AnsweredAssumed Answered

All simulation tabs disappeared (6Gb of simulation data made useless)

Question asked by Umberto Zanola on May 22, 2017
Latest reply on Jun 22, 2017 by Umberto Zanola

SW 2016 SP5 with simulation professional here

I was running simulation for a while and I hit some very frustrating bugs in the past. At the moment costumer portal is down for maintenance and I cannot check if there is a SPR for this kind of behavior or not. My VAR hotline is not open yet so I post it here in the hope someone has a workaround for this.


I was running a lot of simulation lately and I noticed that when I surpass a certain number of tabs the program start to act in a strange manner and became sort of unstable, crashing a lot or losing CWR files. On two different machines (identical hardware HP Z230 Win7 certified drivers on both only ram differs 16/32Gb, latest win updated).


This time I was running more of a dozen thermal simulation on electronics and I had to rename a couple, I have all the CWR files on a local disk and temp file from simulations as well. When I load the assembly file I can see the parts files loading in the status bar from the correct path, but I can see only the model and the motion no simulation existed. And this was all of a sudden!

Simulation plugin is loaded, but if I try to unload it a popup asks me if I want to save simulation data.

when I reload the plugin I get "an invalid argument was encountered"


The folder looks fine and it was cloned with (the bug ridden and almost useless) pack and go from another folder that does not exhibit the same problem.

I should add that at some point the property manager started to act strangely... sort of no refreshing the UI. In fact  run into a similar problem in 2015 when you rename the study from the simulation tree instead of clicking on the tabs directly (and it allows to use "invalid" characters for the study name as well...that speaks volumes on the state of UI). When this kind of bug happens it seems the current tab is not the one visualized as selected but another one so if you rename a study tab,  you end up to rename another one.

Anyway when I run in a similar behavior in the past (it was mechanical simulations on the same machines) I didn't lose my data like this time.


And no I dindn't delete all the tabs myself as the 15 CWRs are still on disk with some HTQ file with the names of my studies in place.





Obviously my VAR asked if I deleted the tabs myself. The problem is that when you delete them the simulation files in the folder are gone instantly and this is not the case as even the temporary files are still there. Also the path was correct as I check it every time before starting a new set of simulations to prevent accidental overwriting in other folders.


1 WEEK of work trashed and I lost my face to the boss.



I found proof that something went very wrong.

If I make a new empty study in the corrupted assembly and open the batch interface. In the window I can see the new empty study I just created and another "ghost" study that has no tab in the SW window and the name (000A) correspond to the first study of my lost I think *some* data is still present in the assembly file.

My VAR told that I have to remake everything from scratch and I think it would be worth to analyze my files as I shared my findings with them as  well.


One entire day is passed and no news from my VAR (they told me they would get in touch with dassault)