Couple of scraping questions (Part 1 of ???)

[macardoso]

I would love some clarification on the roughing cycle of scraping. The largest thing I've practiced on has been a 4" vise body so I'm sure the roughing of a large worn out machine is on a whole different level from what I am doing.

I find myself essentially high pointing with a heavier cut and avoiding low regions to rough out material. I am starting to get the feeling however, that the goal is not to aim for the blue spots necessarily but rather to only focus on cross hatching an area to bring the surface down. I tried to do this but I wonder how you can keep yourself from scraping down the valleys even lower when doing this? Not sure how exactly to phrase my question, but what is the methodology of roughing in the part. In my latest example, I had to correct about 0.0030" of bow across the face of the vise (import garbage). What would be a ballpark time for someone to hand scrape this? It took me probably 2-3 hours to get it flat and I expect another hour to hit the high points to bring up the points per inch.

What is the criteria for switching from roughing to going after highpoints?

I see about half of the youtube guys who scrape doing discrete strokes, while others do a push stroke and drag back across the part to start the next stroke, like one would do with a paint scraper. I feel like discrete strokes is the "correct" way, but I also find that I can scrape faster and more accurately when I drag back on the part doing a bunch of strokes in quick succession.

Finally, my scraping is a bit haphazard when roughing in the surface. I follow the 45 degree lines and cross my previous pass in the other direction, but I don't put any effort in spaing out my strokes evenly, keeping them perfectly angled to each other, keeping the stroke length consistent, etc. Should I be? Is this just an aesthetic thing or is there a good reason to try to pay more attention?

Thanks!

Mike

 

[Richard King 2]

You need to grid out the area you will scrape. I am not sure if your hand or Biax Power scraping? Hand is much slower. But once you have measured the part you need to know how much each of your cut removes so you can use math to estimate how many passes you need to scrape the high area's to get down to the lowest point. You have to mark the high area's with a "high lighter" so when you scrape the high area's you know you've scraped there. Many rookies don't use the highlighter and dig holes. I used to use red lead and now Canode Yellow ink diluted with glass cleaner. You also mark the low area's with a magic marker so can avoid the low area's and only scrape off the high area's.

There are different ways to scrape. If you measure your part and it is more then .001" . Grit your part with a marker drawing 1" x1" squares and only scrape the high squares until your with-in .0002".WhenIrough I use an over lapping technique I call "scraping paint" which is 1/8" wide and no more then 1" long. and as you get closer you shorten your stroke and technique and once your down to .0002" you start the cross hatch checker board technique before you blue up the part on a blued surface plate or using a master scraped straight edge. The yellowed grid technique is what I call "step scraping" and "blind scraping" . The blade tip radius and a sharp blade is also important. I will look for some photo's as you should, add them here, so we can look at what we are talking about. I'll do so soon as I am busy doing quotes now. Rich

 

[macardoso]

You need to grid out the area you will scrape. I am not sure if your hand or Biax Power scraping? Hand is much slower. But once you have measured the part you need to know how much each of your cut removes so you can use math to estimate how many passes you need to scrape the high area's to get down to the lowest point. You have to mark the high area's with a "high lighter" so when you scrape the high area's you know you've scraped there. Many rookies don't use the highlighter and dig holes. I used to use red lead and now Canode Yellow ink diluted with glass cleaner. You also mark the low area's with a magic marker so can avoid the low area's and only scrape off the high area's.

There are different ways to scrape. If you measure your part and it is more then .001" . Grit your part with a marker drawing 1" x1" squares and only scrape the high squares until your with-in .0002".WhenIrough I use an over lapping technique I call "scraping paint" which is 1/8" wide and no more then 1" long. and as you get closer you shorten your stroke and technique and once your down to .0002" you start the cross hatch checker board technique before you blue up the part on a blued surface plate or using a master scraped straight edge. The yellowed grid technique is what I call "step scraping" and "blind scraping" . The blade tip radius and a sharp blade is also important. I will look for some photo's as you should, add them here, so we can look at what we are talking about. I'll do so soon as I am busy doing quotes now. Rich

Rich, I appreciate your time to respond. Very new at this so trying to not develop bad habits right at the start.

Right now I am hand scraping. Sharpening is done by hand, although as time permits I drew up plans to build a carbide grinder for sharpening the blade and indexable inserts. I started working with the Sandvik blade as it comes new knowing the radius is too wide. It is slow going, but I have been incrementally working the blade down to a smaller radius. Don't have a control gage, but by guessing if I started at 150mm, I'm probably down to 100-120mm now. I would like to get it to the 60-75mm range.

I need to measure my scraping depth. Do I do that for a single scrape? Or should I scrape the whole surface and see how it it drops in height?

I think I follow your scraping strategy. Let me see if I can recap.

1) Surface should be treated with contrast to help avoid digging holes
2) In early stages of roughing, the surface should be subdivided into 1 inch squares. The high squares are treated in their entirety.
3) In early roughing, bluing against a surface plate might not even be necessary if the high regions are located with an indicator.
4) Once the surface is flat within 0.0002" then a more refined cross hatch scraping is used with bluing against the surface plate.

Below are a few pictures of my first scraping practice. I'm sure it is all wrong but I'm trying to apply everything I've read and watched.

Initial blue up with heavy canode on the plate. Concavity of 0.003"


Roughing in the bottom for flatness


Step scraping to remove transverse taper and convexity (found by hinging)


Initial blue of the top surface. Bolt holes are blown out, and the bed itself is concave by 0.003"


Roughing for flatness, did damage by trying out a bench stone to assist in roughing. Won't try that again.


High spots marked with sharpie. No blue, just a rub on a clean plate to burnish the spots. Calling this the end of roughing and moving to high pointing for better contact.

 

[Richard King 2]

Looks good. A tip, get a sheet thin rubber like a truck inner tube and lay the vise on it. It won't slide as easy. Your % is real low 0 in some places and 30 in others, your looking for 40 to 60%. I was thinking mail the insert to me and I'll grind it for you. How did you measure the the .003" ? Is it co-planar now to each other?

 

[macardoso]

Looks good. A tip, get a sheet thin rubber like a truck inner tube and lay the vise on it. It won't slide as easy. Your % is real low 0 in some places and 30 in others, your looking for 40 to 60%. I was thinking mail the insert to me and I'll grind it for you. How did you measure the the .003" ? Is it co-planar now to each other?

Appreciate the feedback Rich! The rubber is a good idea. I screwed down some 2x4's on my bench to make a cradle and that has been working well. Just had the vise loose on the table before that and I could barely make cuts without it sliding everywhere.

Do you think this was too early to stop roughing and start going after only high points? I've done 4 high pointing cycles since the last picture and have gotten more bearing, but the distribution is not perfectly uniform. Maybe this is OK, but would have like to see it better.

Here is after 3 high pointing cycles.


I appreciate your offer for grinding an insert for me. That would be a great way for me to have a reference as I go forward! If you are willing to PM me your address, I'll get an insert off eBay and send it to you directly. I'd like to hold onto the one I currently have so I can keep practicing.

The 0.003 of concavity was measured by placing the vise in the middle of the surface plate, supporting one end with a shim, and walking around the part with a 0.0001" DTI.

As far as coplanarity, I believe yes the two top rails are coplanar. The bottom is not yet parallel to the top. I started with the bottom hoping that it would be the only surface I needed to scrape, but the top ended up being such a noodle that I had to scrape it flat first in order to be able to measure the parallelism on the bottom. So once I call the top finished, I'll go back and work on the bottom.

Do you have any tips on making smaller scraping tools for fine detail work or dovetails? I can't imagine the Sandvik scraper would easily fit into the tiny dovetails on my mill.

 

[Richard King 2]

Buy a Dapra 15 - 150 blade. how about showing us the surface just blued and not the magic marker. For practice you could go further on the vise, how about scraping something useful like a straight-edge. That vise is 10 time better then new

 

[macardoso]

Buy a Dapra 15 - 150 blade. how about showing us the surface just blued and not the magic marker. For practice you could go further on the vise, how about scraping something useful like a straight-edge. That vise is 10 time better then new

I currently have the Sandvik hand scraper, so I'd need to fab something up to hold the blade. Unless you think it is unusable, I was looking for a Sandvik 620-2525 H10 insert (I have a 620-2025 H10, 20mm width, that came with my scraper but I can't find those on eBay). It's also 1/3 the brice of the Biax.

Here are some blued surfaces. I was posting the sharpie versions because I am having a hard time getting a good shot off the blued surface.

After Highpoint pass #2


After Pass #3


After Pass #5


Thanks to it being cloudy today, #5 is not super visible, but I feel like it has 25% better coverage than pass #3.

Think I'm about to call the top good and start on the bottom.

Scraping the vise was out of necessity as I could not get good parts made for my steam engine. A straightedge is my next project. I'm having a hard time stomaching the cost of the castings for an 18" camelback. Do you have any thoughts on working on this kind for something as short as 18"

If I can start getting the hang of this I'd like to scrape my G0704. Seems silly for a cheap import machine, but it is what I have and there is a LOT of room to improve it.

45° Dovetail Straight Edge – Coreprint Castings

coreprintcastings.com

 

[Richard King 2]

Take a look at minute 5-17 on this show and see how we "Bump" scrape to get 40 PPI . I got a bit silly in the beginning as I never thought it was going on You Tube. lol Bob Goslin one of the scraping host shop owners took some of those 1 x 1 Sanvik blades to a shop with a water jet machine and had it cut in 1/2.

 

[macardoso]

Take a look at minute 5-17 on this show and see how we "Bump" scrape to get 40 PPI . I got a bit silly in the beginning as I never thought it was going on You Tube. lol Bob Goslin one of the scraping host shop owners took some of those 1 x 1 Sanvik blades to a shop with a water jet machine and had it cut in 1/2.

Question for you. Is it only about hitting the correct number of high points per inch, or is there also a requirement for the minimum (or maximum) depth of the low points between high points? I guess The reason I am asking is that the bump scraping seems to leave a much deeper scrape (when I tried it) compared to how I was finishing without any instruction as to the right way to do it. I was holding the scraper near the tip and almost rubbing off the high points, removing nearly zero metal in the process. It was a bit slow going for sure, but did get the number of points I was after. As a consequence of that method, the difference between my high spots and low spots is only about a tenth, very flat - maybe too flat.

I could see on a sliding way, maybe the valleys are too shallow to hold sufficient oil? Maybe this wasn't an issue for the semi-static ways on the vise I was working on. The slides on the moving jaw were left much more coarse. Although I think that is what flaking is about right? Rich, I remember seeing a scan of a worksheet you made some time ago that commented on the pattern and depth of scrape marks, but I can't remember for the life of me what was good or bad.

As a separate question, I haven't found any good videos about hand flaking (either the tool grind or technique). Do you have any references?

 

[Richard King 2]

You need to check your depth...average it out over the entire surface. I usually say check 3 places. I say people need a minimum of .0002" deep and a maximum of .001" If your were deeper you are pressing down. You only need to hold the scraper behind the high spot and when bumping the downward hit it cuts. This technique is only for scraping 30 to 40 PPI. With that said your using a a 40 radius blade tip ground to a neg. 5 degree for cast iron. I'm out of the office for a bit. I only have photo's of the 1/2 mooning. This week I will take a video from my phone during during this weeks training class. One of my students in Norway has several you tube shows one shows hand flaking. search- Jan Sverre Haugjord -on YouTube....got to go....Rich