RE: [ATM] reducing mirror weight by sandblasting

[Dan Feldkhun <glasspusher@mindswell.com>]

My attempt to send this to the list earlier today seems to have been 
intercepted by a wandering black hole, so here goes again...  If the other 
one escapes from the event horizon, proof that internet traffic does not 
obey general relativity will be at hand.  I apologize in advance if some of 
you find that possibility upsetting.

Thanks for all the helpful emails!  I'll try to reply in one swoop ...

>Dan you might want to discuss with Richard Swartz and his effort in
>making lightweight "ribbed" mirrors.  If I recall correctly he had
>"print through" issues during grinding/polishing due to the ribs that I
>think would also hamper your efforts.

I've heard this from several people.  I wonder how thin the mirror was in 
the gaps to see this effect.  I really should chat with Richard.

>I read an article recently about "Liquidmetal(tm)" recently.  It's an
>alloy of metals that creates an amorphous structure (as is glass); can

I checked out their website.  Sounds like a really cool technology, though 
I don't think its available yet (certainly not in the price range of 
ATM's).  Also hard may mean very hard to polish.  I was also pondering 
making a mirror out of low CFE epoxy at one point, but epoxy interacts with 
abrasives very differently than glass and would require different polishing 
technique.  I wonder though if one could cast it using a polished-out tool 
as a mold -- this is how they make copy gratings from 
lithographically-patterned master gratings.

>I think what ever weight you save will be offset by
>the 27 point mirror cell.

I think I could make a pretty light 27-point cell by strategically removing 
aluminum from the trianges, etc.  It's really not that much more metal or 
complexity than an 18-point cell and according to PLOP I would need one 
even if I went with a plain 1/2" blank.

>Do you have one (FEA SW), or have access to it.
>Do you know how to use it
>or do you have somebody to teach you?

I've looked at and used some free FEM packages (GRAPE, LISA, JL Analyzer, 
freeFEM++) when I was doing truss analysis, so I'm aware of the steep 
learning curve that lies ahead if I try to model the mirror.  On the other 
hand, it would be a necessary (not to mention educational) excercise if I 
were to try to understand the effects of the cavities on the 
wavefront.    I think freeFEM++ or freeFEM3D might be the most appropriate 
package to do this.

>Instead of sandblasting the blank you may consider
>ribbed (or posted) blanks as  means to decrease weight  without
>much compromising  stiffness .
Hmm, that might be also be a big timesaver and avoid strain effects of 
machining glass.  There seem to have been some conversations regarding this 
on the ATM list in May '96.  Richard Schwartz was offering to make some for 
ATM's.  Also, Hextek makes some really light gas-infused ribbed 
blanks.  Any idea how much one costs?  Just for perspective, a 
pre-genereated 12.5"x0.5" Pyrex blank would run me about $150.

>Would it be reasonable to accomplish the weight reduction by making a 
>series of more or less flat bottomed holes in the back of the blank? I'm 
>thinking of a drilling tool that would be a tube with a semi circular flat 
>bottom.

I wonder how long it would take to make lots of these holes.  Something 
similar is suggested by Texereau (I think) for drilling out central holes 
in a mirror.  Induced strain can become a major problem, especially if done 
too quickly or with too much force.  I also have a table-top milling 
machine that I could equip with a diamond-coated carbide end-mill. I'm 
thinking of trying this on an old 8" pyrex tool and see how far I get and 
what kind of strain it induces.

>I personally have had many architectural creations carved in tempered glass.
>The process is the only way to work tempered glass. When I started this

How do you carve glass?

>with crossed polarizers. I would think a more important problem to tackle
>would be the thermal calculations necessary in the creation of a pattern
>which would not affect the figure of the mirror at equilibrium. Probably,
>the thinner the glass the quicker equilibrium is reached.

Why is this a problem?  It seems that as long as I wait long enough for the 
entire mirror to come to equilibrium before using it the figure should be 
nominal.  Perhaps this is more of an issue during polishing/figuring though 
-- could the "print through" effect be caused by different parts of the mirror


:Dan


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