Hinges between plates/shells

[johnwang]

triggering nonlinear analysis.

'mean rotation MPC' will not triggering nonlinear analysis.

https://forum.freecadweb.org/viewtopic. ... 30#p516595

[twangrt]

The only working solution for Calculix we have found so far is by changing the original mesh, which is not something I am happy with ... https://forum.freecadweb.org/viewtopic. ... 20#p516277

It would be cool if we would found something without changing the mesh and without triggering nonlinear analysis.

Okay then I dig some more, my thought was that I'll use the EQUATION-card. But as mentioned then the nodes as elements need to be altered or rewritten.

But just of curiosity, why cant we rewrite the nodes/elements? I was thinking a constraint-like function that can be applied on edges that rewrites the already written .inp file. I know, I'm quite naive

[twangrt]

Is there anyone already pursuing this?

Got a window of opportunity and was thinking of trying to give it a serious go.

Tough I'm been poking around the code and it seems like a more complicated task than I initially thought. But you can't blame a guy for trying.

I have uploaded a working version with MPC as hinge for Mystran.

For Calculix, you need to understand the 'mean rotation MPC' I mentioned.

I'll have a look, maybe me that just focus to much on calculix

[user1234]

Source_code_management

Sorry for late reply, i was really busy. One question: is gitlab an option? Since i also wanted to try the CI runner for gitlab.

Greetings
user1234

[bernd]

Source_code_management

Sorry for late reply, i was really busy. One question: is gitlab an option? Since i also wanted to try the CI runner for gitlab.

Greetings
user1234

no problem. A direkt PR will not be possible butsince I can fetch from gitlab it would work too.

[bernd]

But how would you solve this? AFAIK we have not found how to solve it on calculix without rewriting the mesh.

[user1234]

Also not with an equation in the imp file as other suggested? If not, i want do help and look at the lists in the tracker or improve (unify) the usability. At least one mini fix i have found in PartDesign.

Greetings
user1234

[bernd]

We may need to have another closer look. But we will need the this new object anyway. Solver Mystran is in master and solver Mystran is for sure able to do this rather simple. See former post of john. Solver Mystran is my choice ATM for beams and shells. I will spend time to better support them. Thus a new object to define hinges for shells and beams would be needed anyway.

Furthermore we will need to improve the beam section object to be able to define more cross sections. May be even free ones. This means to put in moment of inertia manually or even retrieve area and moment of inertia from a face modeled in FreeCAD. But this is another topice see https://forum.freecadweb.org/viewtopic. ... 20#p520457