[semiOT] looking for help on FEM problem

[cyrille de brébisson]

Hello,

I am in trying to write a small app, that will be designed to help amateur telescope makers make their telescopes...
The app will have a lot of features (in fact, I am planning to put ALL the needed stuff, at this point in time someone like me making scopes has to jungle through 5 or 6 different apps which is frustrating).

I have most of everything covered at this point, however, I am having issues with one of the feature. The "mirror Cell" design.

The problem so solve is: how to place the supports (3, 6, 9, 18, 27 or 36 of them) that "holds" the telescope mirror (a heavy glass disk) so as to minimize the deformation of the mirror under it's own weight (it must not deform by more than 1/20 of the wavelength of light or ~27nano meters!

The 'inputs' to the function will be the mirror diameter, thickness (homogeneous in first approximation, curved in 2nd), the glass young modulus and the support positions.

It might be possible to find an analytic solution, but I doubt it :-(
I was thinking that FEM calculations could do it, but the best I could do/find gave me 3 to 4 minutes to do the simplest calculations (and I would need to iterate on that to find the best placement!)

Does anyone here have any experience in such thing and would be willing to help me?

Thanks!

[Benoit Maag]

Hi Cyrille,

My gut feeling is that an FEM calculation is unnecessary for what you want to do and that what you really need to do is find the distribution of support points which minimizes the distance between them

Otherwise, Timoshenko’s theory of plates and shells (https://archive.org/details/TheoryOfPlat...s/page/n33) has all the calculations for plate deformations but it is not fun to read...