Damage to 1/8" ER16 collet bore under "normal" use

We’ve been milling parts with a 1/8" cutter, 2-flute, in both aluminum and plastic and had this happen to the 1/8" collet we were using with a Velox Quick Change system.

The bore appears to have softened and hollowed out slightly, “flowering” at the tip and into the edges of the collet, resulting in not being able to successfully grip tools.

Example feeds & speeds include 23kRPM, 30ipm, 0.80" DOC, 2 flutes in Aluminum.

Has anyone else seen failures like this? This is completely new to me.

Typically a cause of premature collet failure/wear is due to tools with a short shank being put inside the collet with the tool not all the way inside the collet( typically you want at least 80%) otherwise it causes the collet to flex and get damaged, its the type of thing you can do in a pinch but it will damage the collet over time.

99% sure you’re onto something with this suggestion - we have definitely blindly extended tools out in order to cut at >1" (bottom side of 2x1 drive rails being cut in a single pass…)

With these possibilities considered I would still not expect for the actual end of the collet to soften and flower like that, it is honestly quite crazy to see considering that the steel in collets is usually very hard. It could be that you were at one point running very slowly causing the tool and then the collet to heatup and basically anneal itself. But this is unlikely.

Maybe, actually? We’ve definitely melted aluminum onto our cutters… more than once… mostly when new operators haven’t understood what “A LOT” in “use A LOT of cutting fluid” means.

I think this points generally towards needing to integrate more math and ‘hard number checks’ into our SOPs

Could you have been not removing large enough chips to get rid of heat. We use single flute so that a bigger chunk is taken out which lets it run cooler. Is it possible the tool got so hot that it took away the temper on the collet? Then additionally trying to “reach” with the tool?

1 Like

Yeah, not putting enough shank into the collet will cause the PSI to become to high at where it interfaces and deform. But what you have pictured looks like the tool broke while spinning and mangled the end. I have quite a few of both cases here at the SunPower Davis location we use with our router and CNC mills.

1 Like

Maybe. We mostly run super hot, very low chip-load, with a lot of coolant, because we’ve found our PorterCable ~2hp head doesn’t like running below full blast and our newer operators aren’t ready to run high feedrates. We tend to break endmills frequently enough through operator error (1-2x per season, cutting ~1.5 robots of parts) that dullness doesn’t become an issue.
Our low chipload allows us to take fairly deep passes at least, but I think I’d be willing to sacrifice that to get closer to optimal heat in our cuts…

Well we’ve definitely done that too. Plenty of broken 1/8 mills through that collet. Wouldn’t it need to have the cutter up at that shank level to do that kind of damage? Or is the simple leverage of the steel enough to have that effect?

Are you tightening the collet enough? I crank ours down really hard. Whats the runout of the spindle? Maybe make sure the collet is super clean. If it’s not clean, then you won’t be getting as much clamping pressure as you would if it is clean

We have ER32s at my old work that have been around for 10+ years and don’t look like that. We tighten as much as possible.

What endmills specifically are you using? The feeds and speeds look okay to me. Looks like maybe the endmill is flexing somehow. If it’s carbide, it would just break. HSS might flex leave that mark, however based on the feeds+speeds, you’re probably using carbide

I would invest in higher quality ER Collets. Due to the lack of marking on the collet, It looks like it’s a cheaper chinese one, however, our chinese collets are working fine so far.

1 Like

That’s my “chief culprit” - we have primarily student operators, mixed gender 9th-10th grades, and we’ve been using auto-rounder 9000’s to tighten the collets


So, I’m working on addressing that. I don’t think I can necessarily do much about “crank ours down really hard” given our demographics, but at least I can give them the right tools so they’ll have a shot at it.

I am leery of trying to put a torque control on it. Especially because I don’t want them cranking down like that on our R8’s, which don’t need that treatment. Anyone tried putting a torque number on the collet tightness?

Judging from the damage to the inside of the collet and the marks on the outside, it looks like you spun the tool inside the collet or spun the collet inside the holder or both.

This could have been an annealing problem where your tool spun in the collet, which would have heated then slowly cooled the steel. That could make your hard steel not so hard.

There’s a bit of deformation on the inside hole where it looks like the tool dragged around part of the collet. Do you know if that happened before your last tool change? If so, you would have been much looser on clamping force with the same torque.

1 Like

We have had that happen with tool breakages before. If I remember right the bit broke on entering the metal, the machine wasn’t stopped fast enough, and the result was the collet lowered onto the broken bit shaft.

1 Like

The only way that I can see the metal bending like that is if it heated up enough to anneal it’s self.

1 Like

What brand collet are you using? Would not be surprised if you have pot metal collets, if you’re buying bottom of the barrel ones.

These came from the Velox Quick Change set four years ago, and the damage was only noticable this year, along with the beginning of something similar on our 1/4". If we’d gotten this in the first year after an aliexpress buy, I’d be more inclined to weight manufacturer quality. Not that it doesn’t play a role, but if it does - it’s not a driving force.

I think a combination of under-tightening resulting in breaking the tool while cutting and insufficient insertion depth are much more likely culprits, as we’ve been integrating more and more student operation of the machine into our processes and there is clearly some room for improvement in how we do that - ie twe need to write down things into an SOP for the students, rather than trust “oh yeah we told them that recently” and recently was actually two years ago…