From: "Jan van Gastel" 
To: 
Reply-To: "Jan van Gastel" 


When I made my 20 incher, a 12 points cell would have been good enough. But
I decided to make an 18 points cell, to have a bit more tolerance for
non-accuracy. Even if some points would be about 10 millimiters misplaced
(which they are not of course), the cell would be within tolerances. It
made the building of the cell with non-professional tools, a bit more
relaxed for me.

Jan
http://home.wanadoo.nl/jhm.vangastel/Astronomy/


----- Original Message -----
From: "Jeff Anderson-Lee" 
To: 
Sent: Sunday, June 15, 2003 2:13 AM Subject: Re: ATM Help with plop
analysis of 16"


>
> "Mel Bartels"  wrote:
> > >>>
> > I did a plop analysis of a plate glass 16"  1.125 mirror. I think the
> > results showed that I needed an 18 point cell.  Do others who understand
> > PLOP agree?  I could find no arangement of 9 point cell that would give
me
> > the recomended PV error of less than1.7e -05 mm
> > <<<
> >
> > From previous analyses, a 6 pt support is sufficient for your 16 inch.
> > Sorry I don't have the numbers right in front of me, but Tom Krajci has
> such
> > a cell on his thin mirror 16 and it works very nicely.
>
> OK.  If you allow for Refocus Error Calculation (REC), a 6 point cell is
> perhaps sufficient:
>
> results: f-ratio: 4 r: 0.592923 err: 4.68832e-06
> results: f-ratio: 4.5 r: 0.589932 err: 4.56828e-06
> results: f-ratio: 5 r: 0.587581 err: 4.47652e-06
> results: f-ratio: 5.5 r: 0.585685 err: 4.40415e-06
> results: f-ratio: 6 r: 0.584127 err: 4.34564e-06
>
> I typically use an RMS error of 4.25e-6 as my cutoff and typically
optimize
> with REC off, so I would likely not choose this cell for myself.  But
then,
> I'm design-wise conservative and some say that I tend to "overbuild".
>
> But if Mel says it will work -- it no doubt will.
>
> [I didn't see what your f/ratio way, so the following is based on f/5.]
>
> Jeff
>
> --
> var d  16 * 25.4
> var t  1.125 * 25.4
> var r  0.587581
> diameter d
> thickness t
> density 2.23e-06
> modulus 6400
> poisson 0.2
> f-ratio 5
> n-mesh-rings 20
> rel-support-radii r
> num-support 6
> basis-ring-size 6
> basis-ring-min 0
> optimize r 0.01
> part bar 3 point 0 0 point 0 1
>
>
>
>

--- BBBS/NT v4.01 Flag-4