Surface milling radial surfaces always results in facets

[jgulso]

I have an on going issue with surface milling.  Any "round" surface inevitably turns into a faceted surface.  This is without regard to toolpath strategy, tolerance, tool type, etc.  Even with the surface tolerance set to something extreme like 0.000001" these facets appear.  They are not simply rendering artifacts, if I post this code to the machine, this is how my parts will come out.

Attached are screenshots of the model, and the surface in question was created by drawing a profile sketch and revolving around the center axis, and a screen shot of the simulation machined part showing the facets.  Note also that any straight cylinders in this part were interpolated as well, with "side" features, and they come out round.

[bhorstTGM48]

Can you share your .fm file?

[jgulso]

Sure.  This is 2017.

[AJHanson]

Try 3D milling isoline and turn  on arc line prox. Hope this helps if it does please accept for answer.

[jgulso]

No joy

[AJHanson]

Would this work for you? I created my own surface to mach.

[jgulso]

That calculates to the same faceted surface when I run it?

[bhorstTGM48]

https://www.screencast.com/t/THvWPxkk0E52

This is the way the simulation looks on my screen from AJ's example of isoline. This appearance has nothing to do with the G code or the end result at the machine. If you look at the code from AJ's file it is mostly arc segments instead of all the little XYZ moves. 

[AJHanson · Accepted solution]

Yes that's correct i used arc/line approx. it will turn those segmented lines into arcs.  I also have found out that sometimes using my own surfaces work better than the solid given to me. I just have to make sure both surface and solid are relatively close (+- .001).

[jgulso]

Oh, that's interesting.  Yes I do see the difference in the NC code now.

I still don't understand why none of the strategies produce arc segments instead of linear moves.  Or how he recreated the surface so that it would.

Obviously this particular part would be better turned in a TC but for peripheral reasons like the tooling and time on machines available I'm milling them.  The facets don't even effect the function of the part, it's just an annoyance that's been bugging me since I first encountered it.

I appreciate the help.

[bhorstTGM48]

https://www.screencast.com/t/A43mSqkk

If you want it to look clearer on the screen play around with these. Just be-aware it will slow down the 3D simulation. 

[jgulso]

Thanks.

Ultimately it's just aggravating that (and Fusion 360 does this too apparently, I just tried a few strategies with the original model) a revolved surface gets interpreted as linear moves instead of arc moves without jumping through hoops.

I'll mark it resolved.

[bhorstTGM48]

AJ, I would like to know how you got those 2 separate faces into one nice trimmed up surface?

[AJHanson]

bhorstTGM48,

I took the two faces 43&52 made a curve using isolines turned them into geo. . Then i trimmed lines accordingly top and bottom also trimmed were rad met angled ling. Created one curve by selecting the trimmed rad with the trimmed angle line and joined them together using join curve function. Created a vertical line in center of ID and revolved it 360. I hope this makes sense.