On 31 October 2024 at 14:17 Russell Eberhardt Said:

On 31 October 2024 at 13:11 JasonB Said:

Just be aware that ER collets are not good for short lengths, you really need to fill them to the full length of the gripping area.

 

I’ve never been able to see why this should be true.  While 5C collets only grip at the end, ER collets close parallel and grip all the way along.

…or am I missing something?

Russell

Not in my experience.

A quick video to show the grip is anything but parallel. 1/2″ ER 32 collet with a 1/2″ cutter inserted 3mm into the hole and the nut only tightend by gripping it in my hand- no spanner. It is easy to feel that the nut is going on far more than it would normally. To show how much the other end has closed down I can’t even poke a 12mm dia cutter into the collet so would say the back end is 11.7mm or less. So if the front end is 12.7 then the gripping surface is not parallel so you only get line contact and the part is likely to wobble or at the minimum deflect under cutting loads.

This is why many members say that you need a butt plug in the other end when holding short work in an ER collet to stop the back end collapsing.

We have different ideas of Short.

But I did the same test again. This time a 10mm dia cutter inserted 10mm into an ER32 Collet. Hand tight I could just get a 9.5mm drill shank in the back end. Nipped up with a spanner it went down to 9.3mm so hardly parallel!

I’m not sure what the rest of you make or not but I often hold items by less than 3mm quite often 1mm is all I have to hold. Provided you don’t go silly with stickout or greedy cuts there is little problem with a 5C collet but I would not do it with my various size ER collets without plugging the other end. After all 5C are meant for holding WORK.

Here is an example, Cylinder cover for a 12mm bore engine. 1mm long spigot on the inside and a curved shape spigot on the other.

Hold by the stock dia, turn, face, drill, ream and turn the inner spigot so all concentric.

Cut it off the bar, hold by the 12mm dia x 1mm long spigot to turn the other side

Something a bit longer for the current Clarkson build. 1.6mm dia, 2mm of M1.6 thread, dome the end. Easy enough holding the stock to do that. Now cut off and face the other end back to 4.5mm overall length and dome the end. Move from 5C lathe chuck to 5C collet block to cut the slot 0.5 x 0.5.

Or want to hold a larger cylinder cover for drilling stud PCD. Hold by 0.75mm long spigot and drill. Take care on breakthrough as there is not much room to spare, and don’t lean on the drill too hard. Turning was done with soft jaws

On 31 October 2024 at 18:48 JasonB Said:

We have different ideas of Short.

But I did the same test again. This time a 10mm dia cutter inserted 10mm into an ER32 Collet. Hand tight I could just get a 9.5mm drill shank in the back end. Nipped up with a spanner it went down to 9.3mm so hardly parallel!

I’m not sure what the rest of you make or not but I often hold items by less than 3mm quite often 1mm is all I have to hold. Provided you don’t go silly with stickout or greedy cuts there is little problem with a 5C collet but I would not do it with my various size ER collets without plugging the other end. After all 5C are meant for holding WORK.

Here is an example, Cylinder cover for a 12mm bore engine. 1mm long spigot on the inside and a curved shape spigot on the other.

Hold by the stock dia, turn, face, drill, ream and turn the inner spigot so all concentric.

Cut it off the bar, hold by the 12mm dia x 1mm long spigot to turn the other side

Something a bit longer for the current Clarkson build. 1.6mm dia, 2mm of M1.6 thread, dome the end. Easy enough holding the stock to do that. Now cut off and face the other end back to 4.5mm overall length and dome the end. Move from 5C lathe chuck to 5C collet block to cut the slot 0.5 x 0.5.

Or want to hold a larger cylinder cover for drilling stud PCD. Hold by 0.75mm long spigot and drill. Take care on breakthrough as there is not much room to spare, and don’t lean on the drill too hard. Turning was done with soft jaws

 

On 31 October 2024 at 18:48 JasonB Said:

We have different ideas of Short.

But I did the same test again. This time a 10mm dia cutter inserted 10mm into an ER32 Collet. Hand tight I could just get a 9.5mm drill shank in the back end. Nipped up with a spanner it went down to 9.3mm so hardly parallel!

I’m not sure what the rest of you make or not but I often hold items by less than 3mm quite often 1mm is all I have to hold. Provided you don’t go silly with stickout or greedy cuts there is little problem with a 5C collet but I would not do it with my various size ER collets without plugging the other end. After all 5C are meant for holding WORK.

Here is an example, Cylinder cover for a 12mm bore engine. 1mm long spigot on the inside and a curved shape spigot on the other.

Hold by the stock dia, turn, face, drill, ream and turn the inner spigot so all concentric.

Cut it off the bar, hold by the 12mm dia x 1mm long spigot to turn the other side

Something a bit longer for the current Clarkson build. 1.6mm dia, 2mm of M1.6 thread, dome the end. Easy enough holding the stock to do that. Now cut off and face the other end back to 4.5mm overall length and dome the end. Move from 5C lathe chuck to 5C collet block to cut the slot 0.5 x 0.5.

Or want to hold a larger cylinder cover for drilling stud PCD. Hold by 0.75mm long spigot and drill. Take care on breakthrough as there is not much room to spare, and don’t lean on the drill too hard. Turning was done with soft jaws

 

Yes, that cover is the sort of part I often hold in my ER chuck without plugging the other end and never had a problem.  I have repeated your test with a 12 mm cutter inserted in a 12 – 11 mm collet and yes, the back end closes down, but I only measured about 0.2 mm of closure (perhaps my arthritic hands are getting weak).

My question remains, what is causing the back end to close up? There is no contact at the back between the collet and the socket so the force must come from elswhere.

HYPOTHESIS: Perhaps the problem is caused by the short front taper of the collet not matching that of the nut.  If the initial contact between the collet front taper and the nut taper occurs at the back end of that short taper, then as the chuck is tightened the front of the collet will be pushed out creating a levering effect.

Any other suggestions?

Russell

Is using an ER collet to hold short work a bad idea or not.  Let’s try SODs patent Risk Analysis technique.

What could possibly go wrong?

On a round bar that’s gripped full length by the collet, nothing.   My memory of ‘full length’ is ‘at least 75%’, but I see Mark says: “For toolbits regofix recommend a minimum of 2/3 engagement inside the collet.”   The difference may be related to how well-made the collet is. A genuine Regoflex is more likely to resist abuse than an inexpensive alternative.

The big issue is gripping short lengths, where Jason showed a 3mm front engagement causes severe distortion at the rear.  The problem is that an ER collet is a concentric / coaxial spring, which adapts as pressure is applied to grip the full length of a tool shank.   If the shank is short, the rear will displace, risking two problems:

1. High run out because the rod is no longer parallel to the axis due to the spring deforming.  However, run-out, if present, can probably be corrected by tapping with a mallet.   I think this is low risk and low impact, provided the operator checks the run-out!  (Can’t be assumed because the collet is being used outside it’s normal design envelope.)
2. More dangerous, uneven tightening risks damaging the collet.  By crushing a step into the interior bore of the collet, AND/OR by destroying the set of the spring.  I think the risk of damaging the collet is HIGH, especially if a short item is tightened to specification (ER32 is calls for a lot or torque – 130Nm / 96ft lbs).   But the impact is low because damaged collets are easily replaced for under a tenner!

In other words, don’t fret!  The sky won’t fall in when a machinist wrecks a collet, and that only happens if he aggressively tightens one on a too short workpiece.  Might have to do it several times before the damage becomes apparent.

That said I take care of my ER collets.  It’s because I use the same collet-set to hold cutters in the mill as well as work-holding in the lathe.   I don’t want to find mid milling job that I’ve damaged a collet by careless work-holding.  The cost doesn’t worry me, it’s the stoppage and faff of ordering a replacement.

Dave

And why I said the bit in bold and not the collet length

“Just be aware that ER collets are not good for short lengths, you really need to fill them to the full length of the gripping area.”

Quite a few of my smaller dia collets are the same with a counterbore at the back but gripping area is around the 2/3 collet length mentioned elsewhere as the amount that should be gripped.

It does seem good engineering practice as well as common sense to make sure that the workpiece or cutter is inserted into the collet at least 2/3 of its length.

Anything less than half the depth or length of the collet appears to be asking for trouble, such damaging the collet or failure to hold the workpiece or cutter securely.

On 4 November 2024 at 10:44 mark smith 20 Said:

[…] Glad someone mentioned the collets bored at the rear ,as i have a 5 of er20 in the lot i bought . The collets are approx 31.5mm long ,the back bore is 13mm deep by 9.2mm wide. That equates to only 58.73 %.

I’m happy to be corrected, but I think the general idea is that it keeps the aspect-ratio of the clamped portion within reasonable limits.

MichaelG.