3 axis: Toolpath in Fusion is perfect but machining does not match the part

[floydYPY3E]

We have a few programs that we used a couple months ago that ran perfect back then, but when I try to run (repost process) the same program it seems that the machining is over/undercutting all of the parts. I had a feeling it was a fusion update issue, maybe fusion updated but the individual toolpaths did not receive the same update no matter if I regenerate the pathing. For example we bore a lot of holes, we use cutter compensation to let the machines probe tell the program what the tool size actually is. This should cut a perfect hole every time, but they always come out undersized if the program is outside of the recent update time period. We have found a solution that deleting the tool path and recreating the same exact thing tends to fix this issue, would protecting the toolpath be a good solution to prevent this issue and minimize the need for recreating toolpaths?

[seth.madore]

Are you saying that the toolpaths change when you recalculate these old files? If so, could you open an old file, protect the toolpaths and then share the file here?
File > Export > Save to local folder, return to thread and attach the file in your reply

[floydYPY3E]

Yes, when we try to recalculate the old files on the new version of fusion they tend to have opposite selections for some toolpaths. I made mention about protecting the paths, but got some push back about how the paths will not update if the model is updated so just need a bit more evidence that it will work better. The file that I attached had the inverse selection on one of the 2DPocket toolpaths, but it was not programed that way before the new update. Just looking for a solution to keep those updates from effecting older programs.

[jsormaz]

@seth.madore 

I am having the same issue, except my files are not even that old.

I created a 2D contour toolpath a couple days ago, then updated fusion to 2.0.13162. Now the toolpath still shows as "green", however if I force regenerate, I get a different result. The direction of the curve has changed.

I cleaned up a lot of the attached file so that only the affected toolpath remains.

I'm pretty terrified now. Did I just get lucky catching this one?