Internal thread & taper in same hole

[DellParticipant @dell]

Happy new year one & all I have a conundrum, I have a small four jaw chuck that I want to make a backplate to be able to fit it to my Pultra lathe , the end of the Pultra spindle has a taper & a 7/8” 26tpi thread , has anyone got any suggestions as to the best way to proceed?, I have some 55mm cast iron .
I have added a picture of a carrier for between centre turning because that is the sort of thing I need to make, just a shame I haven’t got 2 of them then I could re purpose one.
Thanks Dell

[JasonBModerator @jasonb]

Provided you put a reasonable run out groove at the bottom of the thread it should not be a problem, the thread is just there to hold the chuck on and the taper takes care of keeping things lined up. The groove will ensure the backplate is being pulled down fully onto the spindle taper rather than going tight at the end of the thread.

[Tony Pratt 1Participant @tonypratt1]

Maybe replicate the lathe spindle thread and taper on a piece of scrap using the carrier as a ‘gauge’ and then use this dummy spindle as the gauge for the internal thread and taper.

Tony

[Clive FosterParticipant @clivefoster55965]

As Tony says you really need to make a dummy spindle end as a test piece so you can see whats going on as you cut the taper.

If you have CAD take the time to carefully draw everything out so you know how much to take out on the final cut or three. Small amounts taken out of taper internal diameters have a disconcerting tendency to generate much greater longitudinal shifts than you expect. Probably because the longitudinal shift is proportional to the change in diameter whilst most of us tend to think in terms of depth of cut. Which is half.

Given the choice and chance I find it much less stressful and generally more successful to figure out everything on paper first and work to the numbers rather than to try and keep track of things on the fly.

A micrometer bed stop makes life much easier as when you get close you can simply generate the amount of longitudinal shift needed by adjusting the stop position and repeating the last cut. Makes life really easy if you have a taper turning attachment and pretty much essential if using a bench lathe without longitudinal feed. If you work by setting the cross slide at an angle arranging things so your cut either starts or finishes on zero, depending on which end has free space, makes it easier to keep on top of whats going on.

I’d bore bottom of taper and thread minor diameter to appropriate depths. Then cut the run-out groove about equi-spaced on thread and taper before cutting the thread. Once the taper is roughed out close enough for the test  piece to screw on about 1/3 rd to 1/2 way finish off using your micrometer bedstop to longitudinally shift the start point of your taper cut.

Assuming you have done your initial figuring out right its simply painting by numbers. About the only gotcha is making sure you taper cut always starts in the run out groove.

As ever the secret is taking time for proper preparation. Both machine and mental.

Clive

[DellParticipant @dell]

Yes I was thinking of making a dummy spindle and someone suggested using the carrier chucked up and indicated in , I think the 7/8”26tpi will be 55deg as the lathe is 80 plus years old.

I forgot I have two axis Rocketronics on my other lathe.

thanks for prompt replies

Dell

[Phil PParticipant @philp]

Dell

That one is obviously a P type………..Do you know if the 1750/70 has the same nose thread and taper ?

Phil

[Phil P]

On Phil P Said:

Dell

That one is obviously a P type………..Do you know if the 1750/70 has the same nose thread and taper ?

Phil

Yes I have just restored a 17/50 and the thread is the same, see attached picture compliments Tony’s website.

[Pete RimmerParticipant @peterimmer30576]

This is less difficult to produce than you might think. You don’t strictly need a matching taper angle if making one is proving difficult. Just turn one the closest you can manage to the angle by measuring then blue up the spindle taper and hand-fit the back plate the rest of the way by spinning it in your other lathe’s chuck and using a dowel and emery to adjust the taper so you get a half-decent contact. All it is is essentually a stop that happens to be tapered, its not a driving taper.

[Pete Rimmer]

On Pete Rimmer Said:

This is less difficult to produce than you might think. You don’t strictly need a matching taper angle if making one is proving difficult. Just turn one the closest you can manage to the angle by measuring then blue up the spindle taper and hand-fit the back plate the rest of the way by spinning it in your other lathe’s chuck and using a dowel and emery to adjust the taper so you get a half-decent contact. All it is is essentually a stop that happens to be tapered, its not a driving taper.

I would have thought that the taper would do the same job as a register on a larger lathe spindle,IE the threads holds it on but the register/taper corrects alignment.

[Dell]

On Dell Said:

I would have thought that the taper would do the same job as a register on a larger lathe spindle,IE the threads holds it on but the register/taper corrects alignment.

It does, but in order to do its job, you would only need line contact between the two tapered parts.

An analogy to what you have there is a valve in an internal combustion engine. When we lap the valves, they have nominally the same taper, but we stop lapping long before a full contact surface is evident.

By a little stretch of the imagination, you can draw a parallel between the spindle thread and the valve spring holding the valve onto its seat.

You could blue up the one you have and investigate the contact pattern.

[Clive FosterParticipant @clivefoster55965]

Have to disagree with Pete and DC31k. You do need a matching  taper if things are to work. As always with taper fixings it’s the taper that does the work. Threaded bits just hold it on. So it needs to fit.

The valve analogy is wrong. A valve merely needs to seal so proper contact over a small area is fine so long as its repeatable contact over a full circle in that nominal contact area. Which may not be in exactly the same place every time.

A non matching taper gives only radial line contact which is unstable and unpredictable in accuracy. Worst case it may damage your spindle. Effectively it’s the same problem of a spherical, or similarly curved, surface appearing to match a short, relatively large diameter taper. Stability of alignment by short tapers under load requires a flange fitting. The common D series  chuck mounts, and its relatives, being a more engineered example of solving the short taper, short contact length problem.

A screw thread is unstable during tightening up so the actual end point and alignment with line or small area contact on taper is inherently unpredictable. A parallel register constrains the thread instability as the chuck is tightened up giving descent results if adequately made. An inherently imperfect but affordable and good enough for lots of things. Historically there have been attempts to improve the system with registers bot in front and behind the thread. Much trouble and effort for gains that appear, in practice, mostly theoretical. It would probably be an interstice exercise in mechanical dynamics to analyse why a parallel register in front of the thread doesn’t work well yet an adequately long, like the Pultra has, taper does  The L-series long taper chuck mount is an elegant way of removing any influence of the holding thread on alignment.

I set up tapers by the old fashioned “inch per foot”, or its equivalent in metric, method. Given a decent indicator, a well aligned test bar to run it on and a sensible amount of travel its very easy to get silly high levels of accuracy.

Clive

[Clive Foster]

On Clive Foster Said:

Have to disagree with Pete. You do need a matching  taper if things are to work. As always with taper fixings it’s the taper that does the work. Threaded bits just hold it on. So it needs to fit.

A non matching taper gives only radial line contact which is unstable and unpredictable in accuracy. Worst case it may damage your spindle.

I set up tapers by the old fashioned “inch per foot”, or its equivalent in metric, method. Given a decent indicator, a well aligned test bar to run it on and a sensible amount of travel its very easy to get silly high levels of accuracy.

Clive

Clive you are spot on, I couldn’t agree more!

Tony

[Pete RimmerParticipant @peterimmer30576]

It’s ok to disagree in fact a lot of productive ideas come from disagreement, so long as you can do it constructively.

With respect though I was offering an alternative methodology to Dell who is clearly troubled by the prospect of producing this part. I submit that vee threads are inherantly self-centring, being nothing more than a spiralling 55/60degree taper. I’ve had screw-on chucks over the years which suffered a very sloppy thread fit and ample clearance on the radial register yet repeated well once they were seated up to the face register so there is some centering effect going on or they would never repeat.

My suggestion was that Dell could turn the taper as closly as his skill and tooling allows. Being turned in the same setup as the thread they would all be concentric. Then he could fit up the part onto a blued spindle and modify the fit if necessary by holding the backplate backwards in another lathe and using emery to modify the taper so that adequate contact was achieved. I didn’t suggest a line contact approach but I dont disagree that it would work it’s just far less than ideal. The guy has to work within the bounds of his personal skill and available tooling.

In any case, whatever approach is used it’s very easy to establish if it’s a success or not. Simply fit the backplate and then turn the OD and face. Remove and re-fit the part ten times and clock it up each time. If it repeats the job is done. If it doesn’t repeat then more work is required.

[Author]

Posts