Running the analysis, the monitor pop up informs that the job failed. The problem is that I don’t know how to detect the root cause, so I don’t know how to fix it.
The goal of the calculation I’m trying to do is check the behavior of a welded structure (Mat S235) in a static step. To do it, I have prepared the model starting from a step file and merging the part in only 1, as bc some surfaces of the part is fixed, and a surface traction is applied in 1 surface.
This message usually means that there’s a mistake in the model definition. It can be also a problem with mesh too dense for the solver to handle. You can try with Pardiso to verify this.
It would be best if you could share the model here or at least attach some screenshots.
Sure, here you have a screenshot of the monitor. Regarding the model, I’m not be able to share here, is it possible to do it by private mail?
Hi jcontreras,
Whenever i come across such problems i use the calculix launcher to see further information why the simulation crashes.
You can download it here :
http://www.calculixforwin.com
When you extracted the file, copy it to your C: Drive
Then also copy your input file there and run it with the launcher.
Best regards
I was rather thinking about the screenshots of the model with BC and load symbols and other settings.
If the file is too large to attach here, maybe you could share a link to Google Drive or Dropbox for example.
I would be interested in what additional information can be obtained in using the Calculix launcehr in comparison with PrePoMax. Maybe I can also include it in the PrePoMax.
Unfortunately i cant share my model but:
I found that with the calculix launcher the output tells me exactly what the error is while Prepomax doesnt show this error.
For example i was able to get this error while i was not able to achive the same error in Prepomax:
Maybe the output in Prepomax is somehow filtered ?
Sorry, the issue comes from I can’t show the design directly on the net. This is why I ask to send it by pivate mail.
All right, you can contact me via private message and I will try to provide further help.
I received and checked your model. I will describe the source of the problem here to let other users know.
Your model is simple and correctly defined in terms of analysis features. However, it’s a thin-walled part with quite complex geometry. Solving it with solid elements is very inefficient and may result in errors - as I said in the first reply, Spooles may fail to handle this mesh. I used PaStiX and it solved successfully - I got the results. However, it would be highly advisable to use shell elements in this case. You need at least 3-4 solid elements in the thickness direction to obtain good results. Such a tetrahedral mesh would be huge. Shell elements will let you overcome this problem. If it happens to be too difficult to prepare this geometry for shell meshing (it has to be done in CAD software), try with PaStiX solver but be very careful when it comes to the accuracy of the results.
I uploaded a convergence study for shell elements to my example collection:
It turns out that S4, S6, S8 and if sufficiently constrained, also S8R are ok, whereas S3 and S4R can’t be recommended.
I don’t expect that single layer tet10 meshes are really that worse than S6 meshes. But I haven’t checked that.
Interesting, PrePoMax runs CalculiX with a simple command like: ccx_static.exe Analysis-1
So I will check if some additional switch can be used to increase the number of output information. Or is it possible to see how the CalculiX launcher runs CalculiX?
