Despite assurances by space agency officials that the flawed mirrors aboard the Hubble Space Telescope were subjected to the most rigorous scrutiny available, independent tests to double-check for defects were never performed, according to optical experts involved in the project and a former employee of the company that built the mirrors.
Astronomers said it is routine to check and double-check telescope mirrors using at least two different, and completely independent, testing techniques. However, such alternate tests were not done on the Hubble mirrors, according to an optical expert who is a former employee of Perkin-Elmer, the Danbury, Conn.-based company that made the mirrors. The expert requested his name be withheld.
A failure to do independent corroborative tests means that if a small error crept into the mirror manufacturing system -- an error as simple as transposing a single number in a set of computer instructions -- it could have been extremely difficult to detect. Like a defective pair of glasses, the mirrors could appear to be perfect, but would be the wrong prescription.
Outside scientists who advised the National Aeronautics and Space Administration on Hubble's mirrors said that as far as they knew, Perkin-Elmer did not perform a second round of independent tests on the mirrors, said William Fastie, an optical expert at Johns Hopkins University and a member of the Science Working Group that advised NASA on the mirrors.
NASA has appointed an independent committee to investigate the flaws in Hubble's optics, the most crucial part of the orbiting observatory, and how they occurred. The panel, using impounded documents and test equipment, held its first meeting last week. Documents published by Perkin-Elmer scientists at the time of mirror fabrication and testing, and now being reviewed by the investigating committee, do not describe testing procedures that include alternative tests.
Officials with Perkin-Elmer, now called Hughes Danbury Optical Systems, have declined to comment on the investigation or provide details of the mirror testing procedures.
The Hubble Space Telescope, completed in the early 1980s after numerous delays and cost overruns, was deployed two months ago by the U.S. space shuttle into a 330-mile-high Earth orbit, placing it well above interference from Earth's atmosphere and enabling it to scan the farthest, faintest corners of the universe for new planets, black holes and quasars.
Last week, NASA officials stressed that problems with the $2.6 billion project were not high-technology bloopers, but an apparent flub during manufacture of one of its two mirrors. "The high-tech works fine, the problem is the low tech," said Lennard A. Fisk, NASA's Hubble project scientist.
But astronomers familiar with the project say Hubble's mirrors are anything but low-technology items. The two finely ground mirrors are the heart of the Hubble observatory, gathering, magnifying and focusing starlight at a tiny spot at the back of the telescope. Because either mirror is apparently mishaped, the entire focus is blurred, producing an image with a bright center surrounded by a skirt of more diffuse light. This degrades the use of almost every instrument on board, severely compromising the science for which Hubble was designed.
Attention focused last week on Hubble's large 94.5-inch primary mirror, which many optical experts suspect is the one with a flaw. The shape and precise curve of the mirror was tested at Perkin-Elmer using an array of smaller mirrors and an equally precise piece of glass called a null lens, according to sources involved in the project.
However, astronomers said the null lens itself possibly was flawed, or that the measurements used to test the primary mirror with the null lens were bungled. Without doing independent tests to corroborate the accuracy of the null lens, a mistake could have slipped through, say astronomers.
"If you only rely on one test and the test is a null lens, there's the very real possibility that you could build an error into the system," said Arthur Vaughan, an astronomer and optical expert at the Jet Propulsion Laboratory.
He and other astronomers said the mirrors of large ground-based telescopes are routinely subjected to several independent tests during construction. The experts said backup tests should have been performed on Hubble and they probably would have caught the flaws.
NASA required only that Perkin-Elmer rigorously test the two mirrors individually, ruling out an "end-to-end" test of the assembled telescope as too costly, according to Charles O. Jones, a NASA executive.
NASA officials have said that a sophisticated end-to-end test with both mirrors in place would have cost "hundreds of millions of dollars." Astronomers, however, contend that a simpler, cheaper end-to-end test could have been made to learn generally the telescope's performance. Sources familiar with the military's spy satellite program also say the Air Force routinely tested satellites similar to Hubble in end-to-end tests.
Astronomers who build ground-based telescopes say aberrations of the type plaguing Hubble are not uncommon, especially with large and sophisticated mirrors.
They cite as an example the telescope at the Canada-France-Hawaii observatory in Hawaii that was found to be off focus by several feet. Last year, the European-built New Technology Telescope in Chile, one of the most sophisticated telescopes ever built, also was found to have an aberration similar to Hubble's.
"It's not that unusual to get a couple of these pieces of glass together and find a problem," said Larry Daggert, manager of engineering at the National Optical Astronomy Observatory in Tuscon.
"This happens quite often on the ground," said Roger Angel, an optical expert and professor of astronomy at the University of Arizona and a member of the five-man Hubble panel.