Facing parallel between centres.

[Graham MeekParticipant @grahammeek88282]

"Facing parallel between centres" is the title of this post. Failure to get a good result is hardly a claim that "it does not work". Especially if the methodology of this process is not fully understood.

I do wonder how much of the of the aversion to this "between centres process", is based on disturbing the tailstock setting. It has always been a big No, No, with some individuals. The same is true when it comes to moving the topslide, when anyone advocates the set over top-slide method of Screwcutting.

Those who know their machine tool history, will know, " Dead Centres" were the basis of all turned work at one time. It was not until later on that the "Live Mandrel" was developed and the faceplate became available. Dead Centres are still used in cylindrical grinding, for most work. The lathe Mandrel on the lathe in the picture above was no doubt finished this way.

Regards

Gray,

[Martin KyteParticipant @martinkyte99762]

parallel faces that are faced slightly concave

I object to the description of parallel being applied to anything other than lines or flat planes. Parallel planes have rotational and translational symmetry.

I'm perfectly happy to describe the periphery of the two ends of the cylinder as being parallel when concave (or convex) by not the face.

regards Martin

[blowlampParticipant @blowlamp]

Posted by Martin Kyte on 14/09/2020 11:57:51:

parallel faces that are faced slightly concave

I object to the description of parallel being applied to anything other than lines or flat planes. Parallel planes have rotational and translational symmetry.

I'm perfectly happy to describe the periphery of the two ends of the cylinder as being parallel when concave (or convex) by not the face.

regards Martin

How about 'evenly spaced' or just 'offset'?

Martin.

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Martin Kyte on 14/09/2020 11:57:51:

[…]

Parallel planes have rotational and translational symmetry.

.

[ my emboldening ]

But surely that is what distinguishes them as planes [a special case of the spherical surface]

MichaelG.

[ZanParticipant @zan]

A while back in one of the magazines was a discussion about producing some form of solid square where both faced and had to be parallel. The method was to bolt down on the miller…..  Not the lathe. 
machine one end flat then invert and machine the other next check the dimensions for parallel and if there’s any error pack up the low side with a shim and re-machine. Repeat until u get the correct result.

this workpiece would need slots or holes in each end to clamp it down and the work would need to be stood on strips clamped to the table . There would need to be stops to locate the x and y positions and after checking for parallel before repositioning, this should be achieved without any significant rotation of the billet. The clamping holes in this case will not be any problem

Edited By Zan on 14/09/2020 12:38:32

[Martin KyteParticipant @martinkyte99762]

Posted by Michael Gilligan on 14/09/2020 12:21:54:

Posted by Martin Kyte on 14/09/2020 11:57:51:

[…]

Parallel planes have rotational and translational symmetry.

.

[ my emboldening ]

But surely that is what distinguishes them as planes [a special case of the spherical surface]

MichaelG.

Yes, exactly. For example I would not consider two concentric circle as having parallel circumferences or stacked cones either. I just think it's important in discussons to use accurate terminology so we all talk about the same thing.

Evenly spaced or offset as suggested works better.

regards Martin

[JasonBModerator @jasonb]

Posted by blowlamp on 14/09/2020 10:45:34:

So am I correct in thinking that you started this thread after you had actually completed the job

Martin.

Yes, if you read it Robin was seeking clarification about the advice previously given, he said nothing about wanting to remachine it between ctrs.

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Martin Kyte on 14/09/2020 12:56:35:

Posted by Michael Gilligan on 14/09/2020 12:21:54:

Posted by Martin Kyte on 14/09/2020 11:57:51:

[…]

Parallel planes have rotational and translational symmetry.

.

[ my emboldening ]

But surely that is what distinguishes them as planes [a special case of the spherical surface]

MichaelG.

Yes, exactly. For example I would not consider two concentric circle as having parallel circumferences or stacked cones either. I just think it's important in discussons to use accurate terminology so we all talk about the same thing.

Evenly spaced or offset as suggested works better.

regards Martin

.

Interesting [‘though not worth debating]:

You say ”Yes, exactly” and then describe how we disagree.

… It obviously worries you, and I respect that.

MichaelG.

[Martin KyteParticipant @martinkyte99762]

Hi Michael

I'm agreeing with you that it's the translational and rotational symmetry that distinguishes planes which is what you wrote.

regards Martin

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Martin Kyte on 14/09/2020 13:47:31:

Hi Michael

I'm agreeing with you that it's the translational and rotational symmetry that distinguishes planes which is what you wrote.

regards Martin

.

Sorry, Martin … I must have read it in haste

[feeble excuse] : I had just been summoned to go out for the afternoon.

Happy to know that we agree

MichaelG.

[Pete RimmerParticipant @peterimmer30576]

Posted by Graham Meek on 14/09/2020 11:28:54:

"Facing parallel between centres" is the title of this post. Failure to get a good result is hardly a claim that "it does not work". Especially if the methodology of this process is not fully understood.

I do wonder how much of the of the aversion to this "between centres process", is based on disturbing the tailstock setting. It has always been a big No, No, with some individuals. The same is true when it comes to moving the topslide, when anyone advocates the set over top-slide method of Screwcutting.

Those who know their machine tool history, will know, " Dead Centres" were the basis of all turned work at one time. It was not until later on that the "Live Mandrel" was developed and the faceplate became available. Dead Centres are still used in cylindrical grinding, for most work. The lathe Mandrel on the lathe in the picture above was no doubt finished this way.

Regards

Gray,

Well said Graham. Moving-jaw chucks were an accessory at one time. Lathes would be supplied with centres and drive dogs.

One should be able to produce exceptionally good parallelism working between centres.

[Robin GrahamParticipant @robingraham42208]

Posted by JasonB on 14/09/2020 13:14:34:

Posted by blowlamp on 14/09/2020 10:45:34:

So am I correct in thinking that you started this thread after you had actually completed the job

 

 

Martin.

Yes, if you read it Robin was seeking clarification about the advice previously given, he said nothing about wanting to remachine it between ctrs.

Thanks Jason, that is it exactly. I started this thread because I had been advised that the best way to skin this particular cat was to turn between centres. Before working on the riser I tried doing that with a piece of 4" diameter bar which I had to hand. The results were disappointing compared to the faceplate method and I wondered why. Not having had any formal training in machining (apart from 10 hours in the postgrad workshop at the Schuster Lab in Manchester, too many years ago!) I have to try and figure things out as I go along and take advice – this forum has been invaluable to me in that respect.

I now have a better grasp of the geometry at least, and shall make further experiments to see what I can achieve with my lathe.

Gray/Pete – thanks for the info about the history of of turning. I shall certainly try to get to grips with between-centres work.

A plane is a space spanned by two linearly independent vectors in my book. My mistake was to think that the surface produced by facing is necessarily planar – obviously not, as Hopper pointed out. Looking back I see that I was trying to view a 3D problem in 2D, and that's where I got confused.

I remember reading somewhere that Roger Penrose, when lacking anything more interesting to do, would occupy his idle hours by mentally tessellating four dimensional planes. Or something. Good God, I struggle to think in 3D!

Robin

Edited By Robin Graham on 15/09/2020 00:25:40

Edited By Robin Graham on 15/09/2020 00:29:59

Edited By Robin Graham on 15/09/2020 00:34:38

[not done it yetParticipant @notdoneityet]

I started this thread because I had been advised that the best way to skin this particular cat was to turn between centres. …

… I have to try and figure things out as I go along and take advice – this forum has been invaluable to me in that respect.

As I wrote easrlier, had you told us what you wanted to do with your chunk of metal, you would quite likely have been advised differently. Facing both ends parallel was clearly not required – it was only the final cut for the sipgot that needed to be parallel with the outer part of the other end.

It would help to figure out the order of operations before starting, not ‘as you go along’ and also consider the precision actually required for the finished article. Unnecessarily seeking perfection costs a lot – in time and effort, as well as the kit to do it with.

[JasonBModerator @jasonb]

Posted by not done it yet on 15/09/2020 06:53:44:

…………………………… told us what you wanted to do with your chunk of metal, you would quite likely have been advised differently………………………….

As per the first link I posted yesterday, the opening line seems to tell us what Robin wanted to do with his piece of metal.

"This arises because I got involved with a project to make a riser for a VMC mill"

[JasonBModerator @jasonb]

Here is a thought for you all to debate this morning.

When turning between ctrs at what stage does the ratio between length and diameter make the workpiece unstable?

For example had Robin's piece of metal been 25mm thick and 150mm dia pressure on the sideof the work from the facing cut would tend to act as a lever moving the tailstock ctr towards the operator if there were the slightest flex in the machine. If it had been 300mm long and 150mm dia this effect would be almost non existent.

Discuss (nicely)

[not done it yetParticipant @notdoneityet]

My point was – why do it if it is not necessary?

Where did this requirement for facing both ends arise? If I were to make one of these it would have been skim the surface to clean it up, face one end (we now have two cleaned up surfaces, hopefully perpendicular to each other), cut the hole in the end and position the holes for attachment. These can be all done without removal of the workpiece from the chuck, even if the hole-drilling cannot? Clearly, I would prefer to use a 4 jaw independent chuck for these operations as a) it has 4 gripping points and b) more reliable gripping power than a self-centring 4 jaw.

That lot done, the other end can be tackled. Lump lined up carefully in the chuck (the 4 jaw independent, of course) to achieve perfect concentricity from end to end and a centre support added after facing enough area near the centre achieve a flat surface for the centre drill followed by support (or concentricity, if removed from, and later replaced in the chuck). Face the outer part, then mark and drill the other set of fixing holes or threads, cut the spigot and the job is done – apart from finishing to diameter, which is more an aesthetic consideration than a dimensional requirement.

Removing and replacing a part in the 4 jaw is certainly more time consuming than turning it round between centres, but just as accurate (with patience and suitable dial gauges) for the precision required.

I’m not a machinist, so maybe someone can point out what my mistakes would have been, please?

With 3 bolt holes already positioned in the first end, I am confident these could be used as a means of simple, safe attachment to a faceplate – if that was the preferred work-holding method at that stage – because we all know “there are more ways, than one, to skin a cat’.

It’s basically how I made my first chuck backplate about 25 years ago. I cut the spigot (and under-cut it) then scribed a circle (with a pointed cutter) at the diameter required for the bolt holes (it took about three attempts, IIRC) and, after dot punching the first mark, set the other two clearance bolt holes with dividers. That chuck was fitted in the proper fashion – none of this ‘loose spigot’ lark like some advocate – although that was when I learned the part must be completely cool, before the final cut for the spigot, as it turned out a close fit, not the tighter fit I had expected.🙂

[Michael GilliganParticipant @michaelgilligan61133]

Posted by not done it yet on 15/09/2020 06:53:44:

[…]

Unnecessarily seeking perfection costs a lot – in time and effort, as well as the kit to do it with.

.

But it is an interesting and useful exercise [both academic and practical] to explore the limits of what one can achieve with the available ‘kit’

That seems quite sufficient to justify what Robin is doing.

MichaelG.

.

Note: Robin has already made his riser, to a standard of accuracy which many here would consider excellent.

… so the job is done.

Surely he should not be lectured for daring to spend his own time and effort investigating how much better he could do if required.

[blowlampParticipant @blowlamp]

The reason to do it between centres is that this way provides an intrinsically 'perfect' solution to the problem.

Assuming the lathe is facing-up correctly and the headstock & tailstock centres are in proper alignment with each other, then this should be a simple job.

I say it's a 'perfect' solution because it requires no setting up with gauges or other tools. Therefore, it can be removed and replaced at will, with no loss of repeat accuracy. The end faces are generated, which is a (theorectically) perfect way to create an object, as it isn't reliant upon things like exact tool geometries or how precisely the tool is aligned to the job.

In contrast to face milling the ends, you would have to be sure the milling spindle axis is set exactly perpendicular to the table movement as well as ensure the table working surface is flat and perpendicular too. The effect of any inaccuracies would reflect in the job by faces not being parallel or slightly hollow due to spindle misalignment.

Martin.

[Graham MeekParticipant @grahammeek88282]

Posted by blowlamp on 15/09/2020 11:55:18:

The reason to do it between centres is that this way provides an intrinsically 'perfect' solution to the problem.

Assuming the lathe is facing-up correctly and the headstock & tailstock centres are in proper alignment with each other, then this should be a simple job.

I say it's a 'perfect' solution because it requires no setting up with gauges or other tools. Therefore, it can be removed and replaced at will, with no loss of repeat accuracy. The end faces are generated, which is a (theorectically) perfect way to create an object, as it isn't reliant upon things like exact tool geometries or how precisely the tool is aligned to the job.

In contrast to face milling the ends, you would have to be sure the milling spindle axis is set exactly perpendicular to the table movement as well as ensure the table working surface is flat and perpendicular too. The effect of any inaccuracies would reflect in the job by faces not being parallel or slightly hollow due to spindle misalignment.

Martin.

Martin,

I could not have put it better myself,

Regards

Gray,

[Robin GrahamParticipant @robingraham42208]

Posted by blowlamp on 15/09/2020 11:55:18:

The reason to do it between centres is that this way provides an intrinsically 'perfect' solution to the problem.

 

Assuming the lathe is facing-up correctly and the headstock & tailstock centres are in proper alignment with each other, then this should be a simple job.

 

I say it's a 'perfect' solution because it requires no setting up with gauges or other tools. Therefore, it can be removed and replaced at will, with no loss of repeat accuracy. The end faces are generated, which is a (theorectically) perfect way to create an object, as it isn't reliant upon things like exact tool geometries or how precisely the tool is aligned to the job.

 

In contrast to face milling the ends, you would have to be sure the milling spindle axis is set exactly perpendicular to the table movement as well as ensure the table working surface is flat and perpendicular too. The effect of any inaccuracies would reflect in the job by faces not being parallel or slightly hollow due to spindle misalignment.

 

 

 

Martin.

Thanks for your detailed explanation Martin – very helpful to a tyro like me.

This whole riser-turning thing has been an education for me on several fronts but I think, from my point of view at least, this thread has run its course – I have plenty to worry about and work on.

My thanks again to those who have made constructive comments.

Robin

Edited By Robin Graham on 15/09/2020 23:24:06

Edited By Robin Graham on 15/09/2020 23:25:16

Edited By Robin Graham on 15/09/2020 23:26:26

[Author]

Posts