ER Collets

[Kiwi BlokeParticipant @kiwibloke62605]

Problems built in?

[Kiwi BlokeParticipant @kiwibloke62605]

The thread 'Arc Euro ER16 runout' got me thinking. Thought I'd start a new thread, rather than hijack that one.

I've just looked at the Schaublin collet chuck that Emco supplied for the FB2 milling machine. I'd rather taken it for granted before – it just did its job, apparently accurately. But now I'm wondering… are these things designed optimally? Clearly, axial alignment cannot be taken for granted and the availability of 'forcing clamps' (dunno what they're really called) to tweak cutter alignment suggests that concentricity may not be as reliable as one would want.

I assume that the collet is aligned by the taper in the body of the chuck. If the collet collapses symmetrically (and is made concentric), alignment should be determined by the collet chuck's accuracy. But this assumes that the collet is not able to be deflected radially as it is tightened. Is this assumption valid, in practice? There are two complications…

Firstly, the short taper on the collet's nose mates with the taper in the closer (nut). The closer is threaded to the body with a 'normal' V thread, so it is 'centred' by the thread. If there is a malalignment between the closer and the body, the closer's taper and the centring action of the thread may fight each other, resulting in a radial force on the nose of the collet. Wouldn't a 'sloppy' square thread (large radial clearance) be better? There would then be no chance of the closer exerting a radial force on the collet. (OK, we really need symmetrical spanner work, but that applies to the conventional design anyway.)

Secondly, for collets with extractor grooves, it might be that, as the closer collapses the collet, allowing the collet to settle deeper into the closer's taper, the closer's extractor (half-) ring 'bottoms out' on the collet's groove, and contact between the closer's taper and the collet nose's taper is lost. Also, the axial force would be off-centre. Wouldn't it be better for the cap to be able to exert only a reliably axial force? Why have a taper in the closer, or is it to ensure the greatest radial force is applied to the cutter at the collet's nose?

Perhaps I should just continue to take it for granted…

[Nick HulmeParticipant @nickhulme30114]

Buying good quality collets and collet nuts is half the solution, ensuring your collet chuck dimensions meet the standard is the other half, then provided you don't need alignment to the last tenth you'll be fine – provided you keep your threads and tapers clean and lightly lubricated (I use a Tungsten Disulfide dry lube) and look up the correct tightening torque for your collet nut size and style and tighten to a significant proportion of that.

The most significant causes of issues with ER collets are too much tightness at the wallet and not enough tightness at the collet nut

[John RuddParticipant @johnrudd16576]

Posted by Nick Hulme on 11/12/2018 09:18:05:

The most significant causes of issues with ER collets are too much tightness at the wallet and not enough tightness at the collet nut

How true this is, and doesnt necessarily apply just to collet chucks/collets…

That said, there is some budget tooling that is 'made' to tolerances acceptable for what is required in model engineering without costing the earth while not being 'cheap and nasty'.

[Neil WyattModerator @neilwyatt]

Posted by Kiwi Bloke 1 on 11/12/2018 08:29:11:

I've just looked at the Schaublin collet chuck that Emco supplied for the FB2 milling machine. I'd rather taken it for granted before – it just did its job, apparently accurately. But now I'm wondering…

This way lies madness

Nothing sows more discontent in our workshops than seeking perfection in tooling rather than perfection in results.

If it was possible for all these devices to be perfect, why is it that industry with easy access to all top-notch most accurately still requires the services of highly skilled tool setters?

Neil

P.S. the extractor ring should not contact the groove of a collet in use, the dimensions for grooves and rings are well defined and not too difficult to machine (yes I have done it!) so it's unlikely you would get a duff nut.

[Chris TriceParticipant @christrice43267]

I've got a Schaublin MT2 ER 25 chuck and nut and it's my go to collet chuck.

[Chris TriceParticipant @christrice43267]

…. however I agree with the OP. I was thinking along the same lines since I mentioned in that other thread that you can tap the cutter with a rubber hammer and influence the TIR. Not so so much with the Schaublin version.

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Kiwi Bloke 1 on 11/12/2018 08:29:11:

The thread 'Arc Euro ER16 runout' got me thinking. Thought I'd start a new thread, rather than hijack that one.

.

I think you might find it interesting to read this previous thread: **LINK**

https://www.model-engineer.co.uk/forums/postings.asp?th=120333

… Although I note that it now starts rather strangely [?], there was a lot of useful discussion.

[ and, predictably enough, a lot of 'soapbox-ing' ]

MichaelG.

[Neil WyattModerator @neilwyatt]

Posted by Michael Gilligan on 11/12/2018 10:53:24:

… Although I note that it now starts rather strangely [?],

If you read down it become clear that it was 'moved across' from another thread to stop the earlier one being bogged down.

Revisiting that thread the goalposts seem to have moved around a bit as it progressed!

Neil

[Trevor Drabble 1Participant @trevordrabble1]

Kiwi , You may be interested in visiting the site of D&J Workholding in Oxon who are able supply ranges of different quality collets . No connection with company other than a very satisfied customer. Trevor .

[Kiwi BlokeParticipant @kiwibloke62605]

Thanks MG for the link. I've read it all (phew!). Some interesting content, but sadly diluted by rudeness, mud-slinging and anti-intellectual attitudes. I'm not impressed by arguments that go like: 'I was taught to do it this way, as an apprentice, and have done it this way for the last 40 years, so it must be right', or: "Just do it, don't waste time thinking about it". Kiwi Bloke's First Rule: The heat under the collar is in inverse proportion to the quality of activity above.

From the linked-to thread, it seems that lack of concentricity of the closer can / will cause the collet nose to be deflected radially. I had assumed that the collet would make contact (well, several line contacts) all along the sides of the chuck body's taper, thus it would be the chuck body that aligned the collet. However, it seems that it's only really held at the narrow end of each taper, so the closer's radial alignment is critical. Without a register, the closer relies on the centring action of a V thread, and will also be deflected by off-centre spanner forces. This probably explains the need to tighten HARD – to maximize the not-very-good centring action of the thread.

These things (most things) are more complicated than at first sight. Clearly, in practice, the things work well enough, resisting sensible radial loading, even for holding work, rather than cutters, and for use with milling cutters. However, some people experience radial run-out. I wanted to start an intellectual discussion as to possible causes and especially whether the problems were inherent in the design. Perhaps, in view of the vitriol in the linked-to thread, I was unwise.

[Kiwi BlokeParticipant @kiwibloke62605]

Oh, forgot to say that I have a Jacobs tapping chuck, which takes Rubberflex collets. These are axially aligned entirely by the chuck body's internal taper. The closer applies only an axial force to the end of the collet and mates to the body via a 'loose' square thread. Seems like a good design… Or is it?

[Chris TriceParticipant @christrice43267]

ER's popularity is based on its ability to hold all sizes with a (relatively) few number of collets. They hold their best and most solidly at their nominal size like conventional collets, but once they are pulled into the chuck to hold something smaller, they are only making contact on staggered points (not line contact as sometimes suggested) because the slots in them are staggered. In practice, this is "good enough" for most people who are balancing cost versus even better accuracy. I can't comment on 5C collets although they are clearly an industry standard, but I can for R8 and Morse Taper collets which you can sense feel more solid than ER although that might partly be because they sit rigidly inside a main spindle bearing. I use ER for convenience (25's and 32's) but R8/Morse when it really matters.

[Kiwi BlokeParticipant @kiwibloke62605]

Aargh! Just noticed brain fart in 3rd line of 2nd para of my first post. Of course, I meant 'radial alignment', or making sure there's no radial displacement: i.e. making sure everything is co-axial. Confusing innit?

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Chris Trice on 12/12/2018 01:46:07:

ER's popularity is based on its ability to hold all sizes with a (relatively) few number of collets. They hold their best and most solidly at their nominal size like conventional collets, but once they are pulled into the chuck to hold something smaller, they are only making contact on staggered points (not line contact as sometimes suggested) because the slots in them are staggered. In practice, this is "good enough" for most people who are balancing cost versus even better accuracy. I can't comment on 5C collets although they are clearly an industry standard, but I can for R8 and Morse Taper collets which you can sense feel more solid than ER although that might partly be because they sit rigidly inside a main spindle bearing. I use ER for convenience (25's and 32's) but R8/Morse when it really matters.

.

A very nice summary, Chris

MichaelG.

.

I do, however, have a slight reservation [offered for civilised discussion] regarding the point vs line nature of the contacts:

In the referenced thread [p7], Nick Hulme showed a reasonably convincing practical demonstration using layout marker; then [p9] Mark C presented some top class Finite Element Modelling which, if he had been encouraged to continue, would have gone a long way towards explaining the situation.

To my great personal regret, the 'practical men' dismissed all of this as irrelevant:

[quote]

Meanwhile the people that can do, do

And the people who can't, waffle on.

[/quote]

[Anonymous]

Posted by Michael Gilligan on 12/12/2018 08:50:39:

To my great personal regret, the 'practical men' dismissed all of this as irrelevant:

[quote]

Meanwhile the people that can do, do

And the people who can't, waffle on.

[/quote]

Sadly that's a feature of this forum. Mention that you've got an engineering degree and all the stories pop up purportedly showing how incompetent engineers are when it comes to making things. And as for people with a Ph.D, dear oh dear, they just live in their ivory towers and must be practically useless!

Andrew

[blowlampParticipant @blowlamp]

Another thing to consider is how little of the tool shank may be supported by the main body of the collet holder.

I have a set of Emco ER25 collets and just did some basic measuring of the 4-3mm.

The back end of this collet is relieved to the point of reducing the length of the 4mm bore to ~ 19mm.

Dropping this collet into a holder and measuring the projection from the start of the thread to the front face of the collet gives a figure of ~ 13mm.

Subtracting 13 from 19 gives us a figure of only 6mm of supported tool shank (by the holder) – no matter how far the tool is inserted – the remainder being supported by the closing nut.

This situation gets more pronounced with smaller bored collets, but disappears with the larger sizes.

Martin.

[Howard LewisParticipant @howardlewis46836]

With regard to the accuracy ACTUALLY NEEDED, perhaps the words "Fitness for purpose" may apply?

Clearance holes for bolts that merely hold two parts together, as opposed to Fitted Bolts where accurate alignment of the two parts is a necessity.

I once had a discussion with a Design Supervisor about the need for a chamfer on the retaining plate for a gear.

He eventually agreed that the chamfer was not essential, since it sat in fresh air without making contact with anything!

Howard

[Chris TriceParticipant @christrice43267]

Posted by Michael Gilligan on 12/12/2018 08:50:39:

Posted by Chris Trice on 12/12/2018 01:46:07:

ER's popularity is based on its ability to hold all sizes with a (relatively) few number of collets. They hold their best and most solidly at their nominal size like conventional collets, but once they are pulled into the chuck to hold something smaller, they are only making contact on staggered points (not line contact as sometimes suggested) because the slots in them are staggered. In practice, this is "good enough" for most people who are balancing cost versus even better accuracy. I can't comment on 5C collets although they are clearly an industry standard, but I can for R8 and Morse Taper collets which you can sense feel more solid than ER although that might partly be because they sit rigidly inside a main spindle bearing. I use ER for convenience (25's and 32's) but R8/Morse when it really matters.

.

A very nice summary, Chris

MichaelG.

.

I do, however, have a slight reservation [offered for civilised discussion] regarding the point vs line nature of the contacts:

In the referenced thread [p7], Nick Hulme showed a reasonably convincing practical demonstration using layout marker; then [p9] Mark C presented some top class Finite Element Modelling which, if he had been encouraged to continue, would have gone a long way towards explaining the situation.

To my great personal regret, the 'practical men' dismissed all of this as irrelevant:

[quote]

Meanwhile the people that can do, do

And the people who can't, waffle on.

[/quote]

The problem is the difference between theoretical perfection and reality. Collets are made from a deformable springy material (they have to be) so there is an element of deformation at points of contact which will broaden those same areas and straight lines starts to become wavy and perfect arcs elliptical depending on the forces of deformation. The collet tries to take on the shape of the rigid tool and the (theoretically) rigid chuck despite being in a position that doesn't match its starting shape. As long as all the forces (and I'd hate to be the mathematician that works them all out) remain in balance all around the tool, the tool should be held true, in theory.

[Michael GilliganParticipant @michaelgilligan61133]

Agreed, Chris … I think we're both on the same wavelength.

Mark C was doing a grand job with the FEM until all the nonsense started on that previous thread; although FEM is, as I have said before; by definition, an approximation.

MichaelG.

Edited By Michael Gilligan on 12/12/2018 15:36:33

[Former MemberParticipant @formermember32069]

[This posting has been removed]

[Mike PooleParticipant @mikepoole82104]

It would appear that D&J have people who have a history back to Crawford Collets so should know a decent collet from a dud.

Mike

[Former MemberParticipant @formermember32069]

[This posting has been removed]

[EmgeeParticipant @emgee]

Posted by Barrie Lever 1 on 12/12/2018 21:01:13:

As promised just been out into the workshop and set a clock up on the cutter and holder that has run 8 hours today on a stainless steel job.

Barrie

Excellent concentricity, always best with a warm/working temperature spindle and bearings.

Emgee

[Former MemberParticipant @formermember32069]

[This posting has been removed]

[Author]

Posts