I have a 2 mm thick high-quality carbon fiber composite sheet. I want to cut a shape in it with a 0.8 mm diamond flute end mill. I need a cut this narrow because of the shape's detail. The shape's total cutting length is about 30 cm, cut depth 2.1 mm (three passes of 0.7 cut depth). The plate is glued to a solid plastic base and pressed with several clamps, then submerged in water.

The mill end breaks after about 10 cm (HSS) to 40 cm (tungsten carbide). Current mill parameters are 40 mm/min, cut depth 0.7 mm, 10 000 rpm, but I tried anything starting at 10 mm/min. Does not seem to change a lot.

The CNC is 3018 upgraded with a 500 W 10 000 rpm brushless DC motor. When looking at a rotating end mill, no soft edges are seen in hard light (there is instead an illusion that the mill is not rotating at all) so I think that the spindle's vibration is low. But anyway, I changed the spindle to a small brushless DC 20 000 rpm and tried it at rpm from 8 000 to 20 000. Same thing, the end mill breaks after some dozen cm.

A 3018 CNC is not top quality for sure, but it seems to work smoothly. It is controlled with upgraded electronics - drivers are TMC2209, firmware is GRBL. Maybe there is a ~0.1 mm backlash when looking at a circular cutout, but I'd guess that a backlash is something which may pause the spindle's XY for a moment, not something which jerks it, at least not at only ~ 30 mm/min.

I had never problems with this machine when using end mills of diameters >= 2 mm (can't say about diameters > 0.8 mm, < 2.0 mm because I have never used them). Also, the said 0.8 mm end mills do not have any problems with a very hard maple wood.

This microscopic photo shows an average quality of the cut, pass depth 0.7 mm:enter image description here

Spindle power consumption at 15 000 rpm, 0.4 mm cut depth as suggested by fred_dot_u:

  • free: 17.3 W
  • isolated groove 0.4 mm deep: 18.2 W
  • isolated groove 1.6 mm deep: 18.7 W
  • adjacent groove 1.6 mm deep: 18.3 W

Any ideas? Is it really that hard to cut a carbon fiber plate with a sub-mm end mill or something is wrong here?


1 Answer 1


I've read your question twice, and may yet have missed one detail of value. How deep is your cut? I see "feed 0.7" but not "feed cut depth."

According to an Inventables web site:

As a general rule, your depth per pass should never exceed half the diameter of your bit. For example, the depth per pass for a 1/4" (.25") bit should not exceed .125" per pass, the depth per pass for a 1/8" bit (.125") should not exceed .0625" per pass, etc.

This means that for your 0.8 mm bit, you would want to make five passes of 0.4 mm (or more passes, reduced depth) in order to meet the above referenced requirement. I will presume that the carbon fiber you are cutting is a composite, rather than raw fiber, which may complicate the circumstances.

Performing the cut in a water bath probably resolves any heating of the epoxy which would certainly complicate the project.

Ensure that your work piece is well secured to the sacrificial sub-board. You very much do not want vibration on the work while cutting. Double stick carpet tape is a common substance and may require heating to remove without breaking the part, but carbon fiber composite can handle temperatures well enough as long as the epoxy resin is not damaged.

  • $\begingroup$ I corrected the question. It is indeed a composite. Cut depth is 0.7 mm. I'll try 0.4 mm tomorrow and write an update. $\endgroup$
    – scriptfoo
    Commented Apr 17, 2023 at 22:18
  • $\begingroup$ Did not seem to help. I am looking now for other problems in the setup. I'll write back. $\endgroup$
    – scriptfoo
    Commented Apr 19, 2023 at 16:10
  • $\begingroup$ Well, I assume water is for dust, but water does complicate things by getting in between the laminate at the edges. $\endgroup$
    – DKNguyen
    Commented Apr 19, 2023 at 17:44
  • $\begingroup$ Note that the recommendation is for a maximum, and if you reduce the depth of cut to perhaps half of what you have, may improve circumstances. Edit added to reflect additional considerations. $\endgroup$
    – fred_dot_u
    Commented Apr 19, 2023 at 21:39
  • $\begingroup$ @DKNguyen, one would hope that the composite was created in a proper manner, effectively creating a layered monolith rather than layers which would split under cutting operations. Vacuum forming carbon fiber ensures this, but is not always a certainty. $\endgroup$
    – fred_dot_u
    Commented Apr 19, 2023 at 21:42

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.