Nomad Cutting Quality

I am testing some cuts on aluminum for a project I’m working on.

There are two issues I am seeing with the precision and quality of the cuts that I’m hoping someone could shed some light on.

The first issue is that I have the tolerance set to .001 and the machining quality set to high, but there is some obvious faceting around the edge of a 2 inch circle. The larger inside cut of the circle doesn’t show any faceting.

The second is that I import the STL file, and everything seems to be the correct size. When I am setting the stock dimensions in MeshCam it defaults to the smallest stock size, which verifies that MeshCam is importing the size I am expecting. My problem is that the diameter of the circle should be 2 inches, but the actual diameter is 1.971 on the flat facet, and 1.983 on the points. This doesn’t seem to fall within the stated tolerances of the Nomad. Any one else run into this?

This was cut with the #102 bit. I babysat this job and cleared chips away after every pass, and drenched it in WD-40.

Picture attached. Any help is appreciated.

The outer portion of the cut doesn’t show any faceting.

The faceting looks quite even — are you sure that it isn’t part of the model? (I can see faceting on the other side of the circle pocket as well)

What tool did you use to create the file? Did you try previewing the STL? Did you try exporting the G-code and looking at the preview?

Can you share your stl? What program did you use to create the stl? Since stl’s are a list of polygons, software that create them have to approximate curves. For example, OpenSCAD you can choose the number of facets.

1 Like

I always use .0001" as the calculation tolerance on actual tool paths. The tolerance isn’t a physical number, but more of an internal “fineness” parameter. I’ll back off to .01" for just checking tool path strategies.

That said, the faceting does look to me like it is in the STL. When I export an STL from Solidworks I use pretty fine export settings–never coarser than 1.5 degrees and .0005 deviation–.0001" on small and/or detailed models.

MeshCAM will faithfully follow even small facets in the model, especially in waterline cuts.


I think the faceting you are seeing on the other side of the pocket may be a reflection of the faceting on the opposite side and the tool marks on bottom of the pocket.

It’s hard to see from a photo, but the outer side of the pocket is pretty smooth.

I used emachineshop software to create the file.

here’s the wireframe preview: (meshacam ignored the notch at the due to my end-mill size)

I’ll try to attach my stl. Created with emachineshop. I may have to look at OpenSCAD.

When it is imported to meshcam, I would think that faceting like this would be visible in the model.

feeder_cup_final_nema14_bottom_plate_large_hole_2-05_inch_surface_only.stl (146.2 KB)

Is your name a ATHF reference?

You can draw this in Carbide Create in a few minutes, or am i missing something?

1 Like

I agree, the faceting doesn’t appear to be that stl, but the reason we’re all jumping to that conclusion is that it is regular in all directions so it’s gotta be that the gcode says to cut it that way. I don’t think you’d see some out of tolerance machine problem show up like that.

Perhaps share the gcode as well?

Though Apollo is right, if you just did cutout toolpaths in Carbide Create it’d also machine a lot faster since you don’re really need all of those 3d finishing passes even if you do want to surface the whole thing.

Yes! it is a ATHF reference! Not many people put that together, or even heard of ATHF.

I created my models over a year ago, and have been making many changes to the design. I never really considered doing it in CC, but my design has become much simpler with more parts. I’ll give that a shot and see how it turns out.

1 Like

OK, I’ve git it setup on Carbide Create. Is there a way to estimate the machining time like in MeshCam?

To estimate the machining time for a Carbide Create job:

The old schoola

Heres the feeds and speeds guide for the Nomad:

What feed rates did you use in meshcam?

I have a screenshot of the toolpath parameters in meshcam in the original post. I basically used the feeds and speeds in the chart for the Nomad, but I changed the feedrate to 6.

Depth per pass = .01"
Stepover = .05"
feed rate = 6"
plunge rate = 1"

Although waterline was checked, waterline passes were not made. (guessing they were most likely not necessary)

1 Like

OK, after running the job that was re-created in CarbideCreate, the faceting problem is gone (see pic below). The job ran almost twice as long, but I think the run time can be cut down by reducing the retract height. It was set to 15mm, and after every pass, it would retract 15mm above Z zero and would follow the plunge rate of 1" per minute which wasted quite a bit of run time.This doesn’t seem to be apparent in MeshCam.

The faceting issue can be addressed, however, the tolerance issue still remains. The circle was supposed to have a diameter of 2 inches. After the job completed, I get a measurement of 1.986". I uploaded my g-code below.

Anyone have any ideas about the tolerance?

feeder_cup_path.egc (295.1 KB)

How did you determine the endmill diameter?

My suggestion would be to cut a slot and measure it, or better still, cut a diamond-circle-square test and work out the runout from that.

That’s a good point. I’m using the #102 end mill from Carbide3D with the default properties.

I’ll take your advice and at least make a single line cut and take some measurements.

Which toolpath option did you choose? No offset, inner, or outer (not the exact terms used but I don’t have CC in front of me)? I know this is a very basic question, but it’s often useful to check for the simplest fix, and I’ve made that mistake myself.

I chose outside/right for the offset direction. The small pictures that preview the operation and the toolpath generated on the screen verified that this was the option I needed. Occam’s razor was what I’m looking for here. It is usually the simple things. I haven’t tested it yet, but I think @WillAdams is on to something with finding the actual the end mill diameter.

Another consideration is to check that there’s no excess runout from the endmill and collet not being seated properly ― did things click? If not, please review the instructions.

Yes, things clicked. Collet is seated.

Note: This isn’t my first attempt at machining with my Nomad. Had other models machined with no issues (not just the intro jobs that get you familiar with the machine).

With that said, other than the intro wrench job,this is the first time I’ve tried working with this particular end mill (#102).