Pros and Cons of the ER collet system

[JasonBModerator @jasonb]

The Fahrion .pdf that Raymond mentions can be downloaded from my dropbox here

[Michael GilliganParticipant @michaelgilligan61133]

Posted by JasonB on 06/09/2016 16:20:18:

The Fahrion .pdf that Raymond mentions can be downloaded from my dropbox here

.

Thanks, Jason

MichaelG.

[Martin ConnellyParticipant @martinconnelly55370]

I think that under large radial loads a tool near the bottom of a collet's range may deflect into the spaces where the collet is not touching the tool (or workpiece if used for work holding). If the tool matched the collet size this will not be happening but I wonder if it is possible for the tool to push parts of the collet away from the split if enough force is applied. This would be more likely if the collet holder and collet were both lightly oiled and have a smooth finish.

Martin

[AjohnwParticipant @ajohnw51620]

Posted by Michael Gilligan on 06/09/2016 18:04:02:

Posted by JasonB on 06/09/2016 16:20:18:

The Fahrion .pdf that Raymond mentions can be downloaded from my dropbox here

.

Thanks, Jason

MichaelG.

It reminds me of a full sized large newspaper whole page spread on the health benefits of eating bread. It listed all of the vitamins and etc in it.

Other people make ones with similar balance levels and egocentricity etc. Fraudian slip there. My spell checker must agree.

John

–

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Ajohnw on 06/09/2016 18:55:43:

Other people make ones with similar balance levels and egocentricity etc. Fraudian slip there

.

… and another one ? [Fraudian]

MichaelG.

[AjohnwParticipant @ajohnw51620]

Posted by Michael Gilligan on 06/09/2016 19:38:43:

Posted by Ajohnw on 06/09/2016 18:55:43:

Other people make ones with similar balance levels and egocentricity etc. Fraudian slip there

.

… and another one ? [Fraudian]

MichaelG.

Yes might be called a double wammy.

John

–

[Raymond AndersonParticipant @raymondanderson34407]

Only Fahrion has bothered to get back to me, so unless there is some late replies it looks as though I am going to be eating me slice of humble pie. as regards the rigidity of the ER system during milling. I will always admit when I was wrong, . and me pie is on the table ready to eat, but it's ok cos I have a good appetite . I am through with this thread unless there are any developments viz a ve Manufacturers answers.

cheers.

[Michael GilliganParticipant @michaelgilligan61133]

Thanks for your efforts, Raymond

There are many expedient compromises that give 'adequate' results, but it's worth looking closely at things occasionally … in the hope that we might get to understand better.

Storage on this forum seems to be 'unlimited' so it's a good place to park information for future reference.

MichaelG.

[Nick HulmeParticipant @nickhulme30114]

Posted by Martin Connelly on 06/09/2016 18:32:47:

I think that under large radial loads a tool near the bottom of a collet's range may deflect into the spaces where the collet is not touching the tool (or workpiece if used for work holding).

Martin

It's an interesting theory, but given that at each end the segments of the collet form an evenly spaced circle of solid spacers between the held item on the inside and the taper at that end on the outside where are you proposing that these "spaces" come from?

Nick

P.S. I'm turning 25mm long parts from 30mm stainless today with a 750mm length through and out the back of the Super 7 head, gently applying a fingernail to the end farthest from the collet with the work spinning at 2500rpm there is run out but it is just barely perceptible  

Edited By Nick Hulme on 16/09/2016 09:21:50

[Martin ConnellyParticipant @martinconnelly55370]

Nick,

if you put a Ø5.1 shank in a Ø6 collet the collet will only touch the shank of the tool at six narrow lines on the shank. Between each of these contact lines is a space where it is theoretically possible, with enough force, to deflect the tool into. I know the forces need to be high and the deflection will be small but that is the extreme situation that is being used in the argument that ER collets are not the best tool holding system for milling. I do not push my milling into these extreme areas and have no problems when using ER collets. I think with my mill there are probably much more weak points than the ER collets.

Martin

[Bill PudneyParticipant @billpudney37759]

Returning to reality. It is, I believe generally accepted that ER collets grip well/best at their "free" diameter. So if we are talking milling cutters, their shanks are going to be of a round number size, let's say 10 mm for the sake of the argument. Clearly that will be a really good fit in a 10mm free diameter sized collet. To me that's the end of the problem. As far as I'm concerned if the cutter has, for example a 3/8" shank, the user should splash out on a 3/8" collet, not worry if it's going to be secure in a metric one.

When I had to work for a living, buying tooling for CNC mills of low, medium and highish speed (5,000 rpm up to 26,000 rpm) the usual collet was an ER series, or a shrink fit for some of the highish speed stuff.

cheers

Bill

[Nick HulmeParticipant @nickhulme30114]

Posted by Martin Connelly on 16/09/2016 09:56:59:

Nick,

if you put a Ø5.1 shank in a Ø6 collet the collet will only touch the shank of the tool at six narrow lines on the shank. Between each of these contact lines is a space where it is theoretically possible, with enough force, to deflect the tool into.

Martin,

You haven't understood the question I posed in response to your post, the material of the collet or the cutter has to be deformed for it to move into the "space" you are referring to.

Each end of the collet acts a bit like like an 8- jaw chuck (but with solid material and no wobbly bits), even if the inner faces of the segments of the collet were flat there are 8 points of contact with a round object, in order for it to move over the taper, collet segment or item held must distort.

There is no "Space" into which the item being held can move without distortion of the material, draw it and you'll undersand,

Nick

P.S. Under-Tightening of ER collets is the most common cause of problems, I borrowed my first collet chuck and collet set from a local model engineer who couldn't get on with them, it turns out he was using them bone dry and only lightly hand tightening. 

I use bearing collet nuts and Tungsten Disulphide Dry Lube on the threads and taper surfaces, always tighten adequately and don't have any issues, even operating close to the lower limit of collet capacity, the torque required does seem greater towards the lower limit but I don't find that a big surprise. 

Edited By Nick Hulme on 16/09/2016 10:49:44

[Raymond AndersonParticipant @raymondanderson34407]

I have actually a couple of days ago got the last of me replies from Albrecht, Guhring and Schunk. Am rained off today, so have time to convert the pdf's into jpeg's so I can upload them to my album entitled Milling chuck. There are a few in already.

The main [ although not the only one ] reason why a specialist milling chuck if far more rigid than the ER system is because of the clamping force. An example is, 20mm Ø toolshank H6 tol, is circa 110 nm for the ER, the specialist milling chuck is 450nm for the same tool. rising to in excess of 800nm at 32 Ø We as hobbyists will come nowhere near to exceeding the limits of the ER system and for our use it is the perfect choice. It is flexible, Accurate, and cheap [compared to specialist milling chucks ] So I am correct in that there are better milling chucks than ER [although not for our needs]. For industry at high feed milling then, yes the ER system is down the list.

This has proved "interesting " for me at least, and hopefully for some others. Will stick some more data up once I have it all converted to jpeg.

Edited By Raymond Anderson on 16/09/2016 11:15:10

[Martin ConnellyParticipant @martinconnelly55370]

Nick, what I am doing is purely speculation as to why someone may say ER collets are not the best system for milling. The reality is that for many of us with home workshops they are the best value for money. I was just pointing out the possible differences between the ER system and the shrink fit or hydraulic type of holder being suggested as the best choice in industry for high performance machines.

The tapered faces on the outside of the ER collets can only be in full contact with the holder and nut in one position. As the collet is pushed further into the holder (nearly 1.5mm for 1mm closing range) or nut the radius at any point on the collet taper will probably not match the radius of the holder or nut where it is touching. The result is gaps where there are two different radii in contact with each other. These gaps between the collet and either tool or collet holder do not cause problems for those of us working within the limits we set ourselves or are set by our machinery. They are however a theoretical weak point that will allow movement under extreme loading. The pressure on the contact points will be high because of the small contact area which will allow elastic deformation that may be significant with high force from extreme milling forces. If the collets are only used at their maximum size eg Ø6 tool in a Ø6 collet then there may be very good contact between holder, collet and tool. I use some 1/4" tooling in Ø7 collets which may give less than perfect contact.

As I say I am only speculating as to why ER collets would not be a choice for high performance machining, I have no problem in using them or recommending them to other home workshop machinists.

Martin

[Raymond AndersonParticipant @raymondanderson34407]

Hi Martin, no need to speculate as to why The ER system is not recommended for High performance machining It's the lack of radial rigidity [ compared to a specialist milling chuck.] There are a few other factors involved, but in the main it is "clamping force". I am in the process of converting some more Data from pdf to jpeg All will be in me album entitled Milling chuck.

I reiterate once again, I am not knocking the ER system for TOOLHOLDING. It cant be bettered for our [hobbyist's needs ] But it cannot cope with high feed milling as well as the specialist systems. Which was ONE of the points of my [ purely academic ] argument.

cheers.

[Nick HulmeParticipant @nickhulme30114]

Posted by Martin Connelly on 16/09/2016 11:43:43:

what I am doing is purely speculation

Martin,

I understand that and what I am doing is offering logical argument as to why some aspects of that speculation may not have a firm premise upon which to sit, for the tool to move in any direction there will be one element of the collet loaded in direct compression or very close to direct compression, how does the object being held move into "the space" under these circumstances?

– Nick

[blowlampParticipant @blowlamp]

Posted by Nick Hulme on 16/09/2016 15:46:44:

Posted by Martin Connelly on 16/09/2016 11:43:43:

what I am doing is purely speculation

 

Martin,

I understand that and what I am doing is offering logical argument as to why some aspects of that speculation may not have a firm premise upon which to sit, for the tool to move in any direction there will be one element of the collet loaded in direct compression or very close to direct compression, how does the object being held move into "the space" under these circumstances?

– Nick

The premise sits upon the fact that small amounts of deformation will occur (under heavy radial load) along the lines of contact between the tool & collet and the collet & holder, because a tool at the lower end of collet capacity is being used, which is what causes the tiny gaps to be present between the tool/collet & holder in the first place.

I drew these gaps in CAD and showed them on page 3 of this thread.

 

Martin (blowlamp)

Edited By blowlamp on 16/09/2016 16:23:15

[Nick HulmeParticipant @nickhulme30114]

Posted by blowlamp on 16/09/2016 16:17:56:

I drew these gaps in CAD and showed them on page 3 of this thread.

Significant elements of your CAD were wrong as was pointed out and you failed to correctly model the full collet, had you done so you may have seen your errors, your use of the phrase "lines of contact" demonstrates a fundamental misunderstanding of the operation of ER Collets at the limit of their compression.

There is a mode of distortion which can come into play but you have not spotted or explained it yet, if you have a close look at an ER Collet you might work it out if you can see how they work, if you can't work it out I will give you a hint in a later post

– Nick

[John Stevenson 1Participant @johnstevenson1]

Well I'm pleased I have been doing something right all these years and my customers don't want a refund other wise the yacht over in Cannes will have to go.

[Nick HulmeParticipant @nickhulme30114]

Posted by John Stevenson on 16/09/2016 21:26:14:

the yacht over in Cannes will have to go.

I thought it was a "Yacht in a Can", kind of a low rent, windowless version of a "Ship in a Bottle"

[blowlampParticipant @blowlamp]

Posted by Nick Hulme on 16/09/2016 21:12:32:

Posted by blowlamp on 16/09/2016 16:17:56:

I drew these gaps in CAD and showed them on page 3 of this thread.

Significant elements of your CAD were wrong as was pointed out and you failed to correctly model the full collet, had you done so you may have seen your errors, your use of the phrase "lines of contact" demonstrates a fundamental misunderstanding of the operation of ER Collets at the limit of their compression.

There is a mode of distortion which can come into play but you have not spotted or explained it yet, if you have a close look at an ER Collet you might work it out if you can see how they work, if you can't work it out I will give you a hint in a later post

– Nick

Nothing of any significance was pointed out about my simplified CAD model of the holder/collet/tool assembly.

The whole assembly will distort to some degree when the closing nut's tightened – how much is anyone's guess, but lines of contact will exist I think in a roughly similar way to those of preloaded taper-roller bearings.

Martin.

[Nick HulmeParticipant @nickhulme30114]

In your CAD model the circumference is shown with gaps, this will not happen as shown in your model when a conical collet is pushed into a conical chuck/nut, this was pointed out in a subsequent response.

Your CAD model omits tapers and also omits that alternate segments are joined at alternate ends of the collet, thus the geometry is bogus as are the conclusions drawn from it

The best way to see what's happening is to play with a collet in a chuck, preferably a chuck or collet block which lets you see both ends of the collet as you tighten it, if you have any ER collets at all?

Regards,

Nick

P.S. When trying to illustrate a technical point, unless you are aware of which details really are negligible "Simplified" = "Wrong"  

Edited By Nick Hulme on 17/09/2016 00:46:14

[Neil WyattModerator @neilwyatt]

Posted by Nick Hulme on 17/09/2016 00:43:18:

In your CAD model the circumference is shown with gaps, this will not happen as shown in your model when a conical collet is pushed into a conical chuck/nut, this was pointed out in a subsequent response.

Your CAD model omits tapers and also omits that alternate segments are joined at alternate ends of the collet, thus the geometry is bogus as are the conclusions drawn from it

The best way to see what's happening is to play with a collet in a chuck, preferably a chuck or collet block which lets you see both ends of the collet as you tighten it, if you have any ER collets at all?

Regards,

Nick

P.S. When trying to illustrate a technical point, unless you are aware of which details really are negligible "Simplified" = "Wrong"

Edited By Nick Hulme on 17/09/2016 00:46:14

As a cross-section of the unsplit ends of the cone it is correct, but in reality the two ends are rotated by 15 degrees (for a 12-slot collet, they vary) so in between the ends there will be two contact points, not one, and the contact line will zig-zag around the collet and probably NOT in straight lines, but the principle that it is a line contact is what the diagram was intended to depict and that is fundamentally correct.

Neil

[MuzzerParticipant @muzzer]

Any high speed, high power (professional) application using ER collets will use the "correct" size ie a 12mm collet for a 12mm cutter. So it will be a snug fit on almost all surfaces including the bore and both tapered faces. In which case talk of contact lines etc becomes rather abstract.

Using a 6mm collet to hold a 5.1mm tool or workpiece isn't in the realms of the "high end" users. And TBH, if I were going to insist on using collets over this range I'd probably keep a second set for accurate work eg milling with the right sized collet / cutter. I doubt (m)any of the brochures specify performance at these extreme levels of compression / distortion.

There is no argument to be won, just opinions to be traded.

[Nick HulmeParticipant @nickhulme30114]

Posted by Neil Wyatt on 17/09/2016 11:16:06:

As a cross-section of the unsplit ends of the cone it is correct, but in reality the two ends are rotated by 15 degrees (for a 12-slot collet, they vary) so in between the ends there will be two contact points, not one, and the contact line will zig-zag around the collet and probably NOT in straight lines, but the principle that it is a line contact is what the diagram was intended to depict and that is fundamentally correct.

Neil

Neil,

If the cylindrical bore is distorted differently at each end won't it cease to be perfectly cylindrical and the radial shift result in 8 point contact at each end?

[Author]

Posts