QTPlasmaC with fourth axis - rotary

So the consensus is that we can do it with QTPlasmaC using Sheetcam. I just need to understand how we deal with the fourth axis in QTPlasmac.

snowgoer540 wrote:

mariusl wrote:

I have a license for the rotary plugin of Sheetcam. I could do a sample if you tell what you want to look at?


I was just curious what that particular table running mach3 required to get there. There’s a few ways, I’m sure.


I go back to wondering how other tables do this successfully.

That machine makes use of the plugin in Sheetcam. I don't think the THC is used or required at all for tube cutting.
I take it if we switch the THC off in PlasmaC we could do it.

My understanding is that Sheetcam runs a wrapped rotary. In other words, the X or Y axis is wrapped around the tube. So it's not really a C axis. You You could achieve this manually by changing the ini file stepscale based on the diameter of the tube and the steps per revolution.
To produce a wrapped rotary job without Sheetcam it's super simple in any CAD system that supports sheetmetal.
Just design your tubular part and then do an extruded cut 0.01mm wide for the full length of the tube. Then flatten your part!
Voila, an unwrapped rotary.

Round tube cutters do not need a THC (or even a Z axis under CNC control) because the torch only needs to be set at a fixed height.
Square and rectangular tube rely on the THC to lift the torch as it rotates. It would not be terribly difficult to calculate the circumference of a square tube with a radius on the corner to treat it as a wrapped rotary. Then bonus marks if you can calculate the rotary feed rate to account for the corners! 

While SheetCam does rely on an unwrapped drawing of the tube it internally wraps that drawing around the tube geometry to generate the code and calculate the rotary angle based on diameter. For square/rectangular it calculates the Z height depending on the tube angle. Relying on THC to handle square tube does not work very well. When cutting on the flats of the tube SheetCam uses X,Y moves.
It also has your bonus points for calculating the feed rates for the combined X,Y,Z,A moves. For small corner radius tubes the machine needs to move very fast to maintain the surface feed. You can see that in the video Marius linked. I generally recommend designing the rotary axis to handle at least 200RPM and X,Y axes at least 10m/min (400IPM).

Does anyone know how QTPlasmaC handles referencing when keep-z-motion = 1?

All you say is correct Rob. It is how Sheetcam does it. If you watch that video the guy explains it in some detail how to determine the square tube circumference and the steps. Only thing is that Sheetcam actually use the Z axis on square tubes so far I understand it.

Sorry I didn't see Les answered above

Well we now have it from the horses mouth. Thanks Les!

The torch will not hit if the Z axis is driven by Gcode like the Sheetcam plugin does.