Issues with Fusion 360 adaptive clearing

[SkyeFire (David McMillan)]

So, the license for MeshCAM that came with my Nomad was my first experience with CAM. And while I agree with people who claim that MC is a great entry point into CAM, I really want to take advantage of some of the features that Fusion 360 CAM offers.

But, since I don’t really know what I’m doing, I decided to start with training wheels. So, I mounted a slab of 2x6 (soft pine, cheap, easy to work with) and tweaked a deep hogging/roughing operation in MeshCam until it was working smoothly. My cutter was a .250" 2-flute up-cut, flat-end mill, at 10k-RPM (everyone who talks about Fusion CAM seems to agree on using max RPM).

Then I changed my origin point, and tried running the same model, but with Fusion’s 3D-Adpative CAM. Same chunk of wood, same cutter, very different cut path.
The helical “plunge” was beautiful – very nearly whisper-silent, I probably could have pushed it a lot harder. Since that was the part I was most worried about, I thought I was past the hard part… but then I hit the bottom of the plunge, at ~90% flute-length engagement of the cutter, and started into the flat outward spiral. And that’s where things got… weird.

I immediately started getting rythmic “pulses” of chatter – very loud chatter for about 250-500ms, followed by 250-500ms of quiet cutting. This kept repeating, very regular.
So, I went back to Fusion and started reducing the Chip Load (which had the side effect of reducing the feed rate), then re-generating the G-Code. But I kept getting this rythmic chatter, on/off. And the rythm didn’t seem to change much, even when I threw my hands up and turned the feedrate down to 10mm/min (no, that’s not a typo). The chatter was so loud, I honestly thought I might be skipping steps on the axis motors, but didn’t have a good way to confirm that.

So… I’m confused. This cutter, in this wood, does good hogging with MeshCAM-style roughing paths, and mostly cutting with the tip of the end mill. My research on 3D-Adaptive suggests that going deep and then cutting outward with the side of the cutter should be easier on the cutter, improve tool life, and give better surface finishes, as long as your chip load per flute isn’t excessive. And at the stupidly low feedrate of 10mm/min, I should have been shaving the wood away a micron at a time. But I still got the darned chatter.

Given how well this cutter/wood combo works on this machine under MeshCAM, I’m inclined against thinking it’s a hardware issue. So obviously there’s something I’m badly misunderstanding about the Adaptive settings. But slowing down (even enormously) didn’t help at all, and I’ve been too much of a chicken to try getting more aggressive yet. The rythmic nature of the chatter also puzzles me – if nothing else, I would have expected the “beat frequency” to change when I made large changes to my feedrate, but it didn’t really seem to have much effect.

[neilferreri (Neil Ferreri Ω)]

Did you happen to capture a video?
Can you share your Fusion file?

[mikep]

The adaptive path attempts to keep the chip load on the cutter about the same at all times when the tool is engaged. On of the big differences here is that the depth of cut can be quite a bit higher than anything meshcam can successfully execute in a single path (the plunge to do so is…scary as all heck).

The “2d pocket” is probably a better place to start. It works much more like meshcam does and doesn’t shoot for absolutely ideal engagement. It does still support the ramp/helix type entries, and should get you something a little more comparable back to back.

[PhilG (Phil Gorsuch)]

First, I salute you for having the courage to go full plunge on your adaptive - I was not so brave and ended up doing multiple passes at shallow step downs to start.

I too have struggled with a similar problem on the adaptive the couple of times I have used it aggressively and look forward to some time to play around and optimize. I will mention that when I look at the surface finish it is far better in one direction than the other which makes me wonder if one specified only climb or only conventional milling during adaptive instead of the default of both if that might make a difference.

[mikep]

It’s intended as a “roughing” strategy, and will by default leave some material as well for a finish pass with another strategy.

[PhilG (Phil Gorsuch)]

Agreed - neither direction is a perfect finish - just noting the chatter is worse in one direction than the other

[enl_public]

Unless you have a good reason not to, you should specify “climb only” for the adaptive (trochoidal) strategies. Though they are a roughing strategy, they can be used for finishing in many cases, but it is usually easier to leave a little stock and perform a finish pass with a finishing strategy.

short detail: in many materials, climbing allows for a better cut and better finish than conventional. It starts the cut where the chip has thickness, so the cutting edges don’t tend to rub as much, which can make a difference in heat that will affect tool life as well as melt (or burn) many materials. It also tends to move swarf behind the cut rather than in front of it, which is sometimes useful if you aren’t/can’t use good chip extraction.

[SkyeFire (David McMillan)]

Good grief, employer has had me running all over the continent the past few weeks. Finally got a chance to make a video of what I’m seeing&hearing:
https://photos.app.goo.gl/kryHquPJnfQnm1gaA

It sounds a lot worse in person that the phone mic picked up.

[PhilG (Phil Gorsuch)]

Sounds vastly different from what I experience on mine. Is that lugging where the RPMs are wandering up and down as well?

[enl_public]

That is a rather unpleasant sound. I have not heard this from my machine, and I have worked it pretty hard on wood, plastic, fibreglass filled nylon and PTFE, and aluminum.

I would guess it is not due to the cut strategy, but is another issue. I don’t know where it is coming from, since I can’t think of where in a Nomad this could be, but it sounds like a dry bushing (friction bearing), or something loose elsewhere.

[mikep]

That sounds a lot like the machine itself is resonating with the chatter. I think what you’re hearing is the machine going into and out of resonance with the chatter (think of it like a beat frequency, as you say). Either way, lots of chatter. What’s the maximum stepover you’re using? Adaptive attempts to keep the endmill continuously engaged at maximum stepover for maximum clearing per revolution, but it can’t actually do that, it just tries to get close. It changes speed as it goes around curves to some degree, etc, so the chatter is likely to change fairly rapidly as well.
The video shows cutting -way- faster than 10mm/min.
You are correct, cutting with the side of the mill should actually be easier on the mill, and it should dull much more slowly. The nomad’s big problem is power. It just doesn’t have much. The cut you show I would do on my shapeoko in a heartbeat, but wouldn’t attempt with my nomad - I’ve never had much luck with slotting (100% engagement) on my nomad.

I ran softwood with a .250" endmill and a .2" DOC, slotting (100% engagement), through gwizard and it says 5ipm, at 10k RPM. If fusion, just set the feed rate, don’t mess with the chip load, all the feed rates (plunge, etc) are on the first panel of the toolpath properties in fusion.