Re: [ATM] Re:Tees instead of triangles

[Mark Holm <mdholm@telerama.com>]

At least, front and back fans are worth making room for in the design. 
If Ken is wrong (I doubt that he is.) and you don't need that much, it 
will be much easier to take one out than make room for one that wasn't 
planned from the beginning.  If you do decide to blow only on one side, 
the evidence is that front fans do more good than  rear ones.  With 
thick glass, front and back fans are very likely to be helpful.  Alan 
Adler and Brian Greer both have done very good and careful work showing 
the usefullness of fans, front fans in particular.

You don't need a hurricane of wind to make a very significant 
difference.  A lot of the poor heat transfer, as well as a lot of the 
image degradation comes from a fairly thin boundary layer of air near 
the mirror surface.  A fairly gentle breeze cuts through a great deal of 
the boundary layer, helping greatly with both aspects of the problem. 
With a relatively thick mirror, improving heat transfer at both front 
and back is useful.

My Windows XP box won't copy the screen of a program running DOS in full 
screen mode, so I can't make screen shots.  You will have to do this for 
yourself.  Download cool.exe from
http://skyandtelescope.com/resources/software/article_328_4.asp

Run it with mirror thickness 3 inches and fan speeds of 6, 12 and 18 
feet per second.  Notice that even a slow fan is much better than none, 
that 12 fps is significantly better than 6 and that 18 doesn't buy a 
whole lot more.

You can probably get a decent estimate of the improvement from front and 
back fans by cutting the thickness to 1.5 inch (the program doesn't have 
a front and back option.

88 feet per second is 60 mph, so 6 fps is 4 miles per hour and 12 fps is 
8.  4 miles per hour is a gentle breeze and 8 a fairly healthy breeze.

Ken Hunter wrote:
> With a 3 inch thick blank, thermal equilibrium is
> going to be a much worse thing to worry about than
> cell induced deformations... Ensure that the blank has
> either a controlled environment (cooled to expected
> nighttime temps) or active cooling that tracks the
> current temp. 


-- 
Mark Holm
mdholm@telerama.com

_______________________________________________
ATM mailing list http://www.atmlist.net/