Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

Spin Polishing - making a perfect edge

  • Please log in to reply
106 replies to this topic

#51 Mark Harry

Mark Harry

    Cosmos

  • *****
  • Posts: 8086
  • Joined: 05 Sep 2005
  • Loc: Northeast USA

Posted 30 July 2008 - 07:50 AM

"But the diffraction edge doesn't photograph well, while the KE inside of focus does,..."
***
I was wondering about that. I haven't spent any time learning how to use a camera to advantage yet. Both tries ended in dismal failure. (during the film era!)
***
200" is a LONG roc! Your results could be interesting; hope you post the results.
"But that can be iffy, and beveling isn't an ideal solution,..."
Even with such a slight width as I note on my 14? You can take a large piece of flat glass, and make a 12 micron grind to accomodate such a narrow TDE in less than a minute. Otherwise, it'll take hours just to eliminate that hairline edge.
But I could be talking out of turn on this, because it isn't finished yet, and it could be less by the time it's fully polished. I'll know a bit more later on this morning.
Regards,
M.
Added note, I polish all mirrors on urethane, for it saves a ton of time. 6" well under an hour, 8" a bit more than an hour, and this 14 is at 8 hours, and counting. To pitch polish these takes several hours with a 6, more for the 8, and the 14, I'd hate to think about! The down side is a certain degree of edge issue as noted in earlier posts, but so far with this 14, it's suprisingly minimal.

#52 rwiederrich

rwiederrich

    Goldfinger

  • *****
  • Posts: 13312
  • Joined: 17 Nov 2005
  • Loc: Gorst Washington(center of the Universe)

Posted 30 July 2008 - 08:44 AM

Added note, I polish all mirrors on urethane, for it saves a ton of time. 6" well under an hour, 8" a bit more than an hour, and this 14 is at 8 hours, and counting. To pitch polish these takes several hours with a 6, more for the 8, and the 14, I'd hate to think about!



(Double take)......You polish on Urethane.. :question:

You don't use a standard pitch lap? Please explain..

Rob

#53 mark cowan

mark cowan

    Vendor (Veritas Optics)

  • *****
  • Vendors
  • topic starter
  • Posts: 9406
  • Joined: 03 Jun 2005
  • Loc: salem, OR

Posted 31 July 2008 - 12:01 AM

No, I didn't mean beveling itself was iffy, just that beveling as a way to finish up in general. Case in point - a 15" f/4.4 that needed a bit of beveling once completely figured, otherwise outstanding. I thought I got it all off (not wanting to take off too much) but the owner has told me that the mirror performs very slightly better with an edge mask on it - but he doesn't know exactly how wide the edge defect is, only that it's extremely narrow. Food for thought, anyway. :) Didn't want to try beveling it himself - I offered to touch it up but he thought the shipping didn't make it worthwhile...

Re polishing times & urethane, etc: I used to get sharp TDEs when using polishing pads, I suspect on account of the way the material reacts as it hits the edge. Urethane is possibly like that, they both have pressure sensitive backings SFAIK. The TDEs and the crazy zones they produced eventually went away when finishing up with pitch. Didn't you have some tools that have the material glued or laminated on somehow, that didn't show that effect?

But I don't get the polishing time comparison - polishing by machine with enough power to run mirrors at the same speed, I find the same substrate in different sizes polishes out in roughly the same amount of time. Takes more work on the bigger ones to do the job, but that's what the machine is for, IMHO.

Could it be geometry? A 14" quartz mirror with a good 12-micron grind polishes out completely for me in about 6-8 hours at 42rpm with an 80% Acculap tool, I run it several times that long just for good measure though.

Best,
Mark

#54 mark cowan

mark cowan

    Vendor (Veritas Optics)

  • *****
  • Vendors
  • topic starter
  • Posts: 9406
  • Joined: 03 Jun 2005
  • Loc: salem, OR

Posted 02 August 2008 - 04:29 AM

Here's some more related to edge zones, as that's the primary goal (at least for me) in adopting any edge-sensitive method of polishing. About 4 years ago I sent a 16x1.65" f/4.65 Pyrex paraboloid to James Mulherin at O** for null and interferogram testing, just out of curiosity. The results showed a 1/2" zone at the edge that appears to be turned up in null tests (like this picture). I hadn't seen that error while figuring the mirror! But knowing it was there I could then see it under Foucault KE testing, albeit with some difficulty. I left the mirror alone while building a Ross null tester and verified the appearance of the zone, then of course refigured the mirror. ;)

Attached Thumbnails

  • 2555709-openfringe null.jpg


#55 mark cowan

mark cowan

    Vendor (Veritas Optics)

  • *****
  • Vendors
  • topic starter
  • Posts: 9406
  • Joined: 03 Jun 2005
  • Loc: salem, OR

Posted 02 August 2008 - 04:32 AM

Here's one of the interferograms of the mirror (double-pass in auto-collimation against a flat):

Attached Thumbnails

  • 2555711-openfringe if.jpg


#56 mark cowan

mark cowan

    Vendor (Veritas Optics)

  • *****
  • Vendors
  • topic starter
  • Posts: 9406
  • Joined: 03 Jun 2005
  • Loc: salem, OR

Posted 02 August 2008 - 04:34 AM

Here's OpenFringe's reduction of the shown interferogram, for the full surface included the (now obvious) TUE:

Attached Thumbnails

  • 2555712-openfringe full surface.jpg


#57 mark cowan

mark cowan

    Vendor (Veritas Optics)

  • *****
  • Vendors
  • topic starter
  • Posts: 9406
  • Joined: 03 Jun 2005
  • Loc: salem, OR

Posted 02 August 2008 - 04:37 AM

And here's the reduction after trimming the outer 1/2" or so of the mirror's surface, in effect eliminating the edge zone (though to my eye this still shows some support-induced astigmatism from the sling support used):

Attached Thumbnails

  • 2555713-openfringe trimmed surface.jpg


#58 mark cowan

mark cowan

    Vendor (Veritas Optics)

  • *****
  • Vendors
  • topic starter
  • Posts: 9406
  • Joined: 03 Jun 2005
  • Loc: salem, OR

Posted 02 August 2008 - 04:50 AM

I'll note that James got better numbers from his testing (Strehl of 0.88 for the full surface and 0.93 for the trimmed surface at 550nm) so I may not be replicating his procedures while using OpenFringe - in any case I'm still on the interferometry learning curve while presently constructing a Shack-cube system. :grin:

The take-home from this for me was that the edge error existed before the mirror went into figuring, and just stayed all the way through. They don't magically fix themselves during figuring. I certainly became more critical of any slope errors near the edge after this - I'm pretty sure that mirror had come off the polisher with this zone on it and I missed it. Notice the fairly dramatic effect on the Strehl.

Best,
Mark

#59 Mark Harry

Mark Harry

    Cosmos

  • *****
  • Posts: 8086
  • Joined: 05 Sep 2005
  • Loc: Northeast USA

Posted 02 August 2008 - 07:12 AM

Looking at the 2 data reductions, I can see the mirror's not in a stable state, though gives coarse indications of similarity.
In the Focogram which side is the knife on? I can see a definite edge issue on both sides, but can't assess what it is without that info.
Can you see the astigmatism in both data reductions and the raw IFgram? Before getting the edge straightened out, I think I'd isolate the cause, whether stand, or glass before worrying about the edge. Some at the bottom looks as if it's due to the way it's supported on the stand, but at the top, the pattern is inclined about 10-15 degrees from horizontal. (definite indicator, IME) It would be much easier to see if the lines were "fluffed out" to only 3-5 line pairs across the whole aperture depicted.
On the right side of the mirror in the dark areas, I note a couple of the "lobes" depicted in the IF reduction pics. A little study can give you some ways to quantify the seriousness of these errors, and their detectability with the "inadequate" KE tester.
Thought I'd throw in my 2 millicents, hope you don't mind.
M.

#60 kfrederick

kfrederick

    Soyuz

  • *****
  • Posts: 3996
  • Joined: 01 Feb 2008

Posted 02 August 2008 - 07:30 AM

hi mark my best edge i made was a 14.7 quartz polished on top of a 17.5 in. medium aculap /long strokes/ the mirror was polished by hand/ only on the oversized lap. the edge polished as fast as the center and spherical i sent the mirror to a optician who is figureing it he said it has no astig . i tried a 14 in quartz before and used a 12 in polisher mirror on top very bad edge lot of astig.have you tried using oversized laps? not sure how they would work on a machine . thank for the info kevin

#61 mark cowan

mark cowan

    Vendor (Veritas Optics)

  • *****
  • Vendors
  • topic starter
  • Posts: 9406
  • Joined: 03 Jun 2005
  • Loc: salem, OR

Posted 02 August 2008 - 03:44 PM

The edge zone was always subtle under the Foucault KE at ROC. It was visible much more easily with my Ross null test. It was turned up, or more accurately, it was pretty much flat (nearly spherical at its own ROC) while the rest of the surface was fully corrected - this makes it look turned up or long. It was also very close to the diffraction edge, which complicates the IFs, the null image (which shows some slow turbulence but I can't be sure), and reading it with Foucault. But it was there and it was symmetrical all around the edge. The mirror obviously had a few other minor issues for which the outside testing was useful.

Just to clarify, these images all document a mirror that was completely reworked in '04, from a sphere, even though James seemed to think it wouldn't be hard to "fix" the edge. My experience with "fixing" edges or any narrow defects has never been satisfying. I have several more IFs from the same test if they're of interest. The two data reductions are from the exact same IF BTW, with just the outer 1/2" left out on the second one.

The test stand astig is similar in nature to every thin mirror I've had tested with IF so far, and James noted it as well as being typical for their setup (vertical support with a sling). He said at the time As mentioned; if not for the TUE the mirror would be quite good so I don't think he saw anything of concern relating to astig. I have 5 IFs that all appear to be taken on the same orientation (unfortunately), when I have some more time I'll trace all of them and average the result as the fringes move a bit and that should improve the accuracy of the analysis.

James was quite helpful in interpreting these results and then waived his usual fee for doing the testing. How cool is that?

In the meantime, here's a Ronchigram in the null test that shows the edge defect as well as the slight undercorrection.

Thanks for your input!

Best,
Mark

Attached Thumbnails

  • 2556437-openfringe ronchi.jpg


#62 mark cowan

mark cowan

    Vendor (Veritas Optics)

  • *****
  • Vendors
  • topic starter
  • Posts: 9406
  • Joined: 03 Jun 2005
  • Loc: salem, OR

Posted 02 August 2008 - 03:57 PM

the mirror was polished by hand/ only on the oversized lap


Yeah, that works very well, it's what I've done for making spheres when the edge has to be good too. I think if you do it by hand that way with quartz and you rotate it regularly it's unlikely you'll fail to get a good astig free result.

Using oversize polishing laps on top with a machine might be difficult though I expect it could be done. I've used full size tools on top on the machine, but oversized tools only MOT and by hand, just like you're describing. You'd almost have to pitch the mirror to a backing plate to work it MOT on a machine, I suppose. :question:

Working with the machine and thin mirrors TOT you need a very flat true turntable and a thin flotation layer (the webbed latex carpet-backer pads or shelf liner rolls sold at home improvement stores work well). The tool can't be allowed to get into resonance with the table rotation, and there can't be any wedge in the glass, plus the optical/physical center of the mirror have to match (no off-center grind). Given all that, working by hand MOT on an oversized tool by hand is probably less trouble if you're only doing a couple mirrors.

Best,
Mark

#63 mark cowan

mark cowan

    Vendor (Veritas Optics)

  • *****
  • Vendors
  • topic starter
  • Posts: 9406
  • Joined: 03 Jun 2005
  • Loc: salem, OR

Posted 02 August 2008 - 10:37 PM

Here's a Zernike average of the 3 best quality IFs for the full surface for comparison. For whatever reason this is a considerably different result than the first frame I showed, and furthermore agrees with James's result for Strehl. It reproduces the null surface image much more closely now, so perhaps turbulence was to blame. Tracing the fringes is slow stuff. ;)

The trimmed version just looks like this with the edge cut off but the Strehl goes to 0.921 - which isn't a big change but shows what a tough to see edge zone can do, which was the point.

Best,
Mark

Attached Thumbnails

  • 2556959-openfringe average ADE.jpg


#64 Mark Harry

Mark Harry

    Cosmos

  • *****
  • Posts: 8086
  • Joined: 05 Sep 2005
  • Loc: Northeast USA

Posted 03 August 2008 - 06:38 AM

"In the meantime, here's a Ronchigram in the null test that shows the edge defect as well as the slight undercorrection."
****
Now is this on James's test rack or yours? the astig is virtually non-existent here. The IF showed it plainly.
M.

#65 Glig

Glig

    Messenger

  • *****
  • Posts: 431
  • Joined: 02 Oct 2005
  • Loc: Baltimore, MD

Posted 11 August 2008 - 01:16 PM

This thread is a favorite of mine and before it sinks into the west for good I'd like to address an old issue. I hope this is not a sore point. Mark Harry I hope you'll forgive me for using such an old quote of yours (From my thread called Handle Grinder)but here goes.

Mark Harry: "...Using a fast spindle is called "spin-polishing". There are some who can figure mirrors that way,(large mirror-small polisher scenario) but if you get carried away, you can really rip an edge that'll take forever to fix. Been there, done that, seen others do that, and thank goodness, I know better!"

I responded: "...I'm wondering how tub polishers could rip the edge? I was told that tales of fast spindles turning down edges are a myth..."

Mark Harry: "...Whoever said it can't happen hasn't done any optical work with machines, period. Anytime the polisher has a tendency to "dwell" at the edge, with some overhang as it changes direction while the spindle has a chance to rotate a slight degree, will turn an edge for sure, guaranteed..."

It seems to me that spin polishing has been a lively topic between Mark Harry and Mark Cowan. I'm been trying to follow it all, but I'm not sure where Mark Harry's ideas about spin polishing and TDE is today. Have we reached a verdict on this? Is there still a fundamental disagreement here? :gve:

#66 Ed Jones

Ed Jones

    Soyuz

  • *****
  • Posts: 3819
  • Joined: 06 Apr 2004
  • Loc: Sin-sin-atti

Posted 11 August 2008 - 07:44 PM

And if you go ahead and remove the spherical and trefoil it'll be a darn perfect mirror! : :roflmao:

#67 Mark Harry

Mark Harry

    Cosmos

  • *****
  • Posts: 8086
  • Joined: 05 Sep 2005
  • Loc: Northeast USA

Posted 11 August 2008 - 08:13 PM

I haven't changed my opinion on that method one bit, Richard.
***
One provision: I believe Mark C is working with quartz, and a polishing medium that lacks resilience. It -CAN- work under those conditions, if you tweak the parameters correctly. (we used "flower-power" with quartz flats, and the spindle speed wasn't too critical, and within the machine's capability) But I think if he changed substrates/pitch, I'd be very suprised if he'd have that same success with minimal edge issues.
***
Richard, I'd say you were trying to 'stirr the pot'......Am I right?
M.

#68 Glig

Glig

    Messenger

  • *****
  • Posts: 431
  • Joined: 02 Oct 2005
  • Loc: Baltimore, MD

Posted 11 August 2008 - 10:34 PM

Richard, I'd say you were trying to 'stirr the pot'......Am I right?
M.

:o :crazy: :mrevil:
O.K. You got me, Mark H. Guilty as charged. But also for another reason. This thread and others like it I'm saving to disc and keeping forever. I'm a seeker of knowledge first and foremost - and this is great stuff.
:bow: :bow: :bow:

#69 Joe Cipriano

Joe Cipriano

    Entropy Personified

  • *****
  • Posts: 8094
  • Joined: 03 Sep 2005
  • Loc: Uh... anyone have a GPS?

Posted 11 August 2008 - 11:07 PM

...Richard, I'd say you were trying to 'stirr the pot'......Am I right?


Annoy sleeping dogs, throw rocks at hornet's nests, etc.

Let's not.

Things have remained civil (kudos to all involved), which can be difficult for a discussion as rigorous as this one. I'd REALLY like to see it continue in the same manner.

Thanks.

#70 mark cowan

mark cowan

    Vendor (Veritas Optics)

  • *****
  • Vendors
  • topic starter
  • Posts: 9406
  • Joined: 03 Jun 2005
  • Loc: salem, OR

Posted 15 November 2009 - 05:35 PM

OK, returning to an old thread. A couple comments above I don't recall seeing before...

Mark H. wrote:

One provision: I believe Mark C is working with quartz, and a polishing medium that lacks resilience. It -CAN- work under those conditions, if you tweak the parameters correctly. (we used "flower-power" with quartz flats, and the spindle speed wasn't too critical, and within the machine's capability) But I think if he changed substrates/pitch, I'd be very suprised if he'd have that same success with minimal edge issues.


Curious - I've used the exact same technique on plate and Pyrex as well as quartz, all with fine results. It's the way I prep any mirror. I've also gotten other people to try this to address edge issues, and some have applied it very well and solved their problems. I've no control over their particular substrate/polisher combos.

What I'm not sure about (at this great remove in time) is "polishing medium that lacks resillence"? AccuLap standard is pretty stiff - you need the hardness to make the technique work, partially because it generates significant heat on the substrate, but mostly because the wire-brushed hot-pressed contact needs to remain intact and operating consistently over the 20-30 minute span of the (repetitive) process.

Ed wrote:

And if you go ahead and remove the spherical and trefoil it'll be a darn perfect mirror! :roflmao:


Uhm, yeah. Later work did manage that trick, at least to my satisfaction.

Mark again:

Now is this on James's test rack or yours? the astig is virtually non-existent here. The IF showed it plainly.


Yeah again. The IF and analysis were by James, on the house, and SFAIK he does arbitrarily remove astig in producing the analysis, due to test stand effects. Can get you in trouble, though, unless you average several orientations to Zernikes and then average those to eliminate test stand effects.

Which brings me up to the present, and a new example to show. For the thread vault here's a link to a YouTube video showing what spin polish finish prep looks like on a 14.7" f/4.5, with a 12" polisher under 25 lbs weight (1/3 lb per square inch working pressure).

spinning.mov

This sequence ran essentially unchanged for about 30 minutes, after which the surface shown in the next post (if there is one) resulted.

Best,
Mark

#71 mark cowan

mark cowan

    Vendor (Veritas Optics)

  • *****
  • Vendors
  • topic starter
  • Posts: 9406
  • Joined: 03 Jun 2005
  • Loc: salem, OR

Posted 15 November 2009 - 06:37 PM

Here's a shot of the edge of that mirror - actual size as taken (though cropped) - it's with the KE past center and well inside the paraxial COC as discussed in Texereau, illumination from a 50 micron precision pinhole. Not much to not like. ;)

Best,
Mark

Attached Thumbnails

  • 3449947-spin example.jpg


#72 Mark Harry

Mark Harry

    Cosmos

  • *****
  • Posts: 8086
  • Joined: 05 Sep 2005
  • Loc: Northeast USA

Posted 16 November 2009 - 07:25 AM

""polishing medium that lacks resillence"? AccuLap standard is pretty stiff -"
Yup, that's what I said, and I still agree with you. It's relatively "hard as a brick" and temps encountered while spin-polishing shouldn't reduce the polisher to a pile of goo. I don't think ordinary pitch would be as forgiving with the same treatment.
Nice edge, btw.
I still use the ol' "double 'G'" (gooey Gugolz) My method works well with that stuff, and has other benefits down the road, such as really good rotational symmetry, and no central artifacts.
Thinking about it, some of that pitch is more than 15 years old- still working well!
************
As to polishing and figuring- I did a first with 2-6" mirrors. (both hand polished & figured) One an F/13.63, and the other F/4.5. (complementary opposites) The cut on the polishers amounted to a total of 4 grooves; very shallow. Similar to what you'd have with a polisher used on flats. Of course, they were conformed to the correct ROC in both cases. Logically, it can be assumed that the "flat polisher" would work well with the long mirror; which it did. The test reduction worked out to about 1/37th PVW, at a rounded off 1.00 Strehl. It was a dead-nuts sphere, with a rolloff at the edge of .015" of 1"+ width. One figuring session lasting 10 minutes at the most.
The other case, not so straightforward. Short ROC, relatively lots of deformation there. -BUT THE SAME POLISHER CUT WORKED ON IT, AS WELL!- 4 very shallow grooves, etc.
I quit with this one when the smooth zone-free undercorrection lied just inside the tolerance envelope. The PVW amounted to a bit better than 1/8th, with .93 Strehl, if I recall correctly. No zones, water smooth finish, just a gradual smooth undercorrection across the whole aperture.
The other interesting aspect, was time involved with this last mirror- 5-6 figuring sessions, about 15-20 minutes apiece. With supper, and just plain diddling with the testing, I finished it in under 4 hours. I'm still trying to wrap my mind around all aspects of this last example. It's feasible to imagine that the method I used here would work in production reasonably well and get very accurate smooth mirrors in a timely fashion -BY HAND!- (could it spell the end of 6 month waits???!!!)

#73 mark cowan

mark cowan

    Vendor (Veritas Optics)

  • *****
  • Vendors
  • topic starter
  • Posts: 9406
  • Joined: 03 Jun 2005
  • Loc: salem, OR

Posted 16 November 2009 - 01:25 PM

Hi Mark,

""polishing medium that lacks resillence"? AccuLap standard is pretty stiff -"


Oops - brain lock! It's them foreign words, Mark, and double negatives. (What) was I thinking?

I'm prepping a quartz 14.7" f/6.7 (second of a bino pair) and I'll post how the edge turns out on that one, as you expressed some concerns once about longer focus results. The mirror needs to finish prep with no artifacts as well - figuring is a very short job. ;)

It's feasible to imagine that the method I used here would work in production reasonably well and get very accurate smooth mirrors in a timely fashion -BY HAND!- (could it spell the end of 6 month waits???!!!)


Once, it's an anomaly; twice, it's a trend; three times - could be! Sometimes they just fall into place, other times you have to work way too hard.

Best,
Mark

#74 Glig

Glig

    Messenger

  • *****
  • Posts: 431
  • Joined: 02 Oct 2005
  • Loc: Baltimore, MD

Posted 16 November 2009 - 02:19 PM

Mark could you start a post about spin FIGURING?
I can't seem to remember you saying a lot about figuring - but then, my addled brain can't remember much anyway. :gottahurt:
Or, you could write about it here on this supercool thread.

:bow: :thankyou:

#75 mark cowan

mark cowan

    Vendor (Veritas Optics)

  • *****
  • Vendors
  • topic starter
  • Posts: 9406
  • Joined: 03 Jun 2005
  • Loc: salem, OR

Posted 16 November 2009 - 03:02 PM

Me? I only use spin polishing for polishing out and mirror preparation for figuring. And my long-standing policy is to never talk about figuring (too many variables), so your addled brain is exactly correct!

Appreciate the comments, though.

Best,
Mark


CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics