Just added to the dropbox:

http://www.metalworking.com/DropBox/...ET_CENTERS.jpg
http://www.metalworking.com/DropBox/...ET_CENTERS.txt

---------------------------------

BEARING BALLS FOR OFFSET CENTERS


This is an idea that sprung to my mind a while ago, although I'm
sure I'm not the first one to think of it.

When offsetting the tailstock for taper turning, or using a special
tailstock fixture for the same purpose, the 60 degree center points
don't fit well in the centerholes of the work being taper turned.

This method needs custom-made lathe centers for both headstock and
tailstock.
The sharp point is turned off for a short distance, and centerdrilled
just as is done for the work being turned.

Hardened steel balls are captured in the centerholes between the lathe
centers and the work, at each end.

The correct centerhole size is important in relation to the bearing ball
diameter.

For a standard 60 degree centerdrill, the opening of the hole at the
ends should ideally be between 88% and 90% of the diameter of the ball.

If larger, there may not be enough clearance between the lathe center
and work to allow any offset.
If the hole's opening is smaller than 87% of the ball's diameter, only
the corner of the hole's opening will contact the ball and the whole
thing may come loose under heavy cutting pressure.

In practical experience, I've had very good results with this technique
while turning morse taper shanks.

For the purpose of accurately setting the tailstock setover, the
effective length of the workpiece is measured between the centers of
each ball.

Just mike the workpiece with the balls in place, and subtract the total
of one half the diameter of each ball.

Be sure to use your favorite tailstock center lube on that end
(I use white lithium grease).

Hope this is useful,

Ken Grunke
West Lima, WI
Jan. 09, 2005


--
take da "ma" offa dot com fer eemayl

In article , Ken Grunke says...

When offsetting the tailstock for taper turning, or using a special
tailstock fixture for the same purpose, the 60 degree center points
don't fit well in the centerholes of the work being taper turned.

Your approach is novel and I'm sure it works well.

However, did you ever wonder why this was not done, back
in the days when tailstock setover was a routine approach
to manufacturing tapered items?

Basically, even though the centers don't seem to fit
well, they still allow a true cylinder to be turned on
the part.

Those old-time folks really knew their stuff.

Jim


--
==================================================
please reply to:
JRR(zero) at pkmfgvm4 (dot) vnet (dot) ibm (dot) com
==================================================

"jim rozen" wrote in message
...

However, did you ever wonder why this was not done, back
in the days when tailstock setover was a routine approach
to manufacturing tapered items?

Basically, even though the centers don't seem to fit
well, they still allow a true cylinder to be turned on
the part.

Those old-time folks really knew their stuff.


At school, we use bell-style center drills for turning tapers with an offset
tailstock. Works very well.

Regards,

Robin

So what drives the part here? It looks like it's free to just stop turning
to me. Or is the dog/driver plate setup just left out of the sketch for
clarity?

GWE

Ken Grunke wrote:

Just added to the dropbox:

http://www.metalworking.com/DropBox/...ET_CENTERS.jpg
http://www.metalworking.com/DropBox/...ET_CENTERS.txt

---------------------------------

BEARING BALLS FOR OFFSET CENTERS


This is an idea that sprung to my mind a while ago, although I'm
sure I'm not the first one to think of it.

When offsetting the tailstock for taper turning, or using a special
tailstock fixture for the same purpose, the 60 degree center points
don't fit well in the centerholes of the work being taper turned.

This method needs custom-made lathe centers for both headstock and
tailstock.
The sharp point is turned off for a short distance, and centerdrilled
just as is done for the work being turned.

Hardened steel balls are captured in the centerholes between the lathe
centers and the work, at each end.

The correct centerhole size is important in relation to the bearing ball
diameter.

For a standard 60 degree centerdrill, the opening of the hole at the
ends should ideally be between 88% and 90% of the diameter of the ball.

If larger, there may not be enough clearance between the lathe center
and work to allow any offset.
If the hole's opening is smaller than 87% of the ball's diameter, only
the corner of the hole's opening will contact the ball and the whole
thing may come loose under heavy cutting pressure.

In practical experience, I've had very good results with this technique
while turning morse taper shanks.

For the purpose of accurately setting the tailstock setover, the
effective length of the workpiece is measured between the centers of
each ball.

Just mike the workpiece with the balls in place, and subtract the total
of one half the diameter of each ball.

Be sure to use your favorite tailstock center lube on that end
(I use white lithium grease).

Hope this is useful,

Ken Grunke
West Lima, WI
Jan. 09, 2005

jim rozen wrote:
In article , Ken Grunke says...


When offsetting the tailstock for taper turning, or using a special
tailstock fixture for the same purpose, the 60 degree center points
don't fit well in the centerholes of the work being taper turned.


Your approach is novel and I'm sure it works well.

However, did you ever wonder why this was not done, back
in the days when tailstock setover was a routine approach
to manufacturing tapered items?

Yup, I have wondered, but I suppose that in a production situation, they
got tired of dropping the balls in a pile of swarf never to be seen
again :-) My solution is to glue them in with sticky grease.


Basically, even though the centers don't seem to fit
well, they still allow a true cylinder to be turned on
the part.

Sure, although there are only two points of contact--one at the outside
edge of the hole, and then at the inside edge, where the 60 deg. cone
ends and straightens out to the pilot hole. I haven't done enough taper
turning to know--do those edges wear into the cone center after a while?

thanks,

Ken Grunke


--
take da "ma" offa dot com fer eemayl

Grant Erwin wrote:
So what drives the part here? It looks like it's free to just stop turning
to me. Or is the dog/driver plate setup just left out of the sketch for
clarity?

Yup--and to save time. Just a quick 3D CAD sketch!

Ken Grunke

--
take da "ma" offa dot com fer eemayl

Robin S. wrote:

At school, we use bell-style center drills for turning tapers with an offset
tailstock. Works very well.

Hmmm, never heard of or seen those. Judging by the name, they must cut a
curved-shaped cone?

Ken


--
take da "ma" offa dot com fer eemayl

"Ken Grunke" wrote in message
...
snip-----

Sure, although there are only two points of contact--one at the outside
edge of the hole, and then at the inside edge, where the 60 deg. cone
ends and straightens out to the pilot hole. I haven't done enough taper
turning to know--do those edges wear into the cone center after a while?

thanks,

Ken Grunke


Yes, they do, and in the process the shaft is constantly creating more and
more clearance between the centers as it cold flows to achieve the form. .
Look closely at centers that have been run offset to see how badly they are
deformed from the center drilled configuration. Further, if, when turning
between centers, if the face on either end of a part is not perfectly
square, it has the effect of creating an out-of-round (oval) turn. This
very concept has been the subject of endless debate, with almost no one in
agreement, but all it takes is a little experience in precision grinding to
put it directly into focus. Be certain to maintain right angles on the
ends of offset turned parts unless you don't mind oval turns.

Your ball turning is a very good concept, for it eliminates that problem,
but the same results can be achieved with center drills that are made with a
large radius in place of the 60° cone. DoAll is one of the makers, but
I'm sure there are others. They have a name, but it escapes me at the
moment.

Harold

In article , Ken Grunke says...


Basically, even though the centers don't seem to fit
well, they still allow a true cylinder to be turned on
the part.

Sure, although there are only two points of contact--one at the outside
edge of the hole, and then at the inside edge, where the 60 deg. cone
ends and straightens out to the pilot hole. I haven't done enough taper
turning to know--do those edges wear into the cone center after a while?

The contact area is larger than you might think in this case.
There was a thread on this a while ago, and I took some photos
of this. It really winds up being a line, and fairly large
contact patch:

http://www.metalworking.com/DropBox/_2001_retired_files/offcenters2.jpg

Unless one is really cranking on the tailstock, they don't wallow
out. And if one is really cranking on the tailstock, they'll
wallow out even on straight turning.

Jim


--
==================================================
please reply to:
JRR(zero) at pkmfgvm4 (dot) vnet (dot) ibm (dot) com
==================================================

In article , Harold & Susan Vordos says...

Yes, they do, and in the process the shaft is constantly creating more and
more clearance between the centers as it cold flows to achieve the form. .
Look closely at centers that have been run offset to see how badly they are
deformed from the center drilled configuration. Further, if, when turning
between centers, if the face on either end of a part is not perfectly
square, it has the effect of creating an out-of-round (oval) turn. This
very concept has been the subject of endless debate, with almost no one in
agreement, but all it takes is a little experience in precision grinding to
put it directly into focus. Be certain to maintain right angles on the
ends of offset turned parts unless you don't mind oval turns.

Your memory is incorrect in this regard. Under normal conditions,
the centers don't open up, and in the case I tested, the end of
the part *wasn't* square to the machine axis, and it *did* turn
a round, not oval piece. To remind folks of the tests that were
done:

http://www.metalworking.com/DropBox/_2001_retired_files/offcenters.txt
http://www.metalworking.com/DropBox/_2001_retired_files/offcenters1.jpg
http://www.metalworking.com/DropBox/_2001_retired_files/offcenters2.jpg
http://www.metalworking.com/DropBox/_2001_retired_files/offcenters3.jpg
http://www.metalworking.com/DropBox/_2001_retired_files/offcenters4.jpg

One of the regulars here at that time tested the roundness of the
turned part, it showed no systematic deviation from round to the
limit of the tallyrond tester.

Jim


--
==================================================
please reply to:
JRR(zero) at pkmfgvm4 (dot) vnet (dot) ibm (dot) com
==================================================

"jim rozen" wrote in message
...
In article , Harold & Susan Vordos says...

Yes, they do, and in the process the shaft is constantly creating more
and
more clearance between the centers as it cold flows to achieve the form.
..
Look closely at centers that have been run offset to see how badly they
are
deformed from the center drilled configuration. Further, if, when
turning
between centers, if the face on either end of a part is not perfectly
square, it has the effect of creating an out-of-round (oval) turn.
This
very concept has been the subject of endless debate, with almost no one
in
agreement, but all it takes is a little experience in precision grinding
to
put it directly into focus. Be certain to maintain right angles on the
ends of offset turned parts unless you don't mind oval turns.

Your memory is incorrect in this regard. Under normal conditions,
the centers don't open up, and in the case I tested, the end of
the part *wasn't* square to the machine axis, and it *did* turn
a round, not oval piece. To remind folks of the tests that were
done:

http://www.metalworking.com/DropBox/_2001_retired_files/offcenters.txt
http://www.metalworking.com/DropBox/_2001_retired_files/offcenters1.jpg
http://www.metalworking.com/DropBox/_2001_retired_files/offcenters2.jpg
http://www.metalworking.com/DropBox/_2001_retired_files/offcenters3.jpg
http://www.metalworking.com/DropBox/_2001_retired_files/offcenters4.jpg

One of the regulars here at that time tested the roundness of the
turned part, it showed no systematic deviation from round to the
limit of the tallyrond tester.

Jim

Chuckle! Or perhaps big belly laugh!!

Yep, I remember, and I commend you for the great pictures, but that's not
what we're talking about. My point is turning a taper with an offset
*tailstock* center, although it's possible I never made that clear in my
original argument. It's not the same thing. The degree of error in
drilling offset centers remains constant and there is no movement of the
part as it relates to the fixed, but *in line* centers in the test you
performed. The machine centers, in your specimen, would pick the high
spots and run there, likely not fully seated, but with enough area of
contact to perform without distorting. When you offset the tailstock,
everything changes. You didn't prove your point originally, I simply quit
talking about it because I had quit following RCM (sort of like not talking
to your family, I discovered).

Try that same test, this time offset the tailstock, and for purpose of
proving whether you're right, or I am, turn a much shorter piece, with a
large offset, so it's exaggerated. Be certain that the faces are not at
right angles to the center, which is a part of my argument. You'll not only
mush the centers, you'll detect an oval. Grinders (the machines, not the
operators) don't lie. By the way, you shouldn't need any special machine
to learn what I'm talking about. Simply measuring the part will disclose
the oval. It will be fairly obvious.

Harold

On Sun, 09 Jan 2005 18:27:48 -0600, Ken Grunke
wrote:

Just added to the dropbox:

http://www.metalworking.com/DropBox/...ET_CENTERS.jpg
http://www.metalworking.com/DropBox/...ET_CENTERS.txt

---------------------------------

BEARING BALLS FOR OFFSET CENTERS


This is an idea that sprung to my mind a while ago, although I'm
sure I'm not the first one to think of it.

When offsetting the tailstock for taper turning, or using a special
tailstock fixture for the same purpose, the 60 degree center points
don't fit well in the centerholes of the work being taper turned.

This method needs custom-made lathe centers for both headstock and
tailstock.
The sharp point is turned off for a short distance, and centerdrilled
just as is done for the work being turned.

Hardened steel balls are captured in the centerholes between the lathe
centers and the work, at each end.

The correct centerhole size is important in relation to the bearing ball
diameter.

For a standard 60 degree centerdrill, the opening of the hole at the
ends should ideally be between 88% and 90% of the diameter of the ball.

If larger, there may not be enough clearance between the lathe center
and work to allow any offset.
If the hole's opening is smaller than 87% of the ball's diameter, only
the corner of the hole's opening will contact the ball and the whole
thing may come loose under heavy cutting pressure.

In practical experience, I've had very good results with this technique
while turning morse taper shanks.

For the purpose of accurately setting the tailstock setover, the
effective length of the workpiece is measured between the centers of
each ball.

Just mike the workpiece with the balls in place, and subtract the total
of one half the diameter of each ball.

Be sure to use your favorite tailstock center lube on that end
(I use white lithium grease).

Hope this is useful,

Ken Grunke
West Lima, WI
Jan. 09, 2005


This is a pretty good technique. It's perhaps
worth emphasising that it removes the length uncertainty
that's always present when turning between centre points.
With centre points the bar pivots about a point a bit inside
the pivot hole so the effective length is always an
uncertain bit less than the overall bar length.

With balls this is not a problem. The measurement
of overall length with both balls in place less 1/2d +1/2d
(d=ball dia) precisely defines the effective length.

With this checked with a decent vernier and the
offset set by gauge blocks pretty precise tapers are
possible.

Jim

In article , Harold & Susan Vordos says...

Yep, I remember, and I commend you for the great pictures, but that's not
what we're talking about. My point is turning a taper with an offset
*tailstock* center, although it's possible I never made that clear in my
original argument. It's not the same thing. The degree of error in
drilling offset centers remains constant and there is no movement of the
part as it relates to the fixed, but *in line* centers in the test you
performed. The machine centers, in your specimen, would pick the high
spots and run there, likely not fully seated, but with enough area of
contact to perform without distorting. When you offset the tailstock,
everything changes. You didn't prove your point originally, I simply quit
talking about it because I had quit following RCM (sort of like not talking
to your family, I discovered).

That's not what it seemed to me. The centers were non-axial, and
the part *was* round. Case closed. You could argue that the test
wasn't extreme enough. I thought it proved the point.

Try that same test, this time offset the tailstock, and for purpose of
proving whether you're right, or I am, turn a much shorter piece, with a
large offset, so it's exaggerated. Be certain that the faces are not at
right angles to the center, which is a part of my argument. You'll not only
mush the centers, you'll detect an oval. Grinders (the machines, not the
operators) don't lie. By the way, you shouldn't need any special machine
to learn what I'm talking about. Simply measuring the part will disclose
the oval. It will be fairly obvious.

Well, the HLVH that was used for the other test can't offset the tailstock.
But my SB at home can. So the features should be:

1) the end of the stock should be cut at an angle

2) the centers should be drilled in the correct way, ie, coaxial

3) the tailstock should be well offset

4) the part should not be that long.

I could ship it to you when it's done, for your inspection. :^)

Jim


--
==================================================
please reply to:
JRR(zero) at pkmfgvm4 (dot) vnet (dot) ibm (dot) com
==================================================

Harold & Susan Vordos wrote:



Chuckle! Or perhaps big belly laugh!!

Yep, I remember, and I commend you for the great pictures, but that's not
what we're talking about. My point is turning a taper with an offset
*tailstock* center, although it's possible I never made that clear in my
original argument. It's not the same thing. The degree of error in
drilling offset centers remains constant and there is no movement of the
part as it relates to the fixed, but *in line* centers in the test you
performed. The machine centers, in your specimen, would pick the high
spots and run there, likely not fully seated, but with enough area of
contact to perform without distorting. When you offset the tailstock,
everything changes. You didn't prove your point originally, I simply quit
talking about it because I had quit following RCM (sort of like not talking
to your family, I discovered).

Try that same test, this time offset the tailstock, and for purpose of
proving whether you're right, or I am, turn a much shorter piece, with a
large offset, so it's exaggerated. Be certain that the faces are not at
right angles to the center, which is a part of my argument. You'll not only
mush the centers, you'll detect an oval. Grinders (the machines, not the
operators) don't lie. By the way, you shouldn't need any special machine
to learn what I'm talking about. Simply measuring the part will disclose
the oval. It will be fairly obvious.

Harold


I have to side with Harold on this one. Intentionally offsetting the
center holes on the workpiece between the headstock and tailstock end
has little to do with what I originally posted, or my subsequent
replies, and gives a different situation from the real world where
center holes are drilled in line with each other on a workpiece which is
to be taper turned between centers using a tailstock offset.

In that real world situation, the contact between the workpiece and cone
center would be only at two points as I mentioned before, but it would
be a 360 degree, fully circular contact inside the center hole and not a
partial one. Of course it relies on the accuracy with which the
centerholes were drilled, whether the shaft was straight enough to begin
with, etc.

I also agree with the faces needing to be at right angles.
Picture a shaft that needs to be taper turned between centers, and also
needs to have an angled end--in the real world, we should be able to
mill that angle after turning but just for the sake of argument say we
can't--and we need that angle to be 10 degrees (exagerated for the sake
of clarity).

As that piece is revolved around between the headstock center and the
offset tailstock center, the slope of the one end's angle will cause a
gap between the cone center and the cone-shaped hole it fits into for
each revolution.

If it weren't for the pressure of the tool making the cut, the workpiece
would flop around the tailstock cone--but the tool pressure pushes the
work against the cone, resulting in what actually is an egg-shaped
profile. That's because the contact of the center's cone with the
cone-shaped hole varies between the outer, larger diameter of the hole
and the smaller diameter of the hole's cone shape at the workpiece end's
slanted face.

Jim Rozen, your point that the old-time masters knew their stuff is
well-taken--using normal cone centers should work just fine as long as
the centerholes are drilled in line with each other. Thanks for making
me think about that.

The special centerdrills that Robin S. mentioned and Harold referred to
(bell center drills, cutting a curved cone hole) do improve contact
between workpiece and cone center, and avoid wear on the cone center
where the workpiece happens to be harder than the cone center or where
multiple workpieces cause cone center wear through repetition.

Ken Grunke



--
take da "ma" offa dot com fer eemayl

wrote:

This is a pretty good technique. It's perhaps
worth emphasising that it removes the length uncertainty
that's always present when turning between centre points.
With centre points the bar pivots about a point a bit inside
the pivot hole so the effective length is always an
uncertain bit less than the overall bar length.

With balls this is not a problem. The measurement
of overall length with both balls in place less 1/2d +1/2d
(d=ball dia) precisely defines the effective length.

With this checked with a decent vernier and the
offset set by gauge blocks pretty precise tapers are
possible.

Exactly!
thanks,

Ken


--
take da "ma" offa dot com fer eemayl

"jim rozen" wrote in message
...
In article , Harold & Susan Vordos says...

Yep, I remember, and I commend you for the great pictures, but that's not
what we're talking about. My point is turning a taper with an offset
*tailstock* center, although it's possible I never made that clear in my
original argument. It's not the same thing. The degree of error in
drilling offset centers remains constant and there is no movement of the
part as it relates to the fixed, but *in line* centers in the test you
performed. The machine centers, in your specimen, would pick the high
spots and run there, likely not fully seated, but with enough area of
contact to perform without distorting. When you offset the tailstock,
everything changes. You didn't prove your point originally, I simply
quit
talking about it because I had quit following RCM (sort of like not
talking
to your family, I discovered).

That's not what it seemed to me. The centers were non-axial, and
the part *was* round. Case closed. You could argue that the test
wasn't extreme enough. I thought it proved the point.

Try that same test, this time offset the tailstock, and for purpose of
proving whether you're right, or I am, turn a much shorter piece, with a
large offset, so it's exaggerated. Be certain that the faces are not
at
right angles to the center, which is a part of my argument. You'll not
only
mush the centers, you'll detect an oval. Grinders (the machines, not
the
operators) don't lie. By the way, you shouldn't need any special
machine
to learn what I'm talking about. Simply measuring the part will disclose
the oval. It will be fairly obvious.

Well, the HLVH that was used for the other test can't offset the
tailstock.
But my SB at home can. So the features should be:

1) the end of the stock should be cut at an angle

2) the centers should be drilled in the correct way, ie, coaxial

3) the tailstock should be well offset

4) the part should not be that long.

I could ship it to you when it's done, for your inspection. :^)

Jim

Cool! That's it. Because you'll be turning a taper, it might not be
real easy to detect the out of round condition with a micrometer. I may
have made it sound like you'd be making an egg, but it's not quite that
extreme. By moving your tailstock back on center and running an indicator
on the now taper machined surface, I can't imagine that you won't be able to
detect some degree of out of round. Should also be able to see some
small amount of center deformation, too. I'm keenly interested in your
findings, Jim.

Harold

"Ken Grunke" wrote in message
...
snip--------

That's because the contact of the center's cone with the
cone-shaped hole varies between the outer, larger diameter of the hole
and the smaller diameter of the hole's cone shape at the workpiece end's
slanted face.


Give that man a cigar!

Harold

Harold & Susan Vordos wrote:
"Ken Grunke" wrote in message
...
snip--------

That's because the contact of the center's cone with the

cone-shaped hole varies between the outer, larger diameter of the hole
and the smaller diameter of the hole's cone shape at the workpiece end's
slanted face.


I have to correct myself before I go to bed, or else I won't be able to
sleep. I was wrong about a slanted end resulting in an egg shape.
Actually, the tool pressure would push the workpiece against the cone
center as I said before, but the contact would only be at the bottom of
the centerhole's cone shape.
I hope a picture is worth a thousand words:

http://www.ken.crwoodturner.com/offsetcenters/

These are closeup top views of a workpiece offset between centers. The
end is cut at 10 degrees to illustrate the hypothetical situation I
mentioned in my previous post.

In view A, the tailstock center is all the way into the center hole but
in view B, you see what happens when the workpiece is turned 180
degrees--there's interference at point X. So the tailstock center has to
back out a bit, resulting in a sloppy fit when the shaft turns back
around to the view A position.

I could have just let this go, but nooooooooooo--I had to spend over a
frickin' hour preparing this message and the pics to explain a
hypothetical situation that would probably never occur for 99% of all
machinists. Oh well, it was kinda fun.

Moral of the story: face off the ends of a shaft you're offset turning
between centers, or something wierd might happen!


Give that man a cigar!

Harold


I could use a beer instead! ;-)

Ken Grunke




--
take da "ma" offa dot com fer eemayl

"Ken Grunke" wrote in message
...
Harold & Susan Vordos wrote:
"Ken Grunke" wrote in message
...
snip--------

That's because the contact of the center's cone with the

cone-shaped hole varies between the outer, larger diameter of the hole
and the smaller diameter of the hole's cone shape at the workpiece end's
slanted face.


I have to correct myself before I go to bed, or else I won't be able to
sleep. I was wrong about a slanted end resulting in an egg shape.
Actually, the tool pressure would push the workpiece against the cone
center as I said before, but the contact would only be at the bottom of
the centerhole's cone shape.
I hope a picture is worth a thousand words:

http://www.ken.crwoodturner.com/offsetcenters/

These are closeup top views of a workpiece offset between centers. The
end is cut at 10 degrees to illustrate the hypothetical situation I
mentioned in my previous post.

In view A, the tailstock center is all the way into the center hole but
in view B, you see what happens when the workpiece is turned 180
degrees--there's interference at point X. So the tailstock center has to
back out a bit, resulting in a sloppy fit when the shaft turns back
around to the view A position.

I could have just let this go, but nooooooooooo--I had to spend over a
frickin' hour preparing this message and the pics to explain a
hypothetical situation that would probably never occur for 99% of all
machinists. Oh well, it was kinda fun.

Moral of the story: face off the ends of a shaft you're offset turning
between centers, or something wierd might happen!


Give that man a cigar!

Harold


I could use a beer instead! ;-)

Ken Grunke

Don't jump to conclusions. Your second illustration shows perfectly what I
implied. The area of conflict has to go somewhere, don't you think? The
area at the point is much smaller than the exterior portion, so it will
likely deform at a greater rate than the exterior. You end up losing your
original center, plus you generate an oval. You drawings clearly indicate
why there's deformation of the centers, and why there's an ellipse generated
in the cut. The angular face of the shaft bears against the center
differently on one side than it does on the other as it revolves, thanks to
the offset of the tailstock.

Drop by, I'll buy you that beer.

Harold

In article ,
Harold & Susan Vordos wrote:

"jim rozen" wrote in message
...

[ ... ]

One of the regulars here at that time tested the roundness of the
turned part, it showed no systematic deviation from round to the
limit of the tallyrond tester.

Jim

Chuckle! Or perhaps big belly laugh!!

Yep, I remember, and I commend you for the great pictures, but that's not
what we're talking about. My point is turning a taper with an offset
*tailstock* center, although it's possible I never made that clear in my

[ ... ]

Try that same test, this time offset the tailstock, and for purpose of
proving whether you're right, or I am, turn a much shorter piece, with a
large offset, so it's exaggerated. Be certain that the faces are not at
right angles to the center, which is a part of my argument. You'll not only
mush the centers, you'll detect an oval. Grinders (the machines, not the
operators) don't lie. By the way, you shouldn't need any special machine
to learn what I'm talking about. Simply measuring the part will disclose
the oval. It will be fairly obvious.

One point to consider. I believe that tailstock centers used in
grinding are spring-loaded, to automatically adjust for thermal
expansion. If so, they can adjust to the varying length between centers
as the offset rotates.

Without that, I suspect that the distortion of the center hole
by the lathe's center may tend to hide the effects -- especially if a
soft workpiece material like aluminum is used -- as was used in the test
mentioned in the snipped text.

With non-spring-loaded centers, I would expect to feel a "lumpy"
resistance as the workpiece is rotated by hand and the tailstock
tightened -- at least until there is sufficient deformation of the
center holes to obscure the problem.

Oh yes -- and I expect a greater error if the center holes are
drilled perpendicular to the slanted workpiece faces, instead of along
the center line between the ends.

Enjoy,
DoN.
--
Email: | Voice (all times): (703) 938-4564
(too) near Washington D.C. | http://www.d-and-d.com/dnichols/DoN.html
--- Black Holes are where God is dividing by zero ---

On Tue, 11 Jan 2005 01:26:13 -0800, "Harold & Susan Vordos"
wrote:


"Ken Grunke" wrote in message
...
Harold & Susan Vordos wrote:
"Ken Grunke" wrote in message
...
snip--------

That's because the contact of the center's cone with the

cone-shaped hole varies between the outer, larger diameter of the hole
and the smaller diameter of the hole's cone shape at the workpiece end's
slanted face.


I have to correct myself before I go to bed, or else I won't be able to
sleep. I was wrong about a slanted end resulting in an egg shape.
Actually, the tool pressure would push the workpiece against the cone
center as I said before, but the contact would only be at the bottom of
the centerhole's cone shape.
I hope a picture is worth a thousand words:

http://www.ken.crwoodturner.com/offsetcenters/

These are closeup top views of a workpiece offset between centers. The
end is cut at 10 degrees to illustrate the hypothetical situation I
mentioned in my previous post.

In view A, the tailstock center is all the way into the center hole but
in view B, you see what happens when the workpiece is turned 180
degrees--there's interference at point X. So the tailstock center has to
back out a bit, resulting in a sloppy fit when the shaft turns back
around to the view A position.

I could have just let this go, but nooooooooooo--I had to spend over a
frickin' hour preparing this message and the pics to explain a
hypothetical situation that would probably never occur for 99% of all
machinists. Oh well, it was kinda fun.

Moral of the story: face off the ends of a shaft you're offset turning
between centers, or something wierd might happen!


Give that man a cigar!

Harold


I could use a beer instead! ;-)

Ken Grunke

Don't jump to conclusions. Your second illustration shows perfectly what I
implied. The area of conflict has to go somewhere, don't you think? The
area at the point is much smaller than the exterior portion, so it will
likely deform at a greater rate than the exterior. You end up losing your
original center, plus you generate an oval. You drawings clearly indicate
why there's deformation of the centers, and why there's an ellipse generated
in the cut. The angular face of the shaft bears against the center
differently on one side than it does on the other as it revolves, thanks to
the offset of the tailstock.

Drop by, I'll buy you that beer.

Harold


No this is too good to let go. All along I've known what your talking
about. Still , first off from the grinding I've done it seems like its
grabbing and distorts the work to out of round if not corrected in the
process. I asked you once before and you won't answer about specifics.
That's ok , I won't tell you some things also.

Anyhow, just for fun I looked into it and they say it is an ellipse
from the cross section of a cone. Plus algebra that used to be fun to
me , but I guess never fully understood. I was close to the nexus
when I ran out of money. There should be some rocket scientist that
can settle this here. I think Jim knows exactly what your talking
about , but still wants to drag it out.

I'm still having problems with the less than 1 , one , and greater
than one. For some reason I fall to the egg shape instead of equal
sided ellipse. And that should be the profile of the round hole of the
work and not that of the cone.

I don't like the point problem in the neat pictures , that should be
avoided at all costs. For some reason I can't see the cone and center
being equal an thus reinforcing Hal's point.