What's new
What's new

Scraping saddle on universal mill— t-slot alignment

Brandenberger

Hot Rolled
Joined
Sep 19, 2017
Hi,
I’m (still) working on a VN22LU universal mill.

I had about .007” wear on the saddle ways mating the table,
Mostly as you would expect at the ends and less in the center.

There is also about .0015” wear on the flat undersides of the table.

I judged the overall wear on the saddle based on a small patch of
Scraping that was still visible.

I milled and then have scraped the saddle to .0005” or so.

My question is:

In the connelly book, they discuss the importance of maintaining
Alignment of the universal table pivot and the center t-slot. Indeed
Taking .007” off the saddle will move the center t-slot rearward by
By a few thou.

I could build up the saddle again, but with turcite I would want enough
Thickness for oil grooves cut in but not through it, so that would mean a lot more Metal removal.

A very reasonable question being “will I ever need to cut helical
Features which require t-slot centering?”

Maybe or maybe not, hard to say. I will also need to scrape the knee
But can build up the mating surface to offset. So I think the
T-slot center error would only come from saddle wear.

Picture of saddle prior to cleaning and scraping.

D868F3E5-2AF0-4B72-9F16-0D59E13F2503.jpg
 
I would put alignment of a T-slot with a universal mill table pivot point very, very low in my priority list. You're never going to use your universal mill table like a rotary table, and for any given fixed table rotation the positioning of the T-slots is irrelevant. You can cut helical grooves equally well with a universal dividing head bolted to the front, back or center T-slots.
 
Moreover, nobody would prevent you from widening the central T-slot in the opposite direction, to bring it back to center, if you really want.

Paolo
 
By allignment didnt they mean parralell to the movement of the X ??
That would be relevant

Peter


Yes, I mean the center slot would be parallel to
But offset to the rear of the table X movement.

I believe this would introduce some error in
Features milled with the table set at an angle
And for instance a DH and tailstock keyed into
The center slot.

Of course the table and/or T-slot keys could be modified
To bring the table center back into alignment with the
Pivot.

Phil
 
If it was my machine, I'd "match" the dividing head to my table. IMO, they SHOULD be part of the machine anyway, not an accessory to be interchanged with other machines or put on a shelf and sold separately by a thoughtless auction company(:angry::wrong:).

You could cut alignment keys to offset the head and tail-stock by as much as the table shifted over when it was re-fitted.
 
Yes, I mean the center slot would be parallel to
But offset to the rear of the table X movement.

I believe this would introduce some error in
Features milled with the table set at an angle
And for instance a DH and tailstock keyed into
The center slot.

Of course the table and/or T-slot keys could be modified
To bring the table center back into alignment with the
Pivot.

Phil


Perhaps you can describe the operation and setup in which the offset T-slot would introduce a problem. Looking at the actual geometry might help sort it out.

For an example indirectly related to this question. I bought one of the cheap 5C indexers some years ago. The two parallel edges of the base were in the as-cast condition only painted. Since I often just flop it in the vise, I wanted them parallel to the center and to each other. One parallel to center so it registered on the fixed jaw and thus centerline of the fixture also parallel to the X-axis of table motion. The opposite flat didn't matter except for secure clamping. UNLESS I reversed it in the vise for some setups meaning the opposite base edge must now register parallel.

Okay, that means both edges parallel to center for flexibility of use. Is it also important that they are both EQUALLY SPACED from the centerline? In most cases it's irrelevant. With which ever direction I've got it set, I find the center of the workpiece and do the job. One minor advantage to having equally spaced edges would be if I measured and noted exactly the distance from edge to center so I never had to actually indicate the workpiece if I had the Y-axis zeroed at the fixed jaw and could just calculate any movements from there.

The only other advantage to equal spacing I can imagine would be if some job required the indexer to be reversed in the vise during a single job and i wanted assurance the the Y-location was exactly the same when reversed. Obviously X direction is in the wind and will have to be reset.

So, what might be the conditions for setup where shifting the center of rotation off the center of the T-slot would create a problem?
 
In the test report of the milling machine is there a tolerance given for the maximum offset ??
I have never noticed that it was measured in a test report in any milling machine

Peter
 
By allignment didnt they mean parralell to the movement of the X ??
That would be relevant

Peter


Well, it would be irrelevant on a universal mill, since you orient the X axis. Therefore It isn't anything intrinsic to the machine.

Similarly to TGTool, I have hard time imagining any operation where you would rotate the table in the middle of the machining.

Paolo
 
I believe this would introduce some error in
Features milled with the table set at an angle
And for instance a DH and tailstock keyed into
The center slot.
Please find an example of this, as I don't believe such an offset is relevant in any way. I am willing to be convinced that I've overlooked something, but really don't see how the positioning of the T-slots relative to the universal table pivot point affects anything you might want to do assuming that you are not attempting to use your universal table as a rotary table to mill circular arcs. And that is not something I would ever attempt to do.
 
Connelly Section 27.75 gives a brief description:

We will now take up the subject of offset of pivot axis as promised in an earlier paragraph. The situation briefly is this:

The Lower section and the upper section of the universal saddle are centrally connected by a pivot. The upper section swivels horizontally about the axis of this pivot. The pivot may take the form of a bushing through which a shaft projects. The shaft extends from the knee member and transmits motion to power the table member with automatic feed.

Helical milling is possible on a machine equipped with a universal saddle. However, for the milling cutter to generate a true and undistorted profile, it is essential that:

1. The axis of the spindle
2. The axis of the pivot
3. The center T-slot of the table

be in line with respect to each other within close limits. ...

Only a small bilateral tolerance is permissable in this alignment, if undue distortion is to be avoided. Asa a consequence, rescraping and alignment of a universal saddle presents a particularly difficult problem to the scraper.​

He goes on to say some machinery rebuilders will just skip this step, assuming the impact to helical milling won't be too bad.

In section 27.81 he indicates the permissible tolerance for offset of center T-slot to the axis of pivot is .002".

Having taken .007" off my saddle height, that means about a .007" rearward movement of
the center t-slot (these are 50degree dovetails, but for simplicity assume they're 45degree).

It seems this is not just relevant to if one were to try to pivot the table during an
operation (which does not seem to be useful with any precision).
 
I'm sorry, but I still cannot understand why.
For helical milling you that the projections of the axis of the spindle and the axis of the dividing head intersect exactly at the cutter.

All this must be precisely adjusted by moving the table on the Y axis.

Of course, if the axis of the dividing head does not pass through the pivot point of the universal saddle, you need to crank the Y screw each time you change the angle of the X axis. But, you would need to do it anyhow each time you change the cutter.

I still think that Conelly's alignment constraints could be potentially good shortcuts during the setup in an ideal world, but are completely irrelevant for (almost) anybody.

Paolo
 
I'm sorry, but I still cannot understand why.
For helical milling you that the projections of the axis of the spindle and the axis of the dividing head intersect exactly at the cutter.

All this must be precisely adjusted by moving the table on the Y axis.

Of course, if the axis of the dividing head does not pass through the pivot point of the universal saddle, you need to crank the Y screw each time you change the angle of the X axis. But, you would need to do it anyhow each time you change the cutter.

I still think that Conelly's alignment constraints could be potentially good shortcuts during the setup in an ideal world, but are completely irrelevant for (almost) anybody.

Paolo


Perhaps that is the conclusion then? That if the pivot points are aligned, I can center the cutter
with the table at 90degrees, and then I could swing the angle arbitrarily and the cutter would
remain on centerline of the t-slot.

However, if my table is shifted backward on the saddle, I have an angle-dependent correction to bring the center t-slot back to the centerline of the cutter, and precisely locating that is more involved.
 
I am going to stick my neck out and say Connelly (who was a compiler/editor rather than a practicing machine rebuilder) did not know what he was talking about in this situation. It may certainly be true that some machine tool makers put a specification on alignment of central T-slot and table pivot for their particular machines. But this is not in any way critical or essential to the proper operation of the machine.

I'm going to repeat what I said way up at the top of this thread. You can bolt your universal dividing head on to the front, back or central T-slot and it makes no difference whatsoever to cutting proper, well-formed helical grooves, provided your dividing head gearbox/banjo will reach the added distance. If inches of offset don't matter, neither do thousandths.

However, if my table is shifted backward on the saddle, I have an angle-dependent correction to bring the center t-slot back to the centerline of the cutter, and precisely locating that is more involved.
Please pardon my bluntness, as I don't want to be offensive. However, I do want to be quite definite. If you are not locating your cutter to your workpiece after you have rotated the table, you are "doing it wrong". There are just too many possible sources of setup offset, the T-slot to pivot alignment being just one, to assume that the workpiece is going to rotate directly over the table pivot.
 
Connelly Section 27.75 gives a brief description:

We will now take up the subject of offset of pivot axis as promised in an earlier paragraph. The situation briefly is this:

The Lower section and the upper section of the universal saddle are centrally connected by a pivot. The upper section swivels horizontally about the axis of this pivot. The pivot may take the form of a bushing through which a shaft projects. The shaft extends from the knee member and transmits motion to power the table member with automatic feed.

Helical milling is possible on a machine equipped with a universal saddle. However, for the milling cutter to generate a true and undistorted profile, it is essential that:

1. The axis of the spindle
2. The axis of the pivot
3. The center T-slot of the table

be in line with respect to each other within close limits. ...

Only a small bilateral tolerance is permissable in this alignment, if undue distortion is to be avoided. Asa a consequence, rescraping and alignment of a universal saddle presents a particularly difficult problem to the scraper.​

He goes on to say some machinery rebuilders will just skip this step, assuming the impact to helical milling won't be too bad.

In section 27.81 he indicates the permissible tolerance for offset of center T-slot to the axis of pivot is .002".

Having taken .007" off my saddle height, that means about a .007" rearward movement of
the center t-slot (these are 50degree dovetails, but for simplicity assume they're 45degree).

It seems this is not just relevant to if one were to try to pivot the table during an
operation (which does not seem to be useful with any precision).

Connelly was editing what he had been told.

He's missed the actuality of the geometry.

The pivot angle having been set, it is whatever it is.

The table traverse will be on that angle. The pivot-point no longer matters.

EX: A USMT Quartet makes these cuts without any universal table at all.

The Horizontal spindle is mounted in a turret capable of 360-degree rotation - mindful of the power cords to the VERTICAL spindle's motor sitting mid-ram atop the same turret. Of course.

Downside? AS the turret is rotated, it gets further out of optimal position relative to the entire knee. One will be working "biased" toward one end of the table as well as biased to the side closest to the turret.

BFD.

For the modest angles involved? It can still get the job done even with reduced useful X & Y travel.

Page Two

A,N, Other approach to lack of a universal tabe is to perch a linear slide atop the conventional tabe, set it to the desired angle with clamps and tedious measure, do all traverse on that angle, "primary" table locked.

Not theory. I have a salvaged line-shaft B&S Number 1 Universal miller table with hardly a nibble-mark anywhere on it. It will fit atop the Quartet's table, and "not only". Bodine or Bison DC gearmotors can provide the powered traverse.
 
I am going to stick my neck out and say Connelly (who was a compiler/editor rather than a practicing machine rebuilder) did not know what he was talking about in this situation. It may certainly be true that some machine tool makers put a specification on alignment of central T-slot and table pivot for their particular machines. But this is not in any way critical or essential to the proper operation of the machine.

In other words, Connelly made a very useful contribution by compiling the book, but he wasn't God.

Or the Arab proverb, "Trust in God, but tie your camel to a tree." Read, listen, but think it through and digest it before taking someone else's words as gospel.
 
In other words, Connelly made a very useful contribution by compiling the book, but he wasn't God.

Or the Arab proverb, "Trust in God, but tie your camel to a tree." Read, listen, but think it through and digest it before taking someone else's words as gospel.

Thanks all. I was concerned that I was missing something in not
correcting the alignment. This discussion has helped to understand.

I went back to the Schlesinger book and it too mentions these same tests,
but with no discussion of why this alignment is considered important.

Phil
 
I can see instances where that alignment would be important. But simply knowing how much it is off would be sufficient. For the table t slot alignment to be considered important, it would also be important for the spindle centerline to be exactly over the center of table rotation as well.

In times of flintlocks it was "trust in God and keep your powder dry".
 
I can see instances where that alignment would be important. But simply knowing how much it is off would be sufficient. For the table t slot alignment to be considered important, it would also be important for the spindle centerline to be exactly over the center of table rotation as well.

Umhhh.. you don't have traverse in X & Y? Hard to utilize a rotab without that.

If Tee slots otherwise matter, just sacrifice some "Z". Drop an inexpensive shiney-wood extrusion atop and clamp rather than alter locating keys on a DH and its centre.
 








 
Back
Top