Fitting a new chuck

Hello All. I hope you had a great Christmas and 2018 is a great year.

I was given a new Shars 5" chuck and mount for my 9" South Bend. The mount has to be machined for the new chuck. On closer inspection the threaded boss that seats against the spindle is too long and so only allows about 2 threads to engage at most. I am going to screw it on back to front to machine it down to size with a washer between the mount and spindle shoulder as the mount face does not run up against the shoulder. My old Buck chuck screws on and seats against the spindle shoulder. I am going to machine the thread relief on the new mount to match the old chuck.

Am I going the right way or are there any alternative ideas I should be doing?

Cheers

You got it.
But make your own "doughnut" to use as the spacer. Make it a nice sliding fit and wide enough so that you can cut down that hub as far as possible, and even represent the register if the plate is thick enough.

You got it.
But make your own "doughnut" to use as the spacer. Make it a nice sliding fit and wide enough so that you can cut down that hub as far as possible, and even rebore the register if the plate is thick enough.

Save your donut for future use.

I measured the new plate. The 5" dia. part is 19mm/3/4" thick - on the Buck it is 12.7/1/2".
The 'register is 3mm/1/8". I assume I want to machine things so that the spindle face sits just behind the face of the mounting plate.

Screw on back to front fix ..then you may need to skim the the face at finish.. that is a snap.
Any way you can make it an adjust true..
I don't fully get what you need and intend so be sure before you cut..

Seems a protruding shoulder is long towards the spindle nose and so only allows 2 threads to engage.

Sounds like iwana10k knows what you are doing...

he is doing what IMO should be done with every chuck backplate.

the spindle nose on the 9" is only about 7/8" long. It has about 5 full threads(8tpi) and the register is about 1/4"

Most universal backplates are about 1.250" thick overall so there is a lot of hang out past the spindle nose...reducing that thickness down to just enough to fully cover the spindle(~.875"ish) reduces that overhang to the minimum.

I should add that when you cut down the hub to also take a very light truing cut over all the surfaces so there's no chance of seeing a slight wobble when you flip it back around the right way, and screw it on and off a few times to "seat" it in before fitting up the chuck to it.

And perhaps blue it up to see it hits same all the way round at the face.

Seems like a lot of work to do and end up with a shars chuck.

Actually I am betting he got the one with the 2 piece jaws, my buddy got one with one of their AXA clones and it ain't bad at all.

Not what you'd want on a Hardinge but on a SB 9" that's all you need...better to spend the money on a nice 4jaw.

I just ran into this the other day.

I recently bought a faceplate(TPP-100N) for my 10K, literally out of a trunk in a parking lot. I was concerned about the number of threads engaged with the spindle.

I bought the lathe with the SB 5" 3 jaw, a Skinner for South Bend 4 jaw, a dog plate, and a Bison Set-tru 5C collet chuck. So I pulled them all out and started taking rough measurements (Mitytoyo caliper). The results were somewhat disturbing:



The 3 jaw, the new face plate, and the dog plate all seemed in the same ballpark with over 4 threads of engagement. The 4 jaw only has 3.15 threads engaged, and the heavy. overhung 5C collet chuck only has 2.9 threads of engagement.

I've used the 4 jaw extensively, and it appears by the invoice to have come with the machine, so apparently 3.1 threads engaged is OK.

It is the chuck with reversible jaws. My lathe is not in mint condition. The bed is worn,as is the cross-slide leadscrew(possiblythe next upgrade).The spindle and bearings are in pretty good shape for a 1951 machine. The Buck chuck jaws are very worn and I do not have the outside jaws. Working out when to replace it is part of the problem. Right now it fits my budget and in my part of the world used lathes are in short supply and not cheap. When I read some of the posts I get envious of the machines you blokes have.

Mr.Wheels . 3 to 4 threads is all you need to develop full strength - look at how many threads are in a nut. When calculating the number of bolts for blind flanges we sized the bolts to hold the pressure using the bolt diameter. No allowance was needed for the threads or nuts. When I was an apprentice one of the trick questions was to be asked which bolt under tension would break first - the one with one or the one with two nuts.

gwelo62, thanks! Thread strength is an interesting topic. There are a lot of rules of thumb out there that give surprisingly low numbers of thread engagement for full strength. The simple formulas I've found suggest that the nut thickness should be .47 times the diameter for balanced material properties, and the same multiplied by the ratio of the bolt yield strength divided by the nut yield strength if the materials are unbalanced. For steel/cast iron I think this multiplier is about 3.

The interface between the chuck and the spindle is more complicated than a bolt. We're not just pulling lengthwise on the threads. The threads center the chuck on the spindle, and develop the centering force by the wedging action of the threads against the resistance of the shoulder. There is a big mismatch between the material strength (hardened spindle vs. cast iron), and the spindle is hollow, so it can compress more than a solid form would compress.

There's a calculator at http://www.boltplanet.com/ If you use 8UN and 1 1/2" dia, three threads have a shear stress area of .967 inches. Using three thread engagement, and a 5000 lb preload, the shear stress is about 7500 psi, about a third the shear modulus of CI. It takes about 185 ft-lb of torque to develop this tension. I don't think that even my most badly stuck chucks took this amount of torque to remove.

It also follows along on the simple formula, as it suggests some pretty long engagements to protect the cast iron nut on a steel bolt.

South Bend must have thought that three was enough, as my 4 jaw Skinner that came with the lathe according to the invoice has only 3.1 threads engaged. But it is a pretty light chuck.

It would be fascinating to have someone with FEA skills take a look at this situation. What is the effect of the tool force on the thread stresses, particularly when turning without a tailstock center? What loss in concentricity do you get as cutting forces increase due to movement along the incline of the threads? What is the compressive force against the spindle shoulder? The stress and strain in the spindle?

All this discussion really makes me jealous of the folks with big taper or D series mounts.

Building broach bars the screw thread length was a serious problem. The broach inserts being long cutting toothed cutter bars were top ground to less height at sharpening so the length of hold down bolts would get too long. Time was important so one could not study everything like engineering the package. One trick I employed was to count turns when tightening.. Often 7-11 was the factor.. Sp 7 to 11 turns would make good most imperial screw sizes. 28-16,1/2 -13.5/8-11 ¾ 10.. yes I would mot measure every bolt..But at the start of a section I would line up the bolts ,then count the first bolt in that area.. Yes for speed we would use a spin gun to fast put the bolts in place.

*I understand that one can not go buy the best possible answer to every situation..That is the glory of a decent lathe as it can make almost anything with having some stock and some talent.

Also having at least one HSS tool bit..One can make just about any odd-ball special bit that is not in any catalog.

OT: Broach bars: I would use the surface grinder and the bench grinder to add corner radius and change the lead teeth cutting width and add notches on lead teeth in the first to dull or first to fail area..so often getting 20 to 60% increased broach package life..Yes a broach bar might have 50 different cutters in the package, with every one perhaps +- .0005 to .004" to height and location with all tight and snug to what ever holds it from sliding..And having two bars for a machine often they would be slightly different, so that had to be figured in...

my reasons for doing it had nothing to do with thread engagement and everything to do with reducing overhang.

Just sayin.

Progress report. I took note of another thread and stripped my new chuck and washed it in paraffin. I washed out about a teaspoon of swarf! I used Vactra way oil when I reassembled it. I have started machining down the back plate.
I still don't understand why our Chinese vendors will not take an extra hour or two to 'finish' the job.

gwelo62 said:

I still don't understand why our Chinese vendors will not take an extra hour or two to 'finish' the job.

Click to expand...

Why should they care as long as you bought it?

Mr. Wheels - a question on 'registers.'
I assume 'register' is the portion of the spindle between the point at which the thread ends and the spindle shoulder. Looking at your table, The register in the chucks is longer than that on the spindle which allows the chuck to 'run up' against the shoulder before the thread 'tightens.' Am I on the right track?
Where about on the last thread do you measure from to determine the register length?

1/4-3/8"gives you plenty. Just gotta leave enough room so a chip doesn't wedge in at the end of the thread.

The chuck is now on the lathe. Thanks for all the input. I never realised how much longer the new chuck is compared to the old one - almost double the length. I think the reversible jaws are the reason. The next question is will the overhang affect the bearings?