View Full Version : Small Carbide end mills

Bruce Griffing
09-18-2003, 07:43 PM
I have several small carbide 4 flute end mills that I use for odd materials like circuit boards as well as for milling odd shapes like square holes in the end of shafts. I was using one that is 3/32" in diameter. I had the mill running at full speed (about 2500 rpm - my best but too slow) and I was cutting Aluminum. The depth of cut was 0.030". Since it was a rectangular cut I was using the power feed on x and hand cranking y. I left the job and came back after a phone call and accidently turned the power feed on full speed - about 15"/min. I quickly turned it off fully expecting to have to replace the end mill. But it cut just fine and was undamaged. This showed me that I don't really have a clue as to appropriate depth of cut, feed speed, chip load, etc. Any suggestions on where to look for this?

09-18-2003, 08:56 PM
Hi Bruce:

Well, as you know the spindle speed was WAY slow for 3/32 carbide in aluminum. But the feed was conservative at about .0015/tooth at 15 ipm. Almost all of the materials suppliers can provide speed/feed information. Also see shopswarf.orcon.net.nz/endmill.html


Bruce Griffing
09-18-2003, 10:57 PM
It would seem logical that the tooth loading should go down somewhat as the mill gets very small. Torsional and lateral ridgidity would seem to me to be a limit, but I haven't found a reference that reduces chip load for very small mills.

09-18-2003, 11:06 PM

A good inference and definitely true if mechanical strength was the basis of optimum feed rates. But these feeds are a compromise between stock removal and cutter life between sharpenings.

Tooth loadings for normal materials are pretty insignificant in terms of mechanical stress/deflection on the cutter.


09-18-2003, 11:48 PM
I like this forum,it reminds me of stuff we all should build,but never do,one is a high speed aux.quill,the things are but simple,That does it,I'm gonna build one next week,Damn the torpeados,full speed ahead!

09-18-2003, 11:51 PM
wierd, if you do build a high speed spindle, tell us about it. I've been thinking about a way of attaching a trim router (with speed control) to the mill spindle ... for obvious reasons, I haven't pursued the project.

Bruce Griffing
09-19-2003, 09:18 AM
Ok - I can accept that the guidelines for chip load are based on a compromise between tool wear and cutting speed. But there has to be a limit. Can you take a 3 mill chip per flute with a 1/64 mill? This must break down at some small diameter.

09-19-2003, 10:27 AM

My intuition supports what you say - almost any generalization breaks down when taken to an extreme. Dunno where the .003 chip load breaks down. At 1/64 diameter, .003 is a significant percentage of the physical dimensions of the cutter.

Advice I've heeded on speeds, feeds and depth of cut is to use published data as a STARTING point. Use your best judgement from there to modify the information to your current needs.


09-19-2003, 09:20 PM
Bruce, your post was effective in that it got us thinking about those little carbide end mills and how to efficiently use them.

Today I made a "proof of concept" prototype, something that I've wanted to try for a time. I mounted a trim router to my mill spindle and used an SCR controller to vary the speed of the router.

The prototype is made of ash, braced and screwed/glued together. Even though it's made of wood, it works pretty well. Couldn't try the 3/32 carbide mill - need to make a collet for the router.

But 3/16 cutter worked on aluminum. Tomorrow I'll make a small collet and if the concept works well for the tiny carbide cutters (and milling steel), then I'll make a permanent attachment, probably from aluminum jig plate.


Bruce Griffing
09-19-2003, 09:59 PM
The 3/32 end mills generally have 1/8 shanks. If they did not I would not be able the use them as 1/8 is my smallest collet. Enco has TiN coated carbide ones on sale (or at least did) and I bought several. I have used them in my router with a jig for cutting steel. They work well. I have a 1/4 to 1/8 Porter Cable collet adapter that works in any router.

09-19-2003, 10:12 PM

Yes, my 3/32 carbide mills also have 1/8 shanks. Impressive that you cut steel with the router and just a fixture ! How did you adjust the spindle RPM ?

I bought a cheap HF trim router to evaluate the concept, it has a plastic housing :( If all goes well tomorrow (after I've made the collet), I'll buy a good trim router, do you know of any with a metal housing ?


09-19-2003, 11:33 PM
Ya,I'll tell all,but the first thing I got to do is find a cheap DA collet chuck and collets,ebay,here I come!

Bruce Griffing
09-19-2003, 11:53 PM
I hate to admit it, but when using the router to make shallow cuts in steel, I did not bother to adjust the speed. I figured that since I generally am too slow, why not try too fast for a change! It worked fine.

09-20-2003, 12:04 AM
Bruce, that's funny :) and thanks.

09-20-2003, 12:29 AM

With a 1/8" endmill SF/M should be 100 sfm @ 3k rpm, if you halve that diameter it drops to 50, so 1/64" you get 12.5 sfm. To cut Aluminu effeciently it should be 1000-2000 sfm - hence the need for speed.

At 3K rpm, a 4 flute .oo1" fpt relates to 12ipm feed, so if you want a .005" fpt then you would use 60 ipm feed.

The smaller you get the biggest problem becomes rigidity of the tool. At 25 microns diameter endmill at 500K rpm maximum chip load becomes 1 micron thick - the tools tend to bend before the edge fails - completely opposite to "normal" milling tools. They use a microphone and DSP to determine if a endmill fails due to edge failure or bending in nano-machining otherwise it takes 8 hours (per pass) to find out you botched the job or not. Sandia Labs is making these 25 micron endmills - interesting stuff.

Bruce Griffing
09-20-2003, 12:47 AM
This is indeed interesting. As you suggest, the chip load must go down as the diameter is reduced. But where are the tables that cover this????

Rich Carlstedt
09-20-2003, 12:54 AM
I run my very small endills in a Dremel tool
Took a piece of 1 1/4 ' thick Aluminum plate shaped like a colon. Bored out the big end for the BP Quill and the smaller end for my dremel with about 5" C/L distance. Put a slot between the bores and crossed drilled and tapped for 3/8 cap screw.
Slip it on the quill and drop in the dremel and tighten one screw and I have a ridged high speed spindle (25,000 RPM)
Gee maybe I should write an article for HSM?

Bruce Griffing
09-20-2003, 01:21 AM
Does the dremel give you good tolerances? My impression of routers and dremel tools was that the bearings were not very good. But the proof is in the pudding. I have a nice Dewalt laminate router I could try....

09-20-2003, 01:21 AM

Yes, you should write an article for HSM, it's an ingenious idea.

There are things that sort of trouble me about using the trim router (and which may also apply to the Dremel tool). Two of these concerns are the ability to absorb thrust and deformation of the plastic housing.

Found an article, several months ago, that described a lathe toolpost grinder made from an inexpensive pneumatic die grinder. I thought that was clever - these little tools are mostly metal, not plastic - and I considered applying the same concept for a high-speed milling spindle (but in my home shop, it would be too loud).

I'm thinking that a nice high speed spindle needs to be custom-designed and fabricated, darn it, much as I like the idea of using an inexpensive high speed device such as the trim router or Dremel tool.

Anybody else have ideas about this ?

Paul Alciatore
09-20-2003, 03:45 AM
I've been thinking about making a tool post grinder for my SB-9 with my old Unimat headstock. The Unimat spindle has speeds from about 300 to over 10,000 RPM. It is all metal and has bearings that are intended for use as both a lathe headstock and a milling head so it should take the loads. I just need to check the dimensions and make a mount. I've already made a grinding wheel arbor and purchased several wheels.

Seems like it could also be attached to the spindle of a vertical mill. Some kind of "L" bracket might work. But then, it may be too much weight off center.

There are Unimats available on E-Bay almost all the time and a "junker" with rusted ways would likely have a spindle in usable condition as the lubrication would have been contained on the inside.

Paul A.

09-20-2003, 10:30 AM
Thanks Paul, that's an excellent suggestion ! Good luck on your conversion.