What is Scraping? A way to make a metal object REALLY flat! Richard King DVD: ebay.to/2kF3GG4 Filmed with amzn.to/2jh8oh0 ! Music copyrighted by John Saunders 5 Reasons to Use a Fixture Plate on Your CNC Machine: bit.ly/3sNA4uH
Strange subject but this video is so well done. Zero time wasted on anything. That alone makes this video great. Great to the point talking. And he talks you through the subject just as you would talk through it to yourself. It is a new and difficult subject to communicate and he did a fantastic job. Really awesome presentation this time.
John that was an excellent summary of scraping, well done. Some people confuse "flat" and "smooth", a properly scraped surface is flat, but not smooth. A polished surface is smooth but may not be flat. A well surface ground object may be relatively flat and relatively smooth. A properly lapped surface (eg a surface gauge) will be both flat and smooth. Each type of technique and the finish obtained has their place, and one isn't necessarily better than another for a particular application. For anybody thinking of getting in to this, try to take one of Richard's classes, failing that his video is an excellent introduction. He is also exceptionally generous with sharing information on how to improve for those who show a willingness to learn. Any of us who are doing what we love and earning a living from it are genuinely blessed in my opinion. Rich is clearly doing what he loves and it shows. Beware though, scraping IS an addictive process :)
And so begins the slide, never again will you look at anything and not question it's flatness. It's a great challenge, almost like a chess game, to take your new knowledge and apply it to qualifying a machine. Enjoy the torture, I personally love it! :)
Looking back 40 years +, we learned on the job. Everybody in rebuilding dept. was required to scrape. At that time- no power scrapers. Then about 20 years ago A change of jobs -power scrapers. Wow what a relief. Still a need for hand scraping but not as much. It does take time to learn and years to be good at it, especially scraping a whole machine for alignments.
bcbloc02 Imagine how the old timers thought of use when we first started. They probably waited until we got out of hearing range and laughed their butts off. 😊 Thanks, John
Brilliant! You are so much more than CNC! That's why I like Mr Crispin. So many people now have none of the hand skills! Using the test indicator stand is the best way too rather than mag base.
The fact that we have all these ridiculously accurate machines and we can still achieve such a high level of accuracy by touch, feel and technique is mind boggling. For instance, I work on guitars...I had one of my newer instruments plek'd (cnc machine that levels guitar frets) and found that it was still pretty buzzy when trying to get the action I was after. I levelled the frets with a reasonably straight piece of aluminium i beam (.001") and it played so much better than it did after going through the miracle machine. Goes to show there's still room for handwork after all.
Thanks!!!! It was like a light bulb going off! Reminds me of when going thru radar principles and finally understanding how a Hybrid-T was able to work.
Surface grinders do a pretty good job of keeping things flat. The reason your parallel was low in the middle is because the wheel of a surface grinder wants to grab and pull the metal. All part of the nature of abrasive machining. Super cool videos on this stuff, though!
Great video. Never thought I'd get so excited to learn about scraping! I just thought it was there to look good! (not really, but it does look awesome)
In the machine shop display at the Smithsonian Institute they show a guy standing on a plank and three guys cranking him in with a pulley . He is the scraper and is cutting with a metal wedge through the plank. That is how they kept the early machines flat and accurate.
In days long ago when dudes formally did operations like your talking about they called it FLOWERING and believe me it took a lot of time to complete large surfaces. A good flowered Bridgeport table was a work of art.
Do you have a video about proper blueing? I found that many people make a lot of mistakes, especially with smaller parts that have the tendency to roll over the reference plate instead of rubbing in parallel.
Is there some optical / laser mechanism you could use instead of repeated blueing? Mostly just wondering for the sake of wondering. I'm thinking something that throws a laser at a very very closely controlled height over the surface?
This makes me question my home-built CNC project that I am starting. I had thought to start with a 1" steel table top so that I could do some sort of scraping or grinding to get flat so that I can begin building linear bearings and axis upon that. But now I am am wondering if using a thicker gannet to sit upon might be more stable platform to build on top of?
You can use a two colour method. First you rub a red colour on the piece, then wipe it off so only in the valleys is red paint.. Then you go to the granite blueing to see the highs. Then you scrape until blue is gone and you hit red in the valleys.
In which way is metal scraping better then draw the piece on a completely flat surface (a little thicker sheet of ordinary limeglass is enough) with a sandpaper on that?
Lapping is the Best way to make a surface flat. But two Lapped Surfaces will have SO MUCH surface to Surface contact, that they won't slide. The Surface Adhesion will be so great, that you will need to apply so much force, that when the adhesion is overcome, you will likely Tear one, or both Surfaces and cause Gauling. Flaking is used to Create intentional valleys in one of the Surfaces to retain an anti-Adhesion lubricant. The Lubricant will help to prevent the amount of adhesion and lessen the force needed to overcome it. Also, One surface can be bonded with a coating that resists adhesion and increases the slip. Where the Surface is too large to fit on a lapping machine. Progressive surface scraping can be used to achieve a nominal overall Flatness. It will never be as Flat as a Lapped surface, but you can get progressively closer and closer, by progressively scraping away less and less with each progression. your Demonstration using the gauge block on top of the scrapped surface is a perfect example. The Block is precision lapped to be flat. it's flat surface makes contact with the crests of all of the High spots of the scraped surface. Within any given area, there should be 40% to 60% of evenly distributed high Spots. this will provide adequate contact area for the other Flat surface. Scraping always leaves surface imperfections, a distribution of Valleys and Crests. This is why precision Optics are Lapped and not scraped.
So how do they get those granite reference surfaces so flat? Everything I've seen involves tools of at least the same level of precision as you want to reach. How do you use tools that are good to +- 1 thou to get +- .5 thou?
The deviation on that straight edge might be even less than that, due to the fact that as you held the gauge block it expanded from the heat of your hand thus throwing off the measurement. Nice job on that casting by the way. Thanks for the videos.
NYC CNC We'll get you down to having to use a laser interferometer and optical flats and measure flatness within waves and fringes. 😆 Ps. 1 Fringe = 0.3 microns (.0003mm) . Have done 1/4 fringe spec optics 😁
gredangeo Microns aren't even remotely close to being fine enough for optics. Fringes and waves is measurement using light waves. Fringes are short for interference light fringes. When you pass a single light wave through and optical flat ontop of an object, the light will reflect within the tiny space between the optical flat and the object. Via reflection dark and light bands are generated due to the gap between the optical flat and the object you're measuring. Since the light used is a single wave length, divergence within the gap is predictable and measurable. The greater the distance within the gap, the more divergence is created. And interference lines will appear as dark and light bands. The more lines the more peaks and valleys, and the straighter the lines, the flatter the part. You can even see concave or convex sections within an object with it. Which will show up as arcing light and dark bands It's for measure really really super flat, or super higher accuracy dimensions. You can look up Optical Interference if you really want to know about it. It's been around since 1801. And still used today to measure super flat surfaces.
Occams Sawzall But you only said that 1 Fringe is about 1/3 of a Micron. Just 1/3. That is not a big jump to a higher level of precision. So instead of saying "30 Microns" you say it's "90 Fringes"? Not a big deal as far as I'm concerned. I say it's better off to stick with the more common units, and then when needed go to the Unit that is 1000 times finer than the Micron, and be done.
gredangeo It's how optics and super flat parts are measured. I didn't invent it. No one in optics measures surface deviations in microns. Fringes are a comparative measurement anyway. It's the amount of deviation from a know to unknown surface. OD, sagittia, FOV, center thickness, and annulus dimensions are in microns. When it comes to the physical surface it's fringes and waves.
This kinda starts the whole chicken/egg thing...like do you get something so flat, without having a reference to what is "flat" in the first place?? Like who made the first surface plate without a reference?
Nathan Hamler You actually don't need a reference surface to get something flat. If you take three surfaces, and reference those to each other in succession (1 to 2, 2 to 3, 3 to 1) until they are all three the same the result is perfectly flat. Said another way, the only way that three surfaces can match is if all three are flat. This is how the first reference surface can be produced.
You must have poor surface grinders if you can't get it flatter than scrapping. Scrapping is only used when you can't grind; kinda like refurbishing. It is NOWHERE near as accurate as surface grinders. If you want a superior flat surface to grinding, then you have to LAP. Lapping is what they do to Guage Blocks *AFTER* surface grinding. You will not find scrapped Guage Blocks for the reasons I've stated above.
Trabalhei neste ramo 50 anos me aposentei tenho alguns equipamentos guardados uma régua de aço fabricada na Alemanha com mais de cinqüenta anos deixei rasqueteando e depois fazendo lapidação com 0,001 mm na extensão de 1500 mm.
I'm pretty sure the way you checked flatness with the gauge block was slightly incorrect. Your method was actually the technique used for checking parallelism. I know you were time limited, but flatness requires more points in multiple directions.
It seems to me that scraping is just the old school (and old-school isn't less, in my definition of the word it refers to processes that tend to be at a more basic level (i.e. requiring less tools/technology to make it happen)....like, you can hack saw something by hand, or you can use a motorized hacksaw (or something else) and it all depends on things like; available tools and reosources, what you are wanting to accomplish, how fast, how cheap, how accurate, how...??...whatever....In a nutshell the question then might be, WHAT is "GOOD ENOUGH", "EASY ENOUGH", "FAST ENOUGH" for the project at hand?"
My bet is none. What many call hand scraped is merely hand flaked for some oil pockets on top of a ground surface. This is a 10 minute job for some chinese guy, not a 200 hour precision scraping effort.
[puts on metrology nerd hat] No lie, "It depends on the color and how long you wait." IIRC, blue Dykem is about half of a tenth (or fifty hundredths) when dry. The red stuff is a little thinner. But that's not the factor here: What you're doing is coloring the high spots. When the area you just scraped gets dyed again and has a contiguous, even and unbroken coating you're within that half of a tenth. The only concern is *relative* accuracy, not absolute accuracy. If you were needing it flatter than fifty hundredths, you'd be dialing it in using optical interferometry after the initial scraping with Dykem. But for the home gamer and small job shop machinist both, you probably won't be building for NASA, the ESA, LANL, LLNL, or CERN.
Probably missing something but when you guys say, one 10th, that's one ten thousandth of an inch, right? So wouldn't half a tenth be one twenty thousandths of an inch? Or would no one ever say twenty thousandths because of the position of the number in the decimal and confusing I guess because its actually a 5 and not a 1? What about 5 hundred thousandths? I guess I'm just curious why in machinist language you go right to millionths and skip a decimal term? .5 = 5 tenths or 50 hundredths or 500 thou? .05 = half tenth or 5 hundredths or 50 thou? .005 = 5 thou ... easy one .0005 = 5 ten thou ... ? .00005 = everyone seems to be in agreement its 50 millionths buy why would you just skip hundred thou Your machinist language confuses me, its like you guys discriminate against the word hundredth. If I walked into a machine shop and said 5 hundred thousandths would everyone laugh at me?
.5 = 500 thou or 1/2 inch .05 = 50 thou or 5 hundredths .005 = 5 thou .0005 = 5 tenths .00005 = 50 millionths or 50 mils Most measuring tools only give you a 0 or 5 reading at this scale anyway. Probably has to do with how easy it is to say.
Wow this is so informative video. Thank you. I currently working on my master thesis. And i would like to know more about the history of scraping. If anyone knows about it or how to get access to the source, please help me!!
Thank you. You could have called this: "Everything you wanted to know about scraping but never got a chance to ask?" It certainly answered all my questions. BTW: At 1:04 you point and say "card here to a half an hour video on the Richard King ..." but there's no card and no link. There's no link in the description either. I guess this is that video: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-Aq3tHyRVNys.html
I guess nobody in your scraping class told you that machinist levels aren't flat on the bottom. You could always send it back to Starrett to have them put the arch back into it so you can use it again.
Hand scraping is important process, machine needs an oil pocket to maintain the accuracy of the machine.specially in the small parts like gib and gudeplate zslide and turret slide of the machine parts. The accuracy must be 2 micron .im a scraper for almost 26 years.
Hopefully you'll be able to answer this John. I can't see any of your replies to any videos. I've checked everything on my end to make sure I've got nothing setup strange, but you're the only person I can't see. I see replies of everyone else, just not you. Some of the newer vids I can see replies, but none of the older ones. Any thoughts?
that time we were 12-13 and thought the precision blocks were old pieces of scrap with hammer marks.and used them as anvils for our school project. the Teacher was pulling his hair when he realised what we did.
Nice video. There is a way of making any surface flat to within fractions of a wavelength of light by grinding three flats together. Any two flats will form a concave/ convex spherical surface when ground together, but this is overcome by grinding three flats together which counteracts the tendency for any two surfaces to form a spherical surface. What is formed is actually an infinity sized spherical surface or a plane on all three. Details of how this is done can be found in the book "Amateur Telescope Making".
I've looked at the above and wasn't able to find a definitive answer, but I also don't really know what I'm looking for! I may turn an old file into one.
Would be difficult to scrape a level not knowing whether the Bubble Glass was level with the surface. You could scrape the surface sure, but if your high spot doesn’t make the glass level then you have a flat surface and the bubble level isn’t square with it. If you have a good way to chuck it up with the glass level then I suppose though it could be challenging for certain devices especially if the bubble glass only goes one direction.
yea .0005-.0006 isn't perfectly flat, but for the most part do you ever need a part that's tolerance is less than .001 maybe .0005 unless you're making parts for a jet or NASA?
Jay Brewster Internal parts for this. The motor and bearing housings and the joint flexure systems. The entire system combined has less then .005” total accumulated run out. www.mobiusbionics.com/luke-arm/
One of these days they're going to figure out that lubing sliding surfaces just need scratches like the hone/stone marks need in an engine's cylinders to keep the rings from wearing, and scraping will just be old time machinist masturbation!
Ettore Bugatti hand scraped every single head on his engines. No need for a head gasket. That is another reason why Bugatti cars fetch millions of dollars.