Dear sofistik developers.
On some large models (skyscrapers) we are experiencing significant difficulties in the mesh generating.
Sometimes program does not show that defective elements have been generated in the system. This is revealed only when the calculation is started. Bad elements have to be removed, since the calculation with zero stiffness of these elements is impossible. However, this operation removes all generated loads, which leads to the need to re-generate all loads. It is time, time and time.
Bad elements are mostly quad with concave corners, which could be replaced with two triangs.
The hope for the ctrl mesh 3 + 96 (mixed element shapes allowed) option did not materialize.
Triangs are generated everywhere.
Is it possible to implement diagnostics before generating the mesh so that no bad elements are generated. And fix the command sofimsha del, so that the loads do not disappear.
Kind regards,
Ilyin
Dear Mr Ilyin
I am not sure why you have to delete ‘bad elements’. Perhaps there are some flaws in the SOFiMSHC or SOFiMSHA code?
If you are experiencing these problems regularly, you should ask the support for further help.
I think the command ‘ctrl mesh 3+96’ is not implemented anymore. That’s why you don’t get a combined mesh.
The command ‘del’ (SOFiMSHA) deletes the element and the corresponding load. If all of the loads are deleted, you should insert the command ‘CTRL REST 2’.
Mit freundlichen Grüßen / with kind regards
SOFiSTiK Support Team
Frederik Höller
Dear Mr.Holler, thanks for reply
Defective elements generated in structural elements (sofiplus, sofimshc). This “pops up” on large models with many intersections. There are several (maximum 20 elements, it is very small percent), but ase with a such elements can not be run.
CTRL REST 2 - mean "deletes only if elements are deleted. I.e.it is our case. We are forced to delete elements, all generated loads disappeared with/without CTRL REST 2.
I have previously raised the issue on support.
Thanks, kindest regards,
Ilyin
Dear Mr Illyin
I looked up your problem and saw that you could solve the error just by updating the software.
So I advise to open the SOFiSTiK Application Manager and update all programs to the latest version.
Thanks for your understanding.
Mit freundlichen Grüßen / with kind regards
SOFiSTiK Support Team
Frederik Höller
Dear Mr.Holler!
Unfortunately, in the updated system process finished with the same defective elements and no faster.
Surprisingly, I have to mention sofiload here. As you can see, the program is still 32-bit. And for a task with 39 loads, mainly free area load, on a 3.5Ghz processor, we need to wait 3.5 hours for the result.
I wrote this on the forum with the hope that it might still be worthwhile to think about introducing preliminary (before starting the generation) diagnostics of the model.
Kindest regards,
Ilyin