RE: ATM Spider Design

[epabcc@epa.ericsson.se (Bratislav Curcic)]

Andy wrote :

> I don't think the thickness of the vanes (or anything else) in the
> optical path has anything to do with the amount of diffracted light in
> the image plane, it's a function of the length of the edges. The thickness
> does affect the amount of obscuration because of the increased area.

Hm. It is difficult to isolate which is the predominant effect that
reduces the OTF, but I still think that thicker vanes _will_ produce more
diffraction. If nothing else, the percentage of diffracted light
compared to "useful" light (= contrast) will be worse with thicker
vanes because there is simply less "useful" light. Of course, the
effect is not proportional, and twice as thick vanes won't produce
twice the diffraction. But there is another important issue to tackle.
Thick vanes will have more thermal mass. Spider vanes contribute to
diffraction in two ways : "ordinary" diffraction and "refractive
diffraction". The air around vanes will be of different temperature from
the surrounding, because the vanes themselves will be of different 
temperature (sometimes COLDER than the ambient because of radiation).
This gradient will act as a weak lens and spread the light across in a
fashion that reinforces the diffraction effects.  Thinner vanes will
reach equilibrium far quicker, and better "follow" the ever changing
night temperature. (this may be reason why wire spiders sometimes work
spectacularly well)
This also suggests the use of non metallic spiders. Few planetary
telescopes in Melbourne that are being built at the moment will be
using carbon fibre vanes. I will report the results.

Bratislav