Tailstock Offset Issue

[Martin ConnellyParticipant @martinconnelly55370]

The length of the dumbbell is not too much of an issue. I have a manufacturer's hard centre that fits in the spindle and I can put a chuck on or use a collet to turn a soft centre of whatever length is required so that the dumbbell works. As I have an effective max of about 18" between centres it works as required. I must admit though that I am of the "make it and worry about accuracy only if it doesn't fit" school of thought. It worked through 33 years as support engineer for fitters, electricians, pipe fitters, pipe manufacturing and controls assembly departments. The fact that, as is often stated, accuracy costs money the drive was often to engineer out the need for accuracy in the design and manufacture of parts and assemblies.

I usually only need tight tolerances when required for bearing housings or large rotating parts made for other people. When you are making parts that may be back a few times for modification as designs evolve then between centres machining makes sense. For most other parts I do collets and chucks suffice. I have never levelled or carried out accuracy checks on my lathe but it does weigh about 750kg with its base which only contacts the floor at 3 points.

Martin C

[Dr_GMJNParticipant @dr_gmjn]

Thanks Martin. I was assuming that the quoted accuracy figures meant in all tailstock positions and configurations. I was confused as to how such tight tolerances could be achieved on anything but a new lathe with perfectly set tooling.

Its clear (and should have been obvious to me) that setting up for very accurate work between centres Is similar to setting up a 4 jaw chuck – it needs to be done every time.

same with tailstock drilling. I’ve done some more research and it’s clear that for critical parts this also isn’t a one-time setup (for exactly the same reasons).

Im sure that this is obvious to the old hands, but as a relative novice I thought – it’s a lathe, set it up and it should be spot on (and indeed it will be for most jobs).

Thanks.

[Howard LewisParticipant @howardlewis46836]

I am not convinced that a drill chuck holds things concentric, every time that it is used. Or that every drill chuck is equally accurate.

For reaming, I made floating holder, to carry hand reamers and MT reamers, Hopefully, that removes the risk of an off centre reamer acting as a boring bar and cutting oversize.

If you start adding up all the possible departures from perfect size, or alignment, you are likely to conclude that precision is impossible, especially with the sort of machines, new or old, that most of us use.

(Unless using fitted bolts, we don't ream every hole, drilling is sufficiently accurate for just clamping things together. If needed, dowels will provide more accurate location. )

How many of use use thread gauges when tapping a hole, or producing an external thread? VERY few I suspect.

I don't , and my products do what is required of them.

The important question is: Is it fit for purpose?

If the answer is YES. Go ahead!

My navel is not sufficiently interesting to spend time contemplating it.

Howard

[Dave HalfordParticipant @davehalford22513]

Have you tried measuring the run out on your drill chuck?

[Howard LewisParticipant @howardlewis46836]

Which one?

They seem to breed like rabbits, in my shop!

NO. If, for a moment it was possible that a drill chuck would be precise, I would be upset to think that I had spent money of collets and collet chucks.

When you examine the parts in a drill chuck, most of which have running clearances, the likelihood of one holding work concentric, better than a collet, is small.

Three jaws sliding in slots, moved by a screw thread in a housing that rotates. Say 0.002" clearances, and you are probably looking at 0.006" variation, minimum.

I rest my case

Howard

[old martParticipant @oldmart]

It is difficult to measure the runout of a drill chuck with a Morse Taper. I would chuck up a piece of barstock and turn it to a size to fit the drill chuck. Tighten the drill chuck onto the bar and measure the runout of the MT. We have a lot of chucks at the museum, including several new British made Jacobs. The best tested is a 13mm R8 integral keyless Cutwel branded chuck, with 1/4" it is 0.0005" tir at the jaws and 0.001" tir at 2" out, and with 1/2", it is 0.001" at the jaws and 0.002" at 2" out. The others average at 0.003" to 0.005". You have to test with known true stock.

The tailstock took a long time getting it the correct height, the quill in line with the lathe axis, and the offset true. I turned a piece of 3/8" steel in a collet to a 60 degree end and did not disturb it during the testing. I found that a live centre in the tailstock was not good enough, even though it was a new Rohm. The testing was carried out with a 11" test bar with MT2 at one end, made in India. The test bar is better than my measuring capabilities. Fitting the MT end into the tailstock quill made checking the alignment in both the horizontal and vertical planes easier. The tailstock quill was checked at both ends of its travel. This tailstock quill is a new one that I happened to find on ebay, the taper in the original was not very good.

[Dr_GMJNParticipant @dr_gmjn]

At the risk of opening another can of worms, for routine use I like the idea of setting up using the test bar, with a DTI run along its length and aiming for 0" movement of the needle. Or the standard dumbell test piece which you use as a setting gauge again with a DTI in two saddle positions (ie it doesnt get machined). I can achieve this. Of course, it doesn't take into account slightly dodgy centres or tapers, nor tailstock height errors, but it seems way easier, quicker and – for routine work – effective as a turning test. I know cutting forces aren't considered, but if something was that critical then I should do a turning test as outlined previously. For everything else, the bar/test piece method I'd expect to be OK.

As I said, as a novice it's easy to read or be told about these tests, do them, correct as approprate and assume they're valid once set.

Maybe if any other beginners are out there and ask about these things, the first disclaimer should be that if you do the tests and re-set any aspect of the setup, chances are it will be wrong when re-checked.

Perhaps this was already pointed out at the begining for me, or it may be too obvious for comment, but I simply didn't pick up on it.

I think there's a book to be written on the subject – how about "The Amateur's Knackered Lathe"?

Anyway I really appreciate the comments and help. Thanks.

[old martParticipant @oldmart]

The dumbell test and a test bar compliment one another, both are needed, as neither can do all the tests on its own. If you do have a dead centre for the tailstock that is not perfect, then testing it at different positions can frequently find the sweet spot, and the centre should be marked accordingly. This is where a test bar between centres comes in handy. I have two of 7" and 11" long. They also are useful when centring the rotary table on the mill, and checking chuck runout.

[Author]

Posts