Scraping perpendicular, parallel, and coplanar surfaces

[Builder4wd]

Hi,

I have a basic understanding of the process of scraping to produce flat surfaces, but I was wondering how to create parallel, perpendicular, and coplanar surfaces with scraping. For example, how would I scrape two cubes or rectangles to be the same size with their corresponding surfaces perpendicular, parallel, and coplanar with each other?

Thanks!

[Pressfit]

Scraping? Now that's an art! If I had only taken the guy up on his offer to teach me about 25 years ago, I might be able to tell you.
But Im sure it takes a Surface Plate, High-Spot blue, 123 blocks, Angle Plate, Indicators, time and an extraordinary amount of patience.

[in2steam]

Originally Posted by Builder4wd Hi, I have a basic understanding of the process of scraping to produce flat surfaces, but I was wondering how to create parallel, perpendicular, and coplanar surfaces with scraping. For example, how would I scrape two cubes or rectangles to be the same size with their corresponding surfaces perpendicular, parallel, and coplanar with each other? Thanks!

So how much time(and money or resources) do you have on your hands ?

Although not impossible, if you are talking about GI(other metals would be harder for multiple reasons), it would take quite along time size depending.
some perspective..
When I was taught to scrap we were given a iron surface plate(one plane only) of about 6"x24". We first took that plate which was "made" to be a bad surface(either a fresh casting or some were ground with an angle grinder I think) Our first lesson was to run a simple cnc program to true up 1 side, that gave us real simple cnc crash course. We got those surface plates on the order of about .001-.002 flat. We then breifly draw filed them just to clean them up, and blued them(which is a trick to learn also). From there we used a master surface plate and scraper with, our own hand made carbide insert scrapers, then later after we achieved so many points per inch at a certain size(which escapes me but I would say around 6) we moved on to power scrapers.

So to start once we got them blued properly which was a challenge in of itself to learn. With a normal cross pattern it took about 8-10 days of 2-3 hours a day in class, there was alot grumbling and mistakes. The second part(power) took two 2-3 days of the same, although some were quicker others were longer. I think that our accuracy was a modest .0005 over the whole surface if memory serves the master plate was reg. in at .0002. Although the purpose was to get so many points per inch not a general accuracy of flatness, but by default you are.

Now if you want to do coplanar surfaces thats easy, you just need to blue them to each other. The perpendicular you would need plates that are equally as perpendicular, although I cannot say i have ever thought about it much the hard part would be holding them that way as blueing requires movement.
Parrallel would also rely on the master plates, this would be a little easier then perp I think, but that assumes you can move either the plate or the piece by hand.

One thing you should understand about scraping is not that you are achieving perfect flatness, you are achieving a good bearing surface, many would say that there a modern ways to do this which including surface grinding which are now more accurate. Its quite possible(albiet hard) to over scrap something and in which case you would have so much surface area coplanar that you would stick to what ever you matched it to, and this is what i think you are talking about. In machine repair its a good idea to use both level and flat as reference so scraping goes in hand with it easier to say "flat" even though there could be sag. Scraping is also done for bearings on plain bearing journals, and shafts to some extent still. Scraping also provides the "frost" or "hook" on the surface of a mill or lathe bed ways, its purpose is to hold oil for lubrication not for bearing qualites and would normally be applied after "scraping in".

So to start, if you are doing this with GI, you would machine the stock to near desired size, I would personally use a shaper as it gives the best finish. I then would try and get them ground down, for parallel and perp. Depending upon the grind will get you real close, after that its a challenge, you can do a couple surfaces easily, but getting at least 2 of them done will be a challenge beuase you have to somehow grip 2 sides that you have scraped already without distorting them(you need a good solid vise or beding to do scraping).
You would also need the master plates, the scrapers, and a good way to sharpern it(either carbide or HSS) so the edge is maintained. Stones, files and blueing and cleaner. Most of it is cheap the master surfaces are not, unless you have acces to one that can be blued without hassel(blue stains and does not wash out easy).

Scraping is not like using a mill, articlulation of the hand,posture of the body, speed, pressure, and the scraper itself all play into the surface finish. I scrap hand planes (because I can)all the time, but I am nowhere near the skill of my mentor, who used hand scrap vertical lathes with 10-20' beds(chucks if you will) down to .001 over the diameter(this is not a one or two person affair). For me, I use a engineers scraper set I bought off of enco, and my carbide insert homemade(for roughing) to do the hand planes after I surface them. It takes for instance a number 5 stanley which is about a 3*9" about 1 day if its decent, for a couple of passes and a nice frost. The surface is superior that is why I do it, the flatness is secondary.

So for the time and money i would either buy them, or have them ground, but if you want to learn it will take some understanding of reading a bluing. Scraping is not a one day learning affair and its hard to say that I will ever be confident to say I can do it well. Some of the old timers could scrap and there would literally would be a cloud of iron dust near them, they did this day in and out for years, they would be the ones that could do this if anyone could at all.

chris

[Builder4wd]

Thanks for the detailed response. You mention the use of a surface grinder to get the parallel surfaces. The reason I'm asking about scraping is I would like to create a bar with two parallel surfaces for a machine I'm building. I don't have access to a surface grinder, or I'd use it! I'm not concerned with surface finish or lubrication - I just want it to serve as an accurate mounting surface. I need two opposite surfaces of the bar to be parallel because one surface will mount to the table of the machine and the other will mount one of the axes. The two surfaces must be parallel in order for my machine travel to be parallel to the machine table.

Thanks!

[in2steam]

Originally Posted by Builder4wd Thanks for the detailed response. You mention the use of a surface grinder to get the parallel surfaces. The reason I'm asking about scraping is I would like to create a bar with two parallel surfaces for a machine I'm building. I don't have access to a surface grinder, or I'd use it! I'm not concerned with surface finish or lubrication - I just want it to serve as an accurate mounting surface. I need two opposite surfaces of the bar to be parallel because one surface will mount to the table of the machine and the other will mount one of the axes. The two surfaces must be parallel in order for my machine travel to be parallel to the machine table. Thanks!

I think it will be easier for you to purchase what you need, esp if you are looking for something highly accurate. If you have no experince in scraping you will only cause yourself problems. Scraping was done in the early days after a rough machining, as it was more accurate then the avaliable machining methods of the time. Buy the time you buy what you need to do this you could buy what your end product will be without alot of hassle and wasted time learning(thats not to say you cannot try it, its good to know) and money spent on aleeve. Otherwise getting something ground down is typically not a big deal although size could be a constraint. I think you will find people more then helpfull if you some detail as to what and how you want to do it, and point out any possible flaws or ommissions.

chris

[handlewanker]

Hi Builder, there's two ways you can achieve accuracy, the first is to have an existing surface plate of known accuracy and the second is the three plate method whereby three plates or surfaces are scraped together and used as references for the final result.
Even surface grinding a plate is NOT accurate as most people seem to think, due to wear in the grinder ways, loss of wheel material and cutting ability, and least of all relaxation of the surface of the plate as the stress in the material is released from cutting one side.
The name of the game is COMPARISON.
You are comparing the surface being scraped to a known surface that is certified accurate and has a certificate to verify it, otherwise you might just as well use the welding bench because it has a nice flat looking surface.
The most accurate surface is produced when three faces are scraped together, by using the three faces to reveal the degree of flatness of each other.
Without a master surface plate there is no other way.
The vertical faces are revealed by using a square of KNOWN squareness.
Consider a cylinder turned parallel in the lathe and without moving it, the end face is machined flat or even concave, not convex, that is the face is lower in the middle.
Placed on a surface plate on the machined end, the cylinder is perfectly square to the plate.
You now have two references in two planes that will satisfy any surface requirements for flatness and squareness.
To check for parrallel just needs a micrometer to measure the thickness of each end.
To work to this degree of accuracy is a time consuming job.
We could write an essay on the making and sharpening of a scraper alone before even considering scraping an accurate surface.
When you apply blue to the surface plate it should be barely visible.
The surface being scraped should be scraped alternately at 45 degree to the previous scrape, producing a basket weave effect, and cancelling out a tendendacy to go hollow.
By the way, carbide scrapers are not an option, despite the popular knowledge that they last forever.
How do you sharpen them? Not unless you have a diamond lap or diamond cup wheel on your grinder with a flat adjustable work table.
For scraping cast iron, a scraper with a glass hard cutting edge is easy to maintain with an oil stone, just don't let the edge go dull or you will rub a hard skin that is the very devil to get through.
A surface is considered flat when it has 25 points of reference to the square inch.
Ian.

[Mcgyver]

Originally Posted by Builder4wd Thanks for the detailed response. You mention the use of a surface grinder to get the parallel surfaces. The reason I'm asking about scraping is I would like to create a bar with two parallel surfaces for a machine I'm building. I don't have access to a surface grinder, or I'd use it! I'm not concerned with surface finish or lubrication - I just want it to serve as an accurate mounting surface. I need two opposite surfaces of the bar to be parallel because one surface will mount to the table of the machine and the other will mount one of the axes. The two surfaces must be parallel in order for my machine travel to be parallel to the machine table. Thanks!

IMO, outsource it to someone with a surface grinder, and put your scraping efforts into whatever said parallel bar gets bolted to, the finished mounted piece won't be any flatter than what you torque down to.

if you are trying to scrape surfaces coplanar, like machine ways, get one scraped and then scrape the second by having an indicator on a surface gauge on the first one, guiding scraping on the second. an iterative process. same for scraping something square, get one of the 'accurate to 1/10000 over 6" squares, indicator and surface plate. get one surface flat, then start on the second, constantly checking with the square and indicator.

what scraping does is make something flat (or round) . its the creative use of references, indicators and squares thereafter that let you make things parallel, coplanar etc. imo scraping is ideal for preparing bearing surfaces, restorations, creating a datum on a casting etc - things where grinding isn't an alternative or at least as good an alternative. but its a lot of work just to create tooling more easily done by grinding. grinding a cube perfectly square to 10th's is no walk in the park, scraping would be, well, masochistic. its one thing to want to run a marathon, another to insist on doing it in Greece, between Marathon and Athens barefoot .

I agree with Ian's comments on the carbide, although inexpensive diamonds laps are available from woodworking suppliers (why is another unanswered question, maybe for router bits?) my best scraper was forged from an old file.

[handlewanker]

Hi mac, Some years ago I had to refurbish a 10" X 20" lathe I had, I mean it was worn like big time.
It had .013" wear in the bed from the chuck end to about the middle, which as it was a 1930 vintage wasn't surprising.
I hand planed the raised vees and flats where the saddle and tailstock rode, with a planing device that moved on ball races on the unworn sections of the bed between the raised vees and on the sides, and afterwards scraped the lot together.
The scraper was made from a 14" single cut flat file,one of my best, but I had two so I sacrificed one. I just got it red hot in the fire and forged the end out and thinned it.
The scraper was hardened and left glass hard for the first inch, and I've use it on many occasion since.
Ian.

[Mcgyver]

kind of like the top one? Mr. Wright and Mr. Moore, who's humble effort sits below it, were unable to improve upon it. if its not obvious, and anyone is inclined to have a go, grind off most of the teeth to preserve your skin

[in2steam]

What it sounded like he was doing was making something from scratch, scraping is not the way to go, it would be a sinful waste of time when you can buy materical flat enough to do what hes trying to accomplish. Scraping is far better for restoring or finishing off the machine then trying to make a set of references for references, if you already have the reference there is no need.

Many years I ago I made a scrapper out of a file, and did not grind it down, I payed that price, but it was in a pinch along ways away from my toolbox, since I purchased a set of machine scrapers from enco, they seem stought enough, but still not as good as my carbie, which I have to lap on my diamond since I nolonger have acces to a green wheel. I only really scrap to finish off the buttom of wood planes nowadays, I have run a frost on my shaper recently but it was more for looks then anything else. The true challenge is scraping a round journal bearing in, thats hard, and takes alot of practice.

chris

[handlewanker]

Hi in2steam, as I read it he wanted two square or rectanguler blocks coplanar and parallel to each other, and the same size.
Being classed as a rectangle or cube gives it flat sides parallel and square to each other.
It didn't say that one surface was going to be fitted to another existing possibly not quite flat surface.
So in that case the first rectangular piece would have to be made truly flat on one face, and the other faces made parallel and square to it, with the second item identical.
Unless you have a reference certified flat surface to grind/file/scrape to, you can end up with a banana that is still coplanar.
The definition of a flat surface is a circle of infinite radius, which by it's very description can never be truly flat.
The fact that the ends are equal distant from the same point means that they will meet eventually.
Hi Mac, the scraper I made is exactly the same as the top photo, and as has been said, it's very important to remove the teeth by grinding, but BEFORE heating and forging, otherwise the end will develop hairline fractures across the edge from the teeth.
The other important bit is to thin the edge down when forging to not more than 2.5mm or 3/32" thick, otherwise you'll be wearing a groove in the stone trying to get the edge sharp.
I sharpen the scraper by holding it vertical to the stone and inclined about 1 degree to the side, left hand holding the handle at the top and the right hand holding the blade between thumb and forefinger near the bottom, giving it a forward and back motion, and rotating the point 180 degrees every few strokes, to give two distinct curved flats on the end.
For those that are new to scraping, the blade should be held at a very shallow angle to the work.
There is a tendency for some to raise the handle end to get it to cut when the edge dulls, but this just makes resharpening a longer job.
The scraper is assumed to be sharp when it will raise a scraping when applied to the thumb nail.
Last but not least, you must have complete and utter freedom from outside disturbances, so that concentration and interpretation can be applied.
If you scrape a .001" hollow at any point on the surface, then the WHOLE of the rest of the surface has to be removed to bring it flat.
Ian.

[in2steam]

Originally Posted by handlewanker Hi in2steam, as I read it he wanted two square or rectanguler blocks coplanar and parallel to each other, and the same size. Being classed as a rectangle or cube gives it flat sides parallel and square to each other. It didn't say that one surface was going to be fitted to another existing possibly not quite flat surface. So in that case the first rectangular piece would have to be made truly flat on one face, and the other faces made parallel and square to it, with the second item identical. Unless you have a reference certified flat surface to grind/file/scrape to, you can end up with a banana that is still coplanar. The definition of a flat surface is a circle of infinite radius, which by it's very description can never be truly flat. The fact that the ends are equal distant from the same point means that they will meet eventually. Hi Mac, the scraper I made is exactly the same as the top photo, and as has been said, it's very important to remove the teeth by grinding, but BEFORE heating and forging, otherwise the end will develop hairline fractures across the edge from the teeth. The other important bit is to thin the edge down when forging to not more than 2.5mm or 3/32" thick, otherwise you'll be wearing a groove in the stone trying to get the edge sharp. I sharpen the scraper by holding it vertical to the stone and inclined about 1 degree to the side, left hand holding the handle at the top and the right hand holding the blade between thumb and forefinger near the bottom, giving it a forward and back motion, and rotating the point 180 degrees every few strokes, to give two distinct curved flats on the end. For those that are new to scraping, the blade should be held at a very shallow angle to the work. There is a tendency for some to raise the handle end to get it to cut when the edge dulls, but this just makes resharpening a longer job. The scraper is assumed to be sharp when it will raise a scraping when applied to the thumb nail. Last but not least, you must have complete and utter freedom from outside disturbances, so that concentration and interpretation can be applied. If you scrape a .001" hollow at any point on the surface, then the WHOLE of the rest of the surface has to be removed to bring it flat. Ian.

Ian I don't know how much experince scrapping you have, but two things I noticed in your response that caught my eye, although most people do scrap iron, steel and brasses may also be scraped that changes how you hold the scrapper, atleast in my experience if you hold it too low on steel it just slides, the exact oppiste occurs on brasses(bearings if you will). If you knock off .001 doing scrapping on normal iron you are either staying in one spot too long or you are a brutus, becuase if you take off a couple tenths at a shot with a good scrapper your doing good. It was consider the norm in machinery rebuilding to get something down to .001 and then scrap it in from there, then follow up with a frost or pocket for oil retention. We also blued as often as possible to keep our points in check, if there were any at all. I was taught with a carbide, but I now use steel scrapers for ease of sharpenign, the carbie still comes out when I have some nasty machine marks to bring down.

In the orginal post he asks for a parallel coplaner perpindcular cube, or rectangle, that aleast in my mind means on a flat surface(reference) each one of the same 6 sides are in the same angle and thickness for simplicty.

The other thing with scraping a 4 sided unit of what ever size(which was never specfied) how does one hold it after 3 sides have been scraped, am I asking from a practical standpoint, I have never scraped more then 2 surfaces in realtion to each other and they were permantly mounted ways so they did not need to be held, and then how do you blue them?esp for perpendicular. In the holding aspect I would think on an item lets say 3 inches by 3 inches, that you would mar atleast 2 of the surfaces in an sort of vise or clamp, even with soft jaws or wood, not to mention the strain put on it while scrapping in a vise. Its a little extreme but none the less, it could happen, while I was being taught to scrap in school some second year guys were scraping in a bed way on a old lathe, they started that project on a 90 degree day, the next week it was in the mid 50's they now had machine that was nowhere they needed it, they worked on one way at first then the second the next week thinking they were coplanar. In reality it did not matter but the teacher gave us a stern waring watching that thermostat while doing anything important(while laughing).

chris