Crash when editing a certain sketch & a corupted file

[PinRudolf]

Hi,

I seem to have 2 documents which can completely crash Fusion 360. I have attached the files for review by the development team. Be warned that they will crash Fusion!

One document (attached as Design1) crashes when I make a change in the file. (It does not seem to mather what kind of change as far as I have tested, otherwise change dimensions in Sketch2)

The other seems to be corrupted and makes Fusion crash just by opening it. (So be warned!) I could still download the file through Fusion team but that did not make a change. When I start Fusion this document shows in the data panel with the rotating bar, i cannot open an old revision. 

This one was being worked on simultaniously with Design1. It could be that is was saving when design 1 crashed. It could also be related to the same issue as part of Design2 is a copy of Design1.

Really i'm just reporting a bug here. I do not expect a solution to this. As these are simple models I have not lost much time. 

[ryan.bales]

I'm seeing the same behavior. Does this happen in all versions? 

[PinRudolf]

Hi Ryan,

The component that i can still open only has one revision. The other one is on it's second revision but since Fusion is loading that file continuously I am not able to select a previous revision.

[PinRudolf]

ah no, that's not right. I'm on a new machine and i was able to open the first revision of the second component. Unfortunately that first revision was a copy of the first component which has the issue that I cannot change anything.

[PinRudolf]

Come to think of it... that is unexpected. Since I create a copy of component1 (using Save As) and then changed it. Now when I open that revision I cannot make the same changes that I have made before to get to Revision2. (As it crashes fusion)

[ryan.bales]

I've not found anything about it that is obviously bad or would cause it to crash. I can even flip back through the timeline without issue on the file that opens. So unless we've found some truly bizarre behavior i'm not sure what is going on.

[PinRudolf]

Hi Bryan,

I have drawn the components again and ran into the same problem. (I did remodel them exactly as I had modeled them before) Again, Fusion crashes when editing the models but now I can at least open both files.

Thinking it might be a combination of features and setting; I did  try to make the exact same model again and check the stability with every step. However, this time it did stay stable and I have not found when the error start appearing.

Would it help if I add you to the project so you can check the files?

[ryan.bales]

I redrew the component that works to open as attached and it does not crash. I do notice the most common crashes occur when touching the chamfer on the design 2 you sent along.

[PinRudolf]

Hi Ryan,

When editing your document I received the same crash again. Specifically when editing parameter d7 to 12mm in the second Sketch. 

I restarted Fusion and did try to make the same change to your model. This time I first changed the array to 3 items, then changed parameter d7 to 15mm: Fusion did not crash but the features did not update to the new sketch parameters either. I've attached the file. The model seems to be stable but every time I make a change I will need to update every feature by opening them and accepting them again.

[PinRudolf]

Attached file...

[ryan.bales · Accepted solution]

I see what you mean, and while that does seem to be an issue as we get a crash. We are editing parameters that exist in time way after they are made. You should be able to get this to work by rolling back the timeline prior to the chamfer - then assigning the rectangular pattern spacing to -d7 (or whatever that parameter is). Then it should allow you to compute all and then roll the timeline forward.

I did not get any subsequent crashes with that method. 

[ryan.bales]

I went ahead and marked my post as a solution for the workaround. I did also report the issue to the dev team in regards to the parametric computing forcing the crash when adding parameters after its been modeled. 

[PinRudolf]

Yes sorry, I should have done that! It seems to be the best option. Thanks Bryan!