Å®ÉúСÊÓƵ

A fragment of synthetic film the size of a grain of sand was to blame
for the troubles that afflict the Hubble Space Telescope. The fragment broke
off a calibrating device during the making of the telescope’s primary mirror.
As a result the mirror was ground too flat.

The Hubble satellite, a set of eight instruments built around two mirrors,
was launched in April. At a cost of $1.6 billion (800 million Pounds), it
was to be the ultimate in orbiting optical telescopes. In June, NASA announced
that Hubble’s first images were blurry. A spherical aberration 10 times
greater than tolerable had been ground into the main mirror, NASA discovered.
This means that rays of light striking the mirror close to its centre and
rays striking around its edges are not focused at the same central point,
so causing the fuzziness.

Since July, six experts led by Lew Allen, head of NASA’s Jet Propulsion
Laboratory in California, have been investigating how that could have happened.
Their report published this week describes a chain of errors that began
with a small oversight. But Perkin-Elmer of Connecticut, which made and
tested the primary mirror, and NASA, which was in charge, missed the mistake
because the company’s quality assurance team was understaffed and failed
to enforce its double-checking procedure.

Meanwhile, behind schedule, the optical systems team building the mirror
took short cuts and ignored contradictory findings, the Allen report says.
‘The most unfortunate aspect of this .. failure, is that the data revealing
these errors were available from time to time in the fabrication process,
but were not recognised and fully investigated at the time.’

The original fault happened in 1981 when engineers set up a testing
device meant to measure the curvature of the primary mirror as they finished
its polishing. The device, called a ‘reflective null corrector’ (RNC), consists
of two spherical mirrors and a field lens that are set up in front of the
primary mirror. Perkin-Elmer designed the RNC especially for the Hubble
primary mirror, and considered it more reliable than a similar device it
used previously, a refractive null corrector (RvNC).

Unknown to all at the time, the RNC itself was misaligned. When setting
up the RNC, engineers used a metering rod with capped ends to set the proper
spacings between the mirrors and the field lens. Because of the fine accuracy
required, engineers measured the gap between the rod ends and the RNC’s
elements with an interferometer, which aims a beam of laser light at its
target. An opening in the end caps allowed the beam to strike the rod ends.
The caps were coated with a nonreflective surface so that the beam would
register only when it struck the rod end.

The Allen panel discovered that a small piece of the cap’s coating had
chipped off. When the beam hit the cap it was reflected as if it were the
rod end. The distance between cap and rod end was 1.3 millimetres – exactly
enough to cause the aberration now seen in the Hubble primary mirror.

The mistake could have been discovered and corrected. Tests of the RNC
repeatedly showed that the instrument was misaligned. But technicians blamed
the testing device not the RNC. And in May 1981, the old refractive null
corrector was brought back for special measurement of the Hubble mirror.
When it detected an aberration, however, Perkin-Elmer dismissed it, saying
that the old instrument was imprecise and unreliable.

The Allen panel blames Perkin-Elmer’s ‘complete reliance’ on the RNC
to set the shape of Hubble’s mirror. It also criticises the company’s failure
to verify its alignment. Moreover, there was a ‘surprising lack of participation
by optical experts with experience in the manufacture of large telescopes’
during the making of the mirror. NASA’s scientific advisers were inexperienced
as well, says the report.

The Allen panel also notes that the optical operations division at Perkin-Elmer
worked behind closed doors. The company was doing classified work for the
Pentagon at the same time as the work on Hubble. The panel found no evidence
that the defence department instigated the atmosphere of secrecy. But access
to the Hubble work was strictly limited. Some of the technical staff in
the optical division were ‘deeply concerned’ about the discrepancies in
the optical data, the panel says, but there is no evidence that they took
their concerns to their superiors.