Thread: Cylindrical Square

Hi RC,

What you are saying is correct as far as mounting a "correct" cylindrical square on an offset center.

What happens to run out once you start turning? the run out is removed. The center will no longer be in the center of the work piece. Yes the centers still appear to draw a cone, but the work piece will be wider on one side that the other but the exact amount of the offset.(or more likely double, I havent had coffee yet).

Looks like I'll be turning some cyclinders today.

Stuart

Originally Posted by Stustoys

but the work piece will be wider on one side that the other but the exact amount of the offset.(or more likely double, I havent had coffee yet).

Looks like I'll be turning some cyclinders today.

Stuart

One centre is fixed though.. so you have runout at one end only, hence the cone shape...

Hi RC,

The center at the headstock end is fixed, its just fixed in an orbit around the spindle axis. The part still rotates around the spindle axis, unlike the tailstock end where the center of rotation can be offset from the spindle axis. Once the runout on the outside of the cylinder has been turned off there is now runout out of the cylinder at the headstock end as far as the center is concerned but no taper.

To use your example. 40mm bar with a centers in it placed on an center that is 5mm offset(and assuming we were taking 5mm doc). At the tailstock end you would take 5mm off all the way around. As you moved towards the head stock the cut would become more and more offset(?) until you were taking 10mm off one side(when the center was towards the front) and nothing off the other(when the center was towards the back). No taper but now the center in one end of your cylinder is offset by 5mm.

To put it another way,
Changing the offset from the spindle axis of a center in the headstock, does not change the distance from the center of rotation of the work piece to the tip of the tool.
Changing the offset from the spindle axis of a center in the tail stock does change the distance from the center of rotation of the work piece to the tip of the tool.


Stuart

But then it is not a cylinder any more...

Hi RC
If its round and has parallel sides, how is it not a cylinder?

Stuart

Originally Posted by Stustoys

Hi RC
If its round and has parallel sides, how is it not a cylinder?

Stuart

It doesn't have parallel sides so it's not a cylinder, therefore your question has no answer.

Try it & see, I agree with RC and I've done this myself. Not recently, but I have done it. You generally end up tweaking the TS offset to get rid of the conicity until you do have as close to a perfect cylinder as your lathe is capable of turning. The closer the 2 centres are to aligned in the beginning, the less tweaking you have to do.

Even if your lathe HS axis is pointing off, you can still turn accurately between centres (providing it's not hugely off anyway). BT, DT.

PDW

Originally Posted by PDW

It doesn't have parallel sides so it's not a cylinder, therefore your question has no answer.

Hi PDW

I believe you are wrong(well not all of it, some of it I agree with). I'll try and turn something up tonight but it will be pretty rough, failing that it will have to wait until Tuesday at the earliest.

Stuart

Stuart is 100% correct...in theory
Physically impossible in practice. At the very least from the damage a centre will do to it's mating surface while rotating.

Taking a 5mm cut still won't give concentricity. You will have to exceed 5 mm if we want to get real pedantic about it. Factor in tool wear, machine condition and bearing tolerances along with tool flex and it would be impossible to cut a parallel. As I say correct in theory.

OK Stuart is correct... Seems doing 13 hours of mad cow chasing has made my brain work properly...

Or to put it another way..... If I chuck something in the three jaw held by the tailstock centre it will still turn it parallel... Three jaws never hold anything concentric...

Hi Steamwhisperer,

Originally Posted by Steamwhisperer

At the very least from the damage a centre will do to it's mating surface while rotating.

Which center do you mean?

Originally Posted by Steamwhisperer

Taking a 5mm cut still won't give concentricity. You will have to exceed 5 mm if we want to get real pedantic about it. Factor in tool wear, machine condition and bearing tolerances along with tool flex and it would be impossible to cut a parallel. As I say correct in theory.

Well you could grind the 5mm off. But yes, 5.5mm if you like. I was just trying to demonstrate the maths. I only used 5mm as thats what RC(?) used, I doubt anyone has a center that is offset by 5mm thats its delibrate.

Stuart

Hi All,

Interesting discussion, I think there might be some confusion about the word "center" (centre if you must), so let's avoid the word...

Turning between centers, means that the spindle axis can be skewed a little, but as long as the center point of rotation (not necessarily the "center itself") is exactly the same distance from the ways as the center point of the rotation at the tailstock end, the axis of rotation of the work will be parallel to the ways.

So...

Restating what Stuart is saying, that is if the axis of rotation of the work is parallel to the ways then you will turn a cylinder. (not a cone).

That applies, no matter how the work is held (or driven), if we take an extreme case and offset the work at both ends, we can still turn a cylinder without taper, provided the axis about which the work rotates is parallel to the ways. In practice, getting the tailstock and headstock (center of rotation) aligned parallel to the ways is (as PDW says) the secret, and the hard part..

When you get into micron type precision, the tool and work deflection become major factors, (as Phil has said) I tried to avoid those by polishing the last bit rather than cutting, but you can get weird wave effects depending on the exact details of how you do the polishing...

Unless you have a lathe as good as .RC's, a cylindrical grinder strikes me as the best way to finish the job. (Or go and get some big diesel gudgeon pins.. )

Regards
Ray

PS.. I wonder how a tool post grinder would go..

Originally Posted by Steamwhisperer

Taking a 5mm cut still won't give concentricity. You will have to exceed 5 mm if we want to get real pedantic about it. Factor in tool wear, machine condition and bearing tolerances along with tool flex and it would be impossible to cut a parallel. As I say correct in theory.

Exactly what I was about to post, having just sketched it in Solidworks to confirm I was thinking correctly about it.

If the headstock centre is offset, but tailstock is dead nuts, you will indeed turn a taper UNTIL you reach the point where your depth of cut at the tailstock end EXCEEDS the amount of offset at the headstock end. At the 5mm offset Stuart used for his example, most likely impossible in the real world, for all the reasons Steamwhisperer mentioned. But if we are talking 0.1mm of offset at the headstock end, do theory and practice meet?

Originally Posted by PDW

It doesn't have parallel sides so it's not a cylinder, therefore your question has no answer.

Try it & see, I agree with RC and I've done this myself. Not recently, but I have done it. You generally end up tweaking the TS offset to get rid of the conicity until you do have as close to a perfect cylinder as your lathe is capable of turning. The closer the 2 centres are to aligned in the beginning, the less tweaking you have to do.

Even if your lathe HS axis is pointing off, you can still turn accurately between centres (providing it's not hugely off anyway). BT, DT.

PDW

I'm a bit confused by the last sentence here, but I'm going to assume you mean that by offsetting the tailstock you can compensate for a misaligned headstock?

In which case my question is, in this given situation where the centre is offset at the headstock end, how is offsetting the tailstock any different to advancing the cross slide? The workpiece will still be 'whipping' (for lack of a better word) at the headstock end regardless of where the tailstock is...

Does it really matter as long as the job is true.