ER32 collet chuck criteria for milling

What a load of rot ! Most cannot even measure a tenth of a thou !

Morning, Baron. I'm trying to decide whether I should be offended by the tone of this response. I assure you that I did not make up the information you quoted but perhaps you know more about it than Lyndex and Techniks so why don't you enlighten us so we can correct them?

And yes, I can measure tenths accurately.
 
I don't run a CNC mill. In fact I only have an old Taiwanese RF-31 and a Sherline 5400 mill but both have under 0.0001" TIR at the spindle and both use ER collet chucks that are quite accurate. Since I upgraded to better collets for both mills I can tell that my end mills last longer. However, one experience taught me the value of accurate collets.

I had to cut a Woodruff keyseat in the end of a cross slide lead screw. As you know, the proper fit is an interference fit and that requires an accurate cutter. Luckily, I was practicing on a piece of scrap 1144 steel and was using an import collet that I thought was pretty accurate - specs of under 0.0003"TIR - but the key was loose. I then tried the same cut with a Lyndex collet and got a better but still slightly loose fit. Then I switched to my Techniks collet and the same cutter cut an interference fit in the same material.

For most tasks we may not notice a difference between good and inferior chucks/nuts/collets but there are times when it really shows. I learned that cheap collets are fine for work holding on the lathe but for the mill, I prefer to use good collets.
 
  • Like
Reactions: Z2V
Hi Mikey,

Morning, Baron. I'm trying to decide whether I should be offended by the tone of this response. I assure you that I did not make up the information you quoted but perhaps you know more about it than Lyndex and Techniks so why don't you enlighten us so we can correct them?

And yes, I can measure tenths accurately.

I agree that both you and I may be able to measure to these limits, but most on these forums are hobbyists and would have little ability to do so !

I certainly couldn't define tool life in those terms, not that I would be bothered about getting an extra dozen cuts from a tool that I may only use infrequently. I generally aim for accuracy of a thou or better, so a tenth is neither here or there.

Mitch, made a good point about normal tolerances due to wear and tear on spindle play ! Hitting tenths of a thou is very hard to do repeatably without precision machinery anyway. Making trying to achieve that kind of accuracy extremely expensive ! Way out of a hobbyist pocket.
 
I agree that both you and I may be able to measure to these limits, but most on these forums are hobbyists and would have little ability to do so !

I dunno', Baron, I happen to know a lot of highly skilled "hobbyists" on this forum who are capable of doing professional level work so let's agree not to go there, okay? We wouldn't want to offend anyone.

I certainly couldn't define tool life in those terms, not that I would be bothered about getting an extra dozen cuts from a tool that I may only use infrequently. I generally aim for accuracy of a thou or better, so a tenth is neither here or there.

The point was that the accuracy of the tool holding system matters with regard to tool life. It's pretty well established and accepted throughout the industry and is certainly something even a hobby guy should be aware of, so I raised it. This has absolutely zero to do with the level of accuracy the user is working to. I don't know how to make this any clearer for you, Baron.

Have a look at the tips of the teeth of an end mill. See those little hooks? Typically, the wear you see is not even. One or possibly two will be worn more than the others and this definitely affects the function and accuracy of the tool. That asymmetric wear is an indicator of run out in the tool holding system, and you can be sure this extends to the sides of the end mill as well. How much of an impact this has on tool life, finishes or accuracy depends on the amount of wear that exists. In a hobby shop, tool life may be longer than in a pro shop but shorter life is shorter life as far as I'm concerned.

Your ability to measure how much run out there is or the accuracy levels you aspire to or even the amount of spindle run out the machine has are separate issues that have nothing at all to do with the fact that the more run out the tool holding system induces, the more asymmetrical wear there is and the shorter the tool life will be. That is what the collet makers are pointing out. I was doing the same thing. It is not "rot", or BS as we say on this side of the pond; it is fact.

We come from different cultures. Quoting something someone says and labeling it as a "load of rot" may be acceptable in your country but it will be seen as an insulting challenge here. Just some friendly advice.

As for me, in light of the above consideration, I decided not to be offended. Life is too short for that sort of thing.
 
Hi everybody. Very interesting. This definitely means that the specs are important. It may be more strict for carbide cutters, which I am not planning to use right now. For example, here is a link speaking about tool life. The general rule quoted is keep the runout less than 20% of the chip load.

https://www.cnccookbook.com/spindle-runout-measure-fix-tool-life/

Very nice detailed discussion on the subject. Thanks for posting. Tool like is a significant issue in production and CNC operations, and can significantly impact run costs. Typically in the hobbyist setting it is less significant/measurable because we are not pushing to the same limits as in production, and at least for me, I typically chip a flute from dropping an end mill long before I wear it out. Needless to say, rigidity and TIR are significant factors in tool life and finish tolerances. I will say that I have had significant TIR issues with lesser quality chucks and collets, in similar situations to what Mikey described above. As a self taught machinist, most of my work needs to be 0.001" or better, and setting up both my lathe and mill requires precision levels of 0.0001" resolution. I have a quite a few dial and test indicators, but I only use two (Starrett 25-511 and an a Compac 215GA), both which measure 0.0001". Both of my center finders for the mill needed to be setup to tolerances of less than 0.0004", I had purchased Shar's metric collets to set them up (both R8 and ER32). Both collets had TIR which far exceed these limits by quite a bit. I ended getting a PBA R8 collect and a Technics ER32 collet, which were far better. The ER chucks I use have a TIR of under 0.0001".

Precision is relative to the user, the machine, tooling and project requirements. I do not look at it as being a hobbyist, as most of my projects require finished tolerance of better than 0.001". I Also work with a number of other hobbyist that this also holds true. It is no more time consuming to do so, at least for me, in particular with the tooling/machines I use.

One of the features that makes the ER systems attractive is the system accuracy, relative to say an R-8 or 5C collets, and the range of clamping for the collets. The TIR with an ER system is a function of the collets, chuck and nut used (assuming the machine has minimal runout). If accuracy is not an issue then the whole discussion is a bit moot, use an ER Chuck you want. If you need accuracy, then get a chuck that specifies the tolerance limits and check it. There are many good quality ER chucks and collets at moderate pricing.
 
Since there is so much uncertainty with the online offerings, with the exception of Shars, I was wondering how hard it would be to get decent tolerances if one turns the adapter in house. It seems that the two main sources of error would be the chucking of the adapter, since it will have to be chucked to bore the tapered opening for the collet. The job could be done between centers, then carefully chucked on a parallel section of shank behind the area to be tapered. Would it be reasonable to get 0.0004 TIR with this sequence, assuming a good collet?
 
Back
Top