Atom topic feed | site map | contact | login | Protection des données personnelles | Powered by FluxBB | réalisation artaban
You are not logged in.
Greetings dear colleagues,
I have a problem to make a very simple model of Spherical plain bearing like attached picture.
In Code_Aster it is doable to model such a bearing with spherical contact surfaces, simulation for transitory analysis works as expected.
Problem arises in case of several hundred Spherical plain bearing instances in DYNA_NON_LINE, using the only to my knowledge allowable distribution method "Centralise" for parallel solving, making computation times very long...(cca 1 week+ at e.g. 14 MPI, 128 GB RAM for couple of seconds with very loose precision, resi_glob_rela, resi_glob_max, resi_frot)
So, my question is if exist there any simpler approach to model such a ball joint taking into consideration radius and friction of ball. It would be miracle if this could be done with simple node and beam model.
There is some ball element in Salome Meca for which I cannot find any documentation of its use (docs.salome-platform.org/7/gui/SMESH/adding_nodes_and_elements_page.html).
Any help in this direction would be greatly appreciated.
Thank you for your continuous efforts and stay safe.
Last edited by EnKiNekiDela (2022-09-17 19:50:35)
Offline
There is workaround which is limited (level of compliance is yet to be determined) and requires constant calibrations with same but more complex model with CONTACTS:
1. introduce CABLE element and make it rigid in axial directions
2. on 2 nodes on CABLE element introduce discrete 1D element DIS_T, then apply DIS_CONTACT properties over.
3. Have fun .
Last edited by EnKiNekiDela (2022-09-17 19:53:04)
Offline