Chuck runout and wobble

[Masterjuggler]

So this is more of a back to basics question, which suits my skill level on the lathe.

We seem to use the term runout for both the non-concentric type and the wobbling type. Like we say a lathe 3 jaw chuck has more runout than a collet, but we really mean a collet keeps the work more concentric to the axis of rotation than a 3 jaw. We also say that a drill chuck has runout when the central axis of the drill is skewed from the axis of rotation, causing wobble.

My question is on a lathe with a 4 jaw chuck, how do you mitigate the wobble type of runout when working on an almost finished part? Even high quality chucks become worn with use and can become slightly bell-mouthed, and soft jaws certainly don't help matters. Do you just default to using a tailstock center on parts longer than a few inches if you need to eliminate wobble?

 

[francist]

Neither my 3-jaw nor 4-jaw are new, and with the 3-jaw especially I can get a part that wants to wag around. I find that using my copper jaw liners almost always gets rid of that tendency. I don’t know if it’s because they kind of compress a little or if it’s because they take what might otherwise be a rocking motion in bell-mouthed jaws and turn it into more of a single point contact, it almost always fixes it. Or, maybe it’s just the way I hold my mouth when I use them...

-frank

 

[graham-xrf]

My question is on a lathe with a 4 jaw chuck, how do you mitigate the wobble type of runout when working on an almost finished part? Even high quality chucks become worn with use and can become slightly bell-mouthed, and soft jaws certainly don't help matters. Do you just default to using a tailstock center on parts longer than a few inches if you need to eliminate wobble?

There are lots YouTube videos on how to correct for bell-mouth. Beware, the lessons vary, but I assure you, you can figure up a way.
[YouTube is my favourite educational channel ]

There is the "3-blocks method", and the "disc way up back method), and the "expand into ring method". The last of those is not recommended except for the outer jaws. You want the spiral to be taking up slack the other way to clamp for inner jaws. There is the so-called "inertia method" where you slightly slacken the scroll, then spin at highest speed possible. Find the video, and get into catty critique mode. I would never do it!

The grinder varies also. Everything from fancy toolpost grinders, through mounted die grinders, clamped-on old routers (quite good), to even a Dremel strapped onto a piece of wood. They do all work, though some have more potential to mess up than others.

--> YT on Jaws Grinding
Note:1 In the Vintage Machinery first video, Keith Rucker used the "expand the jaws into ring", which is not the best way for grinding inner jaws, and he mentions it as "cheating a bit" because the chuck was quite new and unworn in its innards.

Note:2 In the xynudu video, he machines a groove into the very front face, to lock a piece of tube onto.

Note3: There is a "Locktite superglue" method of doing much the same trick, but without machining the lock groove into the jaws, so jaws front face stays unmolested.

Note:4 3 Jaw Chuck True-up, Amazing! from Batch562 requires 3 brass spacers milled up first. It apparently worked for him, but the method comes in for some reasoned criticism in some of the other videos. The same method shown in detail on an old Atlas lathe by Wide Vision Metal Fab will encourage you, though I might go for one of the others.

Note:5 How to improve the accuracy of your 3 jaw chuck by Suburban Tool Inc (watch?v=h1lIHmWY-04) is more or less ultimate. Very complete explanation.

As you may gather, I am contemplating doing something like this for my chuck, but at the present, I am just having a good time taking in and comparing what is done by others. That you can achieve excellent results with a DIY Dremel strapped onto ghetto too-post arrangement will have you smiling.

 

[graham-xrf]

@Masterjuggler : I forgot to mention that what is done for 3 - Jaw chucks is more straightforward than for 4 - Jaw chucks.
Clearly with a 4 - Jaw, they being adjustable, clamping them onto a circular reference is a bit moot, but there are two approaches.

One does the jaws one at a time, and the other looks much like the 3-jaw method. There is the imperative to have set the jaws without run-out, not using any of the "worn" parts. I am still checking out the options here.

Here is a video on re-working the 4-jaws one at a time by tying them push against a steel wire with all pressures in the same direction as if clamping onto a part. The video includes his "mistakes". You can see in the picture his final method of swinging the chuck by hand for most of the grind before a rotating final pass.

 

[Tozguy]

My question is on a lathe with a 4 jaw chuck, how do you mitigate the wobble type of runout when working on an almost finished part? Even high quality chucks become worn with use and can become slightly bell-mouthed, and soft jaws certainly don't help matters. Do you just default to using a tailstock center on parts longer than a few inches if you need to eliminate wobble?

Yes that approach can help reduce wobble. It depends on how much play is in the jaw channel. It is not a good idea to 'spring' the part however as it might not come out round.

If the work stays in the chuck for the whole operation then wobble caused by the chuck should not be an issue.

But when reinstalling a almost finished part in the chuck, using a centre in the TS and only gripping with the tips of the 4 jaws should allow you to dial the part back in pretty close without stress in the part. But it is much better to set the work up between centres. This makes it so much easier to remove, replace and flip if necessary.

 

[pontiac428]

The best way I can think of to handle an independent chuck is to regrind the jaws on a surface grinder. I have a 4-jaw that is bellmouthed pretty bad; it was ruining my work. I don't have access to a surface grinder, so I fired up the old computer and found my way to eBay for a new old stock chuck. Fixed all of my problems with the 4-jaw, that's for sure.

I did regrind the jaws of a 3-jaw scroll chuck with my wee little Dumore tomb thumb grinder using the ring method. It was slow work, but it improved things a bit. Still wasn't perfect and I was burnt out on it by then, so I replaced that chuck too. That fixed it complete and total, the new chuck works like a dream.

Replacing chucks is the ultimate solution. I'm not trying to say to go out and spend, but my experience showed me that good as new is the most attainable with a low-mile chuck. Worn out chucks are a compromise, and a well-polished turd is still a... well, you know.

 

[Masterjuggler]

Thanks for the replies guys, I'm definitely saving a lot of this information.

I've actually reground a 3 jaw chuck in the past on an old south bend lathe in the back of my high school woodshop class. I did mention bell mouthed chucks in my first post, but I was really focusing more on the inherent inaccuracy of using a chuck in the first place, whether it be 3 or 4 jaw. My current chucks don't need to be ground in.

Maybe it's just my import chuck's quality, but especially when using soft jaws, I'll indicate in a part in the 4 jaw, move the indicator over a few inches, and it'll be out unless I use a center in the tailstock.

@Tozguy You're right, I definitely have to get some lathe dogs or make them at some point. That's the gold standard of keeping concentricity and wobble away.

 

[graham-xrf]

Maybe it's just my import chuck's quality, but especially when using soft jaws, I'll indicate in a part in the 4 jaw, move the indicator over a few inches, and it'll be out unless I use a center in the tailstock.

Before you go there, check nothing has bumped, moved, stressed the main line out of level, etc. The align of the spindle to the bed.
Best is to remove the chuck, and put a MT? whatever is correct for your lathe test rod, and indicate to it. If you see the same symptom as for something mounted in then check. The chuck might not be entirely to blame.

Moving on the sequence, mount the chuck, and indicate to it as well. If you know there may be "rocking" front to back on a jaw, or if it its the "deep down back there" parts of the jaws that have worn, then sure, it needs the jaws to be sorted out.

One condition is that one jaw can have "front wear", while another can have "rear wear". I understand there is a subtle indication, possibly only good for guys who know the "feel" of the chuck when tightening. That is, a "spongy" feel as the jaws find tight. Good jaws, it seems, feel like a definite hard stop.

 

[Masterjuggler]

Before you go there, check nothing has bumped, moved, stressed the main line out of level, etc. The align of the spindle to the bed.
Best is to remove the chuck, and put a MT? whatever is correct for your lathe test rod, and indicate to it. If you see the same symptom as for something mounted in then check. The chuck might not be entirely to blame.

Maybe I'm thinking about this wrong, but if that were the case, I imagine the indicator would show a different diameter by the needle moving, but testing runout in the new spot would show the same needle movement. It would cause the lathe to turn a taper. Now to be fair that is definitely the case also LOL.

Moving on the sequence, mount the chuck, and indicate to it as well. If you know there may be "rocking" front to back on a jaw, or if it its the "deep down back there" parts of the jaws that have worn, then sure, it needs the jaws to be sorted out.

This is something I haven't thought about and I'll need to check. I always clean the faces of the chuck back and spindle before mounting, but I don't know exactly how well it was machined in the first place.

One condition is that one jaw can have "front wear", while another can have "rear wear". I understand there is a subtle indication, possibly only good for guys who know the "feel" of the chuck when tightening. That is, a "spongy" feel at the jaws find tight. Good jaws, it seems, fell like a definite hard stop.

Wow, I didn't even consider that to be possible. I doubt it's the case with my relatively new chuck, but I'll have to tighten the jaws down on some pieces of stock like in the grinding videos and indicate the inside faces.

 

[graham-xrf]

@Masterjuggler : While I have been reading up, and taking in lots of info on the subject via YT videos, I am not yet experienced in the actual implementation. That is why I am researching! My situation is like yours, in not knowing. You have advantage in having taken on the jaws of an old South Bend, even though it be from high school days! I have 2 lathes, both South Bend 9s, and a whole bunch of 4-Jaw and 3-Jaw chucks, all of which are going to get the checkout.

My posting was about what I would do, especially if an indicate was different near the jaws as compared to further out. The very first thing is to indicate the spindle, and it's alignment parallel to the bed. Even if a spindle has some little run-out, with some skill, that does not matter much. The 4-Jaws indication when mounting into the chuck allows you to get the work onto the spindle center, even if there is some (small) run-out on the chuck and it's back-plate.

Not so if the axis is not pointing exactly parallel to the bed. This can happen over time as any part of the bed support "relaxes", or moves. There are only 3 points that can supply the support, and getting a fourth or more to "share" the weight relies on some sag compliance, and long term stability. You never know which point is changing it's share proportion. The thing can go out of level.

However it happens, if it goes out of level, or develops a twist wind, the steel takes a "set", responding to the stresses. If you are about to take a grinder to a chuck, or bear the expense of a new one, the first thing is to indicate to the lathe test bar without the chuck, just in case you are blaming the wrong thing.

That said, chucks do wear, and the very high probability is that the alignment, and leveling is perfectly fine. It is just worth a quick check.