Question about lapping.

[Flynth]

I'm currently making a set of 3 small mild steel flat laps. They were faced on a lathe, grooved, then surface ground to be flat to within a micron. Then 3 plate method is being used to improve their flatness and surface finish.

However, as the flatness is getting better it becomes difficult to measure progress using anything else than an optical flat. But the surfaces are not reflective enough. So my question is, does anyone know some method to improve surface reflectivity during lapping before moving to the smallest grit size. Will buffing ruin the flatness? Perhaps some way to polish it?

I'm currently using a 14~9 micron grit that results in moderately reflective surface. I tried switching to 3~1 micron and it improved the reflectivity a bit, but not hugely so in the short time I tried it. How long does it take to switch grits like this? This is silicone carbide, not diamond abrasive grit BTW.

 

[benmychree]

I think that steel is not the best choice for a lap, cast iron is much better due to its ability to imbed abrasive.

 

[OTmachine]

I believe optical flats are used with a monochromatic light. Seems that ours was a Helium type light, ( a long time ago).

 

[Flynth]

Although valuable your replies are not answering the question... Have you made flat laps like this before? How did you achieve enough surface reflectivity before using the finish grit? As Tom Lipton said in his video "You gotta now if you're making it better or hurting it(during lapping)". Or did you just carry on hoping for best and it all gets revealed only at the end?

I would like to be able to check progress on my already pretty flat laps (off the surface grinder) with an optical flat. So I'm asking what is the best way to improve surface reflectivity without ruining it's flatness. I assume a buffing wheel will take off material where it shouldn't. Is that right? Or is buffing the way to go? Or if one switches to a fine grit briefly how long does it take to get good reflective surface?

I think that steel is not the best choice for a lap, cast iron is much better due to its ability to imbed abrasive.

Indeed that is true, but in absence of cast iron mild steel is a fairly good alternative providing the parts to be lapped with the lap are hardened.(parts should be harder than the lap). There is no such thing as durabar available for a reasonable price at short lengths (disks) here in my part of Europe. The only ductile cast iron available is from speciality suppliers that charge an arm and a leg for it. Therefore mild steel.

I believe optical flats are used with a monochromatic light. Seems that ours was a Helium type light, ( a long time ago).

The nice warm glowof it is very pleasing

However, any sufficiently monochromatic diffuse light source will do. Low pressure sodium lamps allegedly are very good, unfortunately for some reason they are currently extremely expensive around here ($100+ for a 70W low pressure sodium light bulb!?). Some people use mercury vapor lamps.

Ordinary LEDs have too wide spectrum bandwidth to be of use, but lasers (especially single mode) are a very good modern light source for an optical flat. I found one needs a pretty powerful laser (at least 100mW) and one way to make it work is to remove the lens so the beam diverges naturally, then point that onto a piece of white paper suspended over the part to be viewed slightly off to the side (see pictures of the doall light for a correct angle) and view from the same angle. On a suitably reflective surface (gauge block) the fringes are very good.

 

[homebrewed]

I wouldn't buff the surfaces, there's a good chance that the edges of the flats will become rounded.

You are undertaking a challenging task. You may want to look into metrology techniques used by amateur telescope makers when grinding their own mirrors, or perhaps holographic approaches -- even laser pointers are very bright compared to conventional light sources. It might be possible to coat your flats with a very thin layer of plastic dissolved in a solvent, so the low points are sort of filled in so your optical flat will work better. Look up spin coating for more on that.

And to answer your question regarding how long it takes to polish something to a few wavelengths, again, take a look at information regarding amateurs making telescope mirrors. Some are very good at it.

I also suggest getting a good microscope to inspect your surfaces so you can get an idea regarding your progress in terms of removing scratches left by coarser polishing media. You can always switch to a finer grit but if you do that too soon you will spend a lot of time polishing out the scratches, and run the risk of rounding the edges of your flats.

 

[Flynth]

Thank you for answering the question

I wouldn't buff the surfaces, there's a good chance that the edges of the flats will become rounded.

Yes, that was my worry too, that's why I started this thread to ask.

You are undertaking a challenging task. You may want to look into metrology techniques used by amateur telescope makers when grinding their own mirrors, or perhaps holographic approaches

I've since found a very good resource here: http://bobmay.astronomy.net/johnpierce/part1.htm

It is written for someone making a telescope mirror, but the techniques are the same. I've since found an answer regarding making the surface reflective mid way through grinding. It seems when they make telescope mirrors and other optics they too have to make the surface reflective to check it. The way they do it is with a special "pitch" lap. It is cast from pitch on top of the object being made, Then they use very small polishing grit "for half an hour" to make it reflective enough. For me, when making flat surfaces there is no need for pitch. I can just clean all 3 laps and grind using extremely fine grit. That's what I did, but the grit I used wasn't fine enough (1~3 microns) or I didn't do it for long enough. So I used a precision ground flat stone on the surface at the very last stage and that made it mirror reflective where it contacted the stone.

-- even laser pointers are very bright compared to conventional light sources. It might be possible to coat your flats with a very thin layer of plastic dissolved in a solvent, so the low points are sort of filled in so your optical flat will work better. Look up spin coating for more on that.

And to answer your question regarding how long it takes to polish something to a few wavelengths, again, take a look at information regarding amateurs making telescope mirrors. Some are very good at it.

That's not what I was asking
I meant how long does it take to make it really reflective. My surfaces are within a couple wavelengths to start with (surface ground) so I'm mostly concerned with not making them worse. I'm trying to check progress frequently as I go, because I believe it is very easy to grind it worse in a hurry.

I also suggest getting a good microscope to inspect your surfaces so you can get an idea regarding your progress in terms of removing scratches left by coarser polishing media. You can always switch to a finer grit but if you do that too soon you will spend a lot of time polishing out the scratches, and run the risk of rounding the edges of your flats.

I do have a couple of microscopes, but it didn't occur to me to use one for this. Thank you for a useful idea

 

[Flynth]

Well, it turns out free abrasive rolling between two plates doesn't produce a mirror finish... It is very flat and fine, but matt. Not reflecting. They use the pitch lap in optics and they have it embed the abrasive.

In the videos I saw people used a diamond to get the reflective finish. It isn't clear to me if Tom Lipton in his video Just used a finer slurry on his laps, or did he use his "recharged" diamond lap?

Still, I'd rather not use diamond because of the risk of it embedding.

So did anyone here make flat laps using the 3 plate method and optical inspection? If yes, how did you get the surface reflective enough to inspect it?

 

[homebrewed]

Well, it turns out free abrasive rolling between two plates doesn't produce a mirror finish... It is very flat and fine, but matt. Not reflecting. They use the pitch lap in optics and they have it embed the abrasive.

I wouldn't have expected that, good to know.

I've done polished cross sections for ~30 years but always used polishing pads, usually charged with aluminum oxide slurry, on a rotating polishing wheel fitted with a water drip -- so I never encountered the problem you describe. Cleanliness was paramount, otherwise you never got a good unmarred surface; and you had to replace the pad because it was contaminated by coarser grit. An ultrasonic cleaner with a mild soap solution worked best.

Going back to your discovery, can you clarify what you mean by "free abrasive"? Is it dry powder, or water based/oil based paste? I've seen and used both water and oil-based lapping paste, they both have their place. Water based cleans up more easily but tends to dry out.

 

[Dabbler]

FWIW you might try very light machine oil, applied using a lint free cloth. The lpa should be clean of any contaminants for it to work, of course. It might make it reflective enough.

I've made brass and aluminum laps, but not steel ones, so my suggestion is just theory.

 

[Flynth]

I wouldn't have expected that, good to know.

I've done polished cross sections for ~30 years but always used polishing pads, usually charged with aluminum oxide slurry, on a rotating polishing wheel fitted with a water drip -- so I never encountered the problem you describe. Cleanliness was paramount, otherwise you never got a good unmarred surface; and you had to replace the pad because it was contaminated by coarser grit. An ultrasonic cleaner with a mild soap solution worked best.

Going back to your discovery, can you clarify what you mean by "free abrasive"? Is it dry powder, or water based/oil based paste? I've seen and used both water and oil-based lapping paste, they both have their place. Water based cleans up more easily but tends to dry out.

I have to admit I thought I'm doing something wrong, but then I saw this mentioned on another forum. So I can't quite claim it is entirely my discovery

My "free abrasive" are two things. The 14~9micron is an oil soluble silicon carbide paste made by Diamant (special paste no 800). It is a very thick paste (like grease) so I use deodorised petrol to thin it a lot. I initially used a small container to dissolve the paste, but then I found putting a small dab of it on the lap, spreading it around with a small brush. Then I drip petrol on top. I continue adding petrol As it evaporates.

The other abrasive is a white aluminium oxide polishing stick. On a buffing wheel that paste produces a mirror finish. I don't know the grain size, but I estimate 1~0.3micron. Under the microscope the grains are substantially smaller than the Diamant paste and they are needle shaped. This stick is a very hard crayon. So I scrape some into a mortar, I add some petrol and I grind on it with the pestle for a bit until it dissolves. Then I use a dropper to apply it.

FWIW you might try very light machine oil, applied using a lint free cloth. The lpa should be clean of any contaminants for it to work, of course. It might make it reflective enough.

I've made brass and aluminum laps, but not steel ones, so my suggestion is just theory.

I might try that.

I did two more courses of the polishing paste, then I used one of my precision ground stones on the surface. The stone did make it reflective enough to see good fringes.

I wish I took a photo of that matt surface, but I only took photos of the shiny ones... If you're interested how it looks it is a grey matt color pretty much exactly like here:

Although Tom is using diamond with a copper lap, so I don't know how much "rolling around" and how much embedding.

I'm pretty happy with my laps so far. They are relatively flat, but measuring exactly how flat is not that easy. It is fairly easy to tell if the surface is pretty bad with an optical flat, but when the surface is quite good I find it difficult to estimate the curvature of fringes precisely. Here is few photos. Perhaps someone else can tell more from those fringes.
The surface is split into 3 photos because my light source is not big enough to illuminate the entire surface. The light source is a monochromatic DPSS laser at 532nm.