Rigid link connection between beam and columns

[Anonymous]

Dear Users,

I am modelling a special tower. There is a "beam" between the different floors which is actually made from steel plates. I have defined it as a user defined section. On the figure below you can see the section and how the columns attached with rigid links:

I set the Ux, Uy, Uz, and Rz blocked. However I am not sure in terms of moments and displacements that this is the correct approach to model the beam and the excentric columns' connection/ interaction. It looks like that in the model:

I have attached the example model as well. Please make a review and feedback how you would model such a beam column connection.

Thanks in advance!

[Anonymous]

This might be stating the obvious but how about modelling the 'beam' as a plate element? or even multiple plate elements for a more realistic stiffness, or edit the plate properties for further realistic stiffness?
If you want to investigate local behavior, try modeling the connection as a solid.

[Anonymous]

Thanks for the reply ! Yes, I thought that as well, but it would not be that easy to make a model like that (columns on the surface of the plates) and the model will be larger and more complex, thus I would like to avoid plate elements and the necessary fine mesh as well.

I mean it should be possible to model in this way. Rigid connection is made for these kind of purposes as well. But I have no experience with it, therefore I need some revision and feedback.

[Rafacascudo · Accepted solution]

That is OK , but I would make full rigid links and release the columns ends (if they need to) instead

[Anonymous]

 Thanks for the answer Rafacascudo !

Full rigid links mean that all the Ux, Uy, Uz and Rx, Ry, Rz are blocked, right? Maybe I misunderstood the rigid links, because I thought that with these blocked slave nodes the columns' ends wont be able to move or settle down, which wont be true. But it seems they can move, so what these blocked DoF means for the slave node?

If I release the columns end there are instabilities (Ux, Uz) during the calculation. What do you mean "if they need to" be released?

Thanks for your time and explanation in advance !

[Rafacascudo]

Rigid links Xyz directions are global directions. Blocking them and releasing the bars ends will give you more control on the directions you need to release if there are any. If the connections between the "plate " beam and vertical bars are fully rigid , you don't need to release anything

[Artur.Kosakowski · Accepted solution]

Imagine a rigid link as a bar element which is e.g. 100 times more rigid than the ones it connects. The connection in its master node (let's name it origin) is fully fixed. The "releases" you define in its slave node (end) corresponds to a bar release (but this time you mark fixed these d.o.fs that you want to block) defined in the global coordinate system rather than in the local bar coordinate system. 

If you find your post answered press the Accept as Solution button please. This will help other users to find solutions much faster. Thank you.