QTPlasmaC with six axis

Hi,
Is it possible to run QTPlasmaC with six axis? XYZ as linear and ABC as rotary. Each rotary axis is rotating around its parent linear axis?
If not, what kind of effort will it take to achieve this?
 

Thanks Phill. That would be great. I will work on the kinematics and the post processor for fusion. If we can make it work we will have an very awesome product for plasma cutting. Bevel cutting of parts and holes as well as tubes.

Marius, I had this idea a few days ago for bevel cutting.
I have an abandoned project that reports among other things the arc radius and the tangential heading in real time
github.com/LinuxCNC/linuxcnc/pull/900
Once 2.9 is released I had hoped to find the time to revisit it.
Assuming the following  5 axis design below, The A axis will determine the bevel angle
There will be a XY plane at the pivot point that is a constant Z height regardless of the direction of the bevel.
The distance from the torch tip to the pivot point in X & Y then can easilly be calculated from the A axis Angle
So given we have a tangential heading, we know the direction of the pivot point will be normal to the heading.

So now we just need a component that calculates the required X,Y position for a bevel cut in real time.
This component could then use external offsets to position the torch in the X,Y plane. Phils is the expert there
So this methodology would allow bevel cutting to be applied to a standard 2D DXF file cut without the need for any 5 axis kinematics.
Let me know your thoughts.

phillc54 wrote:

I have pushed a new branch named phillc54/qtplasmac_6_axes which is based on master branch.
github.com/LinuxCNC/linuxcnc/tree/phillc54/qtplasmac_6_axes

The GUI's need a little bit of polish but should be fine for your testing.

Warning: Spoiler!

 

It did make me notice a couple of issues with the fifth axis...

 

Wow you are a champion man. I will install and test on Monday and I will give you some feedback. It will not be precise testing as we are still working on the kinematics for the 6 axis solution. One of these will be a tube so it's a bit more of a challenge. The bevel cutting on flat plate should already be possible with a kinematics model I just completed for a 17m x 4.5m x 1.8m 5 axis router.

rodw wrote:

Marius, I had this idea a few days ago for bevel cutting.
I have an abandoned project that reports among other things the arc radius and the tangential heading in real time
github.com/LinuxCNC/linuxcnc/pull/900
Once 2.9 is released I had hoped to find the time to revisit it.
Assuming the following  5 axis design below, The A axis will determine the bevel angle
There will be a XY plane at the pivot point that is a constant Z height regardless of the direction of the bevel.
The distance from the torch tip to the pivot point in X & Y then can easilly be calculated from the A axis Angle
So given we have a tangential heading, we know the direction of the pivot point will be normal to the heading.

So now we just need a component that calculates the required X,Y position for a bevel cut in real time.
This component could then use external offsets to position the torch in the X,Y plane. Phils is the expert there
So this methodology would allow bevel cutting to be applied to a standard 2D DXF file cut without the need for any 5 axis kinematics.
Let me know your thoughts.

 

Rod that should be possible without any extra modules.We already have 5 Axis kinematics that can be slightly modified ( or could work as is ) to make bevel cutting possible.
The bigger challenge will be the post processor to generate the Gcode. At this point in time our only avenue is Fusion or maybe Freecad although I don't know if Freecad has a working 5 Axis solution yet. Because QTPlamac takes care of the probing, it makes the CAM solution a bit easier. It's a one pass cut on the bevelled edge of your part to generate the Gcode. The trick is that the CAM must produce TCP paths. 

I hope to simulate some simple cutting process in the week. I will most likely hand code the Gcode to see how QTPlasmac handles it.