Lapping a small steel (hrc 56) part to flatness and mirror finish.

[luke8888]

I'm trying to see how flat can I get a gauge block sized hardened steel part (o1 - hrc ~56) while maintaining flatness with hand lapping. I'm after some advice on this.

I have a number of mild steel serrated (and smooth) lapping plates that are flat to 0.5 micron across their size (100mm/4in).

My first try wasn't that encouraging. I used wet lapping with Sic and semi-dry with diamond. I'm after some technique tips as well as when to use wet(free abrasive) vs dry(fixed abrasive, but with liquid lubricant). There is plenty of materials that explain the difference between both processes, but I haven't found any info when to prefer which? Also, when to use serrated and when smooth laps? Likewise with the abrasive choice. Sic or diamond? (we're only talking about hardened steel workpiece and mild steel laps - no cast iron is available to me in sane prices).

The blocks were not ground to a great finish to start with so I used 50 micron silicon carbide (240 grit) first to take out grinding marks. This was in a form of a very small amount of paste. Some light oil was added on the lap to lubricate.

Once done I then used 14~9 micron (500~800 grit) SiC paste in kerosene which resulted in a really nice matt grey finish.

Then I used 3 micron diamond. I pressed it into the lap with a hardened steel block before use,but I haven't cleaned the lap after so there may have been some free diamonds rolling around. (I'll try to embeds it properly with a steel roller next time)

I finished with 1 micron diamond. The finish was nice, but not quite mirror finish. Still it was reflective enough to use an optical flat to check the surface.

Unfortunately my block wasn't flat, but was convex with 2 micron difference between the peak and sides.... So I lapped it at least twice worse than it started with... I already have some ideas where I might have screwed up (too much abrasive, too much pressure when lapping, not cleaning free diamonds ). Also I removed a total of 7 micron (about 3 tenths)? Is this too much to expect it to remain flat within half a micron? The next try will be with much better grind so I can start with 3 micron diamond, but I thought to ask for any tips and advice on the above (when to use dry/wet, which lap, which abrasive, how much material is reasonable to remove and remain flat etc).

I've done a lot of searching on the subject, I read what the machinery handbook has to say as well as the Whitworth book. But some questions still remain.

 

[eKretz]

You've got to get the amount of grit or paste right or you will have flatness issues. Too much will basically float the part. Better too little than too much. I like grooved plates because they help to mitigate that problem, giving the excess somewhere to go. Also, watch your speed. Moving too fast will start to float the part too. I like to run drier, with less grit (if lapping with loose grit) and with slower movement when I'm really shooting for best flatness.

You've also got to watch the geometry of the part at the very beginning. If you have a convex surface to begin with, it can be very difficult to get it to flatten out. I like to start with a flat to slightly concave surface. If I don't have that, I create it by selectively lapping a small hollow in the middle of the part to get a stable contact with the lap.

 

[memphisjed]

How I achieved Ra 0.00054 microns (0.02 microinch)

Hi there! I've made a video about how I achieved a surface roughness of 0.00054 microns - in my attempt to win the 2021 Zeiss #measuringhero award for the smoothest surface. Here's the link to the video: https://youtu.be/6YsMfFBQmJM

www.practicalmachinist.com

 

[luke8888]

You've got to get the amount of grit or paste right or you will have flatness issues. Too much will basically float the part. Better too little than too much.

I'll try a lot less.

I like grooved plates because they help to mitigate that problem, giving the excess somewhere to go. Also, watch your speed. Moving too fast will start to float the part too. I like to run drier, with less grit (if lapping with loose grit) and with slower movement when I'm really shooting for best flatness.

I felt this floating/hydroplaning when trying the figure of 8 movement . I don't normally do figure of 8 movement. I'm doing the movement shown in one of Tom Lipton's videos on lapping. This is a series of small circles while trying to cover entire surface of the lap and rotating the part /switching hands every short while.

You've also got to watch the geometry of the part at the very beginning. If you have a convex surface to begin with, it can b very difficult to get it to flatten out. I like to start with a flat to slightly concave surface. If I don't have that, I create it by selectively lapping a small hollow in the middle of the part to get a stable contact with the lap.

The second time I reground the blocks much more carefully. I've switched off the mag chuck for the last pass hoping it would "unspring" the parts, but I didn't dare clear residual magnetism (my mag chuck has a switch that gives it some AC which demagnetises it almost completely). I was afraid coolant pressure will lift the parts off the chuck and they'll crash into the wheel... So it is entirely possible the parts are very slightly convex when leaving the grinder. I'm not sure if residual magnetism is enough to spring a 9mm thick, about an inch long and 3/4ths wide block of hardened o1 by a micron or two. I know I can't pull the parts off the chuck with no power unless I press the clear button.

So I'll have to measure the parts better before lapping. It is not easy to measure fractions of a micron using contact methods and at that time the parts are not reflective enough for optical flat method.

This time I tried the hardened steel roller and pressing the diamonds in as well as stripping them from the lap afterwards (with a single edge razor-like blade, and a flat stone) just to see the difference.

However, the second attempt ended very similar to the first(slightly better):

I've yet to implement your advice.

How I achieved Ra 0.00054 microns (0.02 microinch)

Hi there! I've made a video about how I achieved a surface roughness of 0.00054 microns - in my attempt to win the 2021 Zeiss #measuringhero award for the smoothest surface. Here's the link to the video: https://youtu.be/6YsMfFBQmJM

www.practicalmachinist.com

This is really impressive :-)

 

[sfriedberg]

It may be worth tracking down your local Durabar or Versabar distributor. You can get either grey cast or ductile iron. Most of these places are in the business of cutting to length, so you could get a 1" slice from a 6" round or square without too much trouble or cost.

 

[luke8888]

It may be worth tracking down your local Durabar or Versabar distributor. You can get either grey cast or ductile iron. Most of these places are in the business of cutting to length, so you could get a 1" slice from a 6" round or square without too much trouble or cost.

Here is my part of Europe they want craaaazy money for cast iron (100eur for a disc like that). Especially ductile cast iron like durabar. Considering mild steel is not that much worse to lap hardened steel (but less resistant to dings as I found out...) I'll be staying with steel or if I need a softer lap I'll be using copper epoxied onto steel or cast lead (for cylindrical stuff).

I'm not entirely sure why that is (ridiculous ductile cast iron prices), but I suspect it may have something to do with the fact it is not used for anything outside very precise metrology lab/scientific apparatus around here for it's stability and even temperature growth/shrinkage. If anyone knows about any kind of small business(in Europe) that uses ductile cast iron in large diameter rounds and has disc offcuts to sell on the cheap please let me know. Why small business though? Because I found large companies don't sell their offcuts. They send them to their recyclers.

 

[memphisjed]

Here is my part of Europe they want craaaazy money for cast iron (100eur for a disc like that). Especially ductile cast iron like durabar. Considering mild steel is not that much worse to lap hardened steel (but less resistant to dings as I found out...) I'll be staying with steel or if I need a softer lap I'll be using copper epoxied onto steel or cast lead (for cylindrical stuff).

I'm not entirely sure why that is (ridiculous ductile cast iron prices), but I suspect it may have something to do with the fact it is not used for anything outside very precise metrology lab/scientific apparatus around here for it's stability and even temperature growth/shrinkage. If anyone knows about any kind of small business(in Europe) that uses ductile cast iron in large diameter rounds and has disc offcuts to sell on the cheap please let me know. Why small business though? Because I found large companies don't sell their offcuts. They send them to their recyclers.

Dumb bells. Like weights to lift.

 

[luke8888]

Dumb bells. Like weights to lift.

These are not from ductile cast iron. The alloy can be anything. You will not know until you buy and you crack it open. If you're lucky and it's not the worst possible form of cast iron with all sorts of grains included "for free" then you don't know when was it cast. Cast iron for precision applications should settle a minimum of a year after it was cast. The process can be accelerated by repeated heating and cooling a little. But not fully. There is a reason why big manufacturers like haas keep their castings for at least 6 months in the yard.

In theory I could source dumbbells that the seller assures me are cast iron(then send them back after figuring out they aren't, rinse and repeat rew times until I do get it) , then I could store them a year just in case. But why? What is so much better about a random Chinese cast iron alloy with a hole in a middle, that may also surprise you with a grain of slag in the worst possible spot as a lap for hardened steel as compared with a known grade of hot rolled mild steel? (genuine question)

Also I'm really just dabbling into this trying to improve my technique. It's going to be a long time before the fact these are steel is going to hold me back.

 

[memphisjed]

Bake at 390 overnight. Aging for longer than 60 days does nothing. Ductile iron is not as stable as grey cast.

https://www1.eere.energy.gov/manufacturing/resources/metalcasting/pdfs/age_strengthening.pdf#:~:text=It%20has%20been%20proven%20that%20aging%20gray%20cast,the%20strength%20or%20the%20yield%20of%20their%20castings.

 

[sfriedberg]

Even with annealing, I still would not try to make anything but ballast from any thing that was cast for use as a dumbbell, exercise weight, sash weight, or counter weight. Squarely in "mystery metal" territory.

 

[luke8888]

It seems I was lapping it too much... I tried a different method. I ground it to a better finish and after grinding I had a slope of 1~1.5 micron rather concavity. Then after just few strokes on 6 micron diamond (different method, more about it below) it is flat to one band, so 260nm with my light (10 micro inches).



Still I think it can be better with better technique.

This time I tried lapping almost dry. After embedding the diamond (with a roller) i used ethanol as lubricant, but wiped most of it leaving just a little in the serrations. Also I tried putting my finger on the center of the part and moving it slowly. As well as turning it a lot less. This did work better. I think the rotation was causing my concavity before (in addition to too much liquid and more).

Also I tried figure 8s. For those of you that do flat lapping by hand, can you describe your technique in detail, please? The only video on YouTube that shows lapping by hand to flatness and measures after is Tom Liptons. He has his own way, I'm interested in knowing how other people do flat lapping.

Do you do figure 8,do you do circles, straight strokes, how often do you rotate the part? Do you put pressure on the centre of the part? Etc.

 

[eKretz]

Speaking solely for myself, when hand lapping I generally do straight lines or small circles while moving in a straight line. Dependent on how much lapping is needed. Pressure always as close to the center of the part as possible. Sometimes I stick a small piece of double sided tape on the part to keep my finger in the right spot and give a little more grip. Occasionally I will attach a weight to the part to help keep the pressure even.

 

[small.planes]

Brake disks are a relatively cheap source of known alloy cast iron.
This might also be of interest:

Small Surface Plate - hemingwaykits.com

Unique projects for the small workshop owner. Please browse our Toolroom and Engine Bay for our latest project kits.

www.hemingwaykits.com

Small cast iron surface plate castings, relatively cheap.
Used surface plates are also generally not too much on eBay if you have the inclination to clean off the rust and flatten them out

 

[jim rozen]

Sash weights are chilled cast iron. The outsides can be harder than the proverbial billy goat's dick.

 

[Conrad Hoffman]

We lap precision parts on a 6" disk of Durabar about an inch thick. Diamond paste is rolled in with a dowel pin and aluminum bar. Way better than diamond lapping film in terms of flatness- we also use a lot of that.

 

[Nobby]

You know cast iron lapping plates (for lapping machines) are hardened? Mehanite is it called if I remember correctly. It is significantly harder than untreated cast iron.

I was reading this thread... and was amazed to see my own video about Ra 0.00054 micron roughness here. I feel flattered!

Lapping is easy... when you know what to do. It's like learning to play a musical instrument. There just aren't many corners to cut. It takes ages.

The picture of your last attempt is impressive: half a lightband! That is 150 nanometers for the non-lapping-humans.

Lapping machines have conditioning rings to keep the lapping plate flat during lapping. With handlapping you don't have that. When you're lapping, you remove material from the part AND the plate. Removal of the plate goes faster when it doesn't have the mehanite hardness. So you might get unflat areas in your plate if you lap in one place.

For lapping I use diamond emulsion - no suspension or paste. That usually is on epoxy lapping plates. For the mehanite plates I use lapping oil with ceramic powder. The oil is rather thin.

The best flatness is achieved by proper lapping. I'd skip the abrisive film. :-)

 

[luke8888]

Speaking solely for myself, when hand lapping I generally do straight lines or small circles while moving in a straight line. Dependent on how much lapping is needed. Pressure always as close to the center of the part as possible. Sometimes I stick a small piece of double sided tape on the part to keep my finger in the right spot and give a little more grip. Occasionally I will attach a weight to the part to help keep the pressure even.

Thank you, this is very helpful. I did see an improvement when I started putting my finger on the center of the part. I can't decide yet if figure 8, straight lines or small circles are better for me, I'm trying all of them at different times. However, I'd prefer not to wear my laps in the middle so figure 8s are probably the least favorable.

About cast iron (multiple posts mention it, so I haven't quoted a specific one). Can anyone say what are actual advantages on using cast iron in manual lapping plates for hardened steel? I have read the same old text ls where they say: "steel works fine, cast iron is best", but they don't actually say why... With my limited experience I can only think (other than more even/predictable bahavior during changing temperatures) about cast iron being better because it's harder to "ding" it. Having a number of them I've already dinged my mild steel laps a few times. Every time I had to stone the dings with the precision ground flat stone, then lap it a little to ensure it doesn't protrude even a little bit. It can really annoy you, having to do that when you're trying to get a part done.

Anyway, I've found a supplier locally that will sell me grey cast iron disks at a price of about 7eur per kilo(that's tool steel prices around here BTW). I might order a few 4in ones in future just to see for myself the difference.

We lap precision parts on a 6" disk of Durabar about an inch thick. Diamond paste is rolled in with a dowel pin and aluminum bar. Way better than diamond lapping film in terms of flatness- we also use a lot of that.

I like that method (rolling in the abrasive into the lap with a roller). I use a cheap ball bearing (roughly 1.5'' OD, half inch wide) with a piece of stainless tube for a handle. It is important to clean all the loose abrasive afterwards to only leave stuff embedded in the lap. This works not just with diamond, but with silicon carbide paste (probably alox too). It can be used dry for a bit too.

I like this method, because it makes the plate wear a lot less and the type of finish is different. (as mentioned before free abrasive vs embedded). Free abrasive results most often in matt finish, embedded is shiny, mirror like.

You know cast iron lapping plates (for lapping machines) are hardened? Mehanite is it called if I remember correctly. It is significantly harder than untreated cast iron.

I didn't know. It is interesting. I might make my own little bench top lapping machine similar to what Stahli makes.

I was reading this thread... and was amazed to see my own video about Ra 0.00054 micron roughness here. I feel flattered!

Lapping is easy... when you know what to do. It's like learning to play a musical instrument. There just aren't many corners to cut. It takes ages.

The picture of your last attempt is impressive: half a lightband! That is 150 nanometers for the non-lapping-humans.

Thank you :-) I've since managed to get it a tiny bit better. It is still near a quarter light band, but rather than a dome-like shape it is flatter in the middle which mattered to me.

Lapping machines have conditioning rings to keep the lapping plate flat during lapping. With handlapping you don't have that. When you're lapping, you remove material from the part AND the plate. Removal of the plate goes faster when it doesn't have the mehanite hardness. So you might get unflat areas in your plate if you lap in one place.

That's on my mind all the time when lapping :-) but with the abrasive embedded in the lap this is less of a problem.

For lapping I use diamond emulsion - no suspension or paste. That usually is on epoxy lapping plates. For the mehanite plates I use lapping oil with ceramic powder. The oil is rather thin.

I would be very interested in seeing some sort of a chart (if it exists) what oil viscosity with which abrasive grain size. However this information is very hard to come by. Most lapping oil manufacturers don't even publicise their viscosity numbers (maybe they tell their customers, but not the public). One that does listed one if their oils as iso 4. This is very thin indeed.

The best flatness is achieved by proper lapping. I'd skip the abrisive film. :-)

Definitely :-)

But there is one thing that helped me tremendously. In one of unrelated Robin Renzetti's videos(about sandpaper "lapping" on an old surface plate) he mentioned "you can't get flat parts on concave lapping plate, but you can on convex plate so it's better to wear your plates convex". Two of my laps were very slightly concave. When I changed them to slightly convex (about half a micron) I got much better results. It is also possible I've had more practice by then which contributed, but anyone making their own laps I'd definitely say to have your plates slightly convex rather than concave.

 

[Conrad Hoffman]

I want to re-emphasize that when you use the rolled-in method of charging a cast iron plate with diamond, you clean off all loose abrasive and paste. Wear on the plate is minimal. We've been using the same plate every day for several years and I haven't had to regrind or surface it yet. It's important to brush/wipe the plate frequently to prevent clogging and crud build-up. Pretty sure we use 9 micron diamond, but I'd have to go look to be sure.

 

[Nobby]

Your last picture really is a beauty. Really shows your true lapping AND cleaning skills.

You said you were disappointed by the roughness? Classical machine lapping is on a (hardened) cast iron plate with ceramic powder (suspended in oil). This gives a matte finish. For inspecting the flatness, parts were polished lightly on sand paper - 2500 grain or so. This sand paper is pretensioned - on a flat table (lapped as well). Commonly 12" square.

Lapping on epoxy plates (with metal particles mixed in the resin) used with diamond emulsion gives a shiny finish. Even 14 microns on hardened steel will give a shiny finish. Maybe even 25 microns does that.

There are hand lapping plates of epoxy on the market too. For getting these flat, I would use machine lapping on a cast iron plate.

My video on lapping (in Dutch) show lapping of hardened parts on a epoxy plate. video about flat lapping (in Dutch - but you'll get the point)

 

[luke8888]

I want to re-emphasize that when you use the rolled-in method of charging a cast iron plate with diamond, you clean off all loose abrasive and paste. Wear on the plate is minimal. We've been using the same plate every day for several years and I haven't had to regrind or surface it yet. It's important to brush/wipe the plate frequently to prevent clogging and crud build-up. Pretty sure we use 9 micron diamond, but I'd have to go look to be sure.

BTW, what lubricant do you use on your lapping plate? Finding the right lubricant (and the amount) can make or brake it IMO. For example if I use the lightest oil I have(iso 10 or kerosene) it is already way too viscous on a 6 micron diamond charged lap. I found a couple of methods that work when I was lapping the smooth sides of my laps on the serrated ones.

One is to use a tiny bit of deodorised petrol(straight gasoline is the same, but much less pleasant to use). Spread it over the part and lap as it begins to evaporate. When resistance gets high stop, clean, reapply this solvent. This works, but one goes through lots of tissues to clean, also gasoline (even deodorised) is not the safest thing.

Second is wd40, but the way to apply it is to put a tiny drop on the part, distribute it with my finger over the surface, then wipe 90% of it off with the edge of my (nitrile gloved) hand. This cuts very fast. You can make only few strokes before it needs cleaning (few more than with gas).

Third, it's to put few drops of my iso 10 oil in 10ml of gasoline, apply few drops of this to the part, spread it around with a tissue, wait few seconds for gas to evaporate. A tiny bit of oil remains. This too speeds up cutting a lot, but loads up very soon.

I also tried deionised water (straight as well with some monoethylene glycol), 95% ethyl alcohol (similar to gas, but less nasty), tetrahydrofuran (I got an idea this might work when I saw it has very low viscosity and relatively high boiling point - no, it was a bad idea, it evaporates very fast, it stinks and it's unhealthy...)

I'm yet to try "polyalkylene glycol monobutyl ether". I saw it as an ingredient in some diamond slurries. It is a synthetic water soluble lubricant. If I can get my hands on some I'd like to try it. I'm always on the lookout for healthier oil/solvent alternatives.

It's good to know for sure what I suspected (that the wear is miniscule if you remove all free abrasive). After pressing the abrasive in I wash the lap very well (while brushing). Perhaps some grains are lost in the process, but it is well worth it. Also during use I wipe the lap fairly frequently with a solvent soaked tissues. It is possible to use a lap like this dry too (briefly) . Probably that's why they call it "a dry method" in some old texts.

Your last picture really is a beauty. Really shows your true lapping AND cleaning skills.

Thank you :-)

You said you were disappointed by the roughness? Classical machine lapping is on a (hardened) cast iron plate with ceramic powder (suspended in oil). This gives a matte finish. For inspecting the flatness, parts were polished lightly on sand paper - 2500 grain or so. This sand paper is pretensioned - on a flat table (lapped as well). Commonly 12" square.

I was initially, before I learned the above. It is very interesting to read how parts are made reflective with 2500 sandpaper. I would've never guessed it, as I would expect it to round corners etc, but being 2500 grit and used with correct technique is probably the key to using it successfully.

Later I found a method using precision ground flat stones to make my laps reflective, but not everyone has these. Sandpaper on a good flat surface sounds like a good alternative. Robin Renzetti's video titled "magic sanding plate" has some tips on sandpaper use while maintaining flatness.

Lapping on epoxy plates (with metal particles mixed in the resin) used with diamond emulsion gives a shiny finish. Even 14 microns on hardened steel will give a shiny finish. Maybe even 25 microns does that.

There are hand lapping plates of epoxy on the market too. For getting these flat, I would use machine lapping on a cast iron plate.

Interesting, it reminds me of the pitch laps I read in context of optical components lapping.

My video on lapping (in Dutch) show lapping of hardened parts on a epoxy plate. video about flat lapping (in Dutch - but you'll get the point)

Good video, thank you. Auto-translated captions work very well.

For people reading this interested in lapping as applied to any material (optics etc) I highly recommend the following youtube channels too:

https://youtube.com/@HuygensOptics

https://youtube.com/@Opticsed