2d Contour Not following Tool Path Around Radius

[eric_konchan]

Hello,

I am having an issue with 2d contours not completely following a chain. The endmill wants to leave the radius early then angles back in along the next wall until its almost to the next radius. I've had this issue in the past and have adjusted both minimum radius and compensation radius to 0, still getting issues. In the past I ended up just switching to a smaller endmill, but I would like to figure out the cause. I did some testing and re-designed this same feature on Fusion. (All the models I deal with typically are imported step files). After doing so I have zero issues with the model I designed in fusion and still keep getting issues with the imported step file. Everything looks the same I don't see why this would cause such issues. The toolpaths are identical, both loaded from the same template. I attached both files, if somebody with more experience could take a look and see maybe why this is happening.

[programming2C78B]

I dont see an issue in either program?

[eric_konchan]

In the "From Step" program it's not wrapping the radius correctly while simulating. It is doing it correct in the "Fusion" program. I attached a couple screenshots. Is it not doing this for you?

[programming2C78B]

no, what you wrote does not explain what I was thinking of.

It has something to with you using In Control. If you change it, it works as expected.

[eric_konchan]

Yeah, I found that changing from In control to either wear or In computer fixed it. That's just not how our machines out setup though. This isn't the first time seeing it. Might be easier to just switch to using In computer or Wear.

Also to note, it has something to do with the work zero. I changed the zero to the center of the window and it worked fine after. I'll have to do some digging and see.

I appreciate you looking in to it.

[programming2C78B]

well, it sounds like a bug and should be reported. I only started using In Control this year because our new machine requires it but havent ran into this issue yet (its all mill turn)

[seth.madore]

I'm not sure I entirely follow. Your images, especially the one of CAM Simulation, don't align with each other. The toolpath I see in the image appears to be the first toolpath, but your position marker in the timeline suggests it's the second toolpath.

Which toolpath is this happening with?

Another note; the toolpath line you see in CAM will be different from what you see in CAM Simulation. The former is showing the edge of the tool, the latter compensates the toolpath to account for tool radius (it places the path in the center of the tool).

Additionally, you're running a .03125" Compensation Radius Allowance, which will also effect how the toolpath enters into corner radii. In the below pic, I duplicated your first toolpath and then adjusted the second one to have zero comp allowance. Notice how the second toolpath is taking more in the corner?

Despite that, it's still going to produce a sharp directional change when it comes time for Simulation, simply due to the stock to leave, cutter diameter (radius) and whatnot.

 

[eric_konchan]

All three of the toolpaths are being effected. However I started moving the work coordinate system zero position to multiple places (on both programs) and I am seeing the simulations differ each time. It works perfectly if I put the zero on the center of the window. I'm trying to compare code to see if this is actually going to give me bad code or if it is just a simulation error. Also the .031" comp shouldn't be an issue on the rough/finish toolpath. Only backs it off a little on the conventional path.

[programming2C78B]

Seth, look at my circled areas. The tool is cutting the straight walls on an "angle" and leaving a wedged-shape material. If you change it to In Computer, the wedge is milled away and the entire wall will be green - i made it cut 1/2 the depth to make it easier to see. 

[seth.madore]

Hey, look at that, now I understand!

[eric_konchan]

I looked over the posted G code, made a sketch with all the points the endmill followed, and the results were that it matches the model.

Is simulation error common? Doesn't exactly leave me with a great feeling any time I see this.

[seth.madore]

There are some edge cases where In-Control doesn't produce the correct visual result, I believe this is one of those cases. I'll get a ticket logged to investigate this (if one's not already created for this issue

[eric_konchan]

Appreciate it!

Thanks for the help as well @programming2C78B