Looking for help with my 4 jaw

[atunguyd]

Ok, so the newish lathe that I got came with a 4 jaw that I seldom use. Today however I found the need and mounted it on the lathe.

I dialed my part in fro zero run out at the furthest position from the chuck with no hassle, but then when I traversed back near the chuck (about 1") my indicator gave me a run-out of about 0.05mm.

I tried a few times but every time the same, runout varied along the length of the part *which had already been turned true yesterday on the three jaw).

So I then removed all the 4 jaws and proceeded to measure runout of the face of the 4 jaw. This measured at about 0.01mm.
I then removed the 4 jaw and mounted only the adapter plate and also measured runout of about 0.01mm on the face of the plate. (Spindle though is perfect).
So I took a light cut on the face of the adapter plate till I could not measure any runout on it.

Remounted the 4 jaw and still the face of it gives me a run out of about 0.01mm. Tried centering my part in the jaws again and this time better but still over 1" I get a varying run out from 0.00 to 0.02mm.

Does anyone have any pointers for me? Thinking this through to me it means that the chuck is not aligned to the spindle, because if the bed was out from the spindle then the runout would remain constant just the value on the indicator would change.

Would it be advisable to take a light grind of the jaws using a toolpost grinder?

Also is it correct to measure the runout of the face of the 4 jaw? I am expecting that this would be parallel to the mounting base, which is inaccessible once the chuck is mounted.

[ctwo]

I was just wondering if you could have clocked the chuck on the mounting plate in different positions to cancel out the run out, but it seems now it would not matter since the plate's face is true.

Are you chucking a smooth rod or a test bar?

I usually indicate first at the jaw just snug, and then back out to the far end and bump it into alignment. Then I'll go back to the jaws for more snugging, then check the far end again. Sometimes I need to go back and forth a few times to get is all set. And I'll sometimes put some soft pads (1/8" aluminum or copper shim) on each jaw.

I think grinding the jaws would be acceptable if done properly.

[atunguyd]

I was just wondering if you could have clocked the chuck on the mounting plate in different positions to cancel out the run out, but it seems now it would not matter since the plate's face is true.

Are you chucking a smooth rod or a test bar?

I usually indicate first at the jaw just snug, and then back out to the far end and bump it into alignment. Then I'll go back to the jaws for more snugging, then check the far end again. Sometimes I need to go back and forth a few times to get is all set. And I'll sometimes put some soft pads (1/8" aluminum or copper shim) on each jaw.

I think grinding the jaws would be acceptable if done properly.

Yes I also thought of trying to clock it but in all four positions it made no difference (was testing by indicating the face of the chuck, not with bar).

The part I was using was a rod that I turned down yesterday on the 3 jaw and then parted off, so I would expect that it should have no runout.

I have a bar that came off of a printer which I will mount tomorrow and test with (it is late here now). I did run a indicator along the front of the jaw tonight and say that all the jaws appear to bell inward which I would expect as under load then will then come up to parallel.

[Carm]

Before removing any metal, find the cause. Your description leads me to think that the jaw registers are not parallel to the back of the chuck.
Depending how the backplate (adapter) mounts to the lathe, grinding jaws could lead to a once-only correction.
Do a little snooping first.

[wlw-19958]

Hi There,

Ok, so the newish lathe that I got came with a 4 jaw that I seldom use. Today however I found the need and mounted it on the lathe.

I dialed my part in fro zero run out at the furthest position from the chuck with no hassle, but then when I traversed back near the chuck (about 1") my indicator gave me a run-out of about 0.05mm.

I tried a few times but every time the same, runout varied along the length of the part *which had already been turned true yesterday on the three jaw).

So I then removed all the 4 jaws and proceeded to measure runout of the face of the 4 jaw. This measured at about 0.01mm.
I then removed the 4 jaw and mounted only the adapter plate and also measured runout of about 0.01mm on the face of the plate. (Spindle though is perfect).
So I took a light cut on the face of the adapter plate till I could not measure any runout on it.

Remounted the 4 jaw and still the face of it gives me a run out of about 0.01mm. Tried centering my part in the jaws again and this time better but still over 1" I get a varying run out from 0.00 to 0.02mm.

Does anyone have any pointers for me? Thinking this through to me it means that the chuck is not aligned to the spindle, because if the bed was out from the spindle then the runout would remain constant just the value on the indicator would change.

Would it be advisable to take a light grind of the jaws using a toolpost grinder?

Also is it correct to measure the runout of the face of the 4 jaw? I am expecting that this would be parallel to the mounting base, which is inaccessible once the chuck is mounted.

I think you are expecting too much from your chuck. Let's first put this into
term that I am familiar with. .05mm is slightly less that .002". So, you are
complaining about .002" runout from one end to the other. This can be caused
by wear and tear on the chuck.

The usual method of mounting stock that has to be indicated precisely is to place
a strip of soft material between the jaws and the work (like copper or lead). Then
you can lightly "bump" the work to get the work into precise alignment from end
to end. Also, the strips will prevent marring of the work by the jaws of the chuck.

Good Luck!
-Blue Chips-
Webb

[Harold_V]

What Webb said. Unless a chuck is fitted with soft jaws, machined for perfect alignment, it is a tall order to expect a chuck, ANY chuck, to grip a part such that it runs true.

As chucks age, or once they have suffered a crash, it's very common to find that the gripping surface of the jaws are no longer parallel to the spindle. That results in pieces being gripped out of parallel, although you may be able to find one place on the extended piece where it appears to run true. Or not. Depends on how it is held, and if the chuck is capable of running true. Most are not.

When you set up a piece such as you described, using shims, even aluminum, get the piece running true near the jaws, but do not grip it tightly, as you will then move to the end, where you'll tap the high side, pushing it towards being concentric. That, of course, will yield a condition whereby the entire piece is now once again not running true.

Repeat this setup procedure until you get identical readings near the jaws as well as at the end of the piece. Only then should you tighten the jaws sufficiently for machining. Recheck once they are tightened, and repeat the procedure if necessary. With care, you should be able to get most any piece within a couple tenths (assuming your indicator is capable).

Harold

[Carm]

My four jaw is old as the hills and didn't start life with a pedigree.
But I'd be upset if it gave me 2 grand per inch.
Shimming to protect the work is great. Would rather separate fly specs from pepper than shim to correct alignment.

[Harold_V]

My four jaw is old as the hills and didn't start life with a pedigree.
But I'd be upset if it gave me 2 grand per inch.
Shimming to protect the work is great. Would rather separate fly specs from pepper than shim to correct alignment.

I agree. Shims are not to correct error, but to protect the surface of the item being chucked, and to allow it to move in the jaws. Correction is accomplished by adjusting the jaws and tapping the piece with a soft hammer.

Harold

[Mr Ron]

.01 mm run out is only .00004" or 4 one hundred thousands. You shouldn't worry about this.

[Harold_V]

.01 mm run out is only .00004" or 4 one hundred thousands. You shouldn't worry about this.

It's actually .0003937", or about four tenths. Not critical in some applications, but double the amount of tolerance at one's disposal for many jobs.

Harold

[atunguyd]

Thanks for the replies, this morning I chucked up a bar that I would regard as a good test piece and dialed it it to have runout that I could not detect near the chuck. I then moved out 6 inches and I now had about 0.06mm runout. I progressed in stages out to the 6 inch and at every stage the runout got worse and was worse at the same point in the rotation of the bar which to me indicates a statistical probability that the runout is caused by either the bar not being straight or the chuck not inline with the spindle as opposed to the bar not being round.

Would the experts here consider this as an acceptable error in the chuck (being 2.3 thousands over 6 inches). Or is this something that should be considered as needing fixing.
Please also note that this is not a worn chuck, it is quiet new and I have used it twice since removing the packing grease that it came with.

I am still concerned about the runout I measured at the face of the chuck - is this reason to be concerned or can one say that the face is not a reference surface and could be out.

I saw on another machinist board a picture that someone posted (not sure if I can link that here). He mounted a bar in hos three jaw (presumable then turned it straight) and the mounted his 4 jaw onto that bar so that the two chucks were face to face. This allowed the back of the 4 jaw to be presented to the lathe and allowed him to do a light cut of the back of the chuck to get it to be orthogonal to the jaw's. This seemed to make sense to me if it is the case that my chucks back is out.

[timvercoe]

Boy I'd be really careful about removing material from the back of the chuck. You could easily make a not so bad problem really worse. You could measure the thickness of the chuck in several different places and see if it is thicker on one side. This may not really mean anything but it could be an indicator of a problem especially if the measurements to along with the run out on the machine. You might try to measure the slots where the jaws mount and see it they check out all the same. A chuck is a pretty complicated machining operation and it seems that it would have some really small acceptable differences from jaw slot to jaw slot that could account for the run out. I am not sure what you would do to correct that. If you knew where the problem with the chuck was, at least then you could devise ways of dealing with it. The use of soft material to chuck your work being a good example of this.

Tim