0
$\begingroup$

I was milling a pocket recently with a 7/16in endmill. It came out 20 thousandths undersized. Worn tool maybe? To try to fix this, I expanded the coordinates by 10 thou and milled the pocket again. Surprisingly, when I measured afterward it had barely changed.

My guess is that the cut was too light and the endmill was deflecting instead of cutting? Is that likely what happened? I didn't try to go farther out of fear that it would start cutting all of a sudden and oversize the pocket.

The material is 7075 aluminum, and I had to clamp only about 0.5in of the endmill in the collet, so that I could reach down far enough into the part. I know that's not ideal. The distance from the collet to the blades was about 1in, and I cut approximately 0.6in depth per pass (just on this operation).

What do you think happened, and how can I remedy this?

Edit: I repeated the original cut using plunge milling followed by a finishing pass to cut off the bumps and that fixed it.

$\endgroup$

1 Answer 1

1
$\begingroup$

Yes, you can definitely take too small a cut. This becomes quite evident when working on a manual machine. The smallest cut you can effectively take depends on the material, the sharpness of the cutting tool, and to an extent what the cutting tool is made of.

Materials have spring to them so taking too light a can cause the material to deform (I'm talking about the workpiece here but I suppose it also holds for the cutting tool) under cutting pressure and not "bite" preventing a proper cut from happening. When this happens the cutting tool instead rubs producing excessive heat made worse by the absence of chips to carry the heat away which kills the cutting tool, or deforms the material instead which works harden the material.

Therefore a sharper tool lets you take smaller cutters since it reduces cutting pressure is required.

All that said, I don't think that should be happening at 10 thousandths in aluminum in a tool of good condition. But I am not familiar with 7075.

Take a cut on a test piece and see how much is actually cut versus the coordinates of the tool. Sometimes it's more and sometimes it's less.

$\endgroup$
4
  • $\begingroup$ A much more palpable example of this can be had by machining some plastic, like Delrin or PTFE. My experience with this is in turning. Even with a well-supported workpiece, you have to have a sharp cutter and you have to make a big enough cut (I can't remember exactly through the trauma, but I'm thinking something like 0.005" minimum). Unlike something well-behaved like 2024 or 6061 aluminum, you can't edge up to a good fit with a bunch of 0.0005" cuts -- you need to make a dramatic cut that ends up at just the right size, or you'll end up cutting the piece undersized. $\endgroup$
    – TimWescott
    May 18, 2023 at 19:51
  • 1
    $\begingroup$ @TimWescott Yeah. I had to clean up a lathe where someone tried to turn delrin with carbide tools. Dust everywhere. With sharp HSS it cuts in strips. With PTFE I was turning recently there was a specific number I kept running into. I forget what it was specifically, but it was something the minimum cut always had to be 0.007" to actually cut anything and the resulting cut would always be a bit over by the same amount which I can't remember. $\endgroup$
    – DKNguyen
    May 18, 2023 at 20:17
  • $\begingroup$ I agree 10 thou doesn't seem like an excessively small cut to me either, but maybe it was a combination of factors, worn tool, thin part wall (about 0.1), or something like that. It sounds like maybe with a much shallower cut and a sharper tool it might be ok? $\endgroup$
    – Drew
    May 19, 2023 at 1:10
  • $\begingroup$ @Drew If it's thin walled, the workpiece will deflect which is why you need tools with lower cutting pressure (i.e. sharp, small corner radius, negative lead angle) for thin walls. You need smaller corner radius and sharpness if you're trying to take shaving cuts anyways and thin walls just increases the need. $\endgroup$
    – DKNguyen
    May 19, 2023 at 2:55

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.