Pros and Cons of the ER collet system

[Nick HulmeParticipant @nickhulme30114]

Belay That,

I thought I'd take an ER40 Stevenson's Collet Block, a good 11-10 collet, a length of 10mm silver steel and a layout marker and see what happens.

This is what happens when you apply a 5 degree turn with a pipe wrench with the collet not quite fully locked up but still pretty tight,

– Nick

[Raymond AndersonParticipant @raymondanderson34407]

Nick, There is no need to dismiss Martin's [ blowlamps ] solid model, as what he has drawn is essentially correct. There is only line contact when the ER collet is closed down to the lower end of it's range That is ONE of the reasons why they are unsuited to the High metal removal rates seen in high performance machining. Today's Machine tools be they High performance Mills or VMC'S / HMC'S can far outstrip the capacity of the ER system.

When the ER's were first designed circa 1973 the machine tools then were a far cry from those of today, with their ability to remove material at a frightening pace. If you look at the clamping force on a properly torqued 20 Ø ER collet when clamping a 20mm Ø toolshank H6 tol is circa 110 Nm and a high performance milling chuck is 450 Nm clamping the same tool, tremendous difference. That is one of the reasons Regofix has designed the Powr Grip chuck, which is along the lines of the Albrecht APC and the Guhring.

I doubt very much if anyone on the forum has a top end VMC / HMC, and as such the ER'S suit our needs just fine.

[Nick HulmeParticipant @nickhulme30114]

Read the post above yours for an empirical demonstration of the contact achieved with an 11 to 10 ER40 closed down to 10mm, it's 8 point contact at each end without firm contact in the middle, that's where the flex can occur and an accurate CAD Model could show this but modelling a distorted collet is even harder than working out in your head what's happening and it seems that's also very, very hard, 

– Nick

Edited By Nick Hulme on 17/09/2016 13:54:56

[Raymond AndersonParticipant @raymondanderson34407]

Nick,, I can see at least 2 lines of contact in the pic [ towards the end of the bar ] Also was the collet spotless, and the bar also ? Also noticed the flex in the center Vey interesting. I could have a go at modelling one in Solidedge which also has FEA, but i've never used the FEA module and would'nt know where to start, and what I would probably come up with would be nowhere near the real happenings.

[MuzzerParticipant @muzzer]

Posted by Raymond Anderson on 17/09/2016 13:43:02:

If you look at the clamping force on a properly torqued 20 Ø ER collet when clamping a 20mm Ø toolshank H6 tol is circa 110 Nm and a high performance milling chuck is 450 Nm clamping the same tool, tremendous difference. That is one of the reasons Regofix has designed the Powr Grip chuck, which is along the lines of the Albrecht APC and the Guhring.

The Fahrion presentation MG linked to earlier on your behalf is a bit free and easy on the numbers they quote (which seems not atypical of German marketing materials) but even so, they appear to be claiming 350Nm for their variant of the ER collet (the "CP 225&quot for a 20mm dia tool. Annoyingly they haven't bothered numbering the slides but it's about half way through, titled "Ze Greatest Possible Holding Torque, Ja?".

[Raymond AndersonParticipant @raymondanderson34407]

I think Fahrion have been a bit optimistic on those figures. Ok, the Fahrion has a ground trapezoidal thread form and double guides for the closer but I can't see how that alone would increase the clamping force to 350Nm from the Normal ER at 110Nm. I would lean towards the Albrecht and Guhring figures. Although I don't know, and i'm certainly not calling Fahrion "liars " Maybe those areas where the Fahrion differs from the normal ER do make such a difference. Thats really well beyond me.

[Nick HulmeParticipant @nickhulme30114]

Posted by Raymond Anderson on 17/09/2016 14:04:45:

Nick,, I can see at least 2 lines of contact in the pic [ towards the end of the bar ]

It's a 10mm bar, approximately how close to the end of the bar?

[Raymond AndersonParticipant @raymondanderson34407]

Looks about 20 odd mm from the end. 1 area going towards the far side of the bar looks like partial cleaned, and the other 2 cleaned areas are toward the front.. Thats why I asked if the collet was spotless and the bar also, as any tiny piece of dirt ect will give a false result. Actually there are too many unknowns doing this to get any accurate findings. The Closer should be torqued to the specified amount for the Ø, and there is no certainty that the silver steel is perfectly round, just because it has a ground finish doesn't mean it is round. It can be lobed, it all depends on how accurate the grinder was set up. So I think any findings are irrelevant.

[Nick HulmeParticipant @nickhulme30114]

Posted by Raymond Anderson on 17/09/2016 16:42:46:

So I think any findings are irrelevant.

Good move, on this basis all your theories are worthless as they are derived from a basis of no proven research or practical personal experience or experimentation 😀

 

 

Edited By Nick Hulme on 17/09/2016 17:10:32

[Nick HulmeParticipant @nickhulme30114]

As one of the few with extensive machining experience using ER collets and apparently the only one willing or able to get off their fundament and do something I was motivated by Raymond's correct observation that Silver Steel can be out of round to take a solid Tungsten Carbide endmill, grind a flat on it for a tap wrench to grip and repeat the test.

With the chuck tightened gradually and the endmill rotated a couple of degrees back and forth each increment to the point where it became impossible to move the endmill with both hands on the tap wrench these are the results –

I'm pleased to say that any lines between the contact points are lighting artefacts, thanks to Raymond for inspiring me to use something accurately cylindrical to once again be the only one to get off my fundament and prove this

– Nick

[blowlampParticipant @blowlamp]

Posted by Nick Hulme on 17/09/2016 13:37:08:

Belay That,

I thought I'd take an ER40 Stevenson's Collet Block, a good 11-10 collet, a length of 10mm silver steel and a layout marker and see what happens.

 

 

This is what happens when you apply a 5 degree turn with a pipe wrench with the collet not quite fully locked up but still pretty tight,

 

– Nick

 

That appears to be line contact with a small rotation applied – what else would explain gaps that wide?

The huge area in the middle with no contact at all doesn't look too good though does it!

 

Martin.

Edited By blowlamp on 17/09/2016 17:46:22

[blowlampParticipant @blowlamp]

Posted by Nick Hulme on 17/09/2016 17:42:03:

As one of the few with extensive machining experience using ER collets and apparently the only one willing or able to get off their fundament and do something I was motivated by Raymond's correct observation that Silver Steel can be out of round to take a solid Tungsten Carbide endmill, grind a flat on it for a tap wrench to grip and repeat the test.

With the chuck tightened gradually and the endmill rotated a couple of degrees back and forth each increment to the point where it became impossible to move the endmill with both hands on the tap wrench these are the results –

I'm pleased to say that any lines between the contact points are lighting artefacts, thanks to Raymond for inspiring me to use something accurately cylindrical to once again be the only one to get off my fundament and prove this

– Nick

If you're arguing against the 'line contact' theory, why have posted 3 pictures that seem to reinforce it?

Martin.

[Nick HulmeParticipant @nickhulme30114]

Martin,

It's not a surprise, based on experience it's where I though the "space" would be on an 11 -10 ER Collet tightened down to 10. But as a realith check that's 2 concentric 8 jaw chucks around 30mm apart in a system where all the side load at the contact points is taken by solid metal in compression with no "space" for sideways movement?

How bad do you think that could be?

You could CAD model it for us and give us the figures?

– Nick

[Raymond AndersonParticipant @raymondanderson34407]

All you have proved is what has already been said, that there will be "line contact " As for no experiment , best to let manufacturers R & D departments handle that. where no doubt it will be carried out under correct procedures. As for one of the few with extensive machining experience using ER's Pray, how did you arrive at that conclusion? You seem to know a lot about other forum users and what their experience amounts to . Oh hold on, you must be another of those who sits at gods right hand side. You appear to be the type who could give us the square root of a jar of pickles, but couldn't tell us how to get the lid off.

I will stick to manufacturers findings thank you,. not some "comic cuts " trials.

Edited By Raymond Anderson on 17/09/2016 18:06:07

[Nick HulmeParticipant @nickhulme30114]

Posted by blowlamp on 17/09/2016 17:53:57

If you're arguing against the 'line contact' theory, why have posted 3 pictures that seem to reinforce it?

Martin.

Martin,

I don't expect people to take my word, I've provided everyone with a way to simply conduct their own tests and given my results.

Of course folks would have to be prepared to move from their keyboard and do a bit of thinking and engineering and I understand that's not for everyone,

– Nick

[Nick HulmeParticipant @nickhulme30114]

Posted by Raymond Anderson on 17/09/2016 18:04:02:

I will stick to manufacturers findings thank you,. not some "comic cuts " trials.

Edited By Raymond Anderson on 17/09/2016 18:06:07

Raymond,

Those who can't engineer just read about it!

– Nick

Edited By Nick Hulme on 17/09/2016 18:27:52

[blowlampParticipant @blowlamp]

Posted by Nick Hulme on 17/09/2016 17:59:25:

…You could CAD model it for us and give us the figures?

– Nick

I'm not here to blow anyone's mind, Nick

Martin.

[Raymond AndersonParticipant @raymondanderson34407]

Nick, please enlighten me about the significance of next Tuesday., SIMPLY HAVE TO KNOW. I haven't learnt anything new on this topic but at least iv'e had a laugh. Sorry that's a lie, II have learned something new, we now have a forum member who sits at gods right hand side.

Ps, next time you are having a tete a tete with him put in a word for me,. you have his ear.

cheers.

[Nick HulmeParticipant @nickhulme30114]

Posted by Raymond Anderson on 17/09/2016 18:26:59:

Nick, please enlighten me

Where's your "Manufacturer's" data about holding and contact points at extremes of range for ER collets then?

What power of mill do you have experience of using ER collets on and what were the issues you encountered?

All you've trotted out is the result of the skills of Google, well done, the way you pilot that armchair is a credit to you!

– Nick

[SillyOldDufferModerator @sillyoldduffer]

Chaps,

This thread is getting too tart for my taste. Isn't it possible to discuss ER32 without getting personal?

Regards,

Dave

[Steven VineParticipant @stevenvine79904]

I reckon this has crossed a line, though I do like it when it kicks off!

[Raymond AndersonParticipant @raymondanderson34407]

You will find the info re contact points on the thread entitled Eccentric collet [I think ] Please spare me the crap about engineers engineer and others only read about it . I see very little evidence in you're album that would display "extensive machining experience".

All i'm seeing on here is someone who thinks simple tests [ agricultural might be a better word ] can better proper highly respected manufacturers tests. and who tries to disprove a proven "line contact " occurence and proceeds to post pics that REINFORCE it.

YAWN, This is getting tiresome now I think I will retire with a cup of cocoa and read some more manufacturers data. Although I will take it all with a pinch of salt, because you obviously know better than Albrecht, Schunk, Guhring, Centaur et al You really have go me beat here. I take me hat off to you.

[Raymond AndersonParticipant @raymondanderson34407]

Listen folks, I derive no pleasure in hurling insults and especially not from behind a keyboard.. It's Nick's holier than thou attitude when he tried to take the mickey from another posters nicely done cad model to explain the line contact occurence. and his antagonistic attitude. He is the only poster on this thread that has that attitude. and I for one will fire back. If that attitude was displayed in the street it would lead to more than "verbals "

This not a school playground, there is no need for anyone "taking the mickey " end of.

cheers.

[blowlampParticipant @blowlamp]

Actually, I have learned a bit more in the last part of this debate, because of Nick's photo's and in particular for me, the second one. It shows quite nicely how the gripping points alternate around the bar and from end to end in relation to where the slots finish.

I think the pictures go to show just how stiff the collet must be in these uncut regions and why it makes sense to insert the tool or workpiece as deep as possible into the collet for best support.

Martin.

[Chris TriceParticipant @christrice43267]

The only way to get perfect line contact along the length of a cutting tool from a holder that "crushes" is either a rubber type collet with separate metal blades or a metal collet whose bore exactly matches the cutter and/or the spindle it fits into with negligible "crush". The moment you squeeze an ER collet down in size (or indeed ANY collet), you lose any uniform line contact between the taper or the cutting tool it's holding. In ER's case, you get a grip of sorts at either end not unlike trying to hold a large circular ring in soft jaws that have been machined for a slightly larger ring, i.e. small areas of contact rather than the whole surface. Because these slots are staggered on an ER collets, these patches of contact at either end of the cutting tool are staggered in relation to each other. The upshot is that the collet may hold the cutter accurately on axis but the areas of contact are smaller than if the cutter and collet and taper fit perfectly.

Edited By Chris Trice on 17/09/2016 19:57:07

[Author]

Posts