What's new
What's new

Cutting taper with Tailstock

BoxcarPete

Stainless
Joined
Nov 30, 2018
Location
Michigan, USA
I know that using the tailstock offset from center will cause a bar turned on the lathe to have some diametric taper equal to: (offsetx2)/(distance between centers) when it is turned. I also know that this is best done between centers. What I want to know is how much taper can be done like this? What's the sane man's limit to the amount of taper that should be done before looking into using a taper attachment?

Also, I wonder how steep one could get away with making a taper using a 4-jaw chuck, around 1/2" diameter bar, maybe 6" long or so. It will bend with each rotation, so I'm guessing anything more than .010 or .020 over the 6" would be starting to impact straightness and stress the bar?
 
The limiting amount of taper will depend on the length of the part and the amount the tailstock can be setover. As to turning a taper using a chuck I consider it abuse of your equipment.
 
It is easy enough to get a couple of thousands taper with the chuck and tailstock accidentally. So how far can you go before it is abuse? Probably depends on the material as well as the diameter and length.
 
The limiting amount of taper will depend on the length of the part and the amount the tailstock can be setover. As to turning a taper using a chuck I consider it abuse of your equipment.

My initial thought was to cut an R8 arbor using tailstock setover to cut the taper, but I wondered if that angle would be feasible to do with that method. Sounds like it's not the taper itself that is limited, but there's no way to set up the short part of the arbor so that it can be turned on a taper since the overall length won't allow it.

And as to abuse of equipment I agree that it would be very possible to do so. My example above involves a reaction force of less than 10 lbf to deflect a 1/2" solid bar of mild steel by .005" to cut a small taper. Didn't seem out of bounds for any reasonably sized machine (cutting forces would be quite a bit higher than that), so I was wondering if anyone would do something mild like that via chuck, or go through the extra effort of setting up for turning between centers for any taper at all.
 
A trick I learned from PM many years ago was to use a boring head in the tailstock to achieve the set over. that way you don't lose your lathe settings and have very fine adjustment to dial in the taper. Rather than use centers, use a ball bearing in a countersunk hole, this allows the part to not stress your center. Do the same in the headstock- in fact, chuck up a piece of metal, drill with a standard center drill, do the same to your workpiece, put a ballbearing in between and then drive it with a dog off of one of the jaws.

Pete
 
I like mandrels to run better than .0003 so trying to bend a ½” bar for making a taper mandrel would be out of the question for me. Running off the taper attachment should take less than 10 minutes to set up.
The quick centers method is to chuck up a slug (even in a 3jaw) and turn a point using the compound for the 60* center point angle and using your fish to make the 60* correct…then tail offset the part ½ the desired taper. Yes dog off one jaw.

Yes with once having the chuck center you need only take a tad skim next time you wish to use it , or you can chuck it in your 4jaw if you happen to have your 4jaw on.

The skim-center runs closer than the lathe head-stock taper held center, and closer the a center in a collet.

Knowing how to be quick to go to center (and back to normal) is the/a key to being the good lathe hand. So many kinds of parts can be better/best with a between centers finish skim.
 
For an R8 arbor why wouldn't you just use the compound? Indicate in an existing arbor/collet and then set the compound to it using an indicator. Doing this with the tailstock would seem to be a royal pain. I'd also guess that's definitely too much taper for centers, and perhaps even the ball bearing trick.
 
For an R8 arbor why wouldn't you just use the compound? Indicate in an existing arbor/collet and then set the compound to it using an indicator. Doing this with the tailstock would seem to be a royal pain. I'd also guess that's definitely too much taper for centers, and perhaps even the ball bearing trick.

I think it is 3.5" per foot so yes to JC's good advice using the compound.
 
Buck - You can find, buy, install, and set up a taper attachment in 10 minutes? That's impressive. :D Part of my reasoning for this thread is to help me decide if I have a need for one or not. My answer is still "I'm not sure" but as I get more involved in making things for the lab (and the needs of the lab itself grow with me) I will be better able to make an informed decision if I ever do.

JC - I agree that indicating an existing collet on the compound is probably the best way, but after considering and deciding against tailstock offset as a viable option, it left that nagging question... R8 is too much taper to cut with tailstock, so there is a line to be drawn. Where exactly is it?

Thanks for the replies, keep 'em coming!
 
QY: [Buck - You can find, buy, install, and set up a taper attachment in 10 minutes?]

No just if you have a taper attachment already.
I would put centers in the part to run the whole part between centers. Turn the straight and then cock the compound to turn the taper. leave a radius at all the corners. It would run near dead on...and later years if wore out could be re touched/trued pretty darn quick.

To be fancy would recess the ends 1/32 to help protect the centers from bumps and recess the centers a bit.

I don't like to off set the tail because that is not so good for the part centers but you have to do what you have to do sometimes. Getting tail back to center has never been a problem but some guys don't like that job.

You said [before looking into using a taper attachment?] I thought you had one..my bad.
 
Trying to hold the work in a four-jaw and offset the tailstock for a taper sounds like trouble (to me). You risk:

1) Walking the part out of the chuck as it's torqued around each rotation
2) Fatiguing the stock as it's held rigidly in jaws and forced to oscillate around by the tailstock. I broke a 1 1/2" dia shaft on a pto chipper with just this sort of wobble

Turn between centers only, please.

Limitations on how much? 1st is how far over you can set your tailstock (or a boring head in your tailstock). 2nd is the effectiveness of the center holes in which your centers hold the part. Beyond a certain offset your 60 degree centers are no longer contacting the center holes in the work sufficiently. Keep offsetting further and you'll throw the work out of the centers. How much is too much? Probably depends on the angle and depth of the center and center holes, how aggressive a cut, how well you lube the center, and long you intend to let it wallow around like that.
 
Some taper attachments are way easier to set up than others. Alas, I don't have one at all. The ball bearing trick is a good one. Also look into type R centers as they tolerate offset turning much better. Just harder to find. I'd use the compound, though getting it exactly right can be difficult. At least it is for me no matter what method I've tried. Close is easy, dead on isn't.
 
Another way to cut a taper with no taper attachment is to use a template at whatever angle you need, then put an indicator on it and maintain the zero as you feed. The first time I saw such a set up the fellow was doing an oilfield taper thread. Somewhere around 7 degrees taper, about 4 threads per inch up to a shoulder. How he kept the eye on the indicator and did not crash the shoulder is a mystery to me. He was a lot better man than me. With pipe thread not up to a shoulder it has worked fine for me. I would not want to try a gauge taper on it but sure is handy for less persnicative stuff. The photo is a mock up for demo purposes, but I made a lot of internal pipe fittings that way. This template has the pipe taper ground in on a sine plate, so I do not have to indicate the taper in just put it in the tool holder.
taperthread1RS.jpg
 
Another way to cut a taper with no taper attachment is to use a template at whatever angle you need, then put an indicator on it and maintain the zero as you feed. The first time I saw such a set up the fellow was doing an oilfield taper thread. Somewhere around 7 degrees taper, about 4 threads per inch up to a shoulder. How he kept the eye on the indicator and did not crash the shoulder is a mystery to me. He was a lot better man than me. With pipe thread not up to a shoulder it has worked fine for me. I would not want to try a gauge taper on it but sure is handy for less persnicative stuff. The photo is a mock up for demo purposes, but I made a lot of internal pipe fittings that way. This template has the pipe taper ground in on a sine plate, so I do not have to indicate the taper in just put it in the tool holder.
View attachment 257209

This sounds like the kind of crack-brained scheme I would come up with and then discard as impractical. You've done me a disservice by telling me it's possible! :D
 
I knew a fellow who had precision ball nose followers to mimic his cutters and so would trace a pattern or an existing part with using a manual mill with rotary table, He worked at an areospace company doing this and then opened his own shop doing much the same..
Still I don't think much of following an indicator to trace a taper on a lathe.
The between centers and compound angle still seems best. With not exact taper match and being perhaps <.002 off angle a crocus cloth finish to blue-in would still be closer that any/most other methods. IMHO.
I have done between centers bearing fit diameter with the last .0005,.002 a crocus cloth finish with using a lathe or set up I did not trust for a skim.

Re:[ It will bend with each rotation] is not a good lathe practice. Just the spring back would mean you could not count on a straight part or intended angles when released.
 
This sounds like the kind of crack-brained scheme I would come up with and then discard as impractical. You've done me a disservice by telling me it's possible! :D

Sorry about that Chief! Hope it did not ruin your day. I made a whole bunch of PVDF plastic pipe fittings that way. Might faster to set up even if you have a taper attachment.

Added info. When I was doing those production plastic fittings resetting the zero was necessary when you moved the compound. If the template was mounted on the cross slide the indicator reading would be independent of the compound. I intended to mill a slot on the cross slide to mount the template with a couple of 1/4 20 screws to hold it in place. Did not have a mill in those days and the customer went out of business when I had a mill. Only occasionally make a pipe thread these days so the effort would be lost. My HLV-H already has tapped holes on the cross slide so it would be easy to do there.
 
Last edited:
Re:[ It will bend with each rotation] is not a good lathe practice. Just the spring back would mean you could not count on a straight part or intended angles when released.

Yeah, it'd be useful for cutting a hyperbolic arc if you didn't overstrain and fatigue the material at the root. Most of the time, nobody would want a mildly hyperbolic arc though, and as you mentioned the spring-back would ruin the shape anyway.
 
Another way to cut a taper with no taper attachment is to use a template at whatever angle you need, then put an indicator on it and maintain the zero as you feed.

I've cut pipe threads like this, they sealed fine.

I've also done other shapes like matching ball and sockets that didn't have to be super close. I'm sure the form was within .005". Hit the power feed and keep the needle on zero, it isn't hard.
 
Agree with Nmbmxer.

Have done this using a 2D template held in the tailstock. Clamped a piece of bent rod to the toolpost sticking out at the same depth as the cutter and simply traced the rod against the template by eye as the carriage traversed under power feed.

Was great for making hand wheel handles.
 








 
Back
Top