PDA

View Full Version : Lathe chuck alignment problem



DaveC
03-15-2011, 06:52 AM
Hi all,

I think I have a problem with the chuck in my lathe, although I am not sure what....

You will have to excuse my lack of terminology in the question, I am mostly self taught on the lathe and mill and definitely missing some terminology.

I have a Warco WM-180 lathe, it's not the mini lathe, but more like the next size up. It's just coming up for one year old and has had a fair amount of hobby work done on it, but nothing major and never really been 'put to the limit'....

Right, the problem I have.... Last night, and I have noticed this happening a lot more recently, I put some (aluminium) bar stock in the chuck, 20mm dia and about 60mm sticking out. It's new clean bar stock with no knocks or dents in it. I faced off the end no problem, but needed to turn it down to 18mm. Now I was only taking a light cuts, the first cut along the length starting at the end takes off a lot more on one side then it did the other and this got less and less one sided as I got closer to the chuck. It's as if the chuck is not holding the work parallel. Bit by bit I turned it down to 18mm, by this time it's an even cut all the way and far as I could tell with the micrometer the resulting part was parallel. I also needed to drill a 10mm hole down the centre, I used a centre drill first followed by a 10mm drill, this went fine with no obvious problems with alignment, especially with the centre drill.

I understand this was just bar stock and could be quite easily be not so round as it I would like, but it's just something that seems to happen more and more.

I have in the past had the chuck off and it's held on with 3 bolts. I have made sure the mating surfaces are spotless and clean and also made sure the surface of the jaws are spotless as well.

My next move is to take the chuck off and give it a dam good clean and see what that does, also going to order up a lump of ground steel bar as a test bar and get the DTI on it and see what is going on....

Any other suggestions or anything I have missed or should try..?

Thanks

Dave

brozier
03-15-2011, 07:33 AM
Is there a "register" to align the chuck with it's back plate? (an accurate step with matching step in the chuck)

If the register is not there or not tight you can tune out a certain amount of eccenticity by having the chuck bolts just nipped up and adjusting it with a soft faced hammer.

Do you have a DTI aka indicator or clock? it will make adjustment much easier.

Three jaw self centring chucks always have a certain amount of run out and this gets worse with age/abuse as the scroll wears unevenly.

For maximum accuracy the Four-jaw independant chuck is better.

Cheers
Bryan

willmac
03-15-2011, 07:39 AM
DaveC -

I assume that you are using a 3 jaw self centring chuck?

A chuck like this will not actually centre bar stock very accurately unless you are lucky, and if it does, it will usually only do so for one diameter of bar. In other sizes it will be a little bit off centre. This effect will be more marked when the chuck is lower quality. This may well account for part of your problem.

However, you say that the off-centre was lowest near to the chuck and progressively greater as you move away from it. That indicates that the chuck is also holding the work at an angle to the lathe axis. That could be caused by a number of different things, one of which is an error in the chuck mounting.

To get to the bottom of this you need to find out where the error is coming from. To do this you will need your DTI. Check the lathe spindle to make sure that it is running true. You will need to take the chuck off to do this. You should see barely a flicker on a sensitive DTI running against the spindle faces and the taper bore. If you do see significant run out then talk to the supplier.

Then clean and carefully re-mount the chuck and clock its back face, OD and front face without the jaws in. You will probably see a very small amount of run out, but if more than that there is something wrong with the chuck or the way you have mounted it.

Finally if all that is OK, clean and refit the jaws in the correct slots and ensure that the close on each other correctly in the centre. Then clock your test bar and see what it tells you.

Chucks like this can accumulate a lot of junk in the scroll area at the back of the chuck. If you think that might be the case, then strip and clean the whole chuck. If you haven't used it a lot, then this is less likely, as is serious wear.

A 3 jaw may seem to be quick to use, but unless you can complete a job in one chucking, the innacuracies that are inherent in it work against you. A 4 jaw chuck is cheaper, grips better and is more versatile. You can learn to centre work in it very quickly and after you have got that sorted out, you won't use the 3 jaw as much.

DaveC
03-15-2011, 07:55 AM
Hi Guys, thanks for the responses...

Yes it's a 3 jaw..

There is a register and also a mark on the side to make sure it's always mounted using the same bolt holes...

I do have a DTI and I am just about to order some ground steel bar so will be able to take a look at what is going on. I will take the chuck off and but the DTI on the spindle as suggested and see what going on there and while the chuck is out I will give is a clean, especially the scroll area...

I do have a 4 jaw and have used it a couple of times but never thought of using it for just normal work, maybe I should give it a go..

Thanks again for the help.

Dave

firbikrhd1
03-15-2011, 09:33 AM
A tiny but or chip in the threads or between the mating surfaces of the spindle and chuck can cause runout. You may already have check for this but I thought I'd mention it.

Carld
03-15-2011, 09:39 AM
If everything is tight it sounds to me you have loose spindle bearings.

Put a dial indicator pointing down on the top of the OD of the chuck.

Then lay a piece of wood on the bed and use a stiff bar of steel to pry up on the chuck.

How much did the chuck move up when you lifted it?

aboard_epsilon
03-15-2011, 12:07 PM
First you need to use the tail stock for support ..
put bar right into the chuck so it only sticks out less than an inch.
drill centre hole for your tailstock centre

pull bar out ..

then watch this vid ..watch it right to the end ..as it tells you how to align bar stock as well

http://www.youtube.com/watch?v=tn7A9PqNftY

you also want to do some adjustments on your tailstock ..so that it is right ..do not assume a tailstock is right ..they probably will not be right out of the box new.

there are other vids on you tube that show how to get your tailstock aligned.

best way i find is just to keep cutting on a thickish bar ..and measuring..and adjusting ..after the initial ruler trick

all the best.markj

PeteF
03-15-2011, 03:08 PM
Dave, how much did you have sticking out from the chuck? You haven't said as much, but I suspect it was too long. Also, even with a reasonable 3 jaw, never mind one that's a bit worn, it's possible to have a piece sticking out from the chuck jaws, but not be clamped perfectly parallel. You can do the jaws up moderately firmly, then knock the work concentric (parallel) at the far end, then fully do up the jaws. On the other hand if you have a long piece sticking out, and do up the jaws just any old how, the chances are it will do precisely what you described.

Pete

Carld
03-15-2011, 03:43 PM
He said the stock is 20mm(,7874") dia. and 60mm(2.3622") sticking out of the chuck. In my experience a piece of aluminum that diameter sticking out of the chuck that far will not bend while cutting. If he cut a taper it is because the spindle bearings are loose.

He said that several passes produced a straight finish cut. The thing that made me chuckle some was, QUOTE "the first cut along the length starting at the end takes off a lot more on one side then it did the other and this got less and less one sided as I got closer to the chuck"

Sooo, it's taking more off one side than the other!!!!!! How can a lathe take more off one side of a shaft than it does the other???? If the DOC is .020" then it takes .020" off the front side but the total taken off the diameter is .040". It can't take more off one side than the other, it is physically impossible even it is cutting a taper.

PeteF
03-15-2011, 06:48 PM
He said the stock is 20mm(,7874") dia. and 60mm(2.3622") sticking out of the chuck. In my experience a piece of aluminum that diameter sticking out of the chuck that far will not bend while cutting. If he cut a taper it is because the spindle bearings are loose.

Sorry Carl, my bad, I was in a rush to get out the door and missed that. No the aluminium doesn't bend while cutting, the taper is as a result of not being positioned in the jaws correctly. In my experience, loose bearings will tend to produce work that is out of round rather than the symptoms that have been described.


Sooo, it's taking more off one side than the other!!!!!! How can a lathe take more off one side of a shaft than it does the other???? If the DOC is .020" then it takes .020" off the front side but the total taken off the diameter is .040". It can't take more off one side than the other, it is physically impossible even it is cutting a taper.

Yes it IS possible and I, and countless others, routinely do it by setting work in my 4 jaw eccentrically. As I said, if the work isn't sitting in a 3 jaw correctly, this is precisely the symptoms one will experience, particularly with a worn chuck at the end of stock. The stock is moving eccentrically at the end, and the symptoms will be precisely as described by the OP. The solution, as I said above, is to ensure the work is set correctly in the jaws, and they are tightened correctly. A good chuck will help pull the work concentric as the jaws are tightened, but a cheap or bell-mouthed chuck needs a bit more care when setting work. It's not a big deal, and simply a good habit to get into regardless.

Pete

lane
03-15-2011, 07:19 PM
My take on the subject. Is the stock sticking out of the chuck was not running true to start with.and wobbling on the outer in. so in cutting it when the wobble is toward the tool,it cut is deeper when it rotates 180 degrees that side is farther away from the tool so less is cut of.. He did not tap far end true before starting to machine. Just sticking a piece of stock in a 3 jaw chuck it will not necessarily be true with out tapping the far end in line.

DFMiller
03-15-2011, 07:20 PM
If the lathe is cutting a taper which it sound like it is there may be some things to look at. I am not expert but I was able to improve my Myford clone's performance a few months ago by doing the following.
One checking preload or sloppiness in the bearings. Helped a bit for me but not the final solution. Changed out the bearings.
Check the chuck or spindle is not flexing.
Check to see if the headstock is aligned with the ways. There are two things that affected my machine.
1) Level the machine. Using a level then shimming if required.
2) Adjust the position of the headstock. In my case there where 4 socket head cap screws inside the headstock. Slacking them off a touch then using a adjuster (Dead Blow Hammer) brought things back into line.
I was happier after doing this and machine does not cut the taper any more.
YMMV
Dave

PeteF
03-15-2011, 07:33 PM
My take on the subject. Is the stock sticking out of the chuck was not running true to start with.and wobbling on the outer in. so in cutting it when the wobble is toward the tool,it cut is deeper when it rotates 180 degrees that side is farther away from the tool so less is cut of.. He did not tap far end true before starting to machine. Just sticking a piece of stock in a 3 jaw chuck it will not necessarily be true with out tapping the far end in line.

Precisely. I think some feel that you can shove material in a 3 jaw like they're sticking a drill bit in a keyless drill chuck and it will turn as concentrically as the chuck will allow. Often that is indeed the case, but not always, and the work should always be checked (and if required adjusted) before fully tightening the jaws.

Mcgyver
03-15-2011, 09:12 PM
He said the stock is 20mm(,7874") dia. and 60mm(2.3622") sticking out of the chuck. In my experience a piece of aluminum that diameter sticking out of the chuck that far will not bend while cutting. If he cut a taper it is because the spindle bearings are loose.
.

or there's twist in the bed....but didn't he say it did not cut a taper?



Sooo, it's taking more off one side than the other!!!!!! How can a lathe take more off one side of a shaft than it does the other???? If the DOC is .020" then it takes .020" off the front side but the total taken off the diameter is .040". It can't take more off one side than the other, it is physically impossible even it is cutting a taper

?? :confused: ....it was the first pass and the work was eccentric so it would be taking more off one side than the other...on the first pass anyway, after that, i agree with you.

Dave, you've got the right idea, get a test bar and start figuring out where the problem is using the indicator, heck could have been bent stock. All three jaw chucks will hold the work somewhat eccentric (you use them when if doesn't matter OR you do all ops at one setting on the 3J...use the 4 jaw or collets or best, between centres when it matters) however even if eccentric, the work axis should still be reasonably close to parallel to the lathe's.

That it turned parallel suggests twist in the bed has not caused the problem so the problem should be in the chuck or chuck mounting or was bent stock...but you need to get the indicator going to start figuring this out. If you have it apart, mount the backing plate and indicate the face of it....if its way off it'll be riding sort of like a swash plate which will produce what you saw - where the run out diminished as it got closer to the head stock. Then try indicated the face of the chuck (when back together), they both should be pretty much zero. If it's not that, or a bit of dirt between the chuck and backplate, then the stock was either bent or the error is in the chuck. Don't forget a chuck could have zero run out....at at one location. Measure a few inches down and if there is still the same run out, things are good, otherwise the work axis is angular to the lathes that isn't good.

Also, quantify where you can....a light cut could have been .001" in which case you'd see that happen just because the stock isn't that accurate :D

darryl
03-15-2011, 09:22 PM
Sounds to me like the workpiece was wobbling before the cutter ever touched it. If that's true, there is no mystery as to why more material came off one side than the other. The mystery would then be why would the OP not have realized this would happen-

So I have to assume that the workpiece didn't wobble to start with, but that something was letting it flex during the cut. So the whole system needs to be looked at in a workable order. Start with the spindle and its bearings. Does it run true, and does it stay in place under side loading- you may need to mount an adapter in the spindle taper so you have something to push against, and you'd indicate on the shoulder. Rotate the spindle by hand and do the up/down/front/rear pressure test at different rotations of the spindle. Ideally, the indicator mounting should be on the headstock somewhere. If it's on the bed, you can be fooled by play between the headstock and the bed. This is the second test- if you determine that the spindle and bearings are tight, as in very little deflection throughout the first test, then remount the indicator on the bed and run the test again. Whatever differences show up are pointing to flexing between the headstock and the bed.

If you get past that test, then you look to see if the backplate mounts well, or if it allows some extra flex that shouldn't be there. There can easily be a particle mashed into a mating surface that interferes with a good solid mating of parts that would mount to the spindle. If you pass that test, then mount the chuck to the backplate and indicate the body of the chuck while doing the up/down/front/rear force test. If all is well, the deviations you read should not have increased by much at all.

Finally you are at the point where you can check the chuck itself. If the body of it runs true, clamp a workpiece into it and true it up. Disregard any uneveness in the cutting until you have arrived at a trued surface. You can do your multi-directional side force tests again, this time indicating on the trued surface you just made. This is where you get to see if the chuck jaws can hold something in position, even if it is offset in some way. If your indications show that you lose centering when side forces are applied, then your chuck is damaged.

You can skip a few steps to start with and begin you testing by indicating the chuck body. If no stray deviations develop and flex is low at this point, you could reasonably say that everything to the left of that is ok. You don't need a test bar for any of this, just some extension that you can use as a handle to apply the various directional forces. Anything round that you can mount in the chuck will do, or if you have to go back as far as the spindle, they you need something that will insert into the spindle, and be held with a drawbar.

I'm trying to think of when I actually needed a test bar- I don't think I ever had a need of one. It will be important at times to know that what you're chucking is actually round- hot rolled or cold rolled is not really round. A shock absorber rod should be pretty good, or anything where you can tell it had to be machined to get to be what it is. What they call shafting is usually pretty good.

So far the premise has been that the workpiece is moving around where it shouldn't be. You could also be having a problem with the cutter moving in ways it shouldn't. Now you're looking at a whole 'nuther string of things. Tool, holder, post, compound, crosslide, carriage. You can do a string of tests on all this by mounting the indicator to the chuck. Without moving the spindle you can exert forces against these other parts to see where some unusual play might have developed, or maybe was there from the start.

Maybe I've gone on a little extra with my ramblings, but I feel it is important to characterize your machines, to know what to expect from the various areas where relative motion can occur. It's definitely worth the time to run through all this.

Carld
03-15-2011, 11:18 PM
Hmm, well since he was using a 3 jaw chuck I assumed it was running fairly true because he didn't mention any runout in the chuck, just that it was turning a taper. Having runout in the chuck jaws or mounting would cause it to more off one side than the other.

He may have off center jaws and the jaws not holding the work straight. If he has that it should show the shaft moving out of round and wobbling at the end.

Is that what it is doing Dave?

DaveC
03-16-2011, 06:15 AM
Wow, thanks guys, there is some very helpful information there, thanks to all for taking your time to respond, I will definitely be going through what has been said.



Hmm, well since he was using a 3 jaw chuck I assumed it was running fairly true because he didn't mention any runout in the chuck, just that it was turning a taper. Having runout in the chuck jaws or mounting would cause it to more off one side than the other.

He may have off center jaws and the jaws not holding the work straight. If he has that it should show the shaft moving out of round and wobbling at the end.

Is that what it is doing Dave?

Yes, this is just about what it is doing.

I found some 15mm silver steel in my rack last night, assuming this was quite straight I mounted this in the chuck, with the DTI close to the chuck there was hardly any movement in the DTI when turning the chuck by hand, moving the DTI along the bed, about 200mm away from the chuck now gives a 0.1mm fluctuation in the DTI.

Did not get time to investigate any further, but this weekend I am going to take the chuck off and check out the spindle and mounting plate with the DTI, but I think I am going to but my money on the chuck and some chips in the scroll area, well I hope so cos I can fix that....

Again, thanks for all you help,

Carld
03-16-2011, 10:04 AM
Dave, it sounds like the flats that hold the stock are not parallel to the axis of the spindle. They are pushing the bar out of alignment with the axis of the spindle. The way to fix that is to grind the jaws ID and there are several threads that describe how to do that. I am sure several will post their way of doing it now so I won't offer one.

Richard Wilson
03-16-2011, 01:03 PM
[QUOTE=DaveC]Wow,I found some 15mm silver steel in my rack last night, assuming this was quite straight I mounted this in the chuck, with the DTI close to the chuck there was hardly any movement in the DTI when turning the chuck by hand, moving the DTI along the bed, about 200mm away from the chuck now gives a 0.1mm fluctuation in the DTI.

OK, so what you are saying is that there is no measurable runout near the chuck, but 0.1mm (.004") at 200 mm from the chuck. Which would mean that your aluminium bar sticking 60mm out from the chuck would have a runout of around .0013" at the extreme end. In my book, this is nothing at all, and I don't understand why it would take several cuts, even light ones to get rid of that amount of eccentricity. Even a cut of .001" would eliminate it on the first pass. Mind you, I wouldn't rely on runout measurements taken 200mm from the chuck using a 15mm silver steel bar. Its too flexible, and yes it is quite an assumption that the stuff is dead straight over that length, especially if its been floating around your shop for a while. You did deburr the end of the aluminium bar before putting it in the chuck didn't you? Did you ever measure the runout on the aluminium bar? Could you see it running out - its quite easy to detect few thou runout just by looking at it while the machine is running. Was the aluminium bar held by the full depth of the jaws, or just the tips?

richard

justanengineer
03-16-2011, 01:52 PM
Sounds to me like you need to use the tailstock and locks and try again. If you still have issues, then I would suspect you have a problem. In my experience its the things you dont think will happen/dont think possible (like material deflection due to cutting forces), and the duh//retard moment factors that will cause the most trouble.

Mcgyver
03-16-2011, 02:49 PM
OK, so what you are saying is that there is no measurable runout near the chuck, but 0.1mm (.004") at 200 mm from the chuck. Which would mean that your aluminium bar sticking 60mm out from the chuck would have a runout of around .0013" at the extreme end. In my book, this is nothing at all, and I don't understand why it would take several cuts, even light ones to get rid of that amount of eccentricity. Even a cut of .001" would eliminate it on the first pass. Mind you, I wouldn't rely on runout measurements taken 200mm from the chuck using a 15mm silver steel bar. Its too flexible, and yes it is quite an assumption that the stuff is dead straight over that length,richard

i agree especially if its .004 TIR (vs him really meaning .008 tir)....and for sure you cannot rely on drill round to be straight or round. Dave, with the AL that was a couple of inches long, how deep a cut were you taking?

Carld
03-16-2011, 05:16 PM
Just make a collar to hold the jaws and grind the damn things and get it over with. All this jacking around is getting you no where.

PeteF
03-16-2011, 06:37 PM
Pulling the chuck apart and checking it's clean sounds like a pretty good idea to me before attacking anything with a grinder :rolleyes: If I understood your initial post correctly, your chuck and lathe is about 1 year old, is that correct? If so, and it's just been used for hobby use, I'd think it unlikely it's excessively wear in the jaws causing the problem and I think you're quite right to look elsewhere for alternative answers to the problem.

Pete