View Full Version : Lathe chuck TIR

Mr Ron
01-10-2011, 02:46 PM
I have a Cushman 8" 3-jaw chuck, plain back mounted on a Sheldon 11" lathe. The TIR is about .002". Is this the best I can hope for or can I get even closer?

01-10-2011, 03:04 PM
0.002" on a 3 jaw is rather decent.

Rich Carlstedt
01-10-2011, 04:04 PM
Two things.
First try each of the three wrench sockets independantly with your test.
Only one will give you the best readings. ( the one used when the jaws were ground )
Mark it , and only use it for tightening (!) SOP !
I paint my master or prime socket white, and the other two red.

Second improvement:
Regrind the jaws, using a good method that loads the tips of the jaw (!)
When you do this , select your socket you want and only use it.
I pick the one nearest the Chuck makers nameplate, so I can easily see it from a horizontal viewpoint....makes it faster to find.

Also realise that a master socket on a 1 inch diameter may not replicate the same results for a larger piece of say 3 inches diameter due to scroll wear. in general they do, but don't count on it.

Jaakko Fagerlund
01-10-2011, 04:08 PM
Rather good for a 3-jaw, although a regrind of the jaws after installment gets you to as close to zero as possible. But usually that sort of accuracy is not needed, as the 3-jaw chuck is just meant for holding stuff, although at work I frequently dial in parts in the 3-jaw chuck with the use of a hefty copper ;)

01-10-2011, 04:16 PM
I am interested in your feedback on the amount of run out? Thanks for the tip on finding which hole gives least run out. I recall being told that you should tighten all three.
I thought also that 3 jaws were not used where accuracy is required?
I am no expert and appreciate how much I can learn here.

Jaakko Fagerlund
01-10-2011, 04:28 PM
Well it depends what sort of accuracy you want, but I do die casting molds for a living and at work I and every other I know of uses the 3-jaw chuck exlusively. You can easily get a part dialed in to under 0.02 mm TIR, which means that it is as best as it can be in there and that sort of TIR doesn't affect anything.

Usually the "TIR has to be zero for it to work at all" -people are the ones who think too much and do much less, as experience in machining would tell them right away if something works or not. This is just my opinion, but I have seen this book-knowledge on many subjects including this one.

01-10-2011, 05:32 PM

What procedure do you follow to dial different sized parts to under .02 mm TIR in a 3 jaw chuck repeatably? I am genuinely puzzled by this - I wonder if we are talking about the same type of chuck?

Bill Pace
01-10-2011, 05:37 PM
I am genuinely puzzled by this - I wonder if we are talking about the same type of chuck?

Me Too!!

I thought the OP saying his Cushman giving .002 was exceptionally good...

01-10-2011, 07:38 PM
I have a Cushman 8" 3-jaw chuck, plain back mounted on a Sheldon 11" lathe. The TIR is about .002". Is this the best I can hope for or can I get even closer?

That's better than most three jaw chucks. If you are working on a piece of metal and turning the diameter down, it will make no difference whatsoever until you remove the part and then reinstall it. At that point, you will have lost the original position.

If you want zero TIR, try a four jaw chuck and spend some time indicating it. You can even take the part out and reinstall it so long as you check the runout again.

This is not anything original with me. It's general knowledge that three jaw chucks will all have some runout. Some will have more than others, but all will have some. .002 TIR is really excellent.

01-10-2011, 07:54 PM
0.002" runout for a 3-jaw chuck is very good. Personally, I wouldn't bother trying to improve it.

For maximum concentricity it's best to do all machining in one setup, but if you do need to remove/replace work, mark the position of the work relative to the chuck jaws and put it back the same way. It won't give you perfect repeatability, but it should be pretty close.

01-10-2011, 08:27 PM
That is pretty good. When dad's 1980's-vintage Victor 1640 lathe was delivered, it came with it's OE 250mm 3-jaw scroll chuck. That chuck runs out about .007"-.008".

I got extremely lucky when I picked up a Pratt Burnerd Super Precision (made in UK) 10" top-jaw style that had been dropped. One jaw is hit hard enough that it's counterbore is closed over the bolt head and you can't take the bolt out of it. My employer at the time was a dealer getting out of the manual machine and accessories business, they sold it to me for $250.

That chuck runs out less than .0003" anywhere in the range with the jaws in the "normal" position. Reversing them for larger-diameter work, it's still under .001". Go figure.

01-10-2011, 09:11 PM
.003 is pretty darned good. Even if you grind the jaws at a dia., it will run out again at other dia's.
Your work concentricity will only be good at the ground dia.
Your chuck sounds nice.
K Liv

Jaakko Fagerlund
01-10-2011, 11:10 PM

What procedure do you follow to dial different sized parts to under .02 mm TIR in a 3 jaw chuck repeatably? I am genuinely puzzled by this - I wonder if we are talking about the same type of chuck?
A regular 3-jaw chuck, nothing special, made by TOS (as is the lathe I use). Chuck a part in using the "zero" screw (the one that gives the best concentricity, but don't tighten it up fully. Then put a DTI on the diameter of the part (or another one at the end if you need to have perpendicular face with a short workpiece), find the high spot and with about 1 kg short copper bar, hit the jaw on the outside edge. It will not destroy the jaws or the chuck, but it will move the part/jaws enough to make it concentric. Then just tighten it up and repeat. Sometimes I get near zero TIR in one go or sometimes it takes a couple of minutes, but never have I done non-working part with this method.

The lathe and the chuck is about the same age than me and still work fine :)

01-10-2011, 11:31 PM

I've seen zeroing done in a 3-jaw that way too. It works pretty well even with old beat up chucks.


Rich Carlstedt
01-11-2011, 12:46 AM
I think .002" is very good for a used chuck
my 50 year old Pratt-Bernard gives me .001 to .002 in most setups.
My Buck Chuck (6 jaw) is .0005" TIR or better

What Jaakko said is very true.
It is hard for many to understand but a former mentor, and the absolute best Lathe man ( I ever met) told me "All 3 jaw chucks are adjustable" and he was right.
I don't use a "copper" but do use a leather mallet (soft blows ) if needed
Let me explain. The jaws will not change, but are engaged in the scroll plate.
The scroll plate has a slip fit in the chuck body. This "slip fit" is maybe .0005" on a good chuck, and maybe .004" (or more !) on a cheap chuck and allows the scroll plate to turn inside the body .
When you use the wrench to tighten, the socket is attached to a bevel gear.
The bevel gear exerts a radial force on the jaws with the spiral grooves, but also almost as much force is used pushing the scroll plate away from the socket, hence the shift in true center of the jaws to move away from the socket side to the opposite side ( and sometimes, the tighter, the more error ! )
Now if the jaws are ground with the master socket used AND a certain torque on the socket, you will get repeatablity for your work. Any other socket will offset the scroll plate in the exact opposite direction, incurring greater error.
If the load bearing surface of the scroll plate is insufficient, variations in torque will result in more error. Notice that torque is an important part of accurate repeatability !
Knowing this, when you have the part setup, you can tap the outside of the "high" jaw, and that will shift the scroll plate in the desired direction, as the socket is not affecting location. That is how Jaakko or me can get a dead nuts concentricty with our "adjustments"
I am not too keen with heavy handed torques after this adjustment, but depending on the job, I have done it. But you need to recheck !.

The technique of tightening all three sockets is an attempt to move the scroll around, and I don't think that is necessary when you find your master jaw,as it is torque variable. You may want to find the high spot, and tighen that socket that is alligned with it. This is somewhat similar to a 4 jaw!

The above effect is well known to chuck manufacturers m and some of the better chucks have only one socket, to eliminate the guess work

Last word. A very important function of grinding jaws, requires that a somewhat substantial load, (not light !) be placed on the jaws before grinding, to match all the effects of normal part capture in a chuck.

Way to go Jaakko !


01-11-2011, 01:44 AM
Thanks for all that great information.
I will add it to my ever growing knowledge.

01-11-2011, 07:11 AM
Jaako -

Thanks for the explanation. My interest was especially in the .02 mm TIR . I think this may have passed some people by - it is 8 tenths, .0008". I will have to give this a try.

Bill Pace
01-11-2011, 09:52 AM
Having the good fortune to become friends with Lane, who seems to really enjoy mentoring poor souls such as me, I picked up this 'trick' of centering a 3 jaw only in the last year or so.

Per Lane: "There aint no 3 jaw chuck gonna be dead accurate " and "in the average home shop, there aint likely to be no (truly) round stock", so you gotta find some work-arounds on these " And the really simple one is to get that piece of Brass, Lead, or ? and keep it near the lathe and do the "bump" that Rich and Jaakko are describing.

My 6" unknown brand of Chinese chuck is pretty decent on an average mount, reading around .003 - .004 thou out, and most often a 'bump' or 2 or 3 or ? will bring it in near or at "0" I've got a piece of scrapyard brass about 1 1/2" x 5" for my 'bumper'