What's new
What's new

A curious case of exploding end mills

trochoidalpath

Cast Iron
Joined
Jan 17, 2016
I had an expensive day today and I am trying to figure out why.

This 6061-T6 part has a number of 0.196" wide slots, either 1.417" or 3.346" long, 0.127" deep. Roughed them out with a ramp down at 2 degrees to full depth, leaving 0.005" per side for a finish pass later on. I was using Helical H40ALV-3 1/8" end mills in either Kaiser NBC with on-size collets, MST shrink-fit collet holders, or ER collet chucks. Running on a Speedio. Very good chip evacuation with flood coolant (Trim 690XT).

I blew up no less than six end mills. Even for me this is egregious. :o

I kept backing off the speed and feed, and ended up at very conservative full slotting numbers (16000 rpm, 19 in/min, 0.060" ramp depth). The broken end mills were pristine -- no broken tips, no visible wear, no chip welding whatsoever. Four of the six broke inside the nose of the collet, which trashed two collets and two shrink-fit collets. :angry:

That ran me out of 1/8" aluminum-specific end mills, and I needed to get the part done, so I grabbed a random 1/8" 4-flute and ran that. It finished the job without complaint.

Maybe a bad batch of carbide from Helical? Those were all ordered at the same time, and have sat in my drawer until today.

I'm just pissed at scrapping a bunch of tooling and I'm trying to figure out why.
 
Check the flutes on the ends for an improper primary radial clearance. If any of the original cylinder was left from the blank they would cut poorly, with excess bending load on them.

Did you notice any "berm" buildup in front of the cuts? Any other unusual behavior?

It's also possible that they had a bad batch of carbide rod with internal cracking. How do the broken ends look? Uniformly granular, or are there shiny areas, or other discontinuities?

Pictures, as ever, would help here.
 
I don't have a way to measure the primary radial clearance, but under 10X magnification the flutes look properly ground to me (not an expert by any stretch).

There was no buildup in front of the cuts. They just exploded. During cutting they had the usual little squeak at entry but otherwise seemed pretty normal. Some made it only a few slots, one did about 50 before it gave up.

I remember at least one broken end has some shiny patches in the cross-section. Usually the cutting end went "ping" and fell out, there was a retained short disc of carbide debris in the collet, and then the remainder of the shank was in apparent good shape.

I have at least a few of the end mills in the Bucket of Shame, so I can grab photos the next time I'm out at the machine.
 
Getting pics will be good, try to make them as large and clear as possible.

On the two bad shrink holders, you can probably trim the ends to clean metal and re-chamfer, as long as there's not a slug left currently that you can't remove. Even then, you could trim back, leaving a thin ring of steel, then try scoring the ring to allow it to break off. It may not be worth the effort, but WTH...
 
Last edited:
Blowing up end mills sounds kinda cool. Like home made fireworks. Mine just break, and fall into the chip conveyor. :(

Ok, seriously, is there enough spindle RPM to do what you want? Maybe you're short on RPMS. The machine sounds pretty capable, but that's a small dia end mill.
 
1/8" 3fl, 19ipm and 16k = ~.0004ipt and ~520sfm. Those seem like pretty safe parameters to me.

I'd guess chip evacuation, maybe the 4fl had enough core strength to suck up the difference? Can you go to a mist arrangement?

How's your TIR? I know you're using good tools and holders in a good machine, but it's worth a look. Even .0002" is 50% of your feed....

PS- I wish I had a good graphic of how TIR changes tooth loading in tools based on eccentricity and timing to a tooth or gullet.
 
We have had quality issues with Helical lately. They are still our goto for finishers, but we got a batch of .125 or .187 40 degree endmills with relieved shanks in the other day where the grind on the shanks was so bad I could see they were garbage from 10 feet away while they were still in the sleeve. :eek:

They did send us out new ones promptly, but they were nowhere near as apologetic or excited about the quality escape as I would have been if we sent out parts that looked like that.

Helical also sells a couple geometries that just don't perform as advertised (looking at you 45 degree AL roughers). Not sure why they keep that crap in their catalog when they know it doesn't work.
 
Might sound too simple, but if you are running soluble oil, is your concentration o.k.?

I checked the concentration and if anything it was a little high. I usually aim for more like 8%, yesterday it was 9.5% (the design concentration for 690XT is 7.5-10% if memory serves).
 
That ran me out of 1/8" aluminum-specific end mills, and I needed to get the part done, so I grabbed a random 1/8" 4-flute and ran that. It finished the job without complaint.

Maybe a bad batch of carbide from Helical?

This ^^^

Given all the other info, this points strongly at a bad batch of endmills.

Regards.

Mike
 
1/8" 3fl, 19ipm and 16k = ~.0004ipt and ~520sfm. Those seem like pretty safe parameters to me.

I'd guess chip evacuation, maybe the 4fl had enough core strength to suck up the difference? Can you go to a mist arrangement?

How's your TIR? I know you're using good tools and holders in a good machine, but it's worth a look. Even .0002" is 50% of your feed....

PS- I wish I had a good graphic of how TIR changes tooth loading in tools based on eccentricity and timing to a tooth or gullet.

I have a TDC coolant manifold, and it was dumping coolant from around 120 degrees. I never saw any chips in the slot, part, tool, anything. No reasonable way to go to a mist setup on the Speedios.

I confess I didn't measure TIR in the spindle. I used separate toolholders, so if it was a TIR problem, it has to be in the spindle... which seems unlikely for a Speedio that hasn't ever been bumped. But worth checking.
 
Before I jump to conclusions based on the endmills breaking inside the collet, let me ask if they broke at the tops of the flutes, broke right across the cylindrical shank, or broke the flutes somewhere below the top. Post #3 sort of suggests the second case, which would be due to bad endmills.
 
Before I jump to conclusions based on the endmills breaking inside the collet, let me ask if they broke at the tops of the flutes, broke right across the cylindrical shank, or broke the flutes somewhere below the top. Post #3 sort of suggests the second case, which would be due to bad endmills.

One broke in the middle of the stick-out, but the flutes were intact.

Four broke at the shank inside the nose of the collet. The broken face was recessed relative to the collet face. The broken end chewed up the collet on the way out.

I hope that helps clarify?
 
I sell Helical as a distributor. We do see quality issues from carbide spring up randomly. If you had that much trouble with such a simple application and variety of holders. It's probably not you as long as your program isn't doing something it says it isn't.
 
Take the tool out of the spindle and clean the inside of the spindle taper. Check how many chips are getting into your carousel and on the tapers on the other tools.

Check the taper on the holder you are using for mashed chips. Check all of your tools in your carousel and clean all tools and carousel of any chips. You are flinging chips at very high speeds. I thought you had a three flute end mill not 4?

It may only be a matter of seeing where there are ops where a lot of chips fly mixed in with coolant. You should step process each op and blow everything off before the tool change and note chip disposition.

Only one thing I would consider first. I always note that kind of thing right off. Simple yet very important usually that can be the case when problems seem to pop up out of nowhere.
 
I had an expensive day today and I am trying to figure out why.

This 6061-T6 part has a number of 0.196" wide slots, either 1.417" or 3.346" long, 0.127" deep. Roughed them out with a ramp down at 2 degrees to full depth, leaving 0.005" per side for a finish pass later on. I was using Helical H40ALV-3 1/8" end mills in either Kaiser NBC with on-size collets, MST shrink-fit collet holders, or ER collet chucks. Running on a Speedio. Very good chip evacuation with flood coolant (Trim 690XT).

ramping at 2° you'd get to full depth in short order, which at .127" is the full dia (1xD). Combine that with .0004 chip load, quoted, it sounds like you have a mess of vibration and rubbing the tool to death.
keep the cut depth under .03 and take an actual chip, is how i'd go.
6061 is sticky and alloyed with silicon, but it doesn't sound like you have a coolant lubricity or flow issue.
 
ramping at 2° you'd get to full depth in short order, which at .127" is the full dia (1xD).

HUH?! You must ramp a whole lot different than I do. My math says a 2' ramp from zero to -.127 will take 3.639"

Anyways, my money is on bad carbide. And, I freaking HATE when they take the collet with them! :angry:
 
Considering they broke in the collet or toolholder makes me think carbide. What flute length? You REALLY need to use .156" stubs for that, you can run them a lot harder than 1/8" mills IME. Harvey would be my current pick for that exact size and flute length.
 








 
Back
Top