[ATM] steel tubing for mirror cell

[Mark Holm <mdholm@telerama.com>]

Jay Kirkland wrote:

>it's possible to microflex a truly wonderful mirror into 
>something ordiinary by the very act of using the presssure of the three 
>collimation bolts, depending on the design and accuracy of the cell 
>manufacture, and especially on the edge support. If the edge support 
>isn't well-nigh frictionless, if it grabs the edge of the mirror or an 
>unbalance portion of the wieght of the mirror and holds it back while 
>the collimation points push  one third of the mirror forward or pull it 
>back, you might be micro-flexing the mirror.
>

Several atm's and pro's have noticed that edge support that applies an 
axial force can astigmatically flex a mirror and that this happens with 
real world cells.  It is not just hypothetical.  Nils Olof Carlson has 
done some simulation using Plop that leads to the same conclusion. 

I think there are other good reasons not to use the support system 
structure for collimation:

1. Your support structure can be made more rigid, and lower profile if 
it doesn't have to be adjusted.

2. Your support structure won't wander slightly as you turn bolts in 
threads that might be slightly off center and loose.  Plop shows that 
many cells are susceptible to small position errors causing significant 
increase in deformation.

3. Your support structure components won't tend to rotate as you adjust 
collimation.  Same benefit as #2.

4. Your collimation adjustments can be spaced farther out, even outside 
the mirror diameter.  This makes collimation adjustment finer and opens 
several possibilities:

a. Collimation mechanism can be partly beside rather than behind the 
mirror.  This makes for lower profile.

b. Collimation can be done from top.  (Just arrange it so your 
collimation tool can't hit the mirror.)

c. As already noted, this makes it easy to do right angle collimation 
adjustment.

d. Makes it easier to use push pull screws .

e. Makes it easier to adopt a "fancy" flotation system such as Don 
Clement's flex hinges or Frederic Gaea's astatic levers.  Actually, I 
think it is required for the astatic system.  Also see Albert Highe's 6 
point cells for a possibility opened up by separating support and 
collimation  http://pw2.netcom.com/~ahighe/12ultra9l.jpg  Note that if 
Albert extended the triangle out beyond the mirror diameter, the 
collimation screws could go up instead of down, lowering profile.  Also 
note that this 6 point cell could easily serve as the basis of an 18 
point cell.

All in all, it is hard to argue for collimation using the support 
components except that one example is very well illustrated and 
documented in Kriege and Berry.

One disadvantage of having the collimation adjustments farther out: more 
linear travel is required to achieve the same angular adjustment.

-- 
Mark Holm
mdholm@telerama.com


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