lots of lathe questions.

[thomas harris]

I have a small old atlas 618 lathe. It tends to turn tapers. I use a 3 jaw chuck and live center at tailstock when turning long stuff. The taper varies from set up to set up. Is this likely the chuck not locking straight onto the stock? I still am searching for a faceplate for it so I can try turning on centers. Perhaps this will alleviate some of the inacurracies. I can eliminate the taper and turn to within a thous. if I tweak the tailstock slightly, but I've heard this can cause the piece to snap back when it's removed, resulting in a "snake" with uniform diameter.

[tattoomike68]

Its hard to say whats wrong, it could be the headstock is not square to the ways and the tail stock is off too.

What happens when you chuck up a good strait peice?

[SteveM]

Can you try turning a long piece between centers? That will eliminate the chuck and point out if the bed is twisted.

Be sure the tailstock is centered. Put centers in the head and tailstock and bring them together. Look at them from the top and adjust until they are in line.

You could try holding the work with a collet, if your lathe takes them.

Headstock misalignment, while possible, is unlikely due to the fact that the headstock sits between the ways. Unless it is loose, that is probably not the problem.

The fact that it is inconsistent may indicate the chuck. Take the jaws out (be sure they are numbered and the chuck body likewise, as they have to go back in the same place). Look at the cotact face and see if it is a straight section or if it is tapered (i.e. a very narrow trapezoid). That would indicate that the jaws are worn (probably more at the front).

Steve

[thomas harris]

what kind of acurracy should I expect from a 3jaw chuck? The taper I turned last night was about .004. on a 12" piece. The tailstock appears to be centered on the chuck now. When I unscrew it, the live center point is dead into the hole in the end. Would this be checked by unscewing the tail, then checking the end farthest from the chuck for run out? I have a four jaw chuck coming in the mail soon. Seeems to me no matter what you do, the turn on centers will provide the most accurate results. The other way will always be limited to chuck accurracy. Is this correct?

[Bill Shields]

I would get a ground mandrel, put it between centers and align the headstock that way...

3 jaws are only as good as the scroll driving them. I have seen some that would hold 0.001 at a specific jaw opening...and some that were 0.030" out here and there.

For really accurate work, use a 4 jaw and an indicator and you can put it where you want.

[Harold_V]

I have a small old atlas 618 lathe. It tends to turn tapers. I use a 3 jaw chuck and live center at tailstock when turning long stuff. The taper varies from set up to set up. Is this likely the chuck not locking straight onto the stock?

No. How a chuck grips material is not related to taper, with the exception being if the chuck is not gripping the material properly and it's moving about in the chuck under the pressure of the cut. That would most likely manifest itself as a three sided turn, not taper. The importance of the chuck is how it relates to concentricity and perpendicularity of machined surfaces as compared to chucked surfaces, but otherwise plays no role in the outcome of a machined part. In other words, a chuck that has considerable runout can still turn out a part that is straight, round and concentric, so long as all features are machined in the same setup. Doing second operations on the opposite end of the same part, or repeated chucking of a part on the same end becomes troublesome when a chuck has runout. That's one of the reasons soft jaws are used. They correct for runout when properly prepared.

I still am searching for a faceplate for it so I can try turning on centers. Perhaps this will alleviate some of the inacurracies.

Turning between centers will insure a part that is not crooked, but it may still yield a taper. That should be addressed by leveling the machine, and setting the tailstock such that error is corrected. Considering machines may or may not be in good condition, the tailstock center may not be in perfect alignment with the headstock, yet it will yield a straight part. In a perfect world, the tailstock should be in dead alignment with the spindle.

I can eliminate the taper and turn to within a thous. if I tweak the tailstock slightly, but I've heard this can cause the piece to snap back when it's removed, resulting in a "snake" with uniform diameter.

That's correct, although when you force the material to run on an artificial center as you've described, it's very obvious in that the quill of the tailstock will usually be constantly moving in a small arc, yielding a part that is not straight, and not concentric with the center. To verify this condition, the tailstock should be relaxed with the part still held in the chuck. The tailstock end will now run eccentric. It is poor practice to run a center on a chucked piece, although there are methods to do so that will yield an acceptable outcome.

Your machine should be checked for level. Could be the bed is slightly twisted. It may also be worn and incapable of turning straight. If it has sleeve bearings, they may be too loose.

Harold

[thomas harris]

Thanks to the above poster. I see Onalaska... man did we catch some panfish on vacation there a few years back!

Back to the lathe... What is the rule of thumb for length that should require a center at the tail. I think someone metioned 3x the diameter once. Does this sound right.

Ooops... We were in WI.

[tornitore45]

Be sure the tailstock is centered. Put centers in the head and tailstock and bring them together. Look at them from the top and adjust until they are in line.

This technique never gave me more than "Close", even with a magnifier I don't think you can get the HS center and the TS center closer than 10 thou or so; may be 5 if the points are sharp.

The procedure I use is to chuck the indicator and indicate the TS hole on half a turn from front to back. This is cumbersome though.

Regarding the lenght to diameter ratio, between 2 and 3 is often quoted as norm. I have worked with 6 to 1 ratios easy with shallow cuts.
Also found that if I need to reduce the diameter, it is safer to reduce in sections from TS to HS but each section all the way to roughing dimension, then one light pass to equlize all the sections and one more to final dimension. This way the diameter stay large toward the HS until you cut away a short section.


Harold, using chuck and TS may not be proper but is often done, is quick and works OK for non critical parts.

You mention that If you have to do it there are techniques do do it right and minimize problems.
Could you please expand on those techniques?

Thanks

[Harold_V]

What is the rule of thumb for length that should require a center at the tail. I think someone metioned 3x the diameter once. Does this sound right.

One can usually coax 5 times the diameter in length, although with a small amount of taper. A lot depends on the material, the depth of cut, the feed rate, and the tool geometry. I suggest, assuming you are not familiar with machining, that you try to stay close to the 3 times diameter formula. That's particularly important if you don't have a clue about the capabilities of machine tools, depths of cut and feed rates. When things go wrong on a long, large diameter piece, you risk considerable damage to the machine. Long pieces are prone to climbing the tool, which usually leads to a major crash. Play it safe until you are familiar with your machine's characteristics.

I mentioned there are ways to run a center in conjunction with a chuck that will yield satisfactory results. You can do that by roughing your part near size, leaving about 1/16" for finish machining. You then chuck the material short, rough, advance the material in the chuck, rough, repeat, until the part is to length. The material should have already been faced when you started the process. Only when you've roughed the part and have it positioned for finish machining should you attempt to drill the center hole. Carefully allow the center drill to start the hole, so it doesn't deflect the part, which is likely to be running out a fair amount due to chuck error and stress having been relieved by the roughing operation. Do not attempt to force the end to run true----that defeats the purpose of this operation. When you have the center started, drill to proper depth, then do the finish machining. Removing the remaining material when you take your part to size won't have much of an effect on the part, so it will remain straight, or much better than it would had you done all your work with the center pre-drilled. If this makes no sense, and you'd like to understand the process better, please ask for clarification.

Ooops... We were in WI.

Three towns bear the Onalaska name, all of them founded by the same guy: Carlisle, a timber baron. His home still stands here in Onalaska, Western Washington, in what remains of the fairly large company town when the mill was active. The old stack and mill pond still remain from the mill. The mill, proper, has been gone for years, long before we moved here in '96.

I'm at a loss to explain his love for the name, but it sure is stupid in my opinion. A town near us has a great name----Mossyrock.

Harold