[How do I?] Thoughts on simply scraping a part flat (efficiently)

Rex Walters

H-M Supporter - Diamond Member
H-M Lifetime Diamond Member
Joined
Apr 21, 2015
Messages
582
I've been away at another of Richard King's scraping classes so I haven't made much progress nor posted any updates recently to my thread about scraping in my lathe.

I didn't get as much done on my lathe during the class as I might have hoped as I was having too much fun running around and "helping out" (AKA "shooting the breeze" and "hanging on Rich's every word").

After three goes at the class, Rich's words are finally sinking in. I thought I'd put down some of my takeaways regarding basic scraping to make things flat.

Scraping something flat is actually pretty easy, but I've seen enough people go through the class now that it seems EVERYONE makes the same mistakes when starting out (its not just me!).

In no particular order:

Hand vs. Power scraping

It doesn't matter if you use a hand scraper or a Biax power scraper. You can do equally fine work with either. A Biax will go a lot faster, of course, and you will likely get more consistent scrapes with one, but (like learning to drive with a stick shifter) it's good to learn with a hand scraper at first. I still hand scrape very small areas like dovetails where it would be hard to control a power scraper. Things can go bad just as quickly with a power scraper as easily as they can go well, though. Scratches are a particular concern. Once you touch a power scraper, though, you'll almost certainly want one. :)

In either case, for cast iron parts you'll want carbide blades ground to about a 2.5" (60mm) radius, honed with about a 5 degree negative rake. You'll need a diamond wheel (ideally a slow speed one like on a Glendo Accu-Finish grinder) to shape (~240 grit wheel) and hone (~600 to ~1200 grit wheels) carbide blades. A cheap demagnetizer is almost mandatory to keep chips from sticking to the tip and scratching lines into the surface of your scrapes.

Mark up vs. Indicating

Ink blue-ups can lie to you, but used properly, indicators never lie. You must consistently use a variety of techniques (like "hingeing" and indicating) to find (and erase!) false markups.

Say you're scraping in a straightedge that is ever so slightly curved ("bellied") in the middle, such that the middle is slightly proud by, say, 0.0002" to 0.0005". When you mark it up on the plate there is no way to avoid slightly tipping the straightedge over the belly, causing the ends/sides to mark up as well as the proud center. If you don't realize that only the middle has true bearing, you're likely to scrape all the blue marks you see and potentially make the problem even worse. Hingeing the part can reveal that the center is high.

If hingeing or an indicator shows one thing, and the blue shows something else, don't believe the blue!

Use the stiffest, most rigid indicator setup you can and measure parts from a surface plate whenever possible. Minimize stick-out of the indicator from the base. 0.0001"/division indicators are almost impossible to use except on a surface plate with a good "transfer" stand. A 0.0005"/division indicator is far and away the best to use for almost all scraping tests. It's easy to read to a half division (0.00025") which is more than accurate enough for most tasks, yet not as bouncy and finicky as a tenths indicator.

Indicate your parts to know how much you need to take off (should you still be roughing, or just going for points?). Beginners rarely take the time to measure how much they need to take off, but they should.

If you suspect false readings during a blue-up, use your thumb to erase those marks so you aren't even tempted to scrape there. It's critical to only scrape the high spots, else you'll just make holes and tipped sides lower. Just because there is blue somewhere doesn't mean it's a high spot!

Roughing to finishing

Roughing uses longer strokes with the edges of each scrape touching or almost touching. Finishing uses much shorter strokes, with each stroke clearly separated from each other. The strokes get shorter and shorter as you progress until eventually you start "pin point" finishing where you are only scraping tiny individual spots of blue.

Rich calls out a variety of techniques for roughing: "circles," "zees," or "heavy crosses" depending on how aggressive you want to be. It's still important to alternate the direction of your strokes to avoid a washboarding effect even when roughing. Typically you'll use a wider blade and blade radius when roughing — the goal is to hog off a lot of metal (typically 0.001" to 0.005") in as few passes as possible.

Mid- and finishing passes use rows of individual scrape marks, with each scrape spaced evenly left to right, and each row separated fore and aft. The separation gap should be about the same width as a scrape marks. The length of the scrapes is gradually decreased between passes as you get closer to finishing.

You use the same force and strive for the same depth of cut whether roughing or finishing! Consistency of your scrapes is more important than almost anything (you'll never get great results unless almost all your scrapes are consistent width, depth, and spacing).

Once your indicator tells you that no point on the surface is lower than about 0.001" of the highest point you should stop heavy roughing and start what I think of as "area" scraping to get to about 5 points per inch (PPI). Note that if you're starting with something that was reasonably flat already, you may not need to do any heavy roughing whatsoever, but it's far more common for beginners to start finishing passes too early when they should still be roughing than vice versa.

One rule you'll not want to break is: Don't scrape low spots even lower! No matter the stage, roughing or finishing, you only scrape "around" where there is true blue (not false blue from tipping parts that aren't hingeing correctly). When roughing or "area" scraping, treat multiple clustered groups of blue as a single area to be scraped. The closer you are to roughing, the more liberal you can be with scraping outside the edges of those blue areas, but avoid any scraping in low areas with no blue at all!

Only after you've finally gotten to about 5 PPI over the ENTIRE surface should you begin scraping individual points rather than scraping rows of scrapes in "areas" of clustered blue spots. This is far and away the most common beginner mistake and can cost you hours of time and innumerable marking/scraping cycles. Five points per inch means there is no area on the surface larger than about a US quarter coin without at least a few blue spots.

Once you get to that point (and only when you get to that point) you begin finishing. Shorten your stroke even more, possibly use a slightly tighter radius on your blade (or simply raise the back of the scraper higher to create a steeper angle of attack) and begin scraping individual blue spots, never scraping more than once on any spot nor allowing your scrapes to touch one another.

Conclusion

Anyway, these are just a few of the basic points. There are many, many more things you need to get right to get good results. Like:
  • Always and constantly wiping to eliminate grit on the parts or the plate
  • Gradually thinning the ink on the plate as you get closer to finishing
  • Learning how to stone the part well to eliminate burrs, but not so roughly that you create problems
  • Early in the process, using "crosses" (overlapping rows at 90 degrees to each other) per pass, then moving to one-way rows and only changing direction between passes when you get closer to the end
  • Learning to both hinge the part on the plate, as well as hingeing a small parallel or gage block on the part to detect bellying and false blue
As long as this post is, the process really isn't that hard. Anyone can learn to do it. It's much easier, of course, if you take the class or have someone show you how to do it. It's impossible to learn how from just this post or from the Connelly book/whatever, nor is it possible to learn just by watching YouTube videos. You must physically get out into the shop and spend a few hours actually scraping to have any hope of learning the process.

Still, I hope this is useful for anyone that is still struggling.

I just purchased an old cast iron surface plate. If there is any interest, I could thoroughly document the process of scraping it in on this forum (with lots of pictures and video showing the exact process of "reading the blue"). I think I know enough now that it'll be a reasonably quick process to get it to gauge quality.
 
Last edited:
I just completed Richard King’s class. Rex helped Rich teach us rookies.
Before I went through the class, I wouldn’t have understood a thing Rex just pointed out.
I made those same mistakes.
Rex,
I have a cast iron plate, I haven’t seen it yet, I would benefit from your tutorial very much.
Thank you.
 
Criminy, Jeff. Commenting at 3:46 a.m.?!! Were you up late, up early, or just unable to sleep?

I’ll post my progress on my plate, but I saw what a beautiful job you did on that angle plate. I know for a fact you can scrape as well as I can.
 
I get up early every morning.
Great day in the morning!
I did sleep in at Jim's scraping class though. Dog tired I guess.
Ah, "I know for a fact you can scrape as well as I can"? I doubt that.
I tell you what though, Jim pointed at that angle plate because I needed something to do. I finished my angle plate.
I was intimidated when I saw the finish on the big old thing. Good quality plate though.
I learned more on his than I did mine.
I had two big holes, I kept pecking away until I was talked into measuring the holes and how deep my scraping patterns were.
That started me on the right track. I had to forget the rules, X down to grade level then begin to work on the flat surface.
When I got home, I placed my angle plate on my surface plate. There is something about flat on flat. You get a suction when it is lifted off.
I did notice the polished points showed some inconsistencies. There is always room for improvement.
Maybe when I get my precision stones this weekend I can touch up the high points?
 
Maybe when I get my precision stones this weekend I can touch up the high points?
Jeff, you, like many, many others, misunderstand what precision ground flat stones are for, and how precision stones work. After Saturday, you will understand...

The precision ground flat stones do a remarkably better job of de-burring scraping work between passes, but will not change anything truly flat at all.
 
Rex, I found your post very "readable" and informative.
I'd be interested in following the process of tuning up your CI surface plate.
Please post a link here, if you start a new thread, so I don't miss it.
 
Rex, I found your post very "readable" and informative.

Thanks! I guess there's a first time for everything. <grin>

I spent a few hours this afternoon roughing it in. I took some photos and time-lapse videos. I'll spend some time tonight editing the video and attempting to put it into some kind of logical order. I'll just post a reply to this thread rather than starting a new one.

In the meantime, here's a thread I wrote after my first time through Rich's class about making some small straightedges. I've just re-read it and still stand by almost all of it.

If you decide to make your own, please don't do what I did! Instead make a longer wooden vee-block "sled" pushed into the bandsaw blade to hold the durabar while cutting (providing support both in front of and behind the blade). What I did with a small metal vee block was way too dangerous.
 
Okay, here's the surface plate as I received it:

293760

I soaked it in Evaporust overnight in a screwed together frame and some trash bags:

293761

I covered it to prevent evaporation and to keep the dew out. I also put it on sawhorses so I could siphon the liquid back into the containers.

It came out pretty nice:

293762

There was only a little bit of pitting.

293763

The deepest pit was about 0.0035" lower than the surrounding area, so I knew I needed to remove a lot of metal.

Here's a wide shot time lapse of me scraping it in:


This is a closer view of the first markup:


And here's a tighter view of the scraping:


After several roughing passes (most of the afternoon) there is still a pretty big (but shallow) hole in the middle. My first several passes with no blue ups took the pits down to about 0.001" deeper than the surrounding area. The pits are toward the middle of the plate, and it was bearing on the outside corners, so I didn't bring it down all the way. Fortunately, the pits are outside of the shallow hole in the middle, so when I finish roughing tomorrow I should have removed the pits altogether.

Here is where I left it:

293764

I'll provide a more detailed description of the process this far tomorrow.
 
I get up early every morning.
Great day in the morning!
I did sleep in at Jim's scraping class though. Dog tired I guess.
Ah, "I know for a fact you can scrape as well as I can"? I doubt that.
I tell you what though, Jim pointed at that angle plate because I needed something to do. I finished my angle plate.
I was intimidated when I saw the finish on the big old thing. Good quality plate though.
I learned more on his than I did mine.
I had two big holes, I kept pecking away until I was talked into measuring the holes and how deep my scraping patterns were.
That started me on the right track. I had to forget the rules, X down to grade level then begin to work on the flat surface.
When I got home, I placed my angle plate on my surface plate. There is something about flat on flat. You get a suction when it is lifted off.
I did notice the polished points showed some inconsistencies. There is always room for improvement.
Maybe when I get my precision stones this weekend I can touch up the high points?
Jeff You Got Mad and started to scrape harder...lol Its a learning curve many have... As I said I can tell someone until I am blue in the face until they get tired of screwing around and in your case lifting it on and off the plate 50 times and not getting anywhere. Your light went on......I asked Jeff to sign the angle plate...I was really proud of it too :)
 
Okay, here's the surface plate as I received it:

View attachment 293760

I soaked it in Evaporust overnight in a screwed together frame and some trash bags:

View attachment 293761

I covered it to prevent evaporation and to keep the dew out. I also put it on sawhorses so I could siphon the liquid back into the containers.

It came out pretty nice:

View attachment 293762

There was only a little bit of pitting.

View attachment 293763

The deepest pit was about 0.0035" lower than the surrounding area, so I knew I needed to remove a lot of metal.

Here's a wide shot time lapse of me scraping it in:


This is a closer view of the first markup:


And here's a tighter view of the scraping:


After several roughing passes (most of the afternoon) there is still a pretty big (but shallow) hole in the middle. My first several passes with no blue ups took the pits down to about 0.001" deeper than the surrounding area. The pits are toward the middle of the plate, and it was bearing on the outside corners, so I didn't bring it down all the way. Fortunately, the pits are outside of the shallow hole in the middle, so when I finish roughing tomorrow I should have removed the pits altogether.

Here is where I left it:

View attachment 293764

I'll provide a more detailed description of the process this far tomorrow.
Rex I think you forgot to lengthen your stroke. and if it's out .003" do some paint scraping circles or heavy no rule scraping using the highlighter and blind / step scrape it. One other thing about those bigger Norton ground stones, they are terrible when you have holes as they only stone off the highest ones and not the lower ones. I had the guys break the stones I gave them in 1/2. Rex show the Japanese stone and the big Norton ones. It's late for me/ I have to get to bed. Night guys
 
Back
Top