A Question of Runout

[macardoso]

This comment is based on ER collets for use in a milling machine. Not the same thing but hopefully some of the methodology is applicable.

I've purchased a number of import collet chucks and collets over the years. I've standardized on ER20 collets for my CNC and have 20-30 toolholders at this point.

My CNC spindle has a TIR on the spindle taper (checked in multiple places using a calibrated Interapid .0001 DTI) of .0002". This isn't particularly good but it is what I have to work with.

The collet chucks are held in a 3/4" precision R8 collet (Tormach TTS) which contributes to the TIR giving me about .0003" when measured on a 3/4" test bar roughly 1" from the spindle. Again not great but it is what it is.

The import chucks are guaranteed to have a TIR <0.0004". Of the 20-30 I've purchased, 6 have been out of spec by as much as 0.0025". These were replaced, but I could have been using garbage if I did not inspect them. Most of the chucks read about .0001-.0002" when spun on a Vee block with a test indicator in the bore.

Depending on how I clock the toolholder in the spindle I can consistently get runout in the spindle of .0002-.0008" or so.

Then it gets bad. The import ER20 collets I've bought are absolute garbage. They are very rough finished and have TIR from 0.001-0.012" (yes, those are thousandths). I've given up on those and only buy Techniks collets at $10 each. I trust them to their 0.0002" TIR.

Finally, the collet nuts. I took my best collet chuck and collet and got them running as true as possible in the spindle. I then tried assembling a test bar with each nut that I owned. The measured runout was highly variable based on the nut used. All were torqued evenly. I have numbered each chuck and nut, and created assemblies which mate the highest precision nut to the highest precision collet chuck.

After all this effort, I have 8 chucks which have excellent assembled TIR (<0.0003") and are good for small tools or anything where the runout really matters. I have roughly 10 chucks which have acceptable (<0.001") TIR and are ok for general purpose use. Finally I have 6-8 chucks with 0.001-0.002" assembled TIR that are only good for roughing tools or drilling.

All this saved me from paying ~$100 per tool assembly ($2-3000 in total) to have guaranteed precision out of the box, but it was a lot of work.

Just sharing my experience,

Mike

 

[T Bredehoft]

TIR giving me about .0003" when measured on a 3/4" test bar roughly 1" from the spindle. Again not great but it is what it is.

.0003 is to die for. Many of us work with as much as .002 and survive. Granted I don't work for NASA, but anything under .001 is bonus. Its hard to see something .001 in diameter.

Chasing tenths is a mind game not worth perusing. Please consider your good tooling excelent, throw away the trash and be happy with what you have.

 

[macardoso]

.0003 is to die for. Many of us work with as much as .002 and survive. Granted I don't work for NASA, but anything under .001 is bonus. Its hard to see something .001 in diameter.

I'm very happy with what I have. Just sharing what it took to get there.

When I work with small diameter cutters (1/32 or smaller) then couple tenths of runout makes a huge difference in the success of machining and not breaking your tool. That is what I save the accurate stuff for.

Edit: Again, this is for milling spindle runout. I don't go through this much effort for my lathe unless it is really needed.

 

[epanzella]

I'm very happy with what I have. Just sharing what it took to get there.

When I work with small diameter cutters (1/32 or smaller) then couple tenths of runout makes a huge difference in the success of machining and not breaking your tool. That is what I save the accurate stuff for.

Edit: Again, this is for milling spindle runout. I don't go through this much effort for my lathe unless it is really needed.

This thread IS about collets in a milling machine.

 

[macardoso]

Whoops, I must've misunderstood. I read the posts above and believed we were discussing collet chucks for a lathe. Principles are all the same but depending on the work you do, you might have different requirements for accuracy in each application.

Hope you find that your milling chuck works well for you. Good quality collets really seem to go a long way. Collet nuts are replaceable if you get a bad one.

 

[epanzella]

Whoops, I must've misunderstood. I read the posts above and believed we were discussing collet chucks for a lathe. Principles are all the same but depending on the work you do, you might have different requirements for accuracy in each application.

Hope you find that your milling chuck works well for you. Good quality collets really seem to go a long way. Collet nuts are replaceable if you get a bad one.

No prob. Threads often get mutated as they chug along and their beginnings become ancient history. My ER32 R8 collet chuck has .002 TIR and it seems to me this would be a good thing to replace. I'm currently looking for suggestions in a decent chuck that won't break the bank. Being just a w/e warrior I don't think spending $375 on a chuck is in my future.

 

[macardoso]

If anyone was curious. These are the collet chucks I've standardized on. They all have a common 3/4" straight shank to be held in a modified 3/4" R8 collet. This is Tormach's TTS Tooling System.

New 10pcs C3/4 ER20A 1.38" Collet Chuck Straight Collet Chuck Out stock 793574701823 | eBay

10pcs C3/4 Er20A extension shank collet holder. New 10pcs C3/4 ER20 A 1.38" Collet Chuck Straight Collet Chuck CNC Milling Lathe. High precision ground machine made. Shank length : 1.35 ".

www.ebay.com

I have also worked with the company directly (sales@igstool.com) to purchase items not listed on their ebay site, such as holders with geometry for use in a tool changer.

For me (a CNC user) having a large quantity of interchangeable tools with predefined tool length offsets is really valuable. I'd imagine even on a DRO it would be as well.