Lathe runout question

[dml66]

After installing angular contact spindle bearings in my HF 7x10 (now a 7x16) and letting it settle in for a couple months or so, I felt the bearings needed a bit more preload which I did yesterday. Indicated off the inside of the MT3 taper, it's back to 0.0002" TIR using a Mitutoyo DTI so all is good.

I realize that 0.0002" TIR at the spindle is unlikely to hold up once a chuck is installed either on, or inside of said spindle. I have a Z Center Live ER20 collet chuck that may well be a unicorn; it reliably turns 0.0002" TIR mounted in the lathe spindle, indicated off the inside collet seat.

Lets say I add a collet which itself has 0.0002" TIR, in a perfect world should I expect the materialized runout of a turned piece to be 0.0002" or 0.0004"?

Somewhere I read that a lathe cannot turn a piece with less runout than it's spindle, I'm wondering if the correct answer is a lathe cannot turn a piece with less runout than the materialized, stacked runout of the spindle and the workholding assemblies?

 

[Janderso]

No expert here.
The cold readings will change when the lathe has a chance to warm up.
.0002-.0004 is pretty darn good man.

 

[markba633csi]

In one orientation the error may well zero out, and at 180 degrees from that the error might stack

 

[extropic]

By reading the runout of only one circular element of a taper, you only know something about the circularity/runout of that circular element and only that one circular element. To know more about the orientation of a taper you need to take readings on at least two circular elements and know how those readings relate to each other (clocking and distance between). That applies to the spindle taper and the ER chuck taper.

 

[dml66]

By reading the runout of only one circular element of a taper, you only know something about the circularity/runout of that circular element and only that one circular element. To know more about the orientation of a taper you need to take readings on at least two circular elements and know how those readings relate to each other (clocking and distance between). That applies to the spindle taper and the ER chuck taper.

Makes sense. The question though, concerns the made part; can any turning assembly with a non-zero materialized runout make a part with less runout?

 

[682bear]

Makes sense. The question though, concerns the made part; can any turning assembly with a non-zero materialized runout make a part with less runout?

Yes.

If your spindle is off center relative to the axis of rotation, any part machined will still be round... or have less runout...

If the runout is caused by radial play in the spindle, then that radial play will be transferred to the part being machined.

-Bear

 

[dml66]

Yes.

If your spindle is off center relative to the axis of rotation, any part machined will still be round... or have less runout...

If the runout is caused by radial play in the spindle, then that radial play will be transferred to the part being machined.

-Bear

Fascinating, I need to think on that a bit.

 

[682bear]

Ok... picture turning a crankshaft journal... the end in the chuck may be offset several inches, but the journal that is being turned would end up round.

Same concept... if your runout is caused by the chuck not holding the work pefectly concentric with the axis of rotation. What you machine will be concentric... and round, barring all other negative factors.

However, if the part is removed from the chuck, turned, then reinstalled, your machined diameter will now show some runout... theoretically, if it is turned 180 degrees, it should show two times the runout of the chuck.

-Bear

 

[markba633csi]

First operations will always be concentric (except for the spindle runout) until removed from the workholder

 

[tghsmith]

I have a LMS 7 x 16, much of my work is with 3jaw chuck, I tried it in the 3 mounting positions with a ground shaft chucked up,, when I found the best spot,, <.0004 I marked the chuck and mounting spindle/plate so the it returns to the same place..