Re: [ATM] Analysis results for a 30" mirror on edge sling

["Robert Houdart" <robert.houdart@cruxis.be>]

Arjan,

> Very good analysis. What did you use to calculate these results?

The open source FE package Z88 for which I've written a program
MirrorMesh3D that automatically generates the mesh, load cases and
constraints of the mirror model. When the FEA is done, it processes the
surface errors (tilting, refocusing, Zernike coefficients).
The program can generate conventional, conical or cellular parabolic
primary mirrors or secondary flats.
See page www.cruxis.com/scope/scope1100_mirrormesh3d.htm that also
contains a comparison to PLOP results.


> This I have done some time ago, using NASTRAN finite element analysis.
> The shear is spread over the bottom, so I expect this should in itself
> have less effect than a sling. What does have effect though, is the
> torque applied by the cell elements.
> See http://home.hetnet.nl/~artm/atm/articles/deformation.html

Very interesting, I did not know your analysis.
For the 30" mirror glued on an 18-point cell I find the following
numbers:
- Surface error RMS:  83 nm or 1/7 wave
- Surface error P-V: 413 nm or 3/4 wave
The results are quite consistent with your work. The surface error is
indeed nearly entirely astigmatic and not acceptable.


> Most people on the list prefer to ignore these results though, and
> happily glue on...

Unfortunately all the analyses of this kind are quite dispersed on the
web and it's difficult to find clear answers to simple questions.
If time permits I'd love to write a comprehensive article or web page
about these issues so people could find all the answers in one location.

Best wishes,
Robert Houdart

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