Machine simulation fails on circular toolpath with polar mode

[elaspol]

Post: Haas Next Generation Post

Machine: Autodesk Generic 5-axis

Operation: 2D Contour

ManualNC: usePolarMode

Hi,

Simulation fails on circular toolpath. Other shapes seem to be simulated well.

Toolpath looks like this:

but simulated like straight lines:

  

[seth.madore]

What build of Fusion are you on, is this an Insider feature you're trying out? Or are you in regular Production trying to get something to work that isn't yet supported?

[seth.madore]

Oh, and your machine/post selection is interesting. You've got a generic machine selected, but you're using the Haas NGC post. We do have a large library of Haas machines, so what's your setup here?

[a.laasW8M6T]

I believe the issue lies with the way machine simulation interpolates rotary moves as straight lines, because the 2D contour there has just two rotary commands(in Gcode) to do the entire circle the tool moves in a straight line between points rather than properly rotating.

I thought this was only a problem with non TCP machine simulations as Brek explained to me.

The reason the Haas post is being used as its the only one that I know of that supports the "USEPOLARMODE" action

[brekmiller]

Seth, does polar mode turn off TCP?  If so, that might explain it.

In the latest development version of Fusion we have implemented non-TCP interpolation for simulation.  We should check to make sure that handles this case correctly.  (Users will get this in an upcoming release; though I think not the next release.)

[elaspol]

We are not part of the insider program.

Version: 2.0.18460 x86_64

I'm trying to implement PolarMode in millplus post and I got this result, so I went back a few steps to check if I get the same on the generic machine with NGC post.

[a.laasW8M6T]

Hi

Just looking at this again

Because the Z axis is vertical  the post doesn't need to use TCP

Currently Machine Simulation doesn't correctly simulate certain Toolpaths properly when not using TCP(EDIT: I should clarify it doesn't simulate toolpaths containing long rotary moves correctly, short rotary/linear moves are approximated OK)

The output code will still be correct though, and should run just fine on the machine

(I know this from experience posting code to my 4th axis using Heidenhain iTNC530)

[brekmiller]

I have opened CAM-52109 for this.

Our internal development version of Fusion can handle non-TCP interpolation.  However, this example still doesn't work.  When the CPS file sees usePolarMode, it turns off TCP.  But it doesn't tell simulation that it has turned off TCP.  CAM-52109 is for fixing usePolarMode.