ER16 Collets from Ebay

Can anyone provide a link to documentation illustrating how concentricity diminishes with lower values of tightening torque, please?

The collets I have that have claimed runout values of less than 0.008mm and 0.005mm have been very good, from ER11to ER40 collets. The best has been seen on the cnc router where cutter life was extended and a much better surface finish too. I don't pay too much attention to the torque of the nut. I do test the nut on the assembly with a known test piece of carbide to see the runout. I have found many bad nuts, and some I have repaired. Others I have thrown away. The ER brand of nuts have all been very good and I have never seen issues with them.

Neil

Me too. But how much value are runout measurements if you don't torque to the manufacturers' specs.?

Like I said, it's been discussed before here several times.

I (think I) vaguely recall someone (possibly Neil) doing a MEW article on this a year or two ago but I'm not sure.

Again, me too but ...

If the tool in the collet is short enough, you could use a deep socket.

A crowsfoot spanner on your torque wrench and a simple calculation also works. As does welding a scrap socket a known distance along a spanner. But these are working mechanics techniques, so probably aren't technical enough for model engineering...

Also, lets not forget that these collets were designed for industrial use not for model engineers. If the requirements for their use are onerous in this latter context, it's up to the ME to adapt.

I'll raise you one technique even lower, and say that you don't need to do a calculation when adding an extender or crowsfoot etc to a torque wrench. Simply set the crowsfoot or extender at 90 degrees to the handle of the torque wrench and use the original specified torque setting. (Source: Snap-On Tools instructions for using their torque wrench extensions.)

So it could be a simple matter of drill and then file a square hole in the handle of the ER collet wrench and hook your torque wrench up to it at 90 degrees.

But I seriously doubt a correct torque setting will correct the OP's 5 thou runout. And I suspect torque wrenches on collets are overkill in the home workshop. Maybe do it once and get a feel for how tight it is and what sort of extension is needed on the original collet spanner to achieve it and then carry on by "tradesman's calibrated arm".

And if you do decide to use a torque wrench then use a set of figures that are correct for your nut, you will not need to apply such a high setting for a bearing nut as a plian one.

Simple Hass chart for low friction nuts, ER16 is 40% of ER32, maybe Peters figures are for plain nuts as they are higher?

I'll just stick to using a spanner (shorter handle for the ER16 one) 

Edited By JasonB on 24/04/2022 07:05:51

And that's with a low-friction ball-bearing nut? Wow. Getting up there for the ER32 at 100 Ft Lbs. That's Harley engine sprocket nut torque and usually needs the long handled 1/2" drive torque wrench. The ER16 at 40 Ft Lbs should be doable with a six-inch collet spanner and both hands.

Absolutely right to question how the measurements were made, because it's difficult. All bets are off in my opinion because the arrangement involves a stack of collet holders, a drill shank of unknown accuracy, and a Dial Indicator.

• The stack accumulates all the errors from spindle to the drill shank; it's not measuring the run-out of the final collet.
• Drill shanks may not be particularly straight or circular.
• How smooth is your Dial Indicator's mechanism? I have two Indicators, almost identical, but one of them has a slightly sticky mechanism. It tends to move the needle in a series of misleading jerks.
• How firmly is the Dial Indicator foot held to the drill shank and is it square? Likewise is the stand as solid as possible, avoiding having the Indicator at the end of a spindly set-up?

The 0.005" error could be explained by the bullet points above. The only thing that can be said with confidence is that the collet run-out is less than 0.005", perhaps a lot less.

I prefer not to confuse myself attempting precision measurements. Metrology is fascinating but hard to get right. Too easy to get wrapped round the axle in a self generated fog of confusion. (Don't ask how I know!) Instead I find it simpler to use the tool to cut metal and diagnose faults only if they matter. I don't worry unless the set-up as a whole fails to do the job. In the event of trouble in the Tulley workshop, too-cheap collets are an obvious suspect, but not as much as a collet chuck held in another collet chuck.

Consider the faulty set-up shown below:

Run-out measured at the chuck-jaws would be much lower than run-out measured at the far end. However, not only does the long rod amplify the true run-out, it adds more due to the rod bending under its own weight, and it will whip around when spun. The risk of whipping in this arrangement is a serious safety issue: bad enough to fly out of the chuck damaging the lathe and spifflicating the operator.

Dave

How many MEs are using their collets in 10hp machines turning £200 tools at 15,000rpm to machine multi-thousand £ castings with the lights out?

How many of us have actually measured the runout of our collets to a reliable standard unless the parts they make aren't up to our relatively basic requirements?

After all, it's not that long ago that most MEs did all of their milling in a vertical slide using basic tool holders(and measuring equipment!) yet they produced useful parts.

Not long at all. Less than a week ago for some of us. And holding the milling cutter in a 65-year-old three-jaw chuck.

You know admitting that will make the Guardians of Myford(formerly the Model Inquisition) come and poke you with long travel DTIs until you repent your sins against Model Engineering?

Oh No! Not ... The Model Inquisition.

Maybe not Dave. If the mill, ER32 and ER16 setup had 15thou runout and the collet a 10thou runout and was only tried in one position you could get a reading of 0.005".

A quick way to get an idea of the collets runout would be to measure the runout inside the ER16 holders taper and without removing it from the machine put in a collet and see what reading you get. If the combination of errors makes the ER16 holder run out by 4thou then there may only be 1 thou runout of the collet.

Surely this all depends on the nominal size of the collet and the size of the tool being held. I can't see a 10mm tool held in a 10mm collet in a clean chuck needing a lot of torque to hold the tool concentrically? You'll need more torque the further the tool held goes away from the nominal size of the collet due to the amount of distortion that needs to occur.

The quoted torque figures are generally for tool holding where you are likely to be using the collet at it's largest size eg a 10mm cutter in a 10-9mm collet.

The closing force to get the collet to close down less than max dia is quiet small compared to the tightening force.

Maybe the tightening torque could effect run-out because the collet is a multi-fingered spring. I suppose a tool in a loose collet might be slightly skewed until all the fingers grip, but I'd expect the effect to be small.

I've always assumed the high-tightening torque is required because of the amount of power industrial machines can put on the tool. Hobby machines aren't in the same league as grown-up machines; it's not unusual for a medium sized Machine-Centre to have 15kW available at the spindle. The amount of force 15kW can put on a collet held tool-shank is enormous and bad things happen if the tool slips. Apart from spoiling the work, the tool will spin and ruin the collet. Even in normal operation, cutters and collets have a short life on busy production machines. In a production setting, ER collets are disposable and chucks are replaced regularly too.

I'm happy if my cutters don't slip in ER collets and the resulting cut is accurate within the limits of my workshop. Cutters only slip in my ER32 chucks when I forget to tighten the ball-bearing collar with a short spanner: seems my 1100W mill isn't powerful enough to overcome that level of friction, and I rarely push the motor to the limit.

Dave

Needing some ER collets and not too happy about taking a chance quality wise I followed Old Mart's advice and went with APT. As he said they sell different run-out tolerance grades.

I'm totally pleased with them.They're all within the quoted spec. and were delivered very quickly. Their website gives a warning, of which I was unaware, that the very highest precision grade should be used with stock close to the collet's nominal size to ensure minimum run out.

The prices I thought quite reasonable.

Michael

I just received these today. The short collets are from the oversized range . I got an ER20 Ø16, ER20 Ø12.7mm(1/2inch), ER20 Ø6.35mm(1/4inch), ER32 Ø24mm, ER32 Ø25mm.

They have the oversized collets in all from ER11 to ER40.

For me the over sized collets allow an overlap of the holders, so instead of using an ER32, I can still use the ER20, especially if I am using any of the Ø16 adaptors I made for the small ER11 collet chuck. The oversized ER32-24mm collet saves having to change out to the ER40 collet chuck. The two that I checked are very good, better than the bearing run out in my mill drill.

Cheers, Neil