Fanuc/fanuc posprocessors: incorrect TWP Euler's angles for G68.2 (P0)

[Anonymous]

Hello everyone

I'd like to report a potentially dangerous error in the "Fanuc / fanuc" postprocessor TWP implementation.

In short: the tests on a Fanuc Robodrill (AC-rotary/swing table)  machine failed - the machine didn't turned the rotary axes as expected (fortunately no collision occurred).

My setup (A0. C0.):

The program starts with the straight op1 (executed successfully) and  tilted op2,  stopped with alarm, no rotary axes move observed.

Op1. straignt, Facing end of stock  Op2., tilted, Facing upper ridge

The output from the "Fanuc / fanuc" postprocessor I tested on the machine:

...
(FACING END OF STOCK)
G69
...

(FACING UPPER RIDGE)
G68.2 X0. Y0. Z0. I-90. J0. K-90.
G53.1
G49
G00 X22.672 Y-13.162
G43 Z37. H01
...

The alarm on the machine was generated by last visible line, G43 Z37.0 H01:   "Z limit exceeded "

It seems that the Euler's angles are calculated incorrectly. Namely, the above TWP command

G68.2 X0. Y0. Z0. I-90. J0. K-90.   generates coordinate system that is not tilted at all, but rotated around Z about -180deg (please correct me if I'm wrong - here is the Fanuc TWP instructory film ).

As far as I know stereometry, the correct values for my case would be : G68.2 X0. Y0. Z0. I-90. J-90. K0.  , or even simpler to understand and check option P3: XZ-vectors:

G68.2 P3 Q1 X0. Y0. Z0. I0.  J-1. K0.  ;new X axis is the old -Y axis
G68.2 P3 Q2             I1.  J0. K0.  ;new Z axis is the old X axis

And here is my kind request to the postprocessor developers: could you please correct me or the "Fanuc / fanuc" postprocessor, the wrong one 😉

Ps.

I tried the Fanuc /Robodrill postprocessor, but it doesn't output any TWP code and, just invalid direct moves (will report these errors in a separate post)

[LibertyMachine]

This doesn't help you with the G68 being incorrect, but does your machine accept A and C moves, such that you could use the Fanuc 4th axis post (modified to output the additional C axis)? Or does it only support Euler angles?

[Anonymous]

uff, that was a quick response!

1. G68.2 is not G68 ! They are completely  different commands
2. The machine (apparently) accepts G68.2, so all the P* option should work (Euler-P0 and RPY-P1 is confirmed physically)
3. I'd rather modify the G68.2 command in Fanuc / fanuc or Fanuc / Robodrill postprocessors than mess up with the direct output (A C values)   - the G68.2 is kinematics/dimension independent  and doesn't require the op points recalculation.

I could even modify the postprocessor myself, but can't see any documentation about it. Where should I look for any resources related to the postprocessors?

[LibertyMachine]

Here is the motherload of info you need:

https://forums.autodesk.com/t5/hsm-post-processor-forum/technical-faq/td-p/7473258

[Anonymous]

Cool, thanks for the links for the postprocessor developers!

I've checked the postprocessor and it seems the getEurler2 method is faulty or missuses (EULER_XZX_R parameter).

I'll try to investigate, but my priority s to produce the part on Friday, so first things first.

Could anyone confirm that my Euler's angle calculations are correct?

Is there anyone who used the Fanuc TWP?

[wgr_eng]

Hello 

I have same issue with the euler angles.

Is there a solution for this problem yet?

Thank you

[pbayldon]

Hey

Did you get anywhere with the "g68.2 p3".

i also want to use this as it is simple.

Phil