A routine, relatively inexpensive test would have caught the error that has crippled the $1.5-billion Hubble Space Telescope, but the test was never conducted, the head of the nation's space agency said Monday.
If the test had been carried out, the telescope would not have been launched into orbit 381 miles above the Earth with an optical system that is no better than ground-based telescopes. "But, in fact, we never did such a test," Rear Adm. Richard H. Truly, administrator of the National Aeronautics and Space Administration, said in a telephone interview.
The test, involving the reflection of light off the telescope's mirror, would have been far simpler than an ambitious effort to test the Hubble as a total optical system before launch--an effort that other NASA officials have said would have cost more than $100 million. Truly said there is little doubt that the smaller test, costing no more than "several million dollars," would have found the flaw.
Although it is not known why the test was never conducted, Truly suggested that the manufacturer of the Hubble's optics may have been more concerned about building a mirror that was perfectly smooth than one that was perfectly shaped. NASA could have ordered such a test, but the telescope was running hundreds of millions of dollars over budget, and the agency has conceded it cut back on some testing to save money.
The flaw in the Hubble's optical system is known as "spherical aberration." It occurs whenever a mirror in a telescope is not ground to the right shape. Experts believe the Hubble's 94-inch primary mirror is slightly flattened near the outer edge, causing light that hits that region to bounce back at a different angle from the rest of the mirror, blurring the image.
The mirrors were built by Perkin-Elmer Corp., now Hughes Danbury Optical Systems of Danbury, Conn. Company officials, who have refused all requests for interviews, declined to comment on Truly's statements.
In discussing the failure to perform the test, Truly acknowledged that what is clear today may not have been so clear nearly a decade ago when the mirrors were built. Although the test almost certainly would have detected the flaw, Truly said that if he had been the NASA administrator a decade ago, "I'm not sure that I would have decided to do it" either.
Truly indicated that the test may not have been considered essential at the time because engineers were more concerned about achieving the correct smoothness of the mirror than grinding it to the right curvature. Shaping a mirror is routine compared to the challenge of manufacturing a mirror that is smoother than any ever built.
"They must have believed the curvature was proper and it was the smoothness that was the tougher requirement to make," Truly said.
In terms of smoothness, the eight-foot mirror had to be nearly perfect. There is every indication that the builders achieved that level of performance. The mirror is so smooth that if it could be magnified to the diameter of the Earth, its highest peak would be no more than five inches tall.
Despite that degree of smoothness, the Hubble's optical system is virtually useless because the mirror is not shaped right and the images the telescope produces are fuzzy. Why that happened no one knows for certain, and there are at least two committees, one headed by Lew Allen, director of the Jet Propulsion Laboratory, trying to determine what went wrong.
Truly said the flaw could have been caught with a fairly simple test.
He said the test would have required the construction of "some sort of test optics" that would bounce light off the completed mirror.
"The reflection of that light would have come back depending on the curvature of the mirror," he said, and the kind of flaw that has caused Hubble's present problems would have clearly shown up.
Truly said he was basing his comments on casual discussions with some of the experts involved in various investigations, rather than on a formal conclusion by the investigating panels. The embattled space agency has been rattled by a rash of recent problems with various programs, especially the Hubble Space Telescope, advertised as the most revolutionary astronomical instrument in four centuries.
"This is turning into a criminal witch hunt instead of a technical probe," he said of news reports concerning the telescope. He was particularly miffed by reports concerning the failure of NASA to use existing military facilities to test the Hubble as a complete optical system.
"Just a molecule of dust would have been severely detrimental" to the telescope, which is designed to study the universe at shorter, ultraviolet wavelengths in addition to visible light. Because of its shorter wavelength, ultraviolet light is particularly sensitive to contamination.
Other experts have contended that it was not necessary to test the telescope as a complete unit, because if the proper tests had been done on its many components, flaws like the one that has crippled the Hubble should have been detected.
"In principle, you could test the telescope without putting the mirrors together," said Harland Epps of UC Santa Cruz, one of the nation's leading telescope designers.
"I can't fault (NASA)" for testing the components separately instead of as a unit, Epps added.
However, somewhere along the way, someone working on the Hubble made what Truly described Monday as "a simple mistake," and either got a measurement wrong or failed to follow directions. Until Monday, only a handful knew that a test that would have caught the mistake had never been carried out.
Who dropped the ball?
The manufacturer was required to build the mirrors to certain specifications, and that performance required adequate testing. NASA was required to certify that adequate testing had been carried out before accepting the mirrors from the manufacturer.
"The (investigative) teams really have not gotten to the bottom of this," Truly said.