Hi Kiwi3D team,
I’m trying to simulate a beam shell element, where black elements are beams and red element is a shell connected to the beams. By applying a force shown in the figure, I want to bend the whole element but when I check stresses and deformed geometry, those are not the expected.
I’ll let the file here and I’ll aprecciate your help.
07_K!-Beam+shell.gh (26.4 KB)
Check this model please, it’s more accurate and organized. But still don´t get the expected results talking about stresses. Can you tell me if I´m modelling everything ok?
Thanks
07_K!-Beam+shell.gh (23.9 KB)
Hey,
beside refining the beam, you have to do the same for the shell (see file below). If this is not the case, the coarse control pint net of the shell dominates the deformation and stresses.
Additionally, I increased the accuracy and added a load curve in order to apply the load incrementally.
07_K!-Beam+shell_mod.gh (27.1 KB)
Hi Anna, thanks for your reply, I applied all your recomendations.
I’m having some trouble with my beam-shell element modelling and simulation, because I’m having very different results depending on the “equilibrium accurancy” and I think that I shouldn’t having that. When I reduce equilibrium accurancy from 0.5 to 0.1, my model explode or is having instability. Do you think am I doing something wrong? How can I trust the results?
I’ll attatch some pictures and the file,
07_K!-Beam+shell.gh (56.0 KB)
Thanks.
An equilibrium accuracy of 0.5 is really high and not recommended. You should at least use 0.001. Otherwise the result is really arbitrary for nonlinear analysis.
Regarding the instability, you can manipulate the load curves and give different load curves to the load components, e.g. first apply the the bending loads and then the dead load. The nonlinearAnlaysis also provides adaptive load control by bisecting non-successful timesteps (Input A). You can also use supportdisplacements or split it into two analyses.
07_K!-Beam+shell_2_mod.gh (86.5 KB)
