Spun Epoxy Mirror

nw_lton · #1 14-10-2008, 09:15 AM

Hi All!

Just thought I'd share my little experiment with you guys, and hopefully gain some valuable feedback / discussions, etc! :-)

A few months ago as I was painting a house I was spinning the bucket of paint in my hand and saw very nice parabola's being formed in the liquid paint in the bucket.

I immediately though of mirrors, and found the spinning mercury one, and the spun epoxy ones as well - it wasn't a novel idea, but I thought I'd have a crack at making one...

So, the conceptual idea was:
Spin a container of liquid epoxy at a constant RPM, at a slow enough speed that a gentle parabola is formed, on a constant axis, until the epoxy sets.

Righto. Let's make it happen...

(Please be aware that this is a proof of concept stage!!!!)

I grabbed an old 7.2v cordless drill, clamped a hook in the chuck, and wired it up for an external supply
http://prism9.com/astro/DSCF5827.JPG

Then found an old power supply, and put a few lamps in series to slow down the motor - waay too fast still
http://prism9.com/astro/DSCF5828.JPG

Tested the setup with a bucket of water - works very nicely actually!
http://prism9.com/astro/DSCF5829.JPG

Drilled holes around the edge of an old cake tin, strung it up, and gave it a quick spin.
http://prism9.com/astro/DSCF5830.JPG

Mounted the drill securely with bits of string, cable ties, and post-it notes
http://prism9.com/astro/DSCF5831.JPG

We then mixed the epoxy on a very cold day, so it'll cure slowly, and began the spinning.
(Note the lid on top of the epoxy to cut down on surface turbulence from slower speed air causing ripples - very high tech!)
http://prism9.com/astro/DSCF5833.JPG

We noticed that the liquid tended to collect on one side of the tin, which caused a wobble, causing more liquid to move to the same spot.

This is the first serious problem we found. Suspended method simply won't work, as the axis of rotation needs to be exact for the whole setting period. This was going to end in disaster, so we cracked open a beer, and said, "Bugger it, let's see what happens!"

We put a few obstacles in the way to stop it wobbling too much.
http://prism9.com/astro/DSCF5835.JPG

The result - we proved the concept. Epoxy will set into a nice parabolic shape if kept spinning at exactly the same speed and on the same axis.

Why our attempt wasn't so successful:
Axis of rotation: Epoxy gathered against one side, caused a wobble, tin swung, and hit one of the objects, which stopped the wobble. Also disturbed the epoxy in the tin.

Speed control: When collision with the stabilising supports (Box of Cat-6, Toolbox) occured, the string from the drill wound up. Once the tin was free, this string unwound, causing the tin to spin faster than the drill speed, deepening the parabola in the dish.

Next time around, I'll be taking a different approach:
Leveling platform onthe floor
Trailer stub axle, with a groove for a belt machined into it's circumference
Motor with tacho-feedback turning the stub axle
Pan or pot with epoxy on top of stub axle, securely mounted so it can't wobble or get out of balance.

Then I might post up the resulting cast - the first one was just embarrasing...

Nige.

bojan · #2 14-10-2008, 10:00 AM

Why don't you use old record player as rotating plate?
I was experimenting with such "out-of-the-square-box"ideas a lot in the past.
Including chamber with lower air pressure behind the thin membrane (mylar).. or high electrical field even.. concept always worked, but that was all.
At the end, you will find that there is a reason why glass is used for mirrors... It is because only glass can keep the shape within required tolerances (1/10 of the wavelength) over wide temperature range etc, while it is reasonably easy to figure it precisely enough [speculum (used in the past for mirrors) is another example of good material... but obviously not as good as glass, otherwise it would have been used already]
Even with glass, if the mirror is not mounted properly, things can go wrong (in optical sense - mirror pinching is notorious, for example.. temperature gradients inside the glass are more subtle but still significant).
This method may be useful though, to obtain the initial shape of the glass for grinding - polishing. I imagine it is used for manufacturing large mirror blanks....

Geoff45 (Geoff) · #3 14-10-2008, 10:56 AM

Quote:

Originally Posted by bojan

This method may be useful though, to obtain the initial shape of the glass for grinding - polishing. I imagine it is used for manufacturing large mirror blanks....

Yep. See here

nw_lton · #4 14-10-2008, 02:42 PM

Quote:

Originally Posted by bojan

Why don't you use old record player as rotating plate?
I was experimenting with such "out-of-the-square-box"ideas a lot in the past.

Basically because of it's limitations with speed control. I'm thinking of spinning an 80cm one next - that needs to spin very slow....

Quote:

Originally Posted by bojan

Including chamber with lower air pressure behind the thin membrane (mylar).. or high electrical field even.. concept always worked, but that was all.
At the end, you will find that there is a reason why glass is used for mirrors... It is because only glass can keep the shape within required tolerances (1/10 of the wavelength) over wide temperature range etc, while it is reasonably easy to figure it precisely enough [speculum (used in the past for mirrors) is another example of good material... but obviously not as good as glass, otherwise it would have been used already]

Hmm, always wondered about the Mylar - saves me trying that! :-)
I was also thinking about coating the final spun cast with a harder layer for final figuring and coating. Perhaps even like the process of case-hardening metals - increasing the carbon content of the outer millimeter or so...most likely impossible without an electrostatic coating method, and/or heating to impregnate the harder substance into the surface for finishing...

Quote:

Originally Posted by bojan

Even with glass, if the mirror is not mounted properly, things can go wrong (in optical sense - mirror pinching is notorious, for example.. temperature gradients inside the glass are more subtle but still significant).
This method may be useful though, to obtain the initial shape of the glass for grinding - polishing. I imagine it is used for manufacturing large mirror blanks....

Hmm, be good if you could build a spinning mould that formed part of a plasma chamber, filled with argon or similar, that melted silica sand in the mould. Saves having a huge oven with high-temperature drives, etc...

Hmm, temperature consistency, etc...

rmcpb (Rob) · #5 14-10-2008, 08:09 PM

Only one word, amazing

Bassnut (Fred) · #6 14-10-2008, 08:54 PM

Rough as guts, but I like it. Can you get the rotation such that (say with harmonics induced by an oscillator fed to an audio pwr amp suppling the drill motor) to get an RC curve?

.

Garyh · #7 15-10-2008, 08:42 AM

Interesting concept but how does epoxy hold its shape and does it have a good expansion coefficient. How does one polish out epoxy?
Keep us posted on your experiment!
cheers

Glenhuon (Bill) · #8 15-10-2008, 08:32 PM

Epoxy mirrors are out there.
My first reflector was an Ebay special, you know the ones, 150 mm, 1400 FL, 700mm tube, with that darn barlow thingy in the focuser that makes them impossible to colimate
I took the mirror to centre spot it, seemed a bit light, and it was reinforced epoxy with a shiny coating.

Bill

nw_lton · #9 16-10-2008, 03:31 PM

Quote:

Originally Posted by Garyh

Interesting concept but how does epoxy hold its shape and does it have a good expansion coefficient. How does one polish out epoxy?
Keep us posted on your experiment!
cheers

It's mixed with the hardner, and gradually sets over a day or two.

The resulting mess was so hard it wouldn't come out of the tin - which also shows there's no shrinkage...

Speed control's an easy fix next time: PicAXE counting pulses from the tacho on a small motor (courtesy of an old 5.25" floppy drive), and PWM drive back to the motor for power adjustment.

Was just thinking - to stop a meniscus forming between the liquid epoxy and the container, I might wind the speed over the planned setting RPM (the one that sets the focal length), so that the epoxy runs up the sides a little, essentially to 'wet the sides', and emilinate the bulge a meniscus might form...

A gross example would be the letter 'M', compared to the letter 'U'
With the M, the parabola forms in the middle, but the epoxy sticks to the sides a little, preventing it lifting high enough to make nice clean edges.

Spin it faster to wet the sides, the slow down, which makes the epoxy fall back down a little, hopefully, theoretically, maybe.. forming a cleaner parabola...

...and who knows, if the air's clean enough, the aluminium coating might fill perfections enough to not need final finishing(high hopes. very, Very high...). Where in/near Tassie can I get silvering / aluminising done though!

...back to Supercheap for more parts!

bojan · #10 16-10-2008, 09:54 PM

Silvering can be done at home.. you just have to fiddle with couple of very dangerous and poisonous chemicals, like cyanides etc.. I have done my first mirror in glass shop, they were also manufacturing mirrors (I had to put quite an effort to persuade them not to paint the silver coating with brown protective paint :-) But that was a long time ago, and in Europe.. It lasted for 6 months before it became too non-reflective...
It is much better to go like this to prove the concept (but, frankly, I think you are wasting your time with this) because aluminizing is not cheap process ($100 - $200 for 10" or so). You will want your mirror to came out good enough for that money, otherwise... it would have been better if you just bought it from vendors (and then it will be GLASS, 1/10 wl guaranteed)

Aluminium coating WILL NOT fill out the imperfections, guaranteed.

Perhaps, you can try to do it in couple of phases: first phase it should be just cast, to have general shape (maybe just mould the concave surface without spinning) .
Then, use this as a dish for second, thin and final coat, this time under controlled conditions (spin, temperature etc).
The third phase could be the metalisation. But you should be able to check the quality of surface before that - this is done with glass mirrors as well, even before final polishing (with wetted surface, to enhance reflectivity): Foucault test or Ronchi to determine if surface is parabolic and how much and where it departs from ideal shape.