Scraper People question

[doctor demo]

I have a Leblond that I acquired several years ago with a few minor problems.
One of the problems until recently was no tail stock, now one of the problems is the tail stock.
I bought one that is in need of a great deal of attention. Re scraping is a must, the quill fits well but points down hill towards the head stock .008 in 4''. The recommended standard by Connelly is up hill .0008 sooo it's out nearly .009.
I've separated the tail stock base from the top and need to scrape the V and flat ways but also have to scrape the top of the base flat first.

Here's the question: Short of chipping a hole in My surface plate , how do I check/blue the top of the base? It has a projection for the transverse adjustment and alignment.

Steve

[Mcgyver]

Here's the question: Short of chipping a hole in My surface plate , how do I check/blue the top of the base? It has a projection for the transverse adjustment and alignment.

Steve

scrape its mate. if one part has a proud section, its mate must be recessive; scrape it first and use it to scrape the base

The top has to fit the base so this op gets done regardless....however I don't think I'd scrape the base to fit the lathe. You could end up with the base perfect to the bed but the barrel low...then you will have to rebore, chrome and grind the barrel (or make a new one) to get things back where they should be

IF the barrel really does fit well, you want to take advantage of that. It might be worthwhile machining a little of the bottom of the base, turciting it, then scraping the turcite to be both a perfect fit to the base AND have the barrel aligned. That's work that is done carefully, when trying to align a couple of things lots of checking and thinking goes into where to scrape

[J Tiers]

You use a pair of parallels that have been scraped-in to be accurate. They "extend" the surface plate upwards.

Look further in Connoley, it's all covered in there. *Everything* (very nearly) is covered in there somewhere.

You can do a lot worse than simply shimming up the front between the base and top casting of the T/S. Done fast, and has little negative impact. Very "un-doable" if that is ever wanted, and since the T/S is not moved that much, there isn't really a reason not to do it.

[doctor demo]

scrape its mate.

The top has to fit the base so this op gets done regardless....however I don't think I'd scrape the base to fit the lathe. You could end up with the base perfect to the bed but the barrel low...then you will have to rebore, chrome and grind the barrel (or make a new one) to get things back where they should be

IF the barrel really does fit well, you want to take advantage of that. It might be worthwhile machining a little of the bottom of the base

The tail stock top is flat, but the transverse way is so snug that using it to blue the base is not possible.
The tail stock base is in poor shape, it looks as if it was dragged behind a truck on a gravel road.
My plan is to scrape the base top flat then scrape the V and flat way paralell to the top. Then I can add a shim between the base and top of the tail stock to make up for how much wear (and scraping) was done to get center height back.
I don't want to shim between a good flat surface and a bad one, that is why I asked My original question .
If I use J.T.'s method of two scraped paralells I'll have to make them I don't have any large enough to blue that size surface. I would rather not do that if there was a different (easier) way.

Steve

[MuellerNick]

The tail stock top is flat, but the transverse way is so snug that using it to blue the base is not possible.

I don't believe that!
Even if you can't move it back and forth, you can insert it. That already transfers blue. Not very much, but enough to get some reading.
NOW! Scrape the base's top flat with some small $whatever (prism, camelback, small surface plate) and check from time to time that you don't generate two randomly positioned planes.

A tailstock is a bitch to scrape in. You have 3 geometric constraints to meet (at the same time, btw.).
Height
parallel to the bed in the horizontal plane
parallel to the bed in the vertical plane


If you just scrape the top and bottom part for contact, you will end in having to scrape the bottom part (with the Vee) to 3 constraints. DON'T DO THAT! Or you will go insane!

Scrape the surface between top and bottom for contact. AND NOTHING MORE!

Scrape the bottom part to the bed and correct for parallelism to the bed*) in the horizontal plane.

Now scrape the upper part (that's the one you can put on your surface plate, right?) for height and parallelism in the vertical plane.


*)
Not really parallel, the left end of the TS has to point to the operator 0...0.01mm/100mm, and point upwards 0 ... 0.01mm/100mm Numbers out of my head.

Nick

[rkepler]

The tail stock top is flat, but the transverse way is so snug that using it to blue the base is not possible.

If it's that snug it's too tight. Clean it up to a nice sliding fit then scrape the top to use as a master on the base. After the base top is scraped flat level the lathe bed then use levels on the top of the base casting to scrape the base to the lathe bed - first for level, then for fit (or if you're good - both at the same time).

My plan is to scrape the base top flat then scrape the V and flat way paralell to the top. Then I can add a shim between the base and top of the tail stock to make up for how much wear (and scraping) was done to get center height back.

You can do that, but in the end you'll have to scrape it to the lathe bed anyway so you may as well have all the alignment tooling set for both operations, otherwise you'll be switching from checking parallelism to checking bed fit back to parallel...

I don't want to shim between a good flat surface and a bad one, that is why I asked My original question .

There's not a whole lotta choice there. You flatten the top of the base casting to mate with the top casting, level the top of the base casting to the lathe bed (while watching to make sure that the taper of the quill of the tailstock is coming in level at the same time - it sounds like you can check that one in advance). After the base is fitted to the bed and the quill is parallel to the bed you can then measure for the shim.

Don't forget: depending on how the offset is managed you may have to oblong some boltholes on the side of the tailstock. You can usually deal with a few thou drop but past 6-8 you have to mill out some relief. Your droop of .008 may mean that the front end of your base is worn .002, so you could be OK (on the "glass half empty" side the whole base could be down .010 and the front .012).

Remember that the taper in the quill may not be parallel to the outside of the quill - about the only way to check this is with a parallel test bar with the appropriate taper.

[MuellerNick]

... then use levels on the top of the base casting to scrape the base to the lathe bed - first for level ...

That is just a waste of time, nothing else.


Nick

[rkepler]

That is just a waste of time, nothing else.

Why do you say that? You need to scrape the tailstock base top so that the quill doesn't rise when extended or moved to offset the point, so it needs to be in the same plane as the lathe bed. Comparing planes is easiest done with levels. You can check it by remounting the top casting and checking the parallel test bar, but on many lathes it's simply faster and easier to check with levels, and you usually get a better baseline for the measurement.

Perhaps you have a point that I'm missing? If so state it - but making blanket statements as you did is the real waste of time here.

[MuellerNick]

You need to scrape the tailstock base top so that the quill doesn't rise when extended ...

No matter how the separating plane is, it doesn't have an influence. The QUILL has to be parallel.

...or moved to offset the point,

Assuming that the separating plane isn't completely off, it doesn't play a role. If it is off by 1°, the influence still isn't worth the effort. If it would be off 1° and you offset the tailstock 10 mm, it would rise/fall by 0.17 mm.
You can assume that it can be left as milled. You can check with the level, but 1 mm / m would be good enough.

so it needs to be in the same plane as the lathe bed.

Wrong. Not parallel to the bed, but parallel to the cross-slide's way.


Nick

[rkepler]

Assuming that the separating plane isn't completely off, it doesn't play a role. If it is off by 1°, the influence still isn't worth the effort. If it would be off 1° and you offset the tailstock 10 mm, it would rise/fall by 0.17 mm.

You can assume that it can be left as milled. You can check with the level, but 1 mm / m would be good enough.

At this point the top casting was scraped and used as a master for the base casting, so the base top is already scraped and flat. We're going to scrape the bottom of the base to the lathe bed, and we need it level fore and aft - at that point it's not a lot of work to make it level side to side. You're going to have to make sure that you don't drop one side too much, if you're checking you may as well make it right. If the top casting was scraped so the quill travel is parallel to it's base all you have to worry about at this point is that the quill travel doesn't shift fore & aft after mounting.

Wrong. Not parallel to the bed, but parallel to the cross-slide's way.

If the cross slide ways aren't coplanar with the bed there's going to be some real problems in other areas (compound travel), so I think we're saying the same thing in different ways.

[MuellerNick]

If the cross slide ways aren't coplanar with the bed there's going to be some real problems in other areas (compound travel), so I think we're saying the same thing in different ways.

I hope, we are not saying the same!
The cross slide has absolutely not to be parallel (looking to tzhe back) of the bed. The bed is not a reverence for this axis. In fact, it is no reference for any axis of a lathe.
A lathe has two axis. X and Z
Z is through the center of the spindle, it is coaxial to the spindle.
X is perpendicular to Z. The physical X-axis (cross-slide) doesn't even cross the Z-axis. Only the mathematical X-axis does intersect Z. Only together with an adjusted tool height, the physical X-axis intersects Z.
There is absolutely no definition how Z is oriented. In most cases, it is horizontal or vertical (pointing upwards, but even sometimes pointing downward). X has only to be perpendicular to Z and that's it. It points to the back or to the front, depending on the machine. Or to the right on a vertical lathe. There is no definition about how it is rotated along Z.
Now to the bed:
The only thing the bed has to be is to be parallel to Z. End of constrains. It can be slanted (as often found on CNCs).
Only as soon as you introduce a physical X-axis (still it may point up/down/front/back/diagonal/whatever) it defines how X is rotated along Z. But that has absolutely no influence on the bed. The first and only definition of how X is rotated along Z is by the guide of X.

in other areas (compound travel)

The compound can be swiveled. If it is swiveled to be parallel to Z it has to be parallel toZ. 90° further, it has to be parallel to X. That has nothing to do with the bed. Absolutely nothing!


There are so many misunderstandings of geometry and how things interact that result in very weird tips.

If the top casting was scraped so the quill travel is parallel to it's base all you have to worry about at this point is that the quill travel doesn't shift fore & aft after mounting.

This only doubles work.



Nick

[rkepler]

I'll just refer folks to Connelly's _Machine Tool Reconditioning_.

[MuellerNick]

I'll just refer folks to Connelly's _Machine Tool Reconditioning_.

Why read when logic can give quicker and better answers?


Nick

[doctor demo]

I don't believe that!
Even if you can't move it back and forth, get some reading. That already transfers blue.
NOW! Scrape the base's top flat check from time to time that you don't generate two randomly positioned planes.
A tailstock is a bitch to scrape in. You have 3 geometric constraints to meet (at the same time, btw.).
Height
parallel to the bed in the horizontal plane
parallel to the bed in the vertical plane
If you just scrape the top and bottom part for contact, you will end in having to scrape the bottom part (with the Vee) to 3 constraints. DON'T DO THAT! Or you will go insane!

Scrape the surface between top and bottom for contact. AND NOTHING MORE!

Scrape the bottom part to the bed and correct for parallelism to the bed*) in the horizontal plane.

Now scrape the upper part (that's the one you can put on your surface plate, right?) for height and parallelism in the vertical plane.
*)
Not really parallel, the left end of the TS has to point to the operator 0...0.01mm/100mm, and point upwards 0 ... 0.01mm/100mm Numbers out of my head.
Nick

I read the tail stock section of Connelly's book a couple of times and it differs slightly with Your description, but I guess there is more than one way to get the same end result.
Now Your steps bring up another question. Let Me try to explain My question as I go, if I blue the t.s. top and transfer blue to the top of the base and scrape for contact only. Then scrape the base bottom to the bed, why would I then scrape the t.s. top to align with ''parallelism in the vertical plane'' instead of scraping the top of the tail stock base to do that?

[J Tiers]

For argument's sake here, I think Nick is correct.

But, I would argue that it is the same thing to some degree..... since both the ways and the X-slide are referenced to the spindle axis..... One is at right angles, the other should be parallel.

If either one is not so, then there is a problem. So in that manner, since the ways should be parallel, the X-slide travel should end up at right angles to a line down the ways.

I think his point is that you look not to the ways, but to the spindle axis in both cases.