RTCP setup for 5axis A/B table/table horizontal mill

[daredevil]

Thanks Tom,
so the rule for the rotary axis is the same as the linear ones: the right point of view is the "pen" so the end mill?

Now my mill is configured as this image:

so according to the "pen" rule I have the rotary axis with opposite direction, effectively in the viewer the linear directions are ok, while the mill graphics moves in the opposite linear directions vs my rotary table.
I had also to change the angular signs (a,b,c) in the kinematics C code (TABLE-TABLE AC & TABLE-TABLE AB), so I believe I should rechange them, in order to be in line with kmotioncnc viewer, correct?

Thanks
all my best
dd

[TomKerekes]

Hi dd,

Sorry I don't completly follow. Are the xyz directions the CAD directions? Or the directions the Table moves? But its probably simpler to try it. I think changing the signs of the angles in the kinematics will mess it up unless you also change the direction the axis moves.

[daredevil]

Hi Tom
the X Y Z directions in the image are the cad ones (the end mill direction), according to right hand rule
while the A & B directions are the table rotation ones (which I believe I need to change)
thanks
best regards
dd

[daredevil]

As you can see from this picture taken from infosys.beckoff.com, I found also this AB horizontal kinematics scheme (the same of Pocket NC):

which I follow for my A & B rotary table, but it does not follow the pen rule for rotary axes,
finally the piece is done but the kmotioncnc viewer moves the A/B mill following the pen rule, correct?

just to understand which are the "correct" rotary table A/B sign directions,

thanks
best regards
dd

[TomKerekes]

I found this link which agrees with the "pen" method. Says use left-hand-rule for Tables

This shows the GViewer with A at 0 and +20degrees moving as right-hand rule movement:

0 Degrees

ZerodegA.png (11.36 KiB) Viewed 1799 times

+20deg

Plus20degA.png (11.33 KiB) Viewed 1799 times

[daredevil]

Thank you Tom,
dd

[daredevil]

Hi Tom,
coming back to the GViewer simulation in kmotioncnc for 5axis RTCP ngc output by the cam sw,
this is my experience:
every 5 axis rtcp milling cycle gcode loaded is always plotted in the same way (CAM correct way) with or without the G43.4 Hxx command,
(as you confirmed the viewer does not know the kinematics loaded, which is driven by the gcodeinterpreter code)

I believe this is due to the fact that the viewer is always in "rtcp mode" cause every angular rotation is made using the tool tip as rotation center,
this could be the reason why the g43.4 is irrelevant in the viewer, is it correct?

If it is correct, does it exist a mode to simulate with Kflop the RTCP gcodes difference with or without g43.4?

do you know/recommend other sw simulators that can do this?
do you know other posts/forums where they discusse about RTCP milling with Kflop?
the G68.2 code management will be available in future Kmotion SW release?

thanks again
all my best
dd

[TomKerekes]

Hi dd,

I believe this is due to the fact that the viewer is always in "rtcp mode" cause every angular rotation is made using the tool tip as rotation center,
this could be the reason why the g43.4 is irrelevant in the viewer, is it correct?

If it is correct, does it exist a mode to simulate with Kflop the RTCP gcodes difference with or without g43.4?

Yes the assumption is that the GCode describes the XYZ positions of the Tool and the angles of the Tool about the Tool Tip. So the Viewer simply plots the specified GCode path regardless. Currently the GCode Viewer considers nothing about the Geometry of the machine or the Kinematics.

I suppose we could add the capability where when we know the kinematics, and RTCP is turned off, and some angular change is made to compute how the xyz tool tip path would actually move and plot it. Its not clear to me why anyone would do this or want this. Maybe to visualize a crash if such a motion is made?

do you know/recommend other sw simulators that can do this?

No I don't. For a simulator to do this it would need to somehow be told about the geometry of your machine.

do you know other posts/forums where they discusse about RTCP milling with Kflop?

Not specifically. If you Google "dynomotion RTCP" there are many hits.

the G68.2 code management will be available in future Kmotion SW release?

Not at this time. Do you have an example of how you would use it?

[daredevil]

Thanks Tom

Maybe to visualize a crash if such a motion is made?

yes it is for this reason
please take a look to this italian company site, they produce similar controllers with included an advanced simulator where you can build your machine graphics according to the kinematics in order to check possible collisions during the rtcp path (which is impossible to view analysing just the gcode as you can do with 3 axis cycles):
https://www.rosettacnc.com/it/download-cnc.html

G68.2 MANAGEMENT Not at this time. Do you have an example of how you would use it?

it could be useful to produce 3+2 indexed 5 axis cycles in RTCP
please take a look to this link:
https://www.linkedin.com/pulse/fanuc-g6 ... -markoski/

thanks again
all my best
dd

[daredevil]

Hi Tom,
I have some issues with 5 axis continous gcodes, in terms of sustainable feed rate:
- please find attached a simple square inclined 5 axis swarf profile gcode, in the path corners the feed rate slows down dramatically depending on the mm/min feed rate indicated in line #8 of the gcode, to a 4% of the requested feed (i.e. from 500 to 20, or from 1000 to 40 and so on).

please note that I need to put F10.000 to achieve around F380 (always mm/min) in the corners when the rotary axes change more their angles, and the mill motors/axis are ok with this FR (around 400, I have set maximum FR for linear axis around 1700 mm/min).

I run Kmotion 435D sw, please find also attached my INIT C program and Toolsetup kmotion cnc screen.

I kindly ask you how can I obtaine a sustenaible feed rate for this kind of 5 axis operations (my goal could be to have a feedrate around 500-800 mm/min), I believe I need to change the axis setup values in terms of acceleration/jerk? (the velocity seems to be quite coordinated between linear and rotary axis) or what else may I try?

Waiting for your indications,
thanks in advance
dd