Beam to Solid coupling hinge

[matsadriaenssen]

Dear all,

I am modelling a steel truss bridge on masonry piers with beam and solid elements. The truss is supported on the masonry piers as a hing/roller support. However, I am do not know how I can simulate these constraints. If I use the Beam to Solid coupling, I cannot decouple specific rotations and everything is fixed. If I use the equalDOF all the rotations are decoupled but I need the out of plane bending and the torsional bending to be transferred to the masonry piers.

I tried to combine the Beam to Solid coupling and the equalDOF by putting an element in between and assigning the truss to the element by equalDOF and then the element to the piers by Beam to Solid coupling. Then I put it on top of each other so the length is 0 but this gave convergence issues. Is there another way?

Thank you for the help!

Kind regards,
Mats

[STKO Team]

In this case you can:

• create an interaction
• assign a zeroLength instead of a kinematic constraint
• assign stiff material only to the DOFs that should transfer loads/couples

[matsadriaenssen]

Dear STKO Team,

Thank you for the help.

Kind regards,
Mats

[matsadriaenssen]

Dear STKO Team,

I tried your recommendations but I still do something wrong since the analysis does not converge in the beginning. I think it might have something to do that I am connecting a line with a solid. I tried to implement the same line in the solid because before it was giving me warnings that the length of the ZeroLengthElement was too long. I do not get that error anymore but as I said it still does not run.

I attached the file, hopefully you can tell me what I am doing wrong.

Kind regards,
Mats

[STKO Team]

Check your model: All your solid elements are separated. You should merge them.

[matsadriaenssen]

Dear STKO Team,

I have merged the model and also checked if all the elements have element/physical properties and conditions. I ran a model with only the steel truss and the analysis is fine. I also ran a model with only the solid masonry elements, which is also okay. Only the full model still does not converge because the "Matrix is Singular Numerically" and I dont know why. I think I narrowed the problem down to the connection of both (in this case the ZeroLength element). I attached the adjusted and merged model. Is there something you see that can be the problem?

Thank you in advance.

Kind regards,
Mats

[STKO Team]

You should also use a material for the 5th DOF of the zero-length.
I understand you left it empty because you have a roller and thought it necessary to avoid moment transfer.

But it is not like this in this case. The solid element already does not have rotational DOFs so it will never transfer moment. But since you are connecting a beam (6DOFs) to a solid (3DOFs) with a zero-length element, and it cannot have different DOFs at its end-nodes, STKO will create and extra node with 6DOFs at the same location of the node of the solid element (end they will be equalDOF'ed in the translational DOFs). So if you don't put a material in the 5th DOF of the zero length, the 5DOF of the extra node will be singular

[matsadriaenssen]

Dear STKO Team,

This was indeed the problem, it is fine now!

Thank you a lot to help me solve this problem.

Kind regards,
Mats

[STKO Team]

You're welcome