Transient buckling analysis under temperature loading

[schleckschling]

Hi,

I'm working on a thin shell model with temperature loading (Input from another odb via *TEMPERATURE, FILE).
Due to high thermal gradients and buckling of thin walls, the analysis has lots of convergence problems. I tried to solve the calculation with *STATIC, STABILIZE and *CONTACT CONTROLS, STABILIZE as well as with *DYNAMIC (explicit and implicit). It's still impossible to get through all 50 steps (transient heating up with a maximaum temperature change of 50°C from step to step).

Can anyone help me to get the analysis running? Validation would be possible by checking the dissipated energy due to stabilization.

Best regards,
Frank

 

[Mustaine3]

It is hard to recommend something, without knowing the exact reason for the nonconvergence.

 

[schleckschling]

Hi Mustaine3,

some steps abort due to a too high displacement correction at nodes in the structure (buckling?). In other steps, the contact constraints do not converge -> too much SDIs.

 

[Mustaine3]

And why is Explicit not working?

Have you also tried *Dynamic, Application=Quasi-Static?

 

[schleckschling]

Hi Mustaine3,

*Dynamic, Application=Quasi-Static leads also to convergence issues.
Explicit is not working at the moment due to modelling issues. I'm currently working on some contact redefinition to get an explicit analysis running.

 

[TGS4]

Is it possible that you convergence issue is because you actually have a bucking instability? Have you considered a Riks analysis to evaluate post-buckling?

 

[schleckschling]

Hi TGS4,

I'm quite sure that the instabilities cause my convergence issues. Unfortunately, a riks analysis isn't possible with temperature loading

 

[TGS4]

Then you issue is structural/mechanical and not analytical, correct?