vapor.
Dutton speculated that a pinhole leak in the bellows had allowed air to seep inside the bellows, where the cold temperatures condensed it into a liquid. When the recirculation pump shut off, hydrogen in the line warmed up and vaporized the condensed air, causing it to stream back out the hole.
Eventually, the buildup of pressure in the bellows caused the rupture of a diaphragm that serves as a safety valve, allowing the pressure to be relieved. The release of pressure caused no hazard to the shuttle because no dangerous gases were involved. There was no leak in the hydrogen line itself.
Sieck said technicians would work through the night examining engine number one and its related circuitry to try to find the source of the problem. The rotating service structure was scheduled to be brought back to the rocket this morning, so that technicians could repair the safety diaphragm.
If the launch is rescheduled for Monday, NASA will have a 43-minute window, beginning at 11:07 a.m. PDT, during which to launch Atlantis. The opening of the window is dictated by the limited amount of fuel available to Atlantis to place its cargo, the Magellan spacecraft, into an energy-efficient orbit to Venus, where it will produce a radar map of the surface.
European Landing Strips
The window closes when it gets dark at the European landing strips that would be used if the mission were aborted during launch.
It was a frustrating disappointment for Atlantis commander David M. Walker, 44; pilot Ronald J. Grabe, 43; Mary L. Cleave, 42; Mark C. Lee, 36, and Norman E. Thagard, 45, who climbed out of the shuttle about an hour after the launch was scrubbed.