[Author Prev][Author Next][Thread Prev][Thread Next][Author Index][Thread Index]
Re: [ATM] Use of bearings in (static) mirror cells
Hi Matt,
> In my opinion, you have to qualify your statement carefully. Your analysis
> is true for certain types of mirror cell configurations and not true for
> others . You are not defining your configuration or boundary conditions so
> to speak, which might lead people to believe the analysis is absolute and
> true for all types of mirror cells.
You are right, that's why initially I hesitated copying this to the list.
I just tried to put the discussion in a practical context: how to see the
severity of stiction problems in the light of the popular method of using a
PLOP designed stack of flotation cell elements which is glued to the mirror.
Anyway, some remarks inserted below:
> Here are some examples where the analysis you did is not applicable.
>
> Assume the support triangles were for example properly supported by
flexural
> stages that only allowed translation along the mirror optical axis and no
> rotation (only piston movement for supports) .
> You are analysing a torque effect (due to mirror COG being in front of
> support pivots, and supports being allowed extra degrees of freedom
> including rotation around pivot) that induces astigmatism obviously. If
the
> supports were restricted to strictly translation instead of allowing
> unnecessary rotation, the problem disappears. In that case, gluing would
> actually be recommended, because it would eliminate mirror shift , and
> astigmatism due to improper edge support.
I can see some practical problems here: what you propose in a sense is a
combination of an astatic and a kinematic cell. If each triangle supported
by a flexing flexure, you will have to tune the forces excerted by each
flexure. Also there is an issue of paraxial mirror positioning and possibly
also vibration. Using a mechanical lateral fixing of the piston locatin
(e.g. by a tube) would cause the same friction/stiction problem the thread
was started for.
> Also consider another case, where the cell has astatic levers . Sum of
> moments around the hinge is null and there is no resultant torque to
induce
> astigmatism (although the supports might be allowed to rotate ).
OK, I think this is a very good solution in general, only more complex to
build and probably not worthwhile for small/medium sized mirrors. But you
are right, the analysis was done for flotation cells only.
> The obvious conclusion being that before deciding for or against a certain
> method , the initial assumptions of the analysis must be carefully stated
> and analysed too , otherwise it's GIGO . People tend to believe what they
> see coming out of computer CAD analyses as absolute .
>
> And last but not least, my remarks were not meant to detract anything from
> the value of your analysis, just to point out some cases when it's
> applicable and when it's not .
>
> best regards,
> matt tudor
>
> -----Original Message-----
> From: Arjan te Marvelde <arjan.te.marvelde@hetnet.nl>
> To: atm@atmlist.net <atm@atmlist.net>
> Date: Friday, April 22, 2005 10:14 AM
> Subject: RE: [ATM] Use of bearings in (static) mirror cells
>
>
> >Hello Jerry,
> >
> >In my opinion, flexures are OK as are the RTV flexures, as long as you
> don't
> >glue them to the mirror.
> >No one seems to care though.
> >
> >Stiction is no big problem if you choose the proper materials.
> >As an example, PTFE (Teflon) on steel has a coefficient of kinetic
friction
> >of only 0.05 or so, and the static cof is only slightly more, probably
0.06
> >or so. Consider a lever in a cell of 75mm long: the force you have to
> excert
> >to overcome stiction is in the order of 3.6 gr, assuming a bearing load
of
> >600gr and a working diameter of 7.5mm (which I consider pretty large).
> >Smaller bearings will have smaller effect.
> >
> >These numbers can be compared with an analysis of deformation of a mirror
> >glued to its cell:
> http://home.hetnet.nl/~artm/atm/articles/deformation.html
> >As you will see, the force couple caused by glueing the mirror to the
cell
> >is way more than this 3.6 gr caused by stiction:
> >120gr when pointing at horizon,
> >60gr when pointing at 60deg elevation,
> >0gr when pointing at zenith.
> >
> >My conclusion is that you will only see the effects of stiction when the
> >mirror slides on its support points. You do need a side support, but the
> >deformation caused by such (properly built) side support is close to
none.
> >
> >Cheers,
> >Arjan
> >
> >> I have noticed Don's zeal urging consideration of the compliant
> >> hinges and/or flex bearings for various telescope parts. This
discussion
> >has
> >> been helpful for me to understand his zeal. I'd bet his would be a very
> >good
> >> mirror cell. I'm surprised there is so much resistance to his
> suggestions.
> >> But then I am not an enginier.
> >
> >_______________________________________________
> >ATM mailing list http://www.atmlist.net/
> >
>
_______________________________________________
ATM mailing list http://www.atmlist.net/