Hand Scraping a Surface Plate to Sub 5 Micron / A - Grade Flatness
Hello and welcome,
in this video I'm showing you how I handscraped my steel surface plate to a flatness of less than 5 microns (0,2 thou).
It ended up beeing somewhere inbetween 3 and 4 microns of flatness.
That's 0,003-0,004mm. So not that much.
All scraping was done by hand.
Roughing with a 30mm/25mm carbide blade with 140mm radius.
Finer scraping was done with a 25mm carbide blade and 90mm radius, final scraping was done with 60mm radius.
For scraping basics check out @StefanGotteswinter 's channel, he has awesome basic videos about scraping. My ghetto carbide grinder is also inspired by him.
• Scraping basics - Scra...
MUSIC:
The Gentlemen - DivKid
Snowy Peaks pt I - Chris Haugen
Snowy Peaks pt II - Chris Haugen
Pink Horizon - Chris Haugen
Soft Feeling - Cheel
Sunday Rain - Cheel
Sunset Dream - Cheel
Blue Dream - Cheel
Lazy Walk - Cheel
Smokey Eye - Cheel
Campfire Song - Chris Haugen
Blue Creek Trail - Dan Lebowitz
00:00-02:55 Intro
02:56-20:50 Rough scraping
20:51-22:37 Carbide grinding
22:38-30:12 Fine scraping
30:13-34:45 Measurement method
34:46-40:55 Metrology theory
40:56-44:45 Measurements and results
44:46-47:02 Bonus
"...I don't have the money to do that and I don't see why." I understand what you meant in context, but I like that phrase out of context a lot more! haha
That's my golden rule as well when I somebody asks me to get a fancy car, but as you said, it may be way more universal than i actually thought :D
Just love the visual image of a micron-reading mechanical gauge used in conjunction with a wooden rule to establish the data for drawing the graphic. Good work.
Very nice job. Plate looks Beautiful. Thanks for sharing. Take care, EM.
Just to be clear, .015mm = .000591"
Very cool project and your determination paid off. It looks amazing...thanks for giving me the bug!!
Fascinating. Love that you did this! Thanks for sharing.
Thanks for watching! :)
Amazing amount of work.
Thx! :)
Nicvto watch, thank for sharing
Really enjoyed every second of your work. Thank you for sharing your experience in creating something that is so mesmerising to watch. Great work!
Thank you so much!
I wasn't feeling well so I re-watched you video, and it's still a beautiful job, and end product.
Thank you so much!
Well done. The air hockey effect at the end was so satisfying. 🎉
Thank you so much! :)
Großartig! Tolle Arbeit! Und großartig präsentiert. Glückwunsch zum Ergebnis! Hat sich wirklich gelohnt.
Hallo und vielen Dank für die lieben Worte! :)
excelent work and explication .
Good job thomas
Hey great job and I like the less amount of ink it makes the high very visible.
Nice work!
Thx! :)
Just found your channel and subscribed. Very nice work. Interesting video
I have never liked Prussian blue for printing, it's very hard to read. Instead, use copper sulfate solution. The approximate formula is 150 ml water, 2 g copper sulfate, and 1 g of weak hydrochloric acid (sulfuric also works). Wipe the solution on clean dry metal plate with a clean paper towel, it will instantly turn lovely dark copper color. When rubbed on the surface plate, the high places are shiny and very easy to see. On the next printing, just wipe solution on the bare metal you have scraped. No worries about your surface plate, It will not damage it in any way.
The heat treating process you're describing at the beginning sounds alot like flame straightening, for anyone interested in the technical term.
Lovely work. Plenty good enough for 99.9% of any work on a steel surface plate.
Looks llike superman scrapin 🙂
Great video, thank you! The half-moon could be where the slab of steel got hit hard with a hammer of other tool late in its manufacturing process. I made my lapped plate using an ultra-precision spirit level instead of a Mahr gage. It came out at about 50 microns, I only needed about 130 microns flatness (for woodworking). It surprised me how accurate the spirit level was, I could have achieved probably 10 microns pretty easily.
It's beautiful
Thanks!
That hard imperfection you describe at the start of the video, might be not fully melted pieces of bearing races/drill bits/reamers/taps/files/etc. that were used as scrap steel raw material to make that plate (I encountered this in "cost-effective" Taiwanese cast iron, as example, that I had to scrape as a machine builder).
Yeah, that might be possible! Good comment! Thanks!
by distinguishing the country where it is made, you could now step up on some politicly sensitive toes, Most of the "cost-effective crap" is coming nowadays from China ( no to confuse that with Taiwan, that would be the same as you say the Mose meat from the USA is crap.... but it came from Canada ) but if they build shit In Taiwan they do deserve the notification....
@@freddykoopmans6985 it arrived from Taiwan, as far as I know, but I have no clue where it was actually made. And I'm not really blaming them for it - who can sort through tons of chips and chunks of metal looking for files and ball bearings before dropping into smelter?
Excellent demonstration of the process! I’m puzzled by the scrapping technique as you’ve shown many times that areas WITH ink are high spots! In my experience with auto body, spots where I had left guide (ink) were supposed to be the low spots. Could you go over that part to explain why scrapping is the reverse? Thank you very much!
@@roadshowautosports It is reversed because the marking medium is put on in reverse. In scraping, the marking medium is put on the perfectly flat reference surface. When the work piece is put on, only the high points touch and transfer the ink. In auto body work, you put the guide coat on the work piece that is not flat. Therefore the guide coat flows into the low spots. Your sanding block is the "perfectly flat" reference and only sands the high points, leaving your guide visible in the low spots.
It is interesting how were made first machine tools when there were no machine tools
Just to add that the youtube channel "an engineers findings" also explored scraping steel and came up with some conclusions on scraping angles that worked (at least for him)
Thanks! Yeah found that after publishing the video! :D
Impressive work!!
Cheers mate, thank you! :)
I would "imagine" to mount the measuring arm on a table in a static vertical position which can reliably rotate 180*. Then you place your plate next to it and make sure the all the 4 corners of it are at 0 height. You can use some bolts to level the plate. But the tricky part is how to bend the measuring arm reliably so that micrometer stays at the same height. I'm sure there is some mechanical trickery you can do but I just don't know about it. Then, you can rotate and bend the arm left and right, up and down, to find the low/high spots and the overall plate bend/radius or even skew/twist. P.S. I don't know anything about this stuff. Just interested to watch (and mby try).
Ich habe nicht alles angesehen, aber das Ende mit gleitenden Metallklotz war verblüffend!
I have experience of scraping a straight edge made of mild steel. If the total flatness of the part is more than 0.02mm it takes days and weeks to make it flat. First of all we must make it as flat as possible with surface grinder and then start scraping.
Really, weeks? 8 hours of scraping every day? I've heard using a surface grinder will make it too "slippery" to get it going. Some prefer even a milled surface
@@valkman761 he wasn't suggesting the grind be the final surface, just to save time when scraping. though I kind of disagree on his time table, that table just isn't that big, 0.2mm is like 8 thou, right? its a lot, but its a few hours, less with a power scraper. I only mention this because a lot of people dont have access to a surface grinder, or the part it too big--in those cases, you can scrape 10's of thous pretty quick...plus its about the only way. Yes, smooth surfaces do tend to be prone to stick slip, but on a surface plate, you really want a smooth (which can be scraped) gauge quality finish. If you got a plate like this ground & it was done with acceptable geometry, you could just do more pattern flaking to make it slide smoother.
If mild steel, take an angle grinder to speed up the process.
I believe the important thing with local flatness is not having the feet of test instruments falling into holes- so test with a 10mm / 3 /8 foot is the standard way to go
The more standard way is to use an instrument holder with a flat lapped base or use a ``repeat-o-meter``... Also, using of gauge blocks is the norm with some measuring methods, if scraped surfaces are in question... The feet are more for general purpose high precision measuring, but for stuff such as this, lapped, instrument grade surfaces akin to gauge block ones are used... The instrument base may be machined as to provide a reduced surface, as in faking feet of larger size, but the end result is lapped... Either way, for madly flat surfaces, you dont even want feet, a large flat lapped surface is preferable, as it will indicate any shift due to any positive extreme anomaly however small on a larger area... Which is what large repeat-o-meters are practically designed for, just like the auto-collimators, but in a different way of doing it, its a bad comparison, but a somewhat adequate one...
Mild steel if it has been cold milled(rolled), you must anneal it first, otherwise will be milling stress left on both sides of surface. This stress trend to curl your surface as you try to wear only one side down.
Yeah, that's pretty much true for all materials. Also for cast iron. The cooldown process is the most critical thing. Warping in my case is not really a problem, because the plate was precision gound on both sides (ground without magnet-clamping on first side). There might be still some stress in the plate, but as long as you don't heat it up locally or release the stress by removing a significant amount of material (by milling for example), it will not affect the plate. This is one of the biggest advantages of scraping: the amount of material removed is very small and almost no stress is induced into the workpiece (at least compared with all other proccesses).
With a liquid level or autocollimator u can check well the flatness..indicator sometime is no the best for do it
Hi there, Yes that's correct. Check out 37:25 where I have also talked about that. Unfortunately I don't have acess to an autocollimator and can't use a liquid level, as the surface plate is not on a steady foundation ;) For my "everyday workshop plate" the indicator method is accurate enough for me.
Nice work , what cutting angle have you use on your carbide tip ? same as regular iron tip ? as steel is quite difficult to scrap .
Thanks! I used a +5 degree angle, which as I found out later is not ideal. You actually would aim for something like a -5 degree angle, as I found out from some very old literature. Nevermind, for me it still felt okay to be honest. I also talked to an oldschool scraping guy, and he used a 0 degree angle for steel, which seemed perfect for him personally. So yeah, -5 degree by the book, but you can get away with other angles as well :)
It is interesting how were made rotary parts like spindles or shafts of the first machine tools in the world. I've been always thinking about it. My question is: is it possible to get a cylindrical part using this method?
Hey there! Good question! Yes, you can also scrape round parts, for example bearings and spindles. The first machines were made from wood, with a primitive spindle that was made by hand, carving a round part out of wood. When steel was invented, they filed and scraped round parts by hand. They again were used for the first lathes. At a certain point it was possible to make more precise parts with the machine, than by hand. From this point on they used machines, and hand work together to build more and more precise machines, that again could build more precise machines. But all precision comes from a flat reference: the surface plate. You can use a surface plate as reference to build everythin you want. And you can make a surface plate by scraping it, using the whitworth three plate method.
@@buildfromsketch8102 it is interesting what methods did they use to make them
Plate is far too thin for serious long term stability. Also get a good electronic level to truely measure flatness, if you're hardcore into it. Used my surface plate for scraping jobs for a few years now and considered it good, until I used a Talyvel check it. It has 50µm low spot on one side, you wouldn't notice that on smaller scraping jobs, other than slightly inconsistent printing. Now, I'm getting some silicon carbide powder to lap it back into shape ;-D
There is a way of measuring two (any) surfaces to the precision of the dial gage, it is called reversal method. Can be done with 2 dimensional surfaces but works best with narrow beams: Put two beams side by side and measure one agains another at equally spaced points. Flip one and measure it from the underneath at the same points. Take the slope out of the two series of measurements. Essentially make the points the deviation from the slope. First beam is the average of the two measurement series. Second is the half the diference of the two measurement series. There is a video on youtube by Dan Gelbart: (Reversal method at the end) kzhead.info/sun/ltudn7misKCmeKM/bejne.html To avoid flipping one beam, 3 beams and a fixture to rest the the dial on the outside ones and measure the one in the center. After the 3 permutations of the center beam there is enough data to determine the 3 of them. Hope I was clear enough.
Hey there, thanks for your comment! Yes, you are right. The way Dan Gelbart is showing in the video, is a great way of doing this. Unfortunately I don't have a surface plate that is big enough to cross-check my hand scraped one. But I will revisit this surface plate in the future and do some more measurements on it! :)
Qual aço e esta base ou é ferro fundido ???
I've got a Mahr Compramess Imperial indicator (0.0001") but never used it as I rarely use inch measurements to that scale (at least not in last 12¬13 years) I haven't worked to micron tolerances for more than 40 years either but I guess learning something new could be fun as I have never considered making a surface plate. Isn't it true that all 'flat' surfaces, even optical flats are curved? Of course, the curve is same as radius of earth so pretty flat to all intents and purposes Still a very interesting video though, I think your being conservative at 30 hours, probably more . The air bearing at the end was cool bonus
Hey there, thanks for your comment. Well that's an interesting thought about the radius of the earth. But actually, that's only true if you use a precision level to check the flatness. I have an upcoming video and I will address this topic there. It's impossible to create a "perfect" flat surface. That only possible in imagination. But not every surface will have earths radius. Yeah, you are probably right, 30hrs might be too conservative. And yeah, I will also show more air-bearing stuff in the future! :D
True flat surfaces are NOT curved at the diameter of the earth. They are compared to themselves with absolutely no regard to where the earth is. Even level on the earth is typically done using lasers and as we all know laser light does not bend around the surface of the earth, it just passes through space on exactly the same plane. Unless of course there is a black hole in your back yard that could deflect the light particles.
I'm looking to rescrape my freshly acquired cast iron surface plate. It has big dings and marks so it can be off quite a lot as of now.
Hi there, that sounds like a lot of work. You may want to consider having it reground before. Anyways, as long you are not aiming to replicate any NIST standards in you home garage, some lowspots are absolutetly ok. The plate I scraped here isn't perfect either. As long as you lnow your plate, that's totally fine for some work you do at home :)
I have the exact same issue, I bought a pair of cast iron surface plates and a camelback at an auction from a former technical college. The camelback was covered in a moderate layer of rust, clearly not having been used in a long while and not having been stored covered in oil, and those surface plates were especially dinged up. We have a lot of projects planned where we need precision references and thus we now have a couple side projects to acquire ultra precise (
@@pieterveenders9793 how about visiting someone who has a good big plate to do the job
@@pieterveenders9793 , Hey there. The issue is that the Whitworth three plate method requires three plates, as the name already indicates. It's also required to have the ability to rotate these plate by 90° in order for the plate to work. Robin Renzetti has a good video, explaining the problem using some kitchen sponges. Nevermind, if that's what you are aiming for, things will require a full regrind first, then copying everything off a known good plate, then Whitworth-scraping it using a third plate. Finally a lapping step may be required to achieve 5µm/1000mm.
You may provide some lifting device to lift the part and not to slide the part on the surface plate. Otherwise you may get false reading.
Yeah thats correct, I later lifted the plate straight up to prevent the false reading. Unfortunately I don't have a crane or the space for one.
gr8 work here, you should continue on this plate with lapping, you will get sub micron if u did so, use a precision machinist level it can sense 1um curvatures but u will need very solid base or measure on bare ground, the dial measure you have used is not really for flatness it is for uniformity, imagine u r using the same setup on a perfect sphere you will still read zeros, also i never liked the neutral/positive rake angles on the blade on that video, it does a lot of biting, even on the video it shows that, not sure why the guy and everybody was happy with that, btw about the diamond wheels, how you control the run out on them? do they come with a convex curve so when you bolt them to the grinder they become flat? how that is addressed?
Hello there, thank you very much! You do have some very good questions. You are right. I did talk about the measurement theory in the video as well: 34:49 The blade angle is something that somebody already suggeested to change to 0° or +5°. I will try this the next time when scraping steel. The runout of the diamond disc is not controlled at all. It's just a cheap diamond wheel off ebay- mounted to a wooden plate. The runout is not critical.
@@buildfromsketch8102 got you, will wait for your experiment with a 0 or +5 rake coz it really did not work with me at all, thanks for your reply, keep up the good work!
@@GraphicManInnovations Thank you for your comment, your questions and your nice words! I will do a follow up video about this in some time!
Beautiful and exhausting job. Congratulations, but be careful with your shoulders because that movements may cause LER (repetitive strain injuries)
some wise guy said: if want something done well you have a few choices: 1# pay somebody to do it, #2 do it yourself, or #3 tell your kids not to do it. For nr 3 i do not trust my kids to get their yogurt in an 8" bowl so forget about something about microns... for pay someone to do it for you without you can check their work......... that is only a good plan if you have money to spend and you know what they do and you can check there a process. heads off his efforts and ill think he did a pretty good job
Thank you sir, highly appreciated!
Great video. What is the radius on the carbide blade?
Thanks! 140, 90, 60 and 40mm radius :)
Hello what type of paint did you use? Thanks.
Hey, "DIAMANT Tuschierpaste" :)
Da Akzentlos, dachte ich du wärst in den USA. Dann hatte ich die Garant Platte gemerk :) Allerdings, schön gemacht!
Danke! :)
I recommend spray on DyKem or similar. Trying to mark up with a permanent marker is just a pain.
amazing... only thing that got my autism going was... were you placing your plate onto the granite surface on the same side and in the same direction every time when applying the ink and does that matter?
Hey there, thanks! Yes, you have a point there. Ideally you would rotate the plate 180° every time, which for me was not practical. My master granite plate during that time was brand new, so crazy flat and no wear in it at all. So in my case didn't really matter, but yes in general it does matter.
What steel you use?Aisi 1045 ?
The steel is unkown, I guess that it's S355. This would be equivalent to A618 steel. Anyway, cast iron would be the material of choice, unfortunately I didn't have that on hand, and I got the steel plate for free. Scraping steel just takes much more effort and is way harder to do. This is good if you want to learn, but bad if you want to be done quickly ;)
@@buildfromsketch8102 i scraping many time Aisi 1044/steel C45,but is no too bad to scraping.becose have carbonium inside.normal steel is too bad,and for scraping that sometime i use HSS,not widia
I looked up the scraper and it's only 300mm long. Isn't that too stiff ?
Well, it is stiff indeed. It worked out for me anyway. Here in Germany I actually have never really seen the more flexible scrapers. But I know they exist. I would like to test them out in the future :)
@@buildfromsketch8102 Thanks for the reply! What's the width of you scraper ?
@@martinnovak3949 It's 25mm. Altough I used a 30mm carbide blade in the beginning (it's sticking out quite a bit). 25mm seems to be a decent allround size, although it's too big for smaller things.
How is the granite plate made flat ?
The granite plate is precision ground on a surface grinder and later precision lapped against a lappingplate or another granite plate using a diamond lapping compound.
Hello there, You are probably now aware that steel is hard and not so nice to scrape as cast iron. The blade angle needs to less, ie . 0 or possibly a few degrees positive. This has a lot to do with how the results become.
Heihei, Yes, scraping steel is no fun at all... I have heard about this, but in all the books I found, I couldn't find anything about a bladeangle for scraping steel. I will try this next time, hopefully it will help :) Thank you very much for the tipp!! :)
@@buildfromsketch8102 Yes, I would try to practise on a steel surface first, as it is not so easy to get a nice surface. There is a description of blade angles for different material in the BIAX literature, as well as some helpful comments. If you have access to this material I would also recommend studying this :) PS! Most "scrapable" surface plates are made in cast iron and are ribbed on the underside plus have a 3 point support, usually studs that are cast-in. The design is therefore especially adapted for resisting any surface deviations from temperature and forces, which makes it better than a steel plate, so to avoid measurement errors I would ensure that this plate is "calibrated" before use, eg. by running the usual checks, eg. with a machinist level. A 3 point support is especially important if you want to measure something heavy as the geometry of the plate can shift substantially.. despite the thickness of the plate and apparant "robustness".
@@jansverrehaugjord9934 Yes I will check that out! So far I am already quite happy as far as precision and surface finish goes. There is room for improvements. I think it's still pretty good for the first "big" project. I will build a support frame for this, which allows a three point contact and expansion due to temperature changes. Anyway, the plate does not flex much under load, 28mm thickness is quite a lot. But for heavy parts I will use my granite plate of course. The steel plate is just a "everyday" use plate.
This could be helpful too, he explains and shows the problems and solutions for scraping carbon steel vs cast iron. kzhead.info/sun/ecuxms19eZqBqYU/bejne.html
@@oscarwilbers Thnks. PS! Biax has made a good brochure showing different materials, blade angles and geometry. I would also recommend this for basic info on scraping from a professional source. It is also worth mentioning that there are Biax scraping scraping classes in Europe also, run by the company Maschratur in Austria. Highly recommended for anyone interested in machine restoration. They are a professional machine shop with much experience and knowledge on scraping etc. can probably also explain the aspects of scraping different materials eg. steel.
Great video and explanation. If you wrote out a script to read it may help quite a bit with your pauses, stuttering, breathing issues and so forth.
Thanks! I'll be working on that! ;)
Then your brother comes along and uses it for a welding table.
*sigh* I'm not crying, you are crying!
But it finally it would be a reasonably precise welding table, you can't argue that :D
Video would be better with out music
Great video! thank you! Still relevant despite time. Russians are used to use your method like in this kzhead.info/sun/n6hsYcWSpmqpdY0/bejne.htmlsi=JH4MRhBy4NOJRt6S&t=951 especially when it comes to improve precision in lathe and drill working.