Official Luthiers Forum!

I've noticed this interesting artifact in a number of different pieces I've made on my CNC, and I'd like to get rid of it.






These are actually pretty subtle. I had to position the camera and the lighting to get them to show up in the photo. They sand-out, eventually.

Because the cutter is actually traversing back-and-forth at 15 degrees from the X-axis, I'm thinking it's some kind of strange artifact in the X-axis of my machine (ShopSabre 3636) - either the bearing ways, or the lead-screw. Or ??

Any and all theories, conjectures and thoughts on the issue greatly appreciated.

Thanks in advance,

Jeff

What packages are using to surface your model with and cut it?
It may be that the software doesn't do a true spline CAD or CAM, but uses a series of short line segments that approximate a spline within some predefined tolerance of deviation.
Sort of like you were cutting a circle with short line segments using G01 instead of interpolating it a G02.
Just a guess.

_________________
Jim Watts
http://jameswattsguitars.com

I agree with Jim on this one. I have seen the same thing using Rhino and Madcam and discovered I did not have my mesh tolerances set tight enough in Rhino. The same thing can happen in you machine from true stl triangle meshes if the mesh density is not heavy enough, only it will show up as triangles instead of what you show in your photos.

Check your cad mesh/model tolerances.

Mike

saw dust on the rails...

_________________
Ken

There may be a function in your software to:

1. Create arcs in lieu of line segments.
2. Reduce the minimum length of these line segments. You'll still get lines but they will sand out easier
3. Create arbitrary line segment lengths, which will eliminate all the points falling in a row.

If you have a toolpath editor you can go through and observe each line and endpoint to make sure this is happening in your toolpath. If you can read your g-code well then you can see it there too. This COULD be happening out at the controller.

I think in any case you're going to find that these lines sand out easier than the tool marks.

_________________
I read Emerson on the can. A foolish consistency is the hobgoblin of little minds...true...but a consistent reading of Emerson has its uses nevertheless.

StuMusic

That area looks like it has the smallest radius of that top surface ? I'm guessing it is a relatively small amount of material, a few thou at most that is being left uncut. It could be chordal tolerance in the cam settings causing it to square off instead of follow the radius. Does your cam simulator show the faceting ? I'd like to think it should. The cam I use shows all the tooling marks and has a function to display excess material not removed by the toolpath that still needs to be removed with a smaller or different tool or strategy. I have had a surface in Solidworks which looked good there and passed checks, which my cam software did not like the quality of. Going back into the model clearing the selections in the dialog boxes for boundary/loft etc and putting the same geometry back in made it recreate the surface better and everything was OK. If the meshing/curvature combs show something complex going on in that area, the geometry might be a little too busy. Like Stuart says, I'd look at the g-code for what it is trying to do in that area, maybe just do a test toolpath with North-South direction so it is easy to look at. Should just be short G01 moves that follow the surface radius. What CAM are you using ?

Thanks for all of your input on this. I appreciate it.
I’ve since had a chance to investigate this with the tech support folks at MecSoft, and it looks as if the problem had to do with my program settings - particularly with regards to "Part faceting tolerances".

Additionally, and maybe just as importantly, if I had set my display settings to "Flat shading", and not "Smooth shading", I might have caught the issue beforehand. Oh well, we live and learn (on good days!).

Here's a screen shot from VisualCADCAM with the faceting tolerances set at .002"



And here's a screen shot with the faceting set at .0001"



I haven't actually machined the new version of this part yet, but I suspect this change should clear things up.
Thanks again for your input.


Better living thru better surfaces!

Jeff

Man, I have to admit I had no idea what you were referring to when I first read this. For some reason, the Bob Dylan song "Blood on the Tracks" was the first thing that popped into my head, and I thought you were obliquely referring to some C&W song I'd never heard of.

However, the coffee kicked in, and I get you now.

My ShopSabre has sealed linear bearings, and I tend to be kind of anal about keeping 'em relatively clean... so I wouldn't tend to suspect this.

Thanks anyways!

Ok good. Mine are not sealed so it can get bumpy if i am not careful





Sent from my iPhone using Tapatalk
Ken

_________________
Ken