Pultra 10mm collet closing torque

[Kevin CobleyParticipant @kevincobley97900]

I have two machines that take 10mm Pultra collets, one a 1750 lather, and the other a Hauser Mill.

On the lathe, its not a big issue, but on the mill I have trouble with end-mills getting drawn from or pushed in to the collet.

I can't find a recommended closing torque mentioned anywhere for this collet system. Does anyone have some info or experience?

The mill has a knurled hand wheel for the collet closer and I'm nervous of using more than hand tightness on this. It does have holes that may be for a pin wrench or tommy bar of some type, but I really don't want to wreck the threads on my collets or draw-bar..

Many thanks,

Kevin.

[Kevin CobleyParticipant @kevincobley97900]

[Michael GilliganParticipant @michaelgilligan61133]

Whatever Hauser has decided to do … I really don't like the idea of plain-shank milling cutters in a smooth collet of that style.

All I can suggest, Kevin, is that you always proceed with caution.

MichaelG.

Edited By Michael Gilligan on 14/09/2022 14:59:02

[Kevin CobleyParticipant @kevincobley97900]

I was trying to be cautious but I have no idea what the limit is. If I tighten up the collet, then its OK, but am I over-tightening it? I have no feel.

[BazyleParticipant @bazyle]

One possible area of difference is in the nose internal angle and where abouts on the collet it is making contact. ie near the outer end or nearer the stem. Not sure which is best but it might be a difference. I take it you are comparing like with like ie using the same sized endmill in the late. Never liked the concept of gripping a dead hard mill with a hard collet both polished as that is just how you design bearings to slip.

[Kevin CobleyParticipant @kevincobley97900]

To be fair, the collet is not a dead hard collet in this case, but a machinable blank that has been drilled to 6mm and slitted, so the collet is definitely softer than the cutter which is carbide.

The collet holds at the nose, but the length of the holding section is not very long – maybe a couple of mm?

The nose internal angle? Not sure there is one. These collets seem to have a gripping range of like +/- 0.1mm.

As for comparisons – I havent done a comparison. The Lathe seems less of an issue since the work is held and the cutter is a point tool, so there is little longitudinal force on the work. Its spiral end-mills pulling as the cut on the side faces or pushing as the cut on the bottom that are the problem.

[Kevin CobleyParticipant @kevincobley97900]

I should add drilled and reamed to 6mm, btw

[Clive FosterParticipant @clivefoster55965]

With the exception of the ER series, nailing down accurate manufacturers specifications for tightening torques of tool holding collets seems to be nearly impossible. After decades of on – off looking, including asking collet makers representatives, I'm coming to the conclusion that even the makers don't actually know!

It seems reasonable to suppose that the actual gripping force generated by a conventional single angle collet is closely related to the taper angle. So given a suitable value for drawbar torque on one breed of collet it ought to be fairly safe to extrapolate torque settings for collets of similar style and not too greatly varying angle. Simple trigonometry should suffice to convert actual clamping force per unit of drawbar pull for a reference collet of known angle to that achieved on a similar collet of different angle. Not forgetting to take into account the helix angle (pitch / TPI) of the respective drawbar threads to convert torque to longitudinal pull. Frictional forces and losses will be the great unknown but its probably fairly safe to assume that such losses will be of sufficiently similar magnitude to be considered a black box constant which need not be evaluated.

At one time Bridgeport suggested that something around 30 to 35 ft lb of torque was appropriate when holding milling cutters in R8 collets. As they invented the R8, and presumably verified its proper operation under real world conditions, it's (probably) acceptable to assume they knew what they were talking about. Unfortunately the reference I found isn't a proper specification. More a side remark in a cutter related tabulation found in a few versions of the manual. Later Round Ram and early Series 1 manuals I think. I really should track it down again.

Anyway 30 to 35 ft lb according to my "sort of calibrated" arm pull works OK for me in my Bridgeport without cutter pull out or drawbar / collet distress. So, in the absence of better information, suitable extrapolation to a (theoretically) similar shank clamp force should be safe.

As I recall it I used a similar pull on my BCA, after converting the knob to a nut, to stop the endemic endmill pull out from the undersize collets. Nothing broke in the 8 months or so it took me to run out of patience with the horrible thing! The BCA might be OK as a small jig borer but, frankly, its a terrible milling machine.

If anyone is into recreational matrix mathematics (eeek!) it might be possible to get some useable data out of this paper :- **LINK**

https://www.mdpi.com/2227-7390/9/5/492/htm

But it does however have a certain air of being a "look how clever I am" publication rather than something of practical use. Having been on the sharp end of "make this work" demands from overeducated bosses with nothing useful to do waving 3 rd generation photocopies of such thing my views are more than a little jaundiced. The real world is distinctly unimpressed by elegant mathematics, even when supported by creative profanity. An entirely reasonable position in my view.

Clive

[Kevin CobleyParticipant @kevincobley97900]

Thanks for the thorough response – I think I'll skip the matrices, at least for now

I will get the calculator out and work out you 30-35ft/lb R8 numbers and go on the odyssey of converting the numbers to my p9 – it will at lest give a ball park!

Your BCA is not a dissimilar size to my Hauser – I have 4" X 3"Y and 2 3/8" Z travel, but it was advertised as a milling machine I think its a 33B (not that they seem ever to have made 2 machines the same..)

I'll be back, in the word of a famous android..

Kevin.

[Anonymous]

I would associate Hauser with jig borers and precision machine tools for the watch industry. So I'd assume you are pushing it beyond the limits expecting it to act as a normal milling machine with, presumably, a 6mm cutter.

Andrew

[Kevin CobleyParticipant @kevincobley97900]

I could be pushing it too hard – yes – I guess I'm still finding out what I can get away with. Yes its for small stuff – I think it was originally bought for the war effort machining bomb fuses. But even when doing light surfacing, I found I had trouble. Hence checking if I'm being too ginger with the collets.

So I had a look at the calculations – wasn't too painful:

R8 is 7/16 – 20TPI and has a mouth angle of 16.5deg – 7/16th is 11.1mm

Pultra 10mm is 9.886mm x 1mm and has a 30 degree mouth angle

So apart from the mouth angle, the puller is similar. If we calculate the pull force from the torque as T=cDF, where D is the diameter of the thread, and F is the pull force, c is the coefficient of friction (0.2 for steel)

firstly: 30 -35 ft-lb == 40-47Nm (I prefer to calculate in metric, just for unit consistency sake – easier for me)

At the lighter end this is 18kN pull force, and thus 64kN holding force through 16.5deg

For that holding force on Pultra we would need a torque of 32Nm, which is a lot more torque than hand applied.

so 32-37Nm which is 24-27 ft lbs.

Human hand max torque is typ about 18NM according to NASA, so I would need to add a lever to the hand wheel to achieve this.

So the question then remains whether this is acceptable torque on that collet system, but at least I know that I expect trouble with just the hand tighten force.

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Kevin Cobley on 14/09/2022 16:00:39:

[…]

The collet holds at the nose, but the length of the holding section is not very long – maybe a couple of mm?

[…]

.

That is a recipe for disaster

Before you start worrying about the tightening torque, get a proper collet and make absolutely sure that the cutter shank is a very close fit. … Pultra collets were probably spec’d at something like +/- 0.01mm

None of your ER latitude available with watchmaker-style collets !!

MichaelG.

[BazyleParticipant @bazyle]

A 6mm endmill seems way too large for a 10mm collet though as I only have 8mm ones I don't have the same sense of their capability. A 3mm on a 6smm shank yes, I think lathework with watchmaker's collets is also different, lower stress, especially if working with a graver. So for turning one might get on ok making a 1/4" winding arbor because you would be going carefully knowing it was way bigger than normal.

[Kevin CobleyParticipant @kevincobley97900]

.

That is a recipe for disaster

Before you start worrying about the tightening torque, get a proper collet and make absolutely sure that the cutter shank is a very close fit. … Pultra collets were probably spec’d at something like +/- 0.01mm

None of your ER latitude available with watchmaker-style collets !!

MichaelG.

As I mentioned, yes I expect no holding tolerance. The collet was a Conflex blank end collet which has been drilled and reamed to closely fit the cutter shank and slitted with a slitting saw in the same manner as the normal collets.

When I say the holding is at the nose – I mean only at the front part of the blank – this is normal for a watchmakers collet no? The front section is not very long. Maybe I wasnt clear with my terminology.

[Kevin CobleyParticipant @kevincobley97900]

Posted by Bazyle on 14/09/2022 17:41:49:

A 6mm endmill seems way too large for a 10mm collet though as I only have 8mm ones I don't have the same sense of their capability. A 3mm on a 6smm shank yes, I think lathework with watchmaker's collets is also different, lower stress, especially if working with a graver. So for turning one might get on ok making a 1/4" winding arbor because you would be going carefully knowing it was way bigger than normal.

OK, I see what you mean – maybe running a smaller cutter and making more passes is the right approach. The closing force is then spread over a smaller area.

As an area thing, then it would scale by the square, so easy to get rapidly into trouble.

[Michael GilliganParticipant @michaelgilligan61133]

Kevin,

Even if the collet is holding decently parallel, 2mm nominal holding length is still ridiculous for a 6mm shank.

Just draw it at 10x and see if it looks in any way ‘right’

MichaelG.

.

Edit: __  to be clear … Yes, I did know what you meant.

Edited By Michael Gilligan on 14/09/2022 18:00:50

[Clive FosterParticipant @clivefoster55965]

Kevin

Your numbers don't look unreasonable although assuming constant losses due to friction and pull to clamp force conversion between R8 and Plutra 10 mm amy be a bit excessive due to the size and angle differences.

Assuming zero friction 30° angle of the Pultra will generate roughly half the clamp force of the 16.5° R8 for any given pull. Out in the real world there will be some sort of relationship between the area of the taper collet to spindle interface and frictional losses.

For starters I'd be inclined to work out the relative taper contact areas and reduce the difference between the NASA nominal 18 Nm hand torque and your calculated 32 – 37 Nm torque in a similar ratio. Stir in some handwavium to taste and you will probably end up around the mid 20s Nm torque.

Which has the right sort of feel to it.

Apparently HSM Adviser will calculate something called torque number for a given set of cutting parameters which can be converted to give an idea of pull out force by suitably arcane methods involving the cutter helix angle. But this is all much more germane to professional level CNC systems aiming to produce more swarf per hour than most of us manage in a decade!

Frankly, if you can cope with the reduced headroom, it might be simpler to just grind flats on the sides of your cutters and make a side-lock holder. Hand tight on the knob will do just fine for that.

Clive

[Kevin CobleyParticipant @kevincobley97900]

Sure I understand what you are saying. I ought to go and actually measure it rather than recollecting from my work office. But sure, its starting to seem that smaller cutters are in order. I have a 3mm shank cutter already so I shall see how that fares. I have a normal collet for that diameter as well.

[Alexander Smith 1Participant @alexandersmith1]

I may be "misremembering" if that's a word, but I think that BCA produced specific collets for milling which were internally threaded near the tail (like a Clarkson autolock collet) to prevent pull out while milling. That would reduce the required holding torque of the collet itself.

Somebody correct me if I'm wrong.

Sandy

[Anonymous]

Posted by Kevin Cobley on 14/09/2022 16:00:39:

…length of the holding section is not very long – maybe a couple of mm?

Irrespective of tightening torque that will never hold a cutter, or work, properly. Looking at my 6mm diameter Pultra 10mm collet the holding length is 8 to 10mm.

Andrew

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Alexander Smith 1 on 14/09/2022 18:43:26:

I may be "misremembering" if that's a word, but I think that BCA produced specific collets for milling which were internally threaded near the tail (like a Clarkson autolock collet) to prevent pull out while milling. That would reduce the required holding torque of the collet itself.

Somebody correct me if I'm wrong.

Sandy

.

You are correct, they did

… more correctly, perhaps, Crawford Collets produced them for BCA

They are superb [unfortunately I could only afford one at the time, and bought the 3/8” version]

MichaelG.

[Kevin CobleyParticipant @kevincobley97900]

Hi – sorry my recollection was way off. The holding face measures ~7.5 – 8mm long on the 6mm collet made from the blank. Apologies for fading grey matter !

Regardless, it seems like a smaller cutter is the order of the day.

Thanks for the guidance.

[Author]

Posts