A bevy of scraping question

[J Tiers]

Actually, as a reference surface they are kinda OK.

You cannot READ spotting on a ground surface for anything, because it is just a smear. But a granite flat is basically a ground surface, and it works OK for "providing the ink" as the reference surface. Because it is smooth, there is ink everywhere on it, and it can blue any bump on the item you are testing, no matter where it is. That is a good feature.

And, the part you are testing, since it is scraped, will pick up the blue in the normal way. So it really is not that bad. (It's a good idea to do an initial light scraping on any surface before you blue it, just to get the initial texture on and let it pick up the blue in the usual way.)

The one issue is that a really smooth polished-ground surface does not "hold blue" as well as the slightly more matte surface of a granite flat. So it WILL tend to smear more, and you will probably have to have a thinner coating of blue on it to get consistent readings.

If you do that, you will be OK, just need to be more sensitive in terms of "reading" the spotting, and don't be afraid to clean off the part you are testing if you don't think the spotting is telling the truth. If it looks smeared or otherwise "not normal", not similar enough to the last spotting, then clean off, re-spread the blue on the reference, thinning it out if necessary, and try again. Try making only short movements when spotting, not long push-pull movements.

Be very careful though. When spotting, always hold the part you are moving (straightedge, or scraped work) by one end, and move it from that end, with the same pressure downward every time. You can switch ends once you have good distribution of spots, but there is a trap you can easily fall into.

You might have a table surface that is slightly convex, and then if you let it, or the reference surface tilt or rock at all when spotting, you may get spotting that does not tell you that the table surface is not straight, because any tilt or rocking that you allow may let "the other side of the bump" pick up blue also, and it can look like you are OK when you really still have a problem, or it can make the problem look like it is in a different place.

Being consistent and keeping slight pressure on it so it cannot "change tilt" will help you not get fooled that way.

There is a Stefan G series of two videos which are pretty good. I think they are under "scraping for flatness". He shows the rocking problem, and interprets the spotting, explaining his reasons.

[Mcgyver]

you can, its just not near as nice. They don't wring together, the issue is the quality of print the blue leaves. Sort of smeared or washed out would be one way to describe it, the high spots aren't as visible or clear. Its imo enough of an issue that you'd never want to grind a reference flat intentionally as a way to finish it, but if you have to you have to. Standard modern lathes for example ground the bed then scraped everything into them using the bed as a reference.

Between Covid cabin fever and Connelly, don't go nuts on us.

[epicfail48]

Back again with another question, one that came up while reading the book on reconditioning. Its mentioned several times in the book that ground surfaces cant be used either as a master for bluing or be blued against a master surface. Anybody whos tried, is that actually the case? If so, why cant you? I could see how you wouldnt be able to blue 2 ground surfaces against each other as they would just wring together if they were flat, would that happen if you were to use one scraped surface against another ground?

Reading about that threw a bit of a wrench in my plans to have the apron on my grinder ground, then use it as a master to scrape the table ways

[mikey553]

Thanks to everybody contributing to this thread. It is very educational and I am going to print it out and save for my scraping project. I do have the "scraping bible" and I have read it a few times, but you guys added nice pictures and clarifications. This board gave me ideas for many of my projects. I have been here for quite a few years mainly as a reader, but when I can I contribute to the board as well. Keep up the good work!

[J Tiers]

.......but its a lot easier with the alignment device's level, built in or otherwise, as per the 3rd photo in post 18 or the 2nd photo in Jerry's post 19, which I don't think it is his photo else he'd be clearer on all this.

That's the example on finding "wind", or non-parallel of the two ways , which I think I mentioned at least twice as an obvious advantage. No it is clearly not my photo, as the K-W in the pic has the levels, and I have said mine does not. I do that with blocks and a separate level, but it's the same measurement.

Is that itself the main big advantage you mean? I figured you had lots more, because that is generally not a big deal for the things I work on. The flat ways allowed the straightedge to be put across in "X" fashion, which shows up twist ("wind") of the ways in a somewhat rougher way, by the "print". It also allows a level to be put across on a parallel to show the same thing. Yes, the K-W would also show it.

The 2.5" distance you describe gives a magnification factor of nearly 5, but with wider ways the magnification is smaller, and becomes actually a "reduction" as the ways exceed a foot, so that the level is reduced in ability to show the "wind" or "twist", since each increment of error produces a smaller error in angle. At a foot, the level is at its normal sensitivity, and at 18" (perhaps a small planer) it is reduced to 2/3 normal. (same ANGULAR sensitivity, but less absolute accuracy as far as actual distance error, how far one way is off in tenths etc.)

With double-V (kinematically over-constrained) ways, the K-W tool just is not suited to the task, levels or no levels, and double V-blocks and a parallel across is needed to do that same thing. I refuse to own an S-B lathe because of the old-fashioned double V. If I had a planer, then I'd have to out up with it, but I do not.

I might mention one other pitfall with the use of a Starrett level across on narrow supports as you show. The middle portion of the level is relieved a bit by the factory to allow the level to sit reliably flat. The relief is not specified, IIRC, and has no profile given and no other specs, so you have no idea what is going on, it is "undefined". The total is "usually" (again IIRC) a couple tenths somewhere in that space between the ends. I had a Starrett document somewhere from a previous discussion that gives it, but don't have it to hand now.

Anyway, it's all OK if you adjust and leave the level in one position. But if you remove and replace it for any reason, with a narrow support, you have to re-level it at the "standard " location on the bed, to make sure you get the same reading as before. Just something to remember. If you have the level supported on the ends, you should be able to remove and replace it no issue.

[epicfail48]

Mikey, if you wanna PM me your email address i can sent you the Machine Tool Reconditioning textbook, its supposed to be the bible for this sort of work

[Mcgyver]

Thanks Mcgyver, I really appreciate your help. Just one more thing - how do you spot the V for scraping? Are you using the straight edge and do one side at a time?

yes, with just the straight edge you do one side of the V then the other. If you must have it to a specific angle, you use a gauge (home made) to check. Then you do the flat way. The trick is to get the imaginary line that would be the apex of the V parallel to the flat - so the V and flat aren't going up and downhill relative to one another. You can do so with a surface gauge and perfectly fitting V block as per the first photo in post 28 (you do want an exact V angle if using a V block, you don't if using the alignment tool) by doing that check along the way.....but its a lot easier with the alignment device's level, built in or otherwise, as per the 3rd photo in post 18 or the 2nd photo in Jerry's post 19, which I don't think it is his photo else he'd be clearer on all this.

as for accuracy, it will be a function of the bed width to some degree, but for a small lathe like in the photo, parallelism is readily obtainable to a tenth of a thou as the level is .0005"/foot but the centre to centre of the ways is only 2.5". Works out that each graduation represents tenth over the 2.5 inches. Even with a much larger bed I don't think would be to difficult to get within a tenth given the graduation spacing. For example the bed shown above has a variance of a fraction of a graduation end to end. On a giant lathe that was say a foot between, I think you'd still get within tenth....and of course the bigger spacing the more dicey the surface gauge approach becomes.

Photo below is a bigger lathe, 5200 lb DSG, the ways are still only 6.5" apart so easy to interpolate to a tenth. I did not scrape that lathe, hardened bed....can be done but not by this guy. To explain the photo, its shown riding on the outer ways just do demonstrate the capacity.....you'd normally use only the inside ways which are 6.5" centre to centre. (normally in the sense that it is most common for the inside ways to be the long ways which are done first)

Once you've got the long ways you switch from using the level to an indicator for the 3 surfaces of the other 2 ways. Both are very useful aspects of the tools use.

[mikey553]

Here'a one way, a level on the alignment tool is far easier and likely more accurate. (I do not get the claims you don't the level for this tool, its fundamental to its use: scraping ways into alignment). This photo was before I made the alignment tool. The V block was scraped and is a perfect fit to the V way, and also perfectly square to the V - done as a matched pair for use in photo 2 on the double V way. The complete reconditioning of this grinder was part of the series article on scraping. None of its rocket science, but you do have to puzzle through a sequence and plan to get everything square and aligned, base, saddle table (and pivoting table on this as its a T&CG)

Thanks Mcgyver, I really appreciate your help. Just one more thing - how do you spot the V for scraping? Are you using the straight edge and do one side at a time?

[J Tiers]

Note to self, never say "level" in a machine shop without being prepared to duck a flying collet wrench...

Nah, it's all good.

I probably misunderstood McGyver anyhow, and thought he was getting all "purist" when he just has a different approach.

With or without the levels (which I agree ARE a useful addition), the basic King-Way tool is one I'd not want to do without when scraping for alignment. They are not hard to make, and so one is likely a good plan for many folks who want to get serious about scraping in a machine.

If you want the levels, they probably will make certain things easier, so the addition is not a bad idea.

You want trouble, post about how "Rollie's Dad's method" is all you will ever need to set up a lathe.

[epicfail48]

Note to self, never say "level" in a machine shop without being prepared to duck a flying collet wrench...

[J Tiers]

I've got a really good idea for you

Instead of being coy about it, and giving me the third degree questioning me as to what I have or have not scraped, to see if any of them "qualifies", why don't you just STATE WHAT YOU USE THE LEVEL POSITIONS FOR....... WHAT EXACTLY they do for you in your work that makes it super-important to have them.

I have stated what I use the K-W for where I do NOT need the level, and in fact it appears that the level would not be very helpful. I can give photographic examples, from posts in the past, or current work.And, it's not BS, that is likely 90% of what I would use the K-W for. I have scraped lathes and lathe parts, mills and mill parts, reference tools, etc, and I plan to scrape an old T&C grinder table and base assy after I get done with other things that are in-line ahead of it.

Just to mention some of the uses again: Getting the knee, saddle, and table aligned to the column ways of a mill, getting the ways aligned with the table slots on a mill, getting cross-slide ways aligned vs a lathe spindle. Getting the opposite side of a dovetail way aligned with the first side, etc, etc. All cases of a relation between one scraped surface and another. Obviously, some of those uses require an auxiliary tool, such as a precision square or the like.

These are all things that are probably illustrated in Connoley, and I could, if I dug out the book, and took some time, probably show practical examples of the KIng-Way tool being used to do all of the things that his illustrations show being done with the multitude of specialized "frames" he shows.

There are things that a level on it would be useful for (I gave some examples), and I have not put one on it, SO, I have to do those things a different way. I still use a level, it just happens not to be on the K-W tool.

REMEMBER, I have NOT said that levels on the thing would be useless, or any such stupid thing. I have said all along that there is a place for them, and such a facility would be useful in at least certain cases. And that inability to put levels on the K-W is not a reason to avoid building or getting one unless you can have levels on it.

Hey, if you buy or build a K-W and have levels on it, great, knock your socks off. I have not had the need strongly enough to build one with levels or spots for them, and you have, which is fine.

I am interested in the things that you find the levels so essential for, and that make up what is apparently the majority of what you do with the King-Way instrument. Maybe each of us can learn something, and give ideas to others.

[Mcgyver]

Flat lathe bed. Simple alignment, lay the level across and see iff you are parallel. Samey-same actual measurement as having the cross-level on the K-W, although having that makes it easier if the bed has a V and Flat, just due to the two contact portions of the instrument fitting the ways in a convenient manner. And it is easier to check many intermediate points, as when doing a "machine survey".

I understand using a level, both for the usual and usefulness in scraping, and have used them. I do not have one as part of the K-W, no. I may someday put levels on it, that's an open question. But since I already have regular levels, and the K-W copy I made is too small for them, I'd need to make the vial holders similar to the way the K-W types are made.
.
One benefit of the K-W level is that you do not need to have the WORK precisely level, you can set the level to BE level at the slope of the work, and it then indicates the surface just as well as if the surface was level. Similar to "level" on a ship when it is on the building ways.... it's off-level, but if the level is corrected to show level at the inclination of the ways, then items can be aligned as to parallel and "level" vs the keel correctly.

And, of course, it is easier to check for "wind", or twist , where the surfaces are at a slight angle off true parallel. Not impossible, as your surface gage and indicator picture shows.

Look, we are agreeing on the use of levels.

The disagreement here is that I simply find the K-W is still extremely useful without them actually integrated into the structure. It seemed that you were incredulous that anyone could find the KW useful without them.

That does not seem to be what you were saying, so that's fine.

I find the levels useful separate from the K-W. As I say, I may put some vials on it, just because there ARE some things that would be a bit more convenient, particularly detecting "wind".

You've probably done more scraping in terms of number of complete machines. I have been actually paid to do it.... We each do things the way we like. Not that it matters a hoot either way.

I just don't want someone to get the idea that they HAVE TO have levels on their KW clone or it is not worth having. It's very very useful even without them, especially with regard to actual alignment of different parts of the machine to each other.

What I disagree with is the BS. That 90% of the use is without the level or that I said it was worthless without the level. I'm fine with being wrong or having someone disagree.....just tone down the hyperbole and maybe ask instead of pronounce occasionally and we're fine.

It doesn't matter the quantity scraped, it matters if you scraped something where real value of this tool becomes apparent. imo that's a requirement before one is in any position to judge what percentage of its value the level function represents. What was it that was a meter long?

[J Tiers]

Flat lathe bed. Simple alignment, lay the level across and see iff you are parallel. Samey-same actual measurement as having the cross-level on the K-W, although having that makes it easier if the bed has a V and Flat, just due to the two contact portions of the instrument fitting the ways in a convenient manner. And it is easier to check many intermediate points, as when doing a "machine survey".

I understand using a level, both for the usual and usefulness in scraping, and have used them. I do not have one as part of the K-W, no. I may someday put levels on it, that's an open question. But since I already have regular levels, and the K-W copy I made is too small for them, I'd need to make the vial holders similar to the way the K-W types are made.
.
One benefit of the K-W level is that you do not need to have the WORK precisely level, you can set the level to BE level at the slope of the work, and it then indicates the surface just as well as if the surface was level. Similar to "level" on a ship when it is on the building ways.... it's off-level, but if the level is corrected to show level at the inclination of the ways, then items can be aligned as to parallel and "level" vs the keel correctly.

And, of course, it is easier to check for "wind", or twist , where the surfaces are at a slight angle off true parallel. Not impossible, as your surface gage and indicator picture shows.

Look, we are agreeing on the use of levels.

The disagreement here is that I simply find the K-W is still extremely useful without them actually integrated into the structure. It seemed that you were incredulous that anyone could find the KW useful without them.

That does not seem to be what you were saying, so that's fine.

I find the levels useful separate from the K-W. As I say, I may put some vials on it, just because there ARE some things that would be a bit more convenient, particularly detecting "wind".

You've probably done more scraping in terms of number of complete machines. I have been actually paid to do it.... We each do things the way we like. Not that it matters a hoot either way.

I just don't want someone to get the idea that they HAVE TO have levels on their KW clone or it is not worth having. It's very very useful even without them, especially with regard to actual alignment of different parts of the machine to each other.

[Mcgyver]

The level can get you close, probably NOT as close as a good straightedge, or there would be no long straightedges, but there are.

That that says you don't you understand application of the level. It in no way replaces the straightedge.

What did you scrape that was a meter long? Trying to understand what geometry requirements you had to scrape to

"Worthless without the level", is of course what it appeared you meant.

Once you understand its use, the level is its greatest value, but hardly its only. It is not understandable how you could assume I thought it worthless when I've shown several shots of doing things without a level. For that matter, I don't think I've seen a Kingway without a level or imagine one who knew how to use it would be satisfied with it in such a hobbled state.

[J Tiers]

Why would you say that? How you could you get to me seeming to claiming the Kingway design is worthless? ........

Have you scraped any lathe beds or long sets of ways? Its maybe the most useful functions of the Kingway which mine provides using an external level. Being able to move a level along is its trick. From someone who's done it (using the level to keep say a longish V and flat parallel), that really is the tool's value.

"Worthless without the level", is of course what it appeared you meant.

A level does not have to be on the Kingway, and that may not be the best place if you have just one level. Use for measuring "level-derived-angle" to check flatness or parallel is not the major value of the King-Way to me. (see last bit below)

And, yes I have, if you want to count a bit over a metre long. My straightedge was somewhat short (775mm), but I stepped it over. I expect that with the reasonably long straightedge it was more accurate than a much shorter level base in many more steps. Then a level set crosswise at the ends to check for "wind". I have a good Taylor Hobson level, it just does not go on the King-Way

The level can get you close, probably NOT as close as a good straightedge, or there would be no long straightedges, but there are.

If you want to get accuracy on longer beds, where your straightedge is too much shorter than the bed, the level is the secondary or "make-do" way, the right way is the autocollimator Yes YOU can DO it with the level, but a good straightedge, or best, an auto-collimator is the "right" way to get accuracy.

The level tells you which way the surface is going, where the level is. it does not tell you how far the surface HAS gone. You have to add up all the intermediate steps to get that. OR, if you HAVE a long straight surface, and the level says it is level along the length (within the error of the level) then you can be fairly sure you are level overall, at least to a few tenths per foot, or perhaps a thou or two over the length of a bed. Then if another surface gives the same result, you can feel reasonably sure they are parallel within a few thou error, depending on length.

The autocollimator also does not give you a distance, but it tells you the angle over the distance, which comes to the same thing. If you had a level that spanned the entire length, that would be a similar measurement to the autocollimator, although not as sensitive. The autocollimator is good to an arc-second or so if it is an old visual type, which is what hobby types might potentially have. The level is calibrated to 8 arc seconds or so, and estimated to a bit closer. (I know people around here who have such things, Talyvels, autocollimators, etc.)

To me, the value of the King-Way is in the way it can be set up to measure the RELATION of one surface to another which it needs to be parallel perpendicular, etc to. In the Connoley book, there are a large number of "measurement setups" using various specialty frames and indicators. The King-Way can duplicate most any of those quickly and easily.