Re: ATM 3-Vane or 4-Vane Spiders

["Nils Olof Carlin" <nilsolof.carlin@telia.com>]

Jim Fitch wrote:

>  I insulated two vanes of a four vane spider with Black Flocked Paper to
> see if I could demonstrate any thermal effects on the diffraction
> spikes. Even though the insulated vanes had a thicker profile the
> corresponding spikes appeared dimmer and were thinner and longer. Unless
> I looked at the wrong spikes (I don't think so) I think the temperature
> difference of the vanes does effect the size of the diffraction spikes.

I believe (until somebody offers a better explanation) that the temperature of the vane is an equilibrium between heat lost by radiation and heat gain from conduction through an essentially stationary boundary layer of air. Both should happen at the surface of the vane, so while insulation would keep the interior from taking part in the heat exchange, it shouldn't (in this model) affect the equilibrium temperature the slightest. The radiative properties of the surface does, though, and "black" cools the most. However, "black" or not refers to the wavelengths responsible for radiation, in this case perhaps some 10s of microns, and the visual appearance is not a good guide here. I would believe polished metal would be safest here - besides, you could better control what is reflected into the light path this way! 
I know there are surface finisheds used for sun-panels that are good absorbers for near-infrared and visible light but reflectors in the far-infrared - this might be ideal for vanes, too. Wonder how carbon fiber laminate behaves in this respect?

So it is far from obvious to me that the flocked vanes would be less cooled than the plain ones. Whether they would be more efficiently warmed by the air, creating a thinner boundary layer, I can't even guess.

Besides, a dew-cap would be even more efficient preventing radiative cooling of the vanes than of the secondary, as the vanes present a very much reduced area to the sky.

For experimenting, it is easy to add "dummy" vanes at an angle to the normal ones, thereby making it easier to identify the ones under test. 

Nils Olof