Re: ATM Thermal behavior of mirrors

["Albert Highe" <ahighe@ix.netcom.com>]

Andrew,
>
> I haven't paid a lot of attention to the cooling fan threads before.  I
have
> always thought, mine is a thin plate glass mirror sitting in a kind of
> oversized mirror box with plenty of ventilation at the rear, so why worry?

I wouldn't worry if your seeing conditions don't allow you to get more than
about 200X and/or you don't observe close doubles and planets very much. For
your routine deep sky work, I doubt you'll see the difference. I have a 1"
thick Pyrex primary in an open structure. Adding a fan wouldn't help me for
95% of my observing. That is why I haven't installed one. Yet, when seeing
permits higher magnifications, I know the warm layer at the front of my
mirror is limiting me.

>
> Maybe I just missed this distinction before, but Albert and Bryan's posts
> have got me thinking there are two separate but not necessarily
independent
> effects that need to be looked at.  The first is the effect of thermal
> equilibrium (or lack thereof) on the physical shape of the optical
surface,
> and the second is the boundary layer problem caused by warm air boiling
off
> the front of the mirror.

A simple calculation shows that the physical deformation due to thermal
expansion may be a culprit. For example, assume that a 1" thick mirror has
one region with a bulk temperature 1 degree warmer than a different region.
The thermal expansion coefficient for plate glass is 10 PPM/ºC, and for
Pyrex is 3.2 PPM/ºC.  That difference in temperarture would create a 1/2
wave of variation in the plate glass, and about 1/6 wave of variation in the
Pyrex. But that is only if the entire 1" thick region is 1 degree warmer
than the adjacent 1" thick region. Likewise, a 10 degree warmer region that
is only 1/10" thick would produce the same surface variation. Anyone know
what the scale and temperature differences are for mirrors during figuring
or coming out of a warm car? A mirror that has a uniform temperature 10
degrees above ambient should show no measureable changes in surface error.
Yet, it will test poorly because of the variations in the warm boundary
layer.

>
> layer.  From separate discussions of mirror testing on the optical bench,
I
> think it must be dM/dt that contributes to "bad figure" (or "temporarily
> different figure").  Off the bench and out in the field, you want fans
front
> and back to bring M(t) towards A(t) as rapidly as possible, and then cross
> your fingers and hope that the air temperature stabilizes so your mirror
> won't have to chase it.

dM/dt contribution to "bad figure" depends on how inhomogeneous the
temperature is. A uniform high temperature should contribute next to
nothing. I suspect "bad figure" is a result of dM/dt due to temperature
inhomogeneity and warm boundary layer. It would be interesting to blow air
across the front surface of a mirror shortly after a figuring session and
see if that compares to a mirror whose tempearture has equilibrated to
ambient.

Albert