The Endeavour astronauts wowed everyone last week as they wrestled and coaxed the Hubble Space Telescope back to good health. But for the mission's top scientists, last week's heady success is tempered by caution and the ghosts of NASA's past " failures.
"It is extremely difficult to keep from getting excited right now," said senior Hubble scientist Dr. David Leckrone, "but I've got to try.
"Hubble's just had eye surgery," he explained. "We think it was successful, but we're not going to know for sure until we're able to take the bandages off."
Unwrapping the bandages will take 15 weeks and occupy hundreds of scientists and engineers at the Goddard Space Flight Center in Greenbelt and at Baltimore's Space Telescope Science Institute.
It is painstaking work, carried out at computer keyboards in windowless rooms. By remote control, Hubble's handlers must gradually nudge its mirrors into place, seeking a precise configuration that delivers the sharpest possible image to the telescope's cameras and spectrographs.
An apt analogy is the way in which a photographer turns the focusing ring on a camera -- first one way, then the other -- until the image of the subject appears sharp.
"It's a very delicate process," said NASA's Hubble program scientist, Dr. Edward J. Weiler. "I can't guarantee success. We've done everything humanly possible to make this succeed."
The process is the same one engineers were working through in June 1990 when they came to the awful conclusion that there was no single configuration of mirrors that could bring the heavens into sharp focus.
Their $1.5 billion space telescope had been launched with a primary mirror ground to the wrong prescription. They later learned that a NASA contractor had ignored test results in 1981 that warned of such a problem.
In many ways, Hubble remained the world's premier astronomical observatory, especially in ultraviolet wavelengths. But scientists concluded that without repairs it could never fulfill its promise to reveal the faintest, oldest and most distant secrets the universe.
With the mirror blunder flashing like neon in their heads, NASA officials vowed this time to ensure the performance of the hardware built to correct Hubble's blurred vision.
Independent review panel
NASA created an independent review panel chaired by Dr. Duncan Moore, a professor of optics at the University of Rochester. Its job was to look at all the work by the many groups that were analyzing Hubble's mirror flaw and designing and building mirrors to correct it.
Everything NASA and its contractors did was reviewed and verified with multiple, independently developed tests.
This time, "we certainly expect everything will be OK," said Dr. Kenneth G. Carpenter, project scientist for operations and ground systems at Goddard. "But everyone is and will be tense, and no one will breathe easily until we get confirmation down that the instruments all work."
Most encouraging to scientists were tests using simulators that generated light bent to Hubble's blurred prescription.
The simulators' light was beamed through the new Wide Field/Planetary Camera and COSTAR, the mirrors that Endeavour astronauts installed to deliver corrected images to three other Hubble instruments.
"And sure enough, we got extremely sharp images, essentially perfect images," said Dr. George Hartig, an optical systems scientist at the space telescope institute.
The focusing work that will spell the success or failure of Hubble's repairs is to begin in two weeks, after Goddard engineers finish tuning up Hubble's guidance systems, including its star trackers and the newly installed gyroscopes.
Focusing will be made easier by more freely adjustable mirrors in the new instruments.
The first job will be to adjust the Wide Field/Planetary Camera's "pickoff" mirror, to make sure it can capture light gathered by the telescope's 94-inch primary mirror and 12-inch secondary mirror, and reflect that into the camera.
Next, scientists will make coarse focusing adjustments with Hubble's secondary mirror, which was not involved in last week's repairs.
"We'll take a series of images with the secondary mirror moved in 9-micron steps, then look at the sequence and find out where it's sharpest," said Dr. Carpenter. A micron is equal to one thousandth of a millimeter, or .000039 inch.
Five weeks from now, after making similar adjustments to an internal "fold" mirror, scientists will begin fine focusing the Wide Field/Planetary Camera, tweaking the telescope's secondary mirror again.
"We'll do this for six orbits . . . take the data and work on it for a week, then see if we need more data," Dr. Carpenter said.
Initial pictures from the Wide Field/Planetary Camera are planned for release after seven weeks.
'First light' pictures
These "first light" pictures will include distant galaxies, star clusters and faint objects in fields of bright stars -- chosen to show off Hubble's better vision. Photos taken before the repairs are expected to be fuzzy in comparison.
Focusing the telescope's other three instruments -- the Faint Object Camera, Faint Object Spectrograph and Goddard High Resolution Spectrograph -- requires adjustments to mirrors mounted on robotic arms in COSTAR.
About four weeks from now, engineers at Goddard will send commands to spin tiny actuator motors on the arms of mirrors serving the Faint Object Camera. They will start with 0.157-inch adjustments, then try still-tinier 0.0098-inch steps.
"We'll be looking at a standard star . . . something just easy and bright so we wouldn't waste a lot of time," Dr. Carpenter said.
Similar work will start at four weeks and eight weeks for the COSTAR mirrors that serve the two spectrographs. Instead of comparing photographs, astronomers will study spectrographic data on computers.
Where there was once a muddle of information from the blurred starlight, "we should see data the scientists wanted all along pop up and be much clearer," Dr. Carpenter said.
If all goes well, the telescope should pack 70 percent of all the light from a star into a central dot. Before last week's repairs, it could focus only 15 percent of the light, with the rest smeared around it.
With all the precautions, Hubble scientists are confident they have avoided the kind of manufacturing blunder that befell Hubble's primary mirror. But they admit to other fears.
"Oh, lots of those," Dr. Hartig said. For a time he worried that vibrations during launch might jostle COSTAR enough to bang its mirrors into the walls of their enclosure, or that insulation might be torn loose as COSTAR was installed.
Early tests and the astronauts' reports have quieted those fears, he said. "But they are the kinds of things that keep me up at night."
Dr. Burrows said he knows of no recent test results, like those discarded mirror tests of a decade ago, that foreshadow problems to come.
"But that doesn't mean that something can't break," he said. "This is why it's not like having a baby born. Childbirth happens all at once. We're having this baby born sort of an inch at a time."