Hello everybody,
I have a big problem since some days now. A calculation has ended on saturday afternoon and since this moment, i can't run another model. Each time i start my calculation, i have a warning message that tell me:
"The analysis has stopped because the solver has exited unexpectedly"
So, I have tried a lot of things like:
-reread the license file from LMTools
-stop, uninstall and restart the server
-delete the file the folder "Simulation CFD 2012" in C:\Users\username\AppData\Local\Autodesk\ to reconfigure the software
And finally I have reinstalled the software.
But nothing works....
Could you help me please?
for more details, when a look into :simulation\calculation computer, I find my work station but with 0 hearts.I think it's not normal....
n.
The core count showing as 0 is normal. For that dialog 0 means to use as many (2^n) physical cores as you have on your machine. If you wanted to throttle the cpu use down (for example on a dual-cpu machine where you have 8 real cores, if you wanted to only run on 4, you could specify that in the menu).
As far as the solver exiting, is it for any model? or just this one?
Have you tried rebooting your machine in case there is something in the Job Manager that is interfering with the new jobs coming in?
Apolo
I have the same problem with several large models. Sometimes it will restart but most times I have to restart at 0. I have done several hundred smaller simulation analysis and I do not have a problem. The larges ones do seem to be prone to this problem.
I have a dedicated WS W7 Professional, 24GB ram i7 processor 3TB didicated hard disk etc. Only Auodesk design and simulation programs installed.
Anybody else with similar problems?
Hi Herman,
I had this problem and even with 128Gb, it happened with a compressible analysis. But did you tried to start your analysis step by step? Is it a compressible analysis? I have fixed this issue in starting with constant gas properties and an incompressible condition for 10 iterations, then continuing 10 more with variable gas, then, X more iterations with compressible condition+gas variable.
Hope it helps
I do not have compressible gas but I do have variable gas and variable solids as this is a high temperature heat transfer simulation.
I am now running an extended W7 memory diagnostic as I am out of town for a few days. I will see tonight if there is any problem If not I will follow your advice and see if it can get to a couple of hundred iterations with no problem.
The reason I am honing in on hardware issues is because Todd at support ran the same simulation on his machine with several boundary conditions and has no problem.
Herman
Hello, I also uses CFD for thermal simulations of large models. I recieve the problem with the closing solver constant. Today it happens again and i found an possibly intresting behavior. The PC I use has 16GB of Ram wich became used for 100% for a few seconds during the volume meshing. Then the solver closes unexpectedly.
I hope this might be helpful solving this.
Martin
Issues like this are typically best handled through support cases unless you are willing to share your support share file on the forums, but that is up to you and your company!
If we resolve the issue we will try to post a solution to the forum to cover issue and resolution as best as possible. If the issue started in the forum and moved into a case.
Thanks,
Royce
Hi I have the same problem, and since my study is of personal use I can send you the support files.
I use my personal PC W7 64bit I5 4MB for this analysis.
I am also having this exact same problem with a Windows Server 2008 OS. I have i7 with 32 GB RAM. Number of iterations until failure varies, but the failure itself is very repeatable.
As the models get larger, the number of iterations until failure is less.
What can be done? Our users are very frustrated with this.
After all types of simple models I create from or import into Fusion close unexpectedly during "solve", I have used Fusion example models as a baseline with success. Still, I have no idea what is incompatible because Fusion example models work and the same parameters with models .
You cannot solve on mapped network drives, for example, B:\ drive mapped to \\company\server\path\folder. However, if you open the design study using the unc path, \\company\server\path\folder, you should be able to solve and get results. We have a bug logged for the mapped network drive issue. Did you try solving on a UNC path?
Herman, my colleague faced solver related similar behaviour for HiTech CFD simulation and analysis. However for smaller CFD simulation and analysis it works fine.
Can't find what you're looking for? Ask the community or share your knowledge.