weird surface finish issues

so i've been having a hell of a time fighting some surface finish issues. tried all kinds of things, varied rpm/feed, amount of material removed etc.



finally a buddy suggested i chart the axis motion in camplete. lo and behold, what does C axis motion look like EXACTLY in the spots the finish is bad?

i had my tolerance cranked up to .00025 on this one, max segment distance of .005 and max angle increment of .25 degrees


loosened up the tolerance to .0015" and looks much better!



increased max segment distance to .025 and angle increment of 1 degree.

yet to run it with these settings but i think it should look much better! will give it a go tomorrow, hopefully try to smooth out the steep ramps while keeping the crazy squigglies away.

camplete and hypermill are some badass tools!

What's the machine you're using? Have you tried the "barrel shape" cutters to lower toolpath requirements? Hypermill is supposed to support them.

Milland said:

What's the machine you're using? Have you tried the "barrel shape" cutters to lower toolpath requirements? Hypermill is supposed to support them.

Click to expand...

matsuura LX160
look closer at the step over on the blades

empwoer said:

matsuura LX160
look closer at the step over on the blades

Click to expand...

Unfortunately the pic is teeny-tiny, and/or my eyes are shot. But from context I take it that's what you're doing now?

Nice machine, by the way - swap my ancient VF-5 for it even-up?

Milland said:

Unfortunately the pic is teeny-tiny, and/or my eyes are shot. But from context I take it that's what you're doing now? yessir

Nice machine, by the way - swap my ancient VF-5 for it even-up?

Click to expand...

thanks! you drive a tough bargain, but i'll have to pass this time :P

It does seem to me that a machine like that and with Hypermill for the programming that you shouldn't have to lobotomize the constraints that much to get smooth-cutting code. Do you have support that you can contact to find out why the dang thing is jerking?

Milland said:

It does seem to me that a machine like that and with Hypermill for the programming that you shouldn't have to lobotomize the constraints that much to get smooth-cutting code. Do you have support that you can contact to find out why the dang thing is jerking?

Click to expand...

i think its a case of the machine being a bit too accurate? its pretty clear to me that with the tolerances getting cranked up, the tiny deviations in the surface made the code jittery and the machine replicates that too good?
i'll know for sure tomorrow after i run it with the smoother code.

If you can, please post some higher-res pics of the before and after, I'd like to see what looks different.

Aren't you also facing issues of compressor efficiency if you go too far from the ideal surface geometry? Smooth to look at doesn't always mean optimal...

Milland said:

If you can, please post some higher-res pics of the before and after, I'd like to see what looks different.

Aren't you also facing issues of compressor efficiency if you go too far from the ideal surface geometry? Smooth to look at doesn't always mean optimal...

Click to expand...

will do.
as with all things, looking for a healthy balance here!

Out of curiosity do you post the code out thru CamPlete or Hypermill?

empwoer said:

so i've been having a hell of a time fighting some surface finish issues. tried all kinds of things, varied rpm/feed, amount of material removed etc.

View attachment 282532

<snip for clarity>

camplete and hypermill are some badass tools!

Click to expand...

Good news is you can make really good replicas of some of Frank Gehry's buildings …

.




___________________________



I'm wondering settings for TCPC (Tool CenterPoint control / Tool poise control and possible conflict with 3d radius compensation (less so 3d comp.)).










TCPC on and off ^^^



^^^ Various differences


_______________________________________________

So I'm wondering with the Matsuura "Meister" layer (s) on the regular FANUC control that the previous owners of the machine may have the TCPC setting set to something completely incompatible with what you are doing... Not sure how Hypermill will be interpreting your tool geometries and TCPC settings + 3d tool (wear) compensation ?

___________

Should call it the Frank Gehry "effect". lol.


_____________________________

* If memory serves me right there are specific codes in certain combinations and orders for TCPC and various methods of resetting / clearing state at the beginning of a program and how that needs to be cleaned up at the end of a program or "abort sequence". Fanuc has a good video on that ** . Not sure that Hypermill would take care of all of that cleanly in a "No brainer" high level way versus whatever the control / machine state really is. Does Hypermill have macros that check what state the machine is in and then generate conditional code on the fly to "Make everything go lovely " ?

** YouTube

^^^ Shows codes and program structure esp. WPESC and ("High speed smooth" )- TCPC etc.

empwoer said:

so i've been having a hell of a time fighting some surface finish issues. tried all kinds of things, varied rpm/feed, amount of material removed etc.
<snip>

finally a buddy suggested i chart the axis motion in camplete. lo and behold, what does C axis motion look like EXACTLY in the spots the finish is bad?
View attachment 282531
i had my tolerance cranked up to .00025 on this one, max segment distance of .005 and max angle increment of .25 degrees

View attachment 282530
loosened up the tolerance to .0015" and looks much better!


View attachment 282529
increased max segment distance to .025 and angle increment of 1 degree.

yet to run it with these settings but i think it should look much better! will give it a go tomorrow, hopefully try to smooth out the steep ramps while keeping the crazy squigglies away.

camplete and hypermill are some badass tools!

Click to expand...

What is or does the "angle increment " effect geometrically in this case ?

Also on your LX it's a 160 ?

And is the other axis a B or an A ?

Do you have plots for that ?

Good luck … I'm sure you will figure it all out / thanks for sharing.

Ta.