Gordon D. Pusch wrote:
(dave schneider) writes:
(Henry Spencer) wrote:
A factor of 100 improvement would bring it down within reach of reason,
but a 1km mirror is beyond what's reasonably practical in the near future.
Eventually, yes.

Is there any indication of how big a bubble could be blown in
microgravity to create an Al or Au sphere, that could be sectioned to
provide several spherical mirrors (yes, I know, there's another conic
section that is better for focussing; ignore the man behind the
curtain for the moment) ?

For a sufficiently large focal length, you don't even need spherical sections;
the individual sections can be optically _flat_, and still not deviate from
the ideal figure by more than a fraction of a wavelength. (IIRC, a 10 km
focal length is sufficient for this to be true.) The primarily problem
then becomes one of _aligning_ the array of mirrors --- not machining.

There was a story in Analog science-fiction magazine some time
ago about exactly such a project.

Subplots abounded, but the major technical problem was the
structure's vibrational modes...


-george william herbert