[garnetram]

Has anyone ever machined and lapped Mic-6 aluminum plates? 8 inch disc from 1/4" to 1/2" thick. The final flatness spec is .0002" Which has proven to be a difficult goal.

Any help would be awesome!

[the4thseal]

my experience has been that aluminum is too soft for lapping. the compound's grit embeds in the aluminum.though you might using the term lapping,differently than me. there are surface grinding discs that will do aluminum. good luck

[Mcgyver]

do you need it .0002 flat, or just .0002 dimensional thickness? The later isn't too bad, the first is a challenge. whats the application that it would require this flatness instead of just dimension? apologies if i'm telling grandma how to suck eggs, but flatness isn't often neccissary as thin parts are usually sandwiched or fastened down so dimensional accuracy might be important but flatness is not..... point being maybe the customer doesnt actually need it .0002 flat.

getting thin things flat is a challenge I agree with 4th on the dangers of lapping AL, BUT you can get non embedding lapping compound that will work....however, how are you going to hold it and lap such that you are distoring the part? it would be very tough to do this via lapping i would think.

thats the challenge, any way you to go hold it changes its shape. if it was steel you could grind on a mag chuck but even then its a royal pita as the force of the chuck will pull it out of shape and you achieve flatness by shimming.

The only way i know that will work is to scrape it in, that way there is no clamping force while the areas requireing material removal are identified. not impossible to hit your spec with that approach, but its time consuming....make sure you charge enough because its like gauge making work instead of part making

interested to see what other ideas come forth...

[the4thseal]

mmmmmmmmmmmmmmmmmmmm... scraping.(think homer and donuts). a topic near and dear.

[scadvice]

...look for a shop in your area. I've had parts done to .0005 = with just DD only. There are shops in Calif. that can do it. The cost the last time I had it done was around 8 bucks ea. for 80 pcs., as I recall. As I said that was just double disc grinding cost not the lapping. From time to time I see and bid on small lappers on ebay but haven't won one yet.
Hard drive makers double lap alum. all the time.
Steve.

[ImanCarrot]

Use true tools and aluminium oxide with water.

I can lap 12" disks flat to a twentieth of a wavelength of light. That's 1/20th of 633 nanometers. ie: 32nm peak to valley across the whole aperture. That's 0.0012".

It can be done, just takes a lot of skill.

See Engis or Hyprez for more advice, or any company that operates a Lapmaster.

Incidentaly, to go lower than a twentieth of a wavelength is difficult since light starts diffracting at that level and acts more like a particle than a wave. ie it's blinking hard to measure to that accuracy.

[garnetram]

We have done some scraping here, but for this application It would not be suitable.....I forgot to mention a few "minor" details! LOL:withstupi

Top and bottom need to be paralell withing .0004" and both surfaces need #4 finish.....

Sorry I forgot that!

On another note we have noticed that as a rule after maching these disc to thickness and a reasonable machined flatness (some of theese disc have through details in then such as pockets and tapped holes) and after sitting a few days they tend to "move" on us. Thus requiring rework. Would stress relieving help? Has any ever stress relieve Mic-6? I have found a little info on the net but have not actually tried it yet.

We would love to do all of this "in-house" if possible. As a rule our customer requires a turnaround too fast to allow for outsourcing, and we have had too many parts such as this come back damaged from either the vendors shipping dept or UPS.

[steve323]

I can lap 12" disks flat to a twentieth of a wavelength of light. That's 1/20th of 633 nanometers. ie: 32nm peak to valley across the whole aperture. That's 0.0012".

Do you mean something close to 1.2E-9 inches? Wow.

Steve

[ImanCarrot]

I make ir 1.2 e-6 inches?

1 twentieth of a wavelength where the wavelength is, say, 633nm (0.000633mm)=

0.000633/20mm

= 3.2e-5mm

divide by 25.4 to get inches

3.2e-5/25.4

= 1.2e-6 inches.



Oh Garnetram, I dunno what mic6 ally is, I useualy use L111 ally which is an aircraft/military spec (I think the commercial equivalent is 6082, but not sure- anyone else know?)

IT IS ESSENTIAL WHEN LAPPING TO THIS LEVEL THAT YOU USE CAST ALLY, NOT ROLLED OR EXTRUDED OR ANYTHING ELSE OTHERWISE YOU WILL GET STRESSES.

Heat treating to anneal the stresses might help but I doubt it on rolled ally which bends like a bannanna when machined.

Oh, how are you holding it? if you're vacuuming it down then you can get print- through from the rear surface or chuck which will afeect the flatness.

You could get it Diamond Machined- that's get your flatness no problem. And thickness tolerance.

[Mcgyver]

I can lap 12" disks flat to a twentieth of a wavelength of light. That's 1/20th of 633 nanometers. ie: 32nm peak to valley across the whole aperture. That's 0.0012".

It can be done, just takes a lot of skill.

.

how thick?

pending some specialize lapping machine, with regular shop equipment, that challenge is how how hold it. by hand, vacuum, holddowns, whatever, its going to move a 1/4" 6" disk more than .0002; ie holding will distort so the warp/bow/whatever in the work reappears when the pressure is released - even very slight pressure

Oh, how are you holding it? if you're vacuuming it down then you can get print- through from the rear surface or chuck which will affect the flatness

how do you hold it? what ever way, unless you map the surface with a very good indicator and then shim will pull a 1/4 disk many thou to its own plane. I agree movements a potential problem, and could be even with cast although less than rolled....you sure want it stabilized before you try to get a flat section to 1/10ths. would also depend on whether you started from flat stock or a slice off a 6 inch bar. in steel I'd know what to do but not sure offhand how to normalize/stabilize/season AL

You could get it Diamond Machined- that's get your flatness no problem. And thickness tolerance.

can you explain how that works so that flatness isnt' a problem - you still have to hold it down, right? what type of machine?

Iman, can you give some more insight into how exactly you're lapping these disks? are doing this work by hand or with some special lapping machine? I'm interested in the process and would like hear more about it

[ImanCarrot]

Lapping by hand although for production quantites you'd probably buy a lapmaster machine. I think you get a better accuracy by hand (less edge roll-off), but you need to touch in your own true tools whcih is an art itself!

www.lapmaster.co.uk (I have no connection with the company)
Engis are good too for consumables
www.engis.com

Diamond Machining- essentialy what you do is hold the component on a vaccuum chuck. this vaccuum chuck has itself been diamond machined flat. The rear of the component is lapped flat and vaccumed down to the chuck. The front of the surface is then cut. No print through, no stresses= perfect opticaly flat surface and an optical mirror finish.

Aluminium is perfectly suited to diamond machining (some folk call it diamond turning), but it's costly. Actualy to give it its full name- Single Point Diamond Machining (wiki it for a fuller expanation).

[Mcgyver]

i know single point diamond machining, once you get one surface flat life gets easy(ier) the question I have is the details on how you establish that first reference flat on the one side via lapping. Can you give some more detail about how you hand lap this? are you using some optical device to show where the work is high and subsequently lapping there? It can't just be relying on the flatness of the lapping plate as the work is so thin hand pressure alone will distort, at least to the tolerance we're talking about. thx

[ImanCarrot]

How to explain. Hmmm... ok....

Every optical manufacturing engineer has three flat "true tools" which he alows no- one else to touch. They last you a lifetime. They are 12" or 6" in dia, about 1" thick and usualy have a 1" female whitworth thread on the back on a sticky out boss (about 1.5" tall). I no longer lap stuff, but mine's were made of mehanite cast iron, gun metal is good, but must be cast.

You work the true tools (for a month or two) to personal taste- I like #1= 2 fringes 'vex, #2= 2 fringes 'cave and #3= less than half a fringe flat.

Right, so you got your true tools (you need to touch them in every week or so).

You also need a test plate which is flat to a twentieth of a wavelength. You can buy them (costly), but I prefer to make my own. I prefer Zerodur as a material- a zero expansion glass ceramic, kinda light amber in colour. Oh, you need a monochromatic light source. I used an old street lamp which is Sodium, but Mercury lamps are good too, slightly different wavelength but who will know

I'll say your part has a back surface and a front (which is going to be diamond machined). You lap the back surface with true tool #1 untill it has a uniform grey. You then brighten it by burninshing dry against the true tool and then test plate it. It should read about 1-2 fringes 'cave. Stick it to one side and have a ciggy break.

Diamond turn the actual sub- chuck flat. Check the chuck using the test plate or an interferometer.

You now have two flat surfaces which to all intents and purposes are a perfect match.

Clean everything spotless and vacuum the back of the component to the sub- chuck on the machine.

Diamond turn the front surface. Check the figure using an interferometer under vacuum (a test plate or talysurf will scratch ally). Re-check off the machine to determine any residual stresses.

If it's an optical mirror stick a high reflectivity coating on it with a protective coating on top of this.

Wrap up in copious amounts of bubble wrap sent to the customer. Await the customer calling to say the courier accidentaly dropped it from a very great height. Start again.

[Mcgyver]

thanks very much for the detailed description - I think i'm getting it .

So initial flatness is from the flatness of the 'true tool' you lapped the work to (I still think 1/4" thick would be a challenge given finger pressure alone will affect how the work contacts the lapping plate) and you have three so they can be checked and touched up when required? .....and the checking of the work is done via an optical instrument, an optical flat?

sorry for the million Q's, optical work is out of my realm and but i'm intersted in how its done

[ImanCarrot]

Finger pressure is indeed a problem. You use your first three fingers of each hand and your thumb-picture placing your pointy finger, middle finger and ring finger in a straight line slightly above the centre line of the component back. Place your thumbs below this. Two pairs of surgical gloves stops heat from your fingers distorting the component.

Depending on how you rotate the component being lapped you can control whether it goes convex or concave. Watch out you don't change the figure too rapidly though otherwise you'll get a roll- off at the edge which will take a while to remove.

Everything nice and slow and easy is the way.

Oh! I nearly forgot! "Prism and Lens Making" by F. Twyman (the guy who invented the Twman Green interferometer) is a good book for all this stuff. Proper engineering. There's another book called "Optical Production Technology" by a chap called Horne which is basicaly a re-write of Twyman's book and in my opinion absolutely no good, it basicaly copies some of Twyman's stuff and misses out the good bits. In addition the maths that it adds is flawed with quite a few mistakes which can totaly confuse the novice.

If you ever get the chance to visit the natural history museum in London they have Newton's original polishers (pitch lapping plates) which are almost identical (albeit a wee bit smaller) than those that us optical engineers use today

Oh and we tend to use cerox rather than rouge these days cos rouge stains everything from your boots to your eyelids red, they used to refuse to let you factory space if you were using rouge cos of the mess it made.

[Edit]Went off at a tangent there, sorry, went prom lapping (or smoothing as we call it) to polishing. Don't use pitch for polishing ally, it won't work. Pitch is for glass and stuff like germanium, zinc slelenide etc. Making your own recipie of pitch is an art in itself. Start with pine pitch, rosin, shellac, beeswax and paraffin wax some linseed oil and a handfull of wood flour. The old chaps who taught me how to do all this used to bite the pitch (no kidding) to see how hard it was- different hardnesses are used at different temperatures and for different materials. Immagine getting that by COSH (or in the US HAZMAT) these days- chewing carcinogenic stuff. Mind you, they use to wash their hands in napthalene before it was replaced by trichloroethane, but that's another story hehe.[/Edit]

[Geof]

....On another note we have noticed that as a rule after maching these disc to thickness and a reasonable machined flatness (some of theese disc have through details in then such as pockets and tapped holes) and after sitting a few days they tend to "move" on us. Thus requiring rework. Would stress relieving help? Has any ever stress relieve Mic-6? I have found a little info on the net but have not actually tried it yet....

I had not looked at this thread earlier and now I find it has some really interesting information.

Regarding Mic-6 I think it is annealed during manufacture and it is the most stress free and stable alloy. It is possible you are pushing the envelope too far; all aluminum alloys gradually change their crystal structure over time and this can cause growth or shrinkage.