GCode for rotary axis

[greg9504]

Hello,

For a rotary axis setup in degrees with a radius (no inverse time mode) can I do gcode that looks like this:

G20
G40
G17 (arc in X Y plane)
G90

G00 X6.000 Y3.00 B90
G03 X7.000 Y4.000 I0.000 J1.000 F50.0 B0 (it's this line that I'm not sure about)

Where B is the rotary axis setup in degrees?

I'm modifying a post processor to put moves in for a pseudo slave axis (the rotary axis in question). The axis does not move the machine or the material, that's why I say it is a pseudo slave. The axis moves guide wheels about the cutting wire, so it is always normal to the cutting direction. This is easy to do for non arc G01 moves, as I just rotate the axis for the angle of vector of the XY. But for arc moves I need the axis to track the arc. I'm just not sure if the G03 line above would result in coordinated movement of the XY and B axis, so that in the above case the B axis would reach 0 degrees at the same time the arc finishes at 7, 4.

Thanks
Greg

on edit, here's a youtube video of a commercial saw that has the feature I'm trying to emulate.
https://youtu.be/n_sBuiirbR4
the axis in question is labeled C1 and C2. It's the one that rotates about the wire.

[TomKerekes]

Hi Greg,

Yes that should work. All XY and angular motion should complete at the same time.

But I don't think you want there to be any radius (specify 0). That way the motion and feedrate will be all based on the XY motion as if there was no angular motion (unless the angular motion is limiting).

Another approach sometimes used for tangential knives is that a KFLOP program monitors the direction of motion and automatically controls the angular axis. This can be problematic if one needs to lift the knife, move somewhere turn the knife to some angle, lower the knife and then go in that direction.

It not clear what would be best in your case regarding reversing. Should the angle quickly reverse or hold the original angle?

The C axis is normally used for angular rotations about the Z axis (XY Plane).

Here is a Video

[greg9504]

Hi Tom,

Thanks for the video!
Nice to see that would work. I had thought that if I had to I could program the axis directly on the kflop, but figured with doing all the 'programming' in the post processor the worst that will happen is the guide wheel is not in the ideal position. Doing it on kflop I figured there's more that could go wrong

WRT to reversing, do you mean if doing a move like:

G00 X0 Y0
G01 X1 Y0
G01 X0 Y0
?
I don't think it would it would matter too much either way. If the machine is moving back through were it already cut the guide position isn't too critical as there will be no resistance on the wire. There's actually two wheels, so if the position is 180 degrees out the wire will still touch a wheel. Right now the wheels don't turn at all and it still cuts fine. The liners just wear faster.

I'll make sure to use the C axis with no radius.
Thanks!

[TomKerekes]

Hi Greg,

WRT to reversing, do you mean if doing a move like:

G00 X0 Y0
G01 X1 Y0
G01 X0 Y0
?

No, I was referring to reverse feedrate. It seems that should work fine.

Good luck