[Author Prev][Author Next][Thread Prev][Thread Next][Author Index][Thread Index]
Re: [ATM] Rotating Resin Mirror
Thank you both for the encouragement and advice. I've had a few thoughts
after an evening of web surfing:
I see that one problem with this method is the poor coefficient of thermal
expansion for resins. For plate glass, alpha=8 to 12x10^-6/Celcius, for
pyrex (borosilicate) alpha=3.3 x 10^-6. However, there does seem to be low
thermal expansion resins available. For example, alpha=12*10^-6/Celcius:
http://www.masterbond.com/tds/ep30lte.html
That's not far from glass. How about adding some glass nano powder. Glass
filled polycarbonate resin (40% glass) has alpha=9*10^-6. That's getting
competitive. see:
http://www.plasticsintl.com/datasheets/Polycarbonate_40_GF.pdf
I haven't found any suppliers of polycarbonate resin so I'm thinking that
the chemical process involved must be a particularly unpleasant one. Does
anyone know if Jo Public can work with polycarbonate resin and where it can
be found?
I have seen "casting resin" readily available which has the properties of
low shrinkage during cure and low viscosity. I'm going to try using this
with a cheap nano-spherical glass like filler that comes out of coal fired
power stations, my reasoning being that the filler particles are very small
so shouldn't leave marks on the surface and being glass they should lower
thermal expansion as well as reducing shrinkage during cure. On the
downside, they will increase the viscosity and so it won't flow as quickly
when spun (end of the hand waving explanation).
The big big problem is deformation during cure. The amount of deformation
must relate to the thickness of the resin so I will try starting with a
stable cylindrical substrate and use as little resin as possible to make
the parabolic surface.
The substrate needs to have very low thermal expansion and strong so I'm
thinking ceramic, limestone portland cement or SiC filled resin might work.
I think an Aluminium or Steel disc would have to be quite thick to prevent
warping and cutting it sounds like hard work.
If the parabolic shape of the mirror forms when the resin is hot then maybe
the mirror will have to be heated when in use. It would be simple to mold
cavities to mount resistors inside the resin to set the temperature of the
mirror.
I'll keep you informed.
Ian
--On 15 February 2006 19:34 -0500 "Dream - Telescopes & Accessories, Inc."
<shane@dreamscopes.com> wrote:
> Ian & Tony,
>
> I agree with Tony that IR wavelength is much more forgiving than visual.
> Submm and radio are even more forgiving. As I've posted previously, (the
> great majority of) resins have very high CTE's. They run the gamut from
> slightly higher than aluminum's CTE to 3-4 times higher than aluminum.
> It also has the habit of shrinking after cure, since the cure is
> typically exothermic and/or needs elevated temps to properly cure.
>
> A gel coat (resin rich layer) is fine for boats and other products where
> surface aesthetics are important (but viewing distant galaxies isn't...).
> But for visual spectrum work the amount of distortion, rippling, print
> through, etc., is just too much.
>
> At the end of last year I posted R&D that Dream has done in this area.
> It can be seen here:
> http://www.dreamscopes.com/pages/proj-05/reflections-01.htm
>
> The surfaces produced then and now are done without the use of gel coats,
> surface veils (thin, tissue-like materials that reduce print through) or
> other low areal weight fabrics. The resin contents are very low. BUT,
> we have still not achieved surfaces fine enough for visual spectrum work.
> Other companies who have been in carbon fiber mirror R&D for 15-20+ years
> still aren't mass replicating carbon fiber mirrors like the technology
> was original thought to allow. Not in visual spectrum at least.
>
> The next technology that is being worked on is nano-fibers. Some are
> hoping this technology will end the print through issue that has plagued
> carbon fiber mirrors meant for visual spectrum work. The technology is
> already being sold and used in many industries. Some applications use
> nano-fillers mixed in with resin to increase the strength of final parts.
> It's a fairly simple way to increase part properties without changing
> much. Fillers have been used by almost every (composite application)
> industry from the beginning. It doesn't take a great deal of the
> nano-filler to improve properties.
>
> Others see the use of this material for injection molded and/or infused
> parts. Typically those parts have high resin contents and thus exhibit
> lower properties versus a part with a higher and longer fiber structure.
> Nano-fillers have enormous potential and gets us back to Ian's original
> post. If these nano fillers and fibers are used, they greatly increase
> part (mirror surface in this case) strength, stiffness and will help to
> reduce the final CTE. They also produce a surface that is isotropic.
> This is similar to how random oriented fibers produce an isotropic
> surface, but on a nano scale.
>
> Does this mean Ian can't make a spun resin mirror? No.
>
> I would advise you to pick your substrate that you are pouring the resin
> into carefully. It's the variables that change things. Document them
> and watch them as closely as possible. Once you understand them, you
> might be able to find a solution. For a time even the Wright's thought
> heavier than air flight by man was impossible... Happy experimenting!
>
> Thank you.
>
> Sincerely,
> Shane Santi - Owner
> Dream - Telescopes & Accessories, Inc.
> http://www.dreamscopes.com
> 610 - 365 - 2833
>
>
> At 01:40 PM 2/15/2006, you wrote:
>
> You'll notice that's an IR mirror, it's not designed for visual
> wavelengths which are much shorter and demand a much higher surface
> accuracy.
>
------------------------------------------------------
1st rule of intelligent tinkering - save all the parts
Ian R. Williams
Department of Mathematics
University of Bristol
email: I.R.Williams@bristol.ac.uk
web: http://www.maths.bris.ac.uk/~mairw/
_______________________________________________
ATM mailing list http://www.atmlist.net/