SAE drill size to ER collet size

[Shotgun]

It is good to know that ER collets have a range that was intended to accommodate variations in tool shank sizing. While they do have a 0.040" range, they were not intended to be used to clamp that full range. They are most accurate, and are able to maintain that accuracy, when used with tools very close to the nominal size of the collet.

How do they loose accuracy? What is the mechanism?

As you tighten the nut, the collet is forced down a tapered shaft that squeezes it in from all sides. The collet has enough slits on both ends so that it can distort evenly along the length and eliminate any space between the tool, collet, and holder wall. The material between the slits makes up slabs that act as anvils between the holder and tool.

When you put the next tool in, it will continue to clamp down until all space is removed. The limit to the squeeze continues to be the thickness of each anvil like slab in the collet. I could see one or two anvils not returning to the correct place if a cheap collet without any spring were to be "oversqueezed", but that would be fixed when returned to the holder which would force them all back to the correct positions.

Where's the hole in my thinking?

 

[mikey]

How do they loose accuracy? What is the mechanism?

My understanding is that they lose accuracy because as the shank size departs from near nominal, the clamping forces focus toward the front of the collet such that the rear of the collet produces less clamping force. IOW, the entire collet does not clamp down evenly on the tool shank. This affects run out.

In addition, all collets experience elastic deformation to some degree when we compress them but if taken to its yield point, as when we clamp down too much on a small object, that elastic deformation can become plastic deformation and the collet may not return to its original form. Presumably, this occurs because the compressive forces are not evenly distributed as above.

My experience supports this. When I tried to clamp down on a 12mm Woodruff keyseat cutter with a 1/2" Techniks Imperial collet I had excessive run out, such that the key was a slip fit. I bought a 12mm Techniks collet and run out dropped to under 0.0002" and the key had the expected interference fit. I have had similar experiences with end mills as well so there is no question, at least in my mind, that squishing down a collet reduces accuracy.

 

[Shotgun]

My understanding is that they lose accuracy because as the shank size departs from near nominal, the clamping forces focus toward the front of the collet such that the rear of the collet produces less clamping force. IOW, the entire collet does not clamp down evenly on the tool shank. This affects run out.

In addition, all collets experience elastic deformation to some degree when we compress them but if taken to its yield point, as when we clamp down too much on a small object, that elastic deformation can become plastic deformation and the collet may not return to its original form. Presumably, this occurs because the compressive forces are not evenly distributed as above.

My experience supports this. When I tried to clamp down on a 12mm Woodruff keyseat cutter with a 1/2" Techniks Imperial collet I had excessive run out, such that the key was a slip fit. I bought a 12mm Techniks collet and run out dropped to under 0.0002" and the key had the expected interference fit. I have had similar experiences with end mills as well so there is no question, at least in my mind, that squishing down a collet reduces accuracy.

Ahh!
Someone correct me if I'm wrong, but we seem to be talking about different collets. The 5C collets you're talking about are only slitted from one end, only compress on the tool near the tip, and are advertised to perform exactly as you stated. The ER collets are slotted from both ends, and compress on both ends as they are pulled into the collet holder. The ER collets are advertised as having a wider holding range.

 

[mikey]

Ahh!
Someone correct me if I'm wrong, but we seem to be talking about different collets. The 5C collets you're talking about are only slitted from one end, only compress on the tool near the tip, and are advertised to perform exactly as you stated. The ER collets are slotted from both ends, and compress on both ends as they are pulled into the collet holder. The ER collets are advertised as having a wider holding range.

Having used ER and 5C collets for about 30+ years, I actually know the differences between an ER and 5C collet systems and I did not misspeak; my comments apply to the ER system.

I assure you that I tested run out on my ER system across multiple chucks. I know how each chuck, collet brand and collet nut that I own performs. After working with the Keo engineers on my little Woodruff cutter issue, I have come to appreciate that the best accuracy of an ER collet is achieved when grabbing a tool shank near its nominal size. The further you go beyond that size, the worse the run out gets. However, that is a separate issue from why I think it is a bad idea to use ER collets to hold drills as Dave was thinking of doing.

ER collets, as least good ER collets, are not cheap. I do not know how far you have to collapse an ER collet before tightening it will cause it to reach its yield point and result in plastic deformation but the potential is there. I have an ER40 3/4" import collet that I use on my lathe that is "loose" now; it was not as loose when I first started using it and I will bet money that it is because I collapsed that collet to hold work that was just slightly smaller than the nominal collet size on more than one occasion. I don't care about that too much because its a cheap collet but I would not risk my ETM or Techniks collets the same way.

So, can you use ER collets to hold a drill? Yeah, you can. Should you do this on a regular basis, No, I don't think it's a good idea. A drill chuck is the proper tool for this job; hence, my recommendation.