Chuck runout at different distances from jaws

[Peter TurveyParticipant @peterturvey44018]

accuracy of lathe chucks

[Peter TurveyParticipant @peterturvey44018]

This is in part a follow on from the earlier Pratt Burnerd 100mm chuck thread – what variance in runout should I expect at different distances from the chuck jaws.

I have checked spindle nose and spindle runout using the internal taper and my 5/8" x 6" Arand test bar, all is on spec.

My 'new old stock' Griptru can be set to give little if any runout at 1" from the jaws on the test bar, but is giving anything from 2 to 6 thou runout at the end of the test bar, about 5 1/2 from the chuck.

With lockdown I can not consult relevant standards. So it this ok or indicative of something wrong with the chuck? Indeed is this sort of measurement useful anyway?

 

 

Edited By Peter Turvey on 12/02/2021 19:29:13

[Dave HalfordParticipant @davehalford22513]

If it varies between 2 thou and 5 thou with the same bar, it's a chucking error on your part. The bar will droop and the chuck will hold it just like that.

[SteviegtrParticipant @steviegtr]

So how are you chucking your part. I always get the item to nearly biting point & then slowly rotate , while tightening. If not then the above post is correct. If you are doing it correct then one or more of your jaws are either worn or sprained.

Steve.

[Oily RagParticipant @oilyrag]

It just so happened that yesterday I checked a 6 jaw SC chuck for run out after making a backplate for it. I used a 5/8" dia hardened 'grinding' arbor. The results I got were 0.0002" next to the jaws, but 0.005" 6" out. I then suspected that the arbor was possibly warped (bent!) so, seeing as it had a centre in the end engaged my tailstock centre and checked again. This time I got 0.0002" at the jaws but 0.0008" at 6" out. I then tightened the tailstock barrel with the barrel clamp and it became 0.0001" at 6". After this last check I then ran the indicator at mid point and got a 0.0016" reading so confirming to my mind the arbor is bent! All readings are TIR.

I then checked again with a 1" hardened arbor and got 0.0003" at the jaws and 0.0002" at the tailstock end as in the final test (i.e barrel clamped), the span was about 8". The arbor runout in the centre area was similar to the ends. Arbour was tapered by 0.001" on diameter from one end to the other. To the OP I would ask is the test bar accurate?

Martin

[Michael GilliganParticipant @michaelgilligan61133]

It’s worth reading Dr. Schlesinger’s notes about test mandrels, before getting too excited:

See p9 [p11 of the pdf] of the book here: **LINK**

http://stubmandrel.co.uk/Reference_Documents/Schlesinger_Testing_Machine_Tools.pdf

MichaelG.

[Michael BriggsParticipant @michaelbriggs82422]

Hello Peter, I would put a mark on the chuck and back plate to reference orientation. Remove the chuck from the back plate and check that the back plate mating surface runs true. That could eliminate or discover a source of error.

Michael

[Martin ConnellyParticipant @martinconnelly55370]

Peter, ask yourself what you are going to be doing where this would matter? If there are simple ways to work around this then don't worry about it. There is always going to be some error in a part placed between the jaws and then just clamped up. It is easy to assume that the closing jaws will straighten up what they are gripping. However since there is friction between the jaws and the workpiece this straightening action stops short of perfection. Having a centre location in the end of the workpiece helps as you can position the workpiece against a correctly positioned centre then tighten up the jaws. Then you start worrying about how well positioned the centre is.

I think most people with a lathe will have placed something in the chuck then, as they have tightened the jaws progressively, checked the runout at the end of the part and tapped it achieve lower runout before final tightening, followed by a final check. Sometimes the part may need repositioning/rotating in the chuck to get the best results.

Martin C

[colin hawesParticipant @colinhawes85982]

For screw on chucks It is vital that the end locating faces of chuck, backplate and lathe spindle are accurate, clean and free from any tiny embedded particles before thinking about test bars. Colin

[Peter TurveyParticipant @peterturvey44018]

Thanks everyone for your comments and guidance. All this started when I bought a screw on Collet Chuck from Eccentric Engineering and found it was giving a far greater runout than expected. Eventually traced most of the problems to a tiny bruise on the spindle nose thread, which has been carefully taken out with a diamond file. The spindle is within tolerance for tests on the register, register end face, internal taper, and runout test bar in the taper. So using the griptru chuck and test bar to see if there might still be an issue here.

[old martParticipant @oldmart]

You could improve the chuck by grinding the jaws. With a griptru, you would have to adjust the body to run concentrically first, and it would also be worth checking the squareness of the front face of the backplate also and correcting it if required. The chuck would have to be dismantled and cleaned before and after any jaw grinding.

Edited By old mart on 14/02/2021 15:22:13

[Peter TurveyParticipant @peterturvey44018]

Postscript, have taken the Griptru apart cleaned and relubricated it again (moly grease on pinions, dry graphite on scroll jaws elsewhere) after doing so a couple of years ago. Getting better results,  5 1/2" from jaws runouts of a thou at most now. So leaving things at that.

Edited By Peter Turvey on 15/02/2021 19:46:43

[old martParticipant @oldmart]

Good result, Peter, 0.001" is very good for any scroll chuck. I am always surprised at how many people shy away from dismantling and cleaning scroll chucks, they are very easy to work on.

[Author]

Posts