It might be that bug with shells again. I’ll go back to this and report on the CalculiX forum soon. Meanwhile, you could try with the same model but using solid elements instead of shells.
Thanks for the advice,
as i am working with composites, I frequently need the *ELASTIC,TYPE=ENGINEERING CONSTANTS keyword. And that is much more convinient to use on shell elements because of their local coordinate system, where the local z-axis is alway perpendicular to the shell and the y-axis lies within the shell.
My current workaround is to alway only solve one step at a time and deactivate the others. A bit annoying but it works for now i think.
You could also try using *Step, perturbation (has to be added manually at the moment). This way the steps will be independent.
1 Like
Thank you for your reply!
I was sick the last few days so I could not answer. I am not used to manually adding keywords to the input deck. Can it be done within the keyword editor in PrePoMax? My first attemts trying it were not succesful…
Keyword editor can only add new keywords, not modify the existing ones. So in this case you would have to export the input file and add the perturbation parameter manually using any text editor.
Ok, thanks.
I did that using the file you shared in the calculix forum (“Shell_multistep_identical.inp”)
The results for the second step is again different to the result of first step. And it is also different from the result I get when run the input file without “Pertubation”.
I’m not quite sure that I understand the “Pertubation” correctly. In the Documentation it seems to be used to carry over loads from a static step to a frequency step. So how does it make steps independent as you wrote before? What am I missing?
Well, it may not help with this bug in CalculiX. I will report it on Guido’s github by the way. But perturbation parameter is useful for solids (there’s no such bug with them). It allows you to perform load case analyses by treating the static steps as linear pertubation ones (perturbation about the base state). This concept is nicely explained in Abaqus documentation.
ok, understood, thanks. So I will stick with deactivating/activating steps when using shells in the meantime.
1 Like