PDA

View Full Version : Help with screwy four jaw chuck



sco
06-20-2012, 04:07 PM
Brand new four jaw chuck - if I put a straight test bar in it and indicate on the middle of the bar till it's concentric then the ends of the bar go round like a baton twirlers baton. Same deal if I indicate on one end of the bar - the other end is miles out.

If I indicate on the outer diameter of the chuck body then it's concentric to 0.02mm - it's difficult to indicate on the front face because of all the markings / bolt holes etc. Does this all suggest that the chuck is concentric but not mounted square or the jaws aren't square to the chuck axis? I tried reversing all the jaws but got the same result.

Assuming the body is on square do I need to correct this by somehow machining the jaws? Chuck is a front mount din 6350 style if that makes any difference.

Cheers for any advice,

Simon.

The Artful Bodger
06-20-2012, 04:13 PM
I suggest as a first step taking all the jaws out and giving it a good clean, maybe there is something in there. Next step, put the jaws in reversed and recheck.

sco
06-20-2012, 04:19 PM
First thing I checked was for swarf, and as I already said reversing the jaws gives the same effect.

The Artful Bodger
06-20-2012, 04:24 PM
Ooops, sorry about not reading carefully enough.:o

ironnut
06-20-2012, 04:51 PM
The axis of a straight test bar when mounted in the chuck should be reasonably parallel with the axis of rotation of the headstock of the lathe. You should be able to traverse a test indicator along the mounted test bar and not have the reading change a great deal. If it does, that suggests that the gripping surfaces of the jaws are not parallel with the headstock axis of rotation and the jaws need to be re-ground.

gordon

Dr Stan
06-20-2012, 04:56 PM
the jaws are not parallel with the headstock axis of rotation and the jaws need to be re-ground.

Your correct in the diagnosis, but since this is a brand new chuck it needs to go back to the distributor for a replacement or refund.

sco
06-20-2012, 05:09 PM
The chuck is new in that it's not been used but I bought it over six months ago - don't think I'm going to have much success returning it.

The chuck is mounted via an adapter plate - I'm wondering if the plate is at fault rather than the jaws - think I need to check it carefully before condeming the jaws or chuck.

Dr Stan
06-20-2012, 05:25 PM
The chuck is new in that it's not been used but I bought it over six months ago - don't think I'm going to have much success returning it.

The chuck is mounted via an adapter plate - I'm wondering if the plate is at fault rather than the jaws - think I need to check it carefully before condeming the jaws or chuck.

That's a good idea as it could just be some dirt or a burr. Did you machine the backing plate while it was mounted on your lathe, or was it already on the chuck? If the latter I'd determine if there is enough material to face off the plate and machine it true to your lathe.

J. R. Williams
06-20-2012, 06:29 PM
Check the spindle nose for damage or run-out, then install the mounting plate and take a light cut on the mounting surfaces. mount the chuck and check the face and OD without the jaws in place. Then you will be ready to install the jaws and a test piece.

TGriffin
06-20-2012, 06:36 PM
Simon,

Take the chuck off the backing plate and run an indicator on the face of the plate to check for wobble. If it runs true then your problem lies with the chuck. To machine the back of the chuck true to the jaws, chuck a piece of round stock in your three jaw and take a clean-up cut on it. Grab the freshly turned round stock with the jaws of your four jaw and indicate it so it runs true radially. Then you can machine the mounting surface of the chuck relative to the jaws to make it run true when re-mounted on the backing plate.

That should cure your wobble issues.

Tom

darryl
06-21-2012, 12:59 AM
Since the chuck is new, it may not have been deburred well enough. I would check each piece and remove any build-ups, etc. You mentioned that the body seems to run pretty close (.02mm), which is good since you don't need any extra vibration from imbalance. If you have to throw that out to get the jaws to align, that would be a bummer.

It would seem that if you can change jaws around and the problem persists, you are largely taking the blame off the jaws. Perhaps one or more of the slots have not been machined parallel to the face of the chuck, or perhaps the jaws are loose in the slots until tightened, in which case they would skew. If you can rule out sloppiness in the jaw fits, then I would start with one jaw and run it in each slot, using a dial indicator to see whether the inside face of the jaw is parallel to the spindle axis. You will probably have to keep some outward pressure on the jaw to somewhat simulate it being tight while you roll the carriage back and forth with the indicator. Be careful as you don't want the indicator tip to catch in the gaps on the gripping edge of the jaw.

Do the same test with each jaw in every slot. From this you'll know if the jaws are off, or if the slots are off.

To be a good test, you'd have to check with the jaw in different positions along the slot. It won't be any good if they grip accurately at say 1/2 inch diameter, but skew the workpiece at another diameter.

It would almost be safe to say that if you do any machining to a jaw, you would then have to dedicate that jaw to that slot.

If you can check the concentricity of the gripping edge to the spindle axis at a few different locations of the jaw along the slot, and find that the error is consistent, then you could theoretically improve things by grinding the gripping edge. And of course this will all get messy right away- you'll have to clean, clean, clean the slots, jaws, etc as you go, otherwise the dust will not only throw the process off- it will end up being lapping compound. If you do end up grinding the jaws, you will have to find a way of simulating that they are tightened against a workpiece, so any tilt or rocking motion possible is 'locked out' before you grind.

sco
06-21-2012, 01:07 AM
Thanks for all replies. The adapter plate was machined on the lathe but I'll check it first. Might also ask the inspection department at work to check the chuck on it's own before machining it's back face to double check.

J. Randall
06-21-2012, 01:15 AM
Simon, can I ask a question, are you experienced at dialing work in with a 4 jaw and this one is throwing you for a loop, or is this your first go round at dialing in?
James

sco
06-21-2012, 03:15 AM
I would say I'm fairly experienced - I'm not very slick at doing it but I can get the required result. The confusing thing was that I could get one part of the bar running concentric but then the ends would be miles off - so it's not like a brick in a washing machine but as I said before like a baton twirler's baton. Reading all the replies and thinking about it a bit more I think for whatever reason the jaws aren't parallel with the lathe axis - just need to work out where the problem lies and the easiest way of correcting it.

sco
06-21-2012, 05:04 PM
Update on this - spent all evening making measurements;

Adapter plate on it's own across the mounting face: 0.02mm
Chuck mounted on the plate across the front face: 0.05mm
Across the chuck jaw running faces: 0.08mm
Chucked up my eight inch test bar and got it running concentric up close to the chuck, at the end of the bar the run out is then just over a 1mm - tightening the jaws an extra amount opposite the high point it comes down to ~0.8mm.

Somebody suggested putting a bar in the 4 jaw chuck and then holding the bar in the three jaw and skimming the chuck mounting face - I've set this up but not machined it - the mounting face moves axially 0.35mm as the spindle rotates. I'm worried that this approach may only reduce some of the problem and might only work for one bar diameter.

Grinding the jaws seems like it might be a better option but how to go about this exactly?

darryl
06-21-2012, 08:53 PM
The reason I mentioned about the chuck body running fairly true, besides the balance issue, is that if you did turn a spud and mount the chuck on it to 'true up' the back, then the body of the chuck will not be running true. Things could go from poor to worse pretty quickly.

Here's a test for you- find two or three different diameters of straight and round stock. Mount each one and center it near the chuck. The outboard end is going to be searching for klingons, or at least for the tailstock center. Mark the high spot and mark the chuck at the same point. Measure the degree of runout at some fixed distance from the chuck. Do this for each piece of stock.

If you find that each piece has the same runout at the same rotation of the chuck, then set up to true the jaws. But if the runout is all over the place, then it suggests that the jaws aren't moving linearly in their slots. In that case, check the pieces that do the adjusting for each jaw (the individual lead screws) You might find that they are not straight and are making the jaws wobble somewhat as you screw them in and out. If some of the problem lies with that, then grinding the jaws won't be the whole answer to the problem.

If it boils down to grinding the jaws, then you first need to find a way to keep the jaws apart while all the screws are tightened. You could clamp down on a ring placed as deep as possible inside the jaws- this will give you something to close down onto, but you won't be able to grind the full length of the jaw. It would be good if there's a hole in the face of each jaw- then you could drill some equally spaced holes in a ring, then pin the ring to the face of the jaws. That would leave the entire gripping surface of the jaws free to be ground true, plus that's about the best way to get the play out of the jaw in the same manner as if you are clamping down on a workpiece.

If there aren't any holes in the jaw- I wonder if they are drillable? It's possible- you don't need a large hole, but you might also consider if it would make sense for you to have some accurately laid out drilled and tapped holes in the face of the jaws. I've certainly had times when I wished I could just bolt something straight on using a hole pattern. This would be for thin stuff usually. If course if the jaws aren't drillable you can't use this method.

Bob Fisher
06-21-2012, 08:59 PM
Since it is a 4 jaw, can you make a fixture to hold the jaw by it's grooves,and grind them square to the groove? At least ,checking the jaws relative to the groove with an indicator might give a clue. If the backplate was machined in the same lathe that it will be used on, the chuck or jaws are suspect. My thoughts, Bob.

ironnut
06-21-2012, 10:41 PM
I assume this is a 4 jaw independent chuck as opposed to a 4 jaw self centering chuck. With the former it is not necessary to to tighten the jaws around a ring to help insure that the jaws center around a work piece. What is necessary is to keep the jaws from moving while grinding them.

I would put a sufficiently large enough ring around the outside of the jaws and move the jaws outward to tighten against this ring. Having the outside of the ring run true would allow you to grind all four jaws at once, which would be nice but it is not a requirement. If the outside of the ring is not concentric with the inside or it doesn't run true with the jaws tightened against it, is not a problem. You will just have to grind each jaw separately until it cleans up across the jaw's clamping surface. This will require you to reset 3 additional times to move each jaw inward to be one that traverses the smaller diameter as the chuck rotates.

The ring needs to be large enough to permit the grinding wheel to pass between the jaws. Plan to take your time while doing this.

gordon

sco
06-22-2012, 04:10 AM
Thanks again for all replies.

If I skim the chuck body mounting face this will put the body out of balance - can I not simply correct this by skiming the outer diameter of the body.

If it comes to grinding the jaws - the only method I can see for doing this is to mount my Dremel type drill grinder in the tool post at centre height and skim each jaw in turn - would this be the right way?

I understand the need for clamping the jaws although they are actually a pretty snug fit in the body - there's no discernible tipping that you can feel.

The jaws are numbered - I did wonder if maybe they had been numbered incorrectly and I should try them in different positions.

Circlip
06-22-2012, 04:35 AM
Even if the chuck is six months old, if you haven't used it, I would still contact the supplier before trying to butcher it true.

Regards Ian.

Jaakko Fagerlund
06-22-2012, 04:45 AM
No no no, do not cut the chuck! You already said you have 0.02 mm variance on the adapter plate, so fix that. First take the plate off, check that your spindle runs without runout and install the adapter plate and check again and correct the plate by cutting.

Forrest Addy
06-22-2012, 04:54 AM
First it's a BIG mistake to expect a 4 jaw chuck to accurately swallow and hold extended work on axis the full length.

In pro shops where special care is taken for long work out of a four jaw chuck the process goes somethen like this: extended work is gripped by the first half inch of the jaws in malleable pads (1/16" to 1/8" copper is traditional.) Dial the work in near the jaws move out to the end and bump it straight, go back near the jaws and make a final adjustment then, return to the far end for a final tap to straighten. Then set up the steady rest, drill a center hole, or do the end detail in the steady.

That's how it's done in the real world. Never expect a chuck to hold work on-axis. Even brand new chucks with precision ground jaws splay a trifle when tightened on swallowed work. The splay will vary from jaw to jaw because of manufacturing tolerances jaw slot to jaw slot. The result is the work will be the most stable at some small angle to the sindle axis. Thus even if the work is tapped "straight," succeeding operations will nudge the chuck's grip to tension equilibrum and the work will "nutate".

Unless the jaws are pre-loaded so the they are restrained at the tips they will grip at the heel. Here's an experiment: dry machine a remnent to a smooth satiny finish without polishing. Without handing it insert it in the chuck and snug the jaws firmly. Relax the jaws, remove the piece, and examine the marks left by the jaws. Most of the grip will be focused at the heels. This is evidence of splay and every chuck unless it's brand new and specially ground will exhibit some splay.

It's odd to consider: a 4 jaw chuck may no longer be precise - in fact it may be considerably worn - but when used with care it is the workholder of choice when tolerances for concentricity are exacting.

sco
06-22-2012, 05:14 AM
No no no, do not cut the chuck! You already said you have 0.02 mm variance on the adapter plate, so fix that. First take the plate off, check that your spindle runs without runout and install the adapter plate and check again and correct the plate by cutting.

Maybe I'm wrong but I thought 0.02 on the adapter plate was pretty good - I expect that's within the tolerance of the cross slide.

0.02mm across 50mm is a tiny angle - by trig this angle would equate to less than 0.1mm runout at the end of my test bar, not 1mm!

If my 3 jaw chuck can hold the test bar concentric to less than 0.1mm at the far end am I wrong to not expect the 4 jaw to be similar - I thought in terms of ultimate precision an independant 4 jaw was supposed to be better than a 3 jaw?

Jaakko Fagerlund
06-22-2012, 07:13 AM
Maybe I'm wrong but I thought 0.02 on the adapter plate was pretty good - I expect that's within the tolerance of the cross slide.
So you didn't actually measure the adapter plate runout but instead cross-slide movement vs. the plate.

You have to measure the adapter plates runout axially and also if there is the alignment boss that the chuck sits in you have to also measure the runout of that radially. This latter is not that important with a 4-jaw, but helps in vibration issues.

The adapter plate should show about nothing in terms of runout when measuring axially.

sco
06-22-2012, 07:45 AM
So you didn't actually measure the adapter plate runout but instead cross-slide movement vs. the plate.

You have to measure the adapter plates runout axially and also if there is the alignment boss that the chuck sits in you have to also measure the runout of that radially. This latter is not that important with a 4-jaw, but helps in vibration issues.

The adapter plate should show about nothing in terms of runout when measuring axially.

Ok I understand - yes I did check the axial runout of the plate as well - just a flicker on the dti needle which I put down to surface finish.

The plate was carefully machined on the lathe so should be spot on - only possible aspect of it that could be squiff is the plate is mounted to the spindle with three bolts and the chuck to the plate with four bolts so maybe the plate distorts unevenly. I clocked the front face of the chuck as I did the chuck bolts up and couldn't detect a change so I think I've ruled that out.

Hoping to get the chuck on the cmm at work today to prove where the problem is.

sco
06-23-2012, 05:44 PM
An update on this - made a whole series of measurements over the chuck body, jaw slots and jaws on a surface table. The chuck body and slots are flat relative to the mounting face to +/-0.03mm. The jaws tightened onto a washer and measuring their clamping face showed a taper of about 0.07mm but what I think was the problem was the taper wasn't in the same direction. Two jaws 'bell mouthed' open as you would expect but two had the taper going inwards - these weren't paired opposite one another but adjacent so the tapers were bushing the bar off axis.

I've ground the jaws with them clamped around the periphory and now the runout at the end of my test bar is less than 0.1mm :)

Dr Stan
06-23-2012, 06:14 PM
Glad you've had success. I do highly recommend that you number the jaws & their respective slots if this is not already the case. Since you have ground them in the chuck to correct the problems in the body you cannot switch them from slot to slot and expect them to be accurate.

Forrest Addy
06-23-2012, 08:34 PM
If you wish to swallow the work the jaws tips should engage first then as you tighten the key, the jaw can deflect until the heels bear. Otherwise there will alway be a little splay and there is no guarantee they will splay equally.

Berofe you grind, you must make a gadget to strongly restrain the jaw tips for grinding. Then when you grind the jaws they will have the needed back-taper - for some time.

Regardless, your next little wreck (busted parting tool, work wrenched out of the jaws, etc) you may have to touch up the jaws again.