Jump to content


Photo

The Future Today Part II: CF Mirrors

  • Please log in to reply
39 replies to this topic

#1 careysub

careysub

    Surveyor 1

  • *****
  • Posts: 1890
  • Joined: 18 Feb 2011
  • Loc: Rancho Cucamonga, CA

Posted 12 January 2013 - 09:26 PM

(I posted a version of this at the end of a long ATM forum thread, but is not really ATM material.)

There is work being done now on carbon fiber/epoxy mirrors that are showing promising results:
http://www.composite...ULTRAOptics.pdf

What they are doing is (in effect) using a glass "inverse mirror" mirror as a mold, applying a layer of commercial pre-impregnated CF fabric/epoxy that is cured at 121 C under 15-30 psi. The composite mirror surface is generated by direct contact with the mold. The CF mirror surface thus produced is then glued to a composite "egg crate" cellular back for support before releasing from the mold.

They perfected their techniques with 16" F/4 mirrors, consistently getting 1/8 or so wavefront error mirrors, and are working on scaling up to a 1 meter mirror for a real telescope to be built.

Interesting side-note, the 16" mirror mold was made by Royce, their first system was a 6" made by Pegasus.

The avoid print through they first coat the mold with a very thin layer of epoxy, cure that, then apply the prepreg fabric.

One interesting thing about this is the total mirror fabrication time should be much shorter, maybe one mirror a day per mold.

Another is that a very fast mirror should be as easy to make as a slow one.

If this technology could be converted into a cost-effective consumer level process, then light 18-20 inch mirrors that are F/3-F/4 might be feasible at lower prices (I am assuming that a producer would aim at a higher end price point than the GSO 16" F/4.5).

A mirror producer might arrange for the production of an Astrotech type inexpensive coma corrector that is optimized for a particular fast focal ratio (say, F/3.5 instead of the current F/4.5) to make a complete inexpensive solution for an ultra-fast cheap big dob.

One thing to look at is the possibility of making a mold from a regular mirror, basically doing the same molding trick twice. The mold would be backed by some sturdy solid fairly cheap casting material. Using a high temperature epoxy/cure cycle for the mold optical surface, and lower temperature one for the mirror should work I think.

A consumer implementation would need to come up with an inexpensive and effective backing material. The egg-crate CF backing shown in the article looks expensive to make, extremely light though it no doubt is. Maybe a Starstone foamed mirror blank would make a good backing? I think a block of end-grain balsa looks pretty good.

[About the idea of using a solid block of end-grain balsa. A block large enough for a 20 inch full thickness mirror would cost $50 and weigh 7 lb, making the entire mirror maybe 8 lb or so. Shaping the sagitta should be quick and easy using a prepared sanding table.]

#2 MessiToM

MessiToM

    Viking 1

  • *****
  • Posts: 886
  • Joined: 21 Dec 2009
  • Loc: Huntingdon PA

Posted 12 January 2013 - 11:51 PM

And I thought carbon fiber toilets were cool.......this is much cooler

#3 sunktank

sunktank

    Lift Off

  • -----
  • Posts: 16
  • Joined: 30 Oct 2011
  • Loc: Switzerland

Posted 14 January 2013 - 01:42 PM

Thanks, that's a really interesting link, as are the other papers published on their website.

Interesting to note that they got NSF funding to conduct their research and already have CF mirrors in use on the International Space Station to name just one.

CF mirrors are not pie in the sky, they are already "eye in (and on) the sky"!

I look forward to reading about the first amateur efforts in this area. Nothing they mention in their technical papers seems insurmountable to a determined amateur.

This French fellow made a balsa-core CF 16" travel newt that weighs 15.4kg (<34 lb) including (glass) mirror.

http://www.astrosurf...abrication.html

(site in French but lots of design and construction pics)

I wonder just how light these things can be made. But then, I live in a 4th floor apartment with no elevator, so I'm biased... ;)

#4 nirvanix

nirvanix

    Surveyor 1

  • -----
  • Posts: 1744
  • Joined: 07 Jun 2007
  • Loc: Saskatoon, SK

Posted 14 January 2013 - 02:20 PM

Amazing future for telescopes. Take a 1m CF/composite dob camping with you, set it up in between burger flips.

#5 sunktank

sunktank

    Lift Off

  • -----
  • Posts: 16
  • Joined: 30 Oct 2011
  • Loc: Switzerland

Posted 14 January 2013 - 02:49 PM

Further to my link above about the 16" dob, it's worth noting that of the 34lb mass, nearly 23lb are for the glass mirror.

Combining Careysub's estimate above of 8lb for a 20" CF mirror, we're looking at a potential "kerb weight" of well under 20lbs for a 16" dob. Compositemirrors got it down to 8kg (17.6 lb), pictured top of page 5 of the linked PDF above. That's single-hand portable. Just amazing.

#6 tezster

tezster

    Viking 1

  • *****
  • Posts: 817
  • Joined: 14 Jul 2009
  • Loc: Missisauga, Canada

Posted 14 January 2013 - 03:54 PM

It would be interesting to see how a traditional truss-dob would look with this type of mirror. Specifically, I'm curious to see how high the rocker box altitude bearings would need to be, as I imagine the balance point would be shifted significantly towards the UTA, considering the weight of the mirror.

#7 sunktank

sunktank

    Lift Off

  • -----
  • Posts: 16
  • Joined: 30 Oct 2011
  • Loc: Switzerland

Posted 14 January 2013 - 04:26 PM

Therein lies a problem...with mass comes stability. I suspect a gust of wind would cause a traditional design truss dob this light to tremble like a leaf.

Keep the traditional design and use sandbags/water bladders inside the rocker box? Might not be so great from a thermal standpoint...how fast would a CF mirror cool to ambient anyway?

#8 careysub

careysub

    Surveyor 1

  • *****
  • Posts: 1890
  • Joined: 18 Feb 2011
  • Loc: Rancho Cucamonga, CA

Posted 14 January 2013 - 04:55 PM

Therein lies a problem...with mass comes stability. I suspect a gust of wind would cause a traditional design truss dob this light to tremble like a leaf.


A new risk for a big dob, having the wind blow it away!

Keep the traditional design and use sandbags/water bladders inside the rocker box?


New engineering approaches would be needed for sure.

I am inclined to think of that the natural pattern here is a genuine Serrurier double truss/strut design, with a high balance point. The Kriege pattern simply does not work with an ultralight mirror, however fast. Base engineering becomes critical. Tall hollow core supports for the bearings. Water bladders in the base are a good idea.

Might not be so great from a thermal standpoint...how fast would a CF mirror cool to ambient anyway?


It might make cooling a problem of the past.

The balsa core I propose is an excellent insulator, and has very little thermal mass anyway, whereas the CF mirror surface is quite thin and has fairly good conductivity through its thickness - 20 or 30 times that of the balsa. Very little heat would be shed from the surface, and since the thermal conductivity of the CF is ten times higher still laterally, it should remain at a nearly uniform temperature everywhere.

#9 Jb32828

Jb32828

    Messenger

  • *****
  • Posts: 479
  • Joined: 01 Aug 2010
  • Loc: Orlando, FL USA

Posted 14 January 2013 - 05:51 PM

Eventually we are gonna be able to 3D print these things and send em off the the coater. Pretty cool tech.

#10 careysub

careysub

    Surveyor 1

  • *****
  • Posts: 1890
  • Joined: 18 Feb 2011
  • Loc: Rancho Cucamonga, CA

Posted 14 January 2013 - 07:22 PM

Further to my link above about the 16" dob, it's worth noting that of the 34lb mass, nearly 23lb are for the glass mirror.

Combining Careysub's estimate above of 8lb for a 20" CF mirror, we're looking at a potential "kerb weight" of well under 20lbs for a 16" dob. Compositemirrors got it down to 8kg (17.6 lb), pictured top of page 5 of the linked PDF above. That's single-hand portable. Just amazing.


Also note that my mirror weight estimate is predicated on the assumption that we are making a full thickness mirror. If I plug the physical data for end-grain balsa into PLOP and run it for an 18 inch F/4 mirror (3 inches thick) I find nearly perfect support (6.3 nM, roughly 0.98 Strehl) with a 3-point "cell", i.e. no cell other than 3 support points. Not needing a cell cuts weight and complexity further.

#11 pbsastro

pbsastro

    Viking 1

  • *****
  • Posts: 534
  • Joined: 21 Mar 2007

Posted 14 January 2013 - 07:30 PM

Reduced mirror weight could bring gravity center issue and wind stability issue.
However I think it will be partially compensated with dobs moving from F/5 to F/3. Shorter dob means much less wing effect and means much less binary from focuser end.

#12 Darenwh

Darenwh

    Mercury-Atlas

  • -----
  • Posts: 2823
  • Joined: 11 May 2006
  • Loc: Covington, GA

Posted 14 January 2013 - 08:22 PM

Don't forget EQ mounting would be quit easy and the weight of the mount would add stability. Make it a casegrain so focus is behind the scope to avoid heat plumes from the body and you are set.

#13 Thomas Karpf

Thomas Karpf

    Surveyor 1

  • -----
  • Posts: 1742
  • Joined: 09 Feb 2009
  • Loc: Newington, CT

Posted 15 January 2013 - 08:21 AM

Keep the traditional design and use sandbags/water bladders inside the rocker box?


I'm not sure I would want the balance to shift as sand shifted.

On the other hand, perhaps wood slabs with lead 'sticks' inserted into holes drilled into the wood, and then sealed with putty would work well. If the slabs slid into brackets attached to the inside of the mirror box, there should be essentially no shift as the altitude of the scope changed.

#14 nirvanix

nirvanix

    Surveyor 1

  • -----
  • Posts: 1744
  • Joined: 07 Jun 2007
  • Loc: Saskatoon, SK

Posted 15 January 2013 - 12:27 PM

Therein lies a problem...with mass comes stability. I suspect a gust of wind would cause a traditional design truss dob this light to tremble like a leaf.


Just need to set up a wind break. Likely this will be a standard thing to do with this type of scope.

#15 sunktank

sunktank

    Lift Off

  • -----
  • Posts: 16
  • Joined: 30 Oct 2011
  • Loc: Switzerland

Posted 16 January 2013 - 03:53 PM


Keep the traditional design and use sandbags/water bladders inside the rocker box?


I'm not sure I would want the balance to shift as sand shifted.

On the other hand, perhaps wood slabs with lead 'sticks' inserted into holes drilled into the wood, and then sealed with putty would work well. If the slabs slid into brackets attached to the inside of the mirror box, there should be essentially no shift as the altitude of the scope changed.


Or just a couple of bricks velcroed to the bottom of the box for that matter.

But my thinking is to have a large aperture hand-transportable scope; lugging lead weights around would defeat the purpose. Whereas water or sand could be obtainable at the observing site, and disposed of immediately after the observing session. Bladders with "bulkheads"(separated compartments) would prevent the sand or water shifting. In oil tanker ships the oil is separated into compartments so that when the ship rolls on the swell, the liquid inside doesn't slosh around and compound the roll. Same idea.

Come to think of it, why not make all the structural elements (rockerbox, UTA, trusses) hollow sealed CF units? - light, transportable, snaps together. Fill it with water at the observing site for added mass and stability, empty it out when done. If you wanted to get really clever you could design it in such way that the water leaks out at a precise rate that alters the balance and pointing of the scope...at sidereal rate... :grin:

I think Careysub could be right, perhaps it's time to look to innovative designs that take into account a much less massive bottom end...

#16 careysub

careysub

    Surveyor 1

  • *****
  • Posts: 1890
  • Joined: 18 Feb 2011
  • Loc: Rancho Cucamonga, CA

Posted 16 January 2013 - 10:37 PM

..I look forward to reading about the first amateur efforts in this area. Nothing they mention in their technical papers seems insurmountable to a determined amateur.
...


I've done a bit of browsing on-line now about "optical surface transfer" technology, and other work done with composite mirrors in the past - and while it is good to know that solutions have been found to all of the problems with doing this, without some advice from the experts who have solved them it looks like a steep hill to climb.

There are two key problems that have to be solved - avoiding "print-through", and getting a clean release from the mold.

Composite Mirrors has solved the first by introducing a controlled thin resin layer (50 microns or so) between the prepreg and the mold. I don't know how they do it, but developing a pre-coating technique seems achievable.

The solution to the second is proprietary and it is not obvious how they do it. Epoxy resins generally stick to stuff, including glass, something fierce.

One possibility that occurs to me is deposition of an extreme thin film of silicone oil on the surface, basically what we do with aluminum to finish a mirror. If this is the case, then a vacuum chamber is needed good enough to evaporate a suitable oil.

Another interesting approach (not Composite Mirrors's) is if you are casting a mold from a mirror (to avoid having to make an optical quality convex surface from scratch) is to use a rigid silicone resin to the cast. This would create a non-stick surface to mold against, and also solve the problem of releasing the mold-to-be from the original mirror.

But rigid silicone resins are little known (and even less available) outside of industry, silicone oils are maybe a bit more accessible. Finding the right products to do this is a challenge.

Think about the problems people have have had with finding an ideal replacement for the original Ebony Star. That is trivial next to finding the right release agents and resins to do this.

Still, the fact that these mirrors are in mass production should provide encouragement to see if they can be brought to the amateur astronomy community.

#17 azure1961p

azure1961p

    Voyager 1

  • *****
  • Posts: 10235
  • Joined: 17 Jan 2009
  • Loc: USA

Posted 16 January 2013 - 11:24 PM

What's doesn't gel here for me is that resin is one hardness and CF entirely something else altogether. At the wavefront level this mismatch of hardness would yield a lumpy inconsistent value. Glass - any glass - is far more homogenous than epoxy and carbon fibers. It sounds ghastly quite frankly. You can't grind soft resin and brittle carbon and expect an even surface smoothness.

Pete

#18 Howie Glatter

Howie Glatter

    Vendor

  • -----
  • Vendors
  • Posts: 858
  • Joined: 04 Jul 2006

Posted 17 January 2013 - 07:34 AM

"You can't grind soft resin and brittle carbon and expect an even surface smoothness. "

The optical surface on these composite mirrors is neither ground nor polished. It is cast in place against a convex glass form.
I understand your concern though, because the inhomogeneities of the structure exist at a much larger scale than wavelengths of light.

#19 Jon Isaacs

Jon Isaacs

    ISS

  • *****
  • Posts: 43401
  • Joined: 16 Jun 2004
  • Loc: San Diego and Boulevard, CA

Posted 17 January 2013 - 08:27 AM

What's doesn't gel here for me is that resin is one hardness and CF entirely something else altogether. At the wavefront level this mismatch of hardness would yield a lumpy inconsistent value. Glass - any glass - is far more homogenous than epoxy and carbon fibers. It sounds ghastly quite frankly. You can't grind soft resin and brittle carbon and expect an even surface smoothness.

Pete


Pete:

That is a concern of mine as well. Carbon fibers have Coefficients of the Thermal Expansion (CTE) but the polymers used for matrix have very high CTEs. Exactly how that plays out on the micro/nanoscale in the mirror as the temperature changes is not so clear.

One thing I did notice in a casual read of the paper:

"Several plate-type mirrors were produced in order to qualify CFRP mirror substrates, this was largely an
exercise in fiber and matrix selection and down selection for the ideal mirror substrate material. Because plate mirrors have shown extreme sensitivity across relatively small temperature bands they are not ideally suited for telescope applications without some tunability of the surface to counteract the thermal deformations."

Jon

#20 careysub

careysub

    Surveyor 1

  • *****
  • Posts: 1890
  • Joined: 18 Feb 2011
  • Loc: Rancho Cucamonga, CA

Posted 17 January 2013 - 02:59 PM

That is a concern of mine as well. Carbon fibers have Coefficients of the Thermal Expansion (CTE) but the polymers used for matrix have very high CTEs. Exactly how that plays out on the micro/nanoscale in the mirror as the temperature changes is not so clear.

One thing I did notice in a casual read of the paper:

"Several plate-type mirrors were produced in order to qualify CFRP mirror substrates, this was largely an
exercise in fiber and matrix selection and down selection for the ideal mirror substrate material. Because plate mirrors have shown extreme sensitivity across relatively small temperature bands they are not ideally suited for telescope applications without some tunability of the surface to counteract the thermal deformations."


One thing to bear in mind is that the mirror undergoes a 100 C temperature change from the time it cures, to the time it cools. If it has a good optical surface at 130 C, and a good one near 30 C, what kind of extreme sensitivity would they be talking about? A change in surface quality should be very bad, but a change in focal length perhaps not so much. (They probably use a special dimensionally stable curing resin, but that doesn't help with the CTE).

It is possible that this type of mirror might need to be thermostatically controlled for best performance. Although the usual solution to mirror thermal issues is to cool it to ambient ASAP, the fact that primary mirror heaters are available (e.g. Kendrick) suggests that continually thermal layer scrubbing might make a "warm" mirror okay. Actually controlling the temperature shouldn't be too hard. The CF composite has a thermal conductivity similar to steel, the mirror surface plate is thin (1-2 mm) and if backed by a good insulator a heating web behind the mirror surface should maintain an even temperature quite well.

Finding exactly the right materials to make these seems pretty important.

#21 Jon Isaacs

Jon Isaacs

    ISS

  • *****
  • Posts: 43401
  • Joined: 16 Jun 2004
  • Loc: San Diego and Boulevard, CA

Posted 17 January 2013 - 03:21 PM

One thing to bear in mind is that the mirror undergoes a 100 C temperature change from the time it cures, to the time it cools. If it has a good optical surface at 130 C, and a good one near 30 C, what kind of extreme sensitivity would they be talking about?



Given the fact that they say that the "plate mirrors" were not of optical interest, my guess is that it is over a much narrower range than 30C to 130C.

I am more curious about what differentiates a "plate" mirror from the mirrors they actually tested. I don't know if plate means flat or monolithic.

Jon

#22 mark cowan

mark cowan

    Vendor (Veritas Optics)

  • *****
  • Vendors
  • Posts: 3981
  • Joined: 03 Jun 2005
  • Loc: salem, OR

Posted 17 January 2013 - 04:23 PM

Well, they're probably talking about substrate, and the need to make the forms out of something that doesn't shift thermally while curing is going on.

Best,
Mark

#23 JM La Galette

JM La Galette

    Explorer 1

  • -----
  • Posts: 78
  • Joined: 24 Nov 2007
  • Loc: France

Posted 08 February 2013 - 02:39 PM

Interesting guys, perhaps to be tried by a motivated amateur...

JMarc

#24 GeneT

GeneT

    Ely Kid

  • *****
  • Posts: 12627
  • Joined: 07 Nov 2008
  • Loc: South Texas

Posted 08 February 2013 - 06:43 PM

I found this to be very interesting. I have wondered for a long time when technology and new materials might provide a break through for mirrors and optics. Still--is this possibility too good to be true?

#25 careysub

careysub

    Surveyor 1

  • *****
  • Posts: 1890
  • Joined: 18 Feb 2011
  • Loc: Rancho Cucamonga, CA

Posted 08 February 2013 - 10:05 PM

I found this to be very interesting. I have wondered for a long time when technology and new materials might provide a break through for mirrors and optics. Still--is this possibility too good to be true?


You can read the material on their website, and there is some coverage providing a few additional insights in optics publications on the web.

One of the key aspects of any breakthrough is simply knowing it can be done. Reminds me of the observation that once something is seen in a big scope, it can be seen in smaller ones.

Unless Composite Mirrors wants to set up its own amateur mirror business, I think the most attractive approach for amateurs is doing a double optical surface transfer - once from an existing mirror to make a mold, then making mirrors from that mold. That way an expensive proven excellent mirror could be replicated without the difficulty of grinding a reverse mirror.

I suspect the trick of getting the composite mirror to release is to deposit a very thin film of silicone as a release agent on the glass inverse mold (or the polymer mold) surface. You will need a chamber to do this for sure to the conditions are clean and steady. However, unlike depositing aluminum with requires a good vacuum (10^-4 torr or better) there are lots of silicone compounds in common use that boil at 200 C or below (some close to room temperature).

Rather than getting hold of pure silicones, or industrial ingredients, I wonder whether identifying a suitable commercial product (like a silicone lubricant) that contains a desirable silicone and distilling it out of the product directly would work. Cyclomethicones for example evaporate at room temperature with vapor pressures around 1 mm, and boil at 200C, and are used in cosmetics.

Experimentation will be required to develop an amateur process.






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics