[rolfinator2002]

Why can they just not surface grind instead of scraping like he was?.

Or am I missing some thing here?

Had the same questions for a long time; here's the answer provided to me by a few trusty sources.
Grinding actions deflects to some extent, more and less depending on the local surface hardness and cooling action. This causes a slightly uneven surface, but avoidable in today's production environment.
The stationary surface called a way, and the moving surface called a slide: For box way systems, the way is usually ground (and hardened) while the slide is ground, scraped, flaked or frosted. Even when Turcite or Moglice is used for the slide, it is flaked. the purpose of scraping/flaking the slide is to prevent hydraulic locking of the slide and way. If they were both ground and left as-is, the slide would jerk and lurch as it moved plus the distance between the slide and way changes with slide velocity. For a mill saddle, it'll make irregular tooling marks and a sloped, stepped or poorly planed surface.
The next time I rebuild a machine tool, I'll be using Moglice or Turcite on the slide, rather the a full rescrape on both way and slide. I'll just scrape or grind the way.

[rolfinator2002]

Whenever a gage is made, the planar surfaces need to be adjusted prior to bringing up points, i.e tilted in one direction or another.
For oblong gages, such as the 90/40/50 degree gage I pictured earlier in the thread, scrape-in the long axis first, then scrape-in the short axes (the 90 degree legs).
Why? An angle coincident to the long axis of a gage is harder to adjust (and drift) since so much more material must be removed to change its angle. If one of the legs of the square, mentioned above, needed to be adjusted much less material is removed; additionally, first scraping-in a short axis risks wasting time since short axes tend to "drift" as they are being scraped-in. The risks and benefits must be weighed for each job. In the case of the gage below, it is important to have all 3 planes flat to within 0.0002" and 90 degree angle to within 0.0006"/feet, so the hypotenuse plane and one other surface are fully scraped to at least 30 points. The last plane is roughed-in to about 5 pts to allow measurement, first tilting the long axis as necessary, then the 90 degree leg is tilted, leaving one end of the plane untouched (similar to a door hinge). Finally bring-up points once the pattern stabilizes.

[rolfinator2002]

In the beginning, I mentioned that I made my prismatic gage to 50 degree to match the dovetail on the supermax mill.
Have since learned that making gages to the exact angle is not always the best idea.
Use a 50 degree angle gage IF machine dovetail is trashed or heavily worn w/o original witness angle.
For most work, make a dovetail gage at an angle LESS than the working angle; allows gage to be tilted into the objective machine way or slide, used to spot the surface, and also used for many different dovetail angles.
Fortunately, finishing all 3 sides is a lot tougher than just 2 sides, but now my gage can be used both for basic dovetail spotting, redo of 50 degree dovetails, and squaring-up box ways and slides. Live and learn.

[rolfinator2002]

Purchased a Starret master level that was a little older, i.e. the cast iron strained.
It was CONCAVE to perhaps .002" at the center and needed a rescrape.
One of the toughest challenges for a newbie (like me) is to avoid CONVEXITY; even a concave part could be scraped past flat and into convex land if good habits are not developed. Hope these tests help folks as they have helped me. The last is a test I modified for better reporting.

Whenever anything is scraped, do a spin test.
Whenever anything is important, also do a paper pull test and if it's really important,
use a water level do a rock test using the water level as indicator.

For a host of reasons, longer items tend to be convex, that is, higher in the center, so the goal is to focus a little more in the center or less on the ends.
Spin test: Put the cleaned piece on a cleaned master, and use your finger to push one end around, then repeat by pushing the other end. If Either or both pivot points are NEAR the center, there is probably some convexity. A flat piece will have different pivot points toward each end and will be somewhat volatile, ie. inconsistent. You will need to practice and judge longer parts differently from 2ft parts. If you suspect a problem, do a pull test.

Paper pull: Use the SAME piece of paper (not different sheets from same ream) and cut identical strips. Use 3 strips for 1 to 2 foot long parts and your judgment for longer parts. Insert a piece at each end and the middle. Pull on each piece and note how much drag you feel, then move the center piece incrementally toward either side and repeat the test. If one strip is notably easier, then there is convexity, depending on middle strip locale, or some unknown issue (paper, technique,etc). You'll need to use a little common sense to compare what should happen with what IS happening. If still unsure and still important, do modified rock test.

Modified rock test: This test is normally a rough test used before the spin test, but with a water level becomes a "last word" type test. If you still aren't quite sure about your work, lightly push down on one end of the piece (cleaned and on the master) and note water level movement. Repeat for the other end. If you have any convexity, a master level will read at least 0.0005"/ft for a 2 foot piece. Ensure that the setup is rigid too.

I did the spin test, like a hundred times before, looked fair, did the paper test, it failed, wondered if it was me, then did modified rock test and found a single high point, scraped it. Repeated test one, perfect, did paper test anyway, perfect, did rock test for giggles, perfect. Now I know for damn sure that master level is good to go.

[rowbare]

You might find this interesting, it seems they are purposely made concave:

http://bbs.homeshopmachinist.net/showthread.php?t=39378


bob

[rolfinator2002]

Issue is that some people report concave master levels. Someone allegedly contacted Starrett and asked; he reportedly received general acknowledgement w/o details. They strain more concave as they age, so here's the support for why they are scraped a tiny tiny bit concave, if at all.
A senior member of the related post stated that he has rescraped Starrett master levels for 50 years... flat, and stated that it was a little annoying that Starrett withheld this alleged information the whole time. He also noted that the concavity had to be less than 0.0005" over the 15" of the level. He was politely disagreeing and he's right.

Starett's statement was taken far out of context and I doubt its necessity altogether.
Why?
1) The intended concavity cannot exceed 0.0005"/15" otherwise it would be impossible to accurately measure surfaces less than 15", and void its accuracy rating, a caution that Starrett would surely note, as it notes proper angularity usage of its dialtest units; Starett provides NO usage cautions for master levels.
2) The frame of a master level is triangular in shape, sooo it has a tendency to strain as it ages to greater concavity.
3) Finally, my level had 0.002" concavity, it would take a propane torch to remove that much bow. It's just not reasonable to assume that my casual hand (nuts)contact with the level would remove 0.002" of concavity.
Conclusion, read Connelley's book regarding levels, yes he covers it, and scrape it flat, with an extra pass in the center; it will preserve required accuracy, allow measurement of shorter surfaces, and avoid convexity risks. I did the same thing on my straight edge on the last side to avoid convexity risk(don't remember if I mentioned it).

[rolfinator2002]

Something that I found useful and not found in any of my resources:

Canode ink works as an excellent spotter, better than traditional spotting inks because it cleans up easily. One problem with probably most inks is gage absorption. When the ink achieves a certain level of viscosity, it does not indicate well at all, and becomes very hard to use. I have noticed that inks get too thick when applied to porous gages made of granite, worse yet for pink granite. I have begun conditioning my granite gages (straight edge, plates, 90 degrees, etc) with mineral oil (baby oil), wiping all sides with oil allows a barrier to form which retards the amount of ink solvent to be wicked out of the ink. Wipe off all residue prior to use and then spot the tool or work piece as usual. I found vast improvement for longer usage, i.e. 2 hours or more in usage.

[rolfinator2002]

After relying on many sources, there was considerable conflict regarding spotting methods.

Do you ink the workpiece or the tool? There are pros and cons for both. Here are some basic observations and simple rules to help me remember the process:
1) Inking the tool is preferred for ease of cleanup.
2) Inking the workpiece allows better accuracy and a more reliable final surface.
3) Ink heavier for initial spotting to protect the gage.
4) Drastically misleading results are obtained if a poor choice in inking is executed.
5) Tool geometry: The second factor that affects accurate spot reading are spots left by successive spottings (tool manipulations). If the workpiece is inked and the tool cannot cover the whole piece, then successive spottings will confound understanding of real high-spots. This can happen if the tool goes back over an already spotted area, excessively rubbing the area, etc.

Rule:
1) At the start of a surface reclamation, always ink the tool. It better protects the tool and speeds cleanup between scrapings.
2) As the surface is brought closer to final stage, i.e. rough pointing, switch to workpiece inking if the tool geometry allows it.
3)Tool geometry: Consider the tool to have a primary (long) and secondary (short) axis. There are 4 rough categories for consideration of whether to ink the tool or workpiece.
a) Primary and secondary at least as big as the work: Ink either.
b) Primary is longer than work and slightly narrower: ex: 12"x18" tool vs. bridgeport base turret way. Ink either but consider tool inking if there are issues. As the tool narrows w/respect to the work, tool inking becomes more important.
c) Primary is longer but very narrow: Must ink tool.
d) Primary and secondary are smaller: Must ink tool.

Justification and caveats:
1) As iron is worked into gages, a good gage ends with inking the piece, oftentimes using nothing more than alcohol with some lampblack as an ink. The desire in all cases is the option to ink the work.
2) Rule 3b: If the tool is almost to size, stroke the tool for and aft while moving sideways. The tool in the example does not need re-inking for each spotting. If the tool is very narrow, the tool will need the ink, and sideways movement will need to be restricted. The tool is then re-inked and placed on the work where you left off. If the work is inked, it is difficult to see where the previous work ended unless the tool was inked.
3) By looking at work in sections, one rule might be used over another.

Takeaway: If the tool is shorter than the piece, must ink the tool, otherwise if tool is as long work, workpiece inking is a possible option. I am NOT an expert, just learned, made some mistakes, and learned some lessons. If an expert want to disagree or clarify, I welcome the opinions.

[rolfinator2002]

There will always be situations where your tool doesn't cover the whole piece. Let's look at an extreme example, a 30"X30" work and a 2" by 36" straightedge.

Apply the ink to the tool, spot the work by stroking the tool 1" or less 4-6 times, with side movement of approx 1". Re-ink the tool, or at least re-wipe the ink to distribute it. Repeat the process where you left off. Repeat until the entire plate has been spotted. DO NOT scrape. Note the spots, use a camera, make a graph, or whatever you like. Clean the work.

Move the tool 90 degrees and repeat the process. Keys are to be absolutely consistent with your technique. Side movement is allowed in so much as the ink can be evenly distributed. If side stroking is desired/needed, consider starting in the center of the side ways path, go from center to left, and then center to right; do anything suitable to promote consistent distribution of ink.

Note the spots, use your knowledge to choose what you scrape. As the work approaches flatness, the first set of spots will look more like the second set of spots, ie. the work will be stabilizing. Scrape the high spots indicated by BOTH sets of spotting.

I have read in Connelley's book that as the work approaches flatness, spotting in only 1 direction is okay. DO NOT do this. Alternate spotting directions even up to your final spottings. If you don't, you'll create a ramped, either concave or convex surface. I learned this lesson the hard way.

[rolfinator2002]

Previously explained some rules and caveats for this type of scraping. Here is a very important point for any scraping project... think, think. Why are the spots appearing this way? It must make sense in your head. Do not assume anything.

Ex: You have disassembled your bridgeport mill and want to scrape the top of the base, the turret way we'll call it. You have inked a 12x18 granite because you are having trouble seeing "points" if you ink the work. Each time you spot the work, you get a non-uniform pattern. When you re-spot 90 degrees to the first, you get the same pattern but 90 degrees to the first. What's wrong????

Nothing. You have achieved flatness. The pattern has achieved stability, meaning that it is repeatable...good, but it is not uniform. The amount of ink available to the work 90 degrees to the direction of stroke is greater per unit area than the work that is rubbed inline. Essentially, this special circumstance provides more ink to some parts of the work, requiring you to catch the trick and not get caught up in blindly following mechanical rules. See the photo for details.

[rolfinator2002]

A quick correction to suggested spotting/inking rules for scraping iron.
I made a mistake with some of the wording. Again just a newbie's suggestion.

Rule:
1) At the start of a surface reclamation, always ink the tool. It better protects the tool and speeds cleanup between scrapings.
2) As the surface is brought closer to final stage, i.e. rough pointing, switch to workpiece inking if the tool geometry allows it.
3)Tool geometry: Consider the tool to have a primary (long) and secondary (short) axis. There are 4 rough categories for consideration of whether to ink the tool or workpiece.
a) Primary and secondary at least as big as the work: Ink either.
b) Primary or secondary is equal to or longer than work with the other axis slightly narrower: ex: 12"x18" tool vs. bridgeport base turret way. Ink either but consider tool inking if there are issues. As the tool narrows w/respect to the work, tool inking becomes more important.
c) Primary is longer but very narrow: Must ink tool.
d) Primary and secondary are much smaller: Must ink tool.

Hope this helps make the inking decision. helps me.

[rolfinator2002]

I'm filming and photographing refurbishment of all the critical surfaces on a bridgeport mill clone.

Here is the face way of the turret; note the shiny spots getting closer to the ends and filling up the voids, not so much the ink intensity. It was warped in the center, scratched and worn on the ends. Warpage put the face way "low" by 0.010" by the gear boss (operating handle to move the mill head in and out. To avoid gear binding between the rack and pinion on the arm, scraping was kept to a minimum, leaving some of the low spot as-is.

It took only 12 passes to bring back this surface to better than 15 pts/sq in. Much fewer than my earlier work.
Tool: biax w/25mm, 90mm radius stiff carbide tool.
Stroke: 7mm to remove most of the material, 5mm to get 10pts, 3mm for the last couple of passes.

[Machine_Rebuild]

I just finished scraping in the carriage on a 22 foot (264") Sirco lathe after grinding the bed on my WMW slydway grinder. I have the facility for all grinding projects up to 32 Feet in length. I can grind 22' (yes that’s FEET) in ONE stroke. My grinder has been sitting on a 9' thick concrete pad for 17 years (Sat for 2 years to allow for movement) I had Toshiba bring a laser in and level it so I am within .0002 over 22 feet which in reality is near impossible to measure. We scrape to fit everything on a perfectly trued bed and add Turcite linings to your carriage and cross slide to eliminate any stick slip and prolong the life of your machine. This costs a fraction of the price than buying new machines.

[handlewanker]

Hi Rollfinator, been reading all your post on this subject, You're truly a genius, Ok, so you give lots of disclaimers about your noobie status, but your achievements speak volumes.

20 years ago I had to renew the bed and saddle ways of a 10" swing Colchester Bantam lathe circa 1920/30 vintage that had .013" wear on the raised Vees halfway down the bed to the chuck which made adjusting the saddle jibs and keepers impossible.

Regrinding was out due to a big lack of funds.

I started off by designing and making a bed hand planer, that used the unworn and unused areas between the raised Vees and the way sides for reference and guide tracks during planing.

The planer was made from a piece of 1/4" steel plate that had various cut outs in it to carry ball races that the planer was guided by and rode on.

On top of the plate I made a small slide that was used to move a carbide tipped tool down the Vee slide faces as I pushed the "planer" along the bed, depth of cut .001"

The bed had two raised Vees for the saddle and one raised Vee and a flat for the tailstock, and all of these had to be recut to start off.

I made a scraper from a large single cut 14" mill file, forging and hardening the end as required.

Long story short, the lathe after adjustinjg for height reduction in the leadsrew and tailstock areas now turns parallel over 6" and faces flat.
Ian.

[kris79]

Hi all.
I start my project of lathe renovation. First i wanted to scrape the compound rest. In photo below you see compound rest.
I have questions about scraping:
Why the scraper sometimes make scrathes on the surface (look at the photo)?
I have scraper from an old drill bit with carbide end , 12 mm wide and 1 mm thick for dovetails. And another scraper with carbide insert 20mm wide and 4mm thick for flat surface. Maybe my grinding technique or scraping technique is bad?
Sorry for my english

[jsheerin]

What is your sharpening technique? A badly sharpened blade could leave scratches.

If you dig in the corners you could leave scratches. If you are progressing across the surface in the same directions your strokes go, you could be picking up previously scraped particles and those could cause scratches.

[handlewanker]

Hi Kris79, probably with a scraper only 12mm wide you are digging in the corners due to the narrow width "rocking" as you scrape.

For one off jobs I would avoid using a carbide scraper, especially a narrow one, as you do not have much control of the tool on the work and have a tendency to dig hollows which leads to pin points of contact and an ever receding contact area.

It's extremely easy to remove material as you scrape, and also extremely easy to dig hollows especially using narrow carbide scrapers.

The rule is to "scrape" the surface as opposed to gouging it, and the broadness of the scraper end ensures you spread the load of the point as you inadvertently bear down on the surface.

Too much radius on the end also results in a digging in and hollows, whereas too little radius results in corner dig in.

You start with the job exhibiting hollows from wear and high points of unworn material which must be removed even though they never had contact due to initial bad fitting procedures if any.

The last thing you want to have is hollows in the unworn areas that are now below the worn areas which leads to false marking and double the work required.

One of the last areas I would thing needed scraping would be the compound slide dovetails......they are slow moving and rarely moved areas of slideway contact....unlike the bed/carriage slides that wear as the years go by.

A badly fitted compound slide indicates the lathe was poorly constructed and finished initially, so any wear on the bed/crosslide areas is also an indication that serious misallignment was built in from day one.

The burning question is....what are you using as a reference to scrape the dovetails to?

Merely offering one dovetail to another, or two flat surfaces to each other with a bit of bluing between is as good as scraping with your eyes closed in the dark.

What references are you using (or not using....gasp) for wear correction and/or allignment?
Ian.

[kris79]

Thanks for yours reply and help.
I answer on the questions:
1. My sharpening technique:
After one passes of scraping I hone my scraper on diamond hone by hand. First I hone cutting edges, and after it I hone flats. This is good sequence?
2. Yes I sometimes progressing across the surface in the same directions my strokes go. Maybe this is the reason of scratches.
3. For compound slide dovetails i must make an angle straight edge from iron or steel (grinded on the grinding machine, angle must be 55 degree) and scraped.
I have only granite surface plate grade AA, precision level (0,02 mm/m) and dial indicator to measure.

I have another question:
What is procedure if my upper part of compound rest is longer than lower part which was scraped earlier? In photo below is my upper part of the compound rest. I slide this from one end to other end to mark high areas. This is good technique for longer part?

[handlewanker]

Hi Kris, in my opinion, the slide arrangement you have for the compound slide is back to front for normal lathe construction.

For clarity we'll call the part where the tool post is attached to... the top part, which moves and has male dovetails.

The part that is attached to the crosslide and is fixed (but can rotate) is the bottom part, and in your lathe has female dovetails.

Normally the top part is a female dovetail slide with a parallel or tapered gib strip fitted to it, and the bottom part has a parallel male dovetail slide, at least on all lathes I have worked on, but your lathe is reversed.

In my opinion you will have to get two parallels (masters) with one side on each at 55 deg.

The parallels need to have the angles at 55 degrees to clear the 60 degree dovetail faces when revealing the flat faces.

(When you do the dovetail faces, the parallels are tilted into the dovetail to reveal the high spots on the dovetail face only).

Place the parallels (masters) on the granite plate, NARROW side down, and use them to reveal the high spots on the bottom of the top slide as in post #38, but do not allow the angles on the masters to contact the inner faces of the dovetails.....it is only for producing a pseudo flat plane for the two flat faces which you cannot directly rub on the granite plate due to the raised section in the middle, and this slide must be done FIRST.

Blue the masters on the top face ONLY, sparingly, and only move the top slide on the masters about 6mm back and forth, 4 movements along the master, to reveal the high spots.

Once you have produced a flat area ( 25 points to the inch) on these two slide faces it is used as a master to produce a mating face to the bottom slide which is done next.

You will have to use rollers in the top part male dovetails to check the dovetails for parallelism and dovetail masters to true the dovetails faces, this time contacting the dovetail face only and not the flat faces which are already finished.

You can then mate the bottom half dovetail (one only) to the top half, and the dovetail on the side with the gib strip adjusting screws (or a tapered face) is left plain.

I assume your bottom FEMALE dovetails are parallel to each other with a parallel gib strip adjusted with side screws in the bottom half, but if'n it's fitted with a tapered gib strip you will have to scrape the gib strip to the other dovetail on assembly.

BTW, the compound slide fit is not all that fussy.....( hear we will hear cries of protest)....but for most of the time the compound slide gib/s is/are tightened up to prevent movement when normal turning is being done, and only loosened to do taper turning, it really has no other function, apart from fine movements when facing.
Ian.

[kris79]

Thank you very much for your reply and help Handlewanker.
When I finish my project, then I show results in this thread.
Where I can buy cheap carbide insert 20-25mm wide and 1-2mm thick? Maybe you can sell me some part.