Nodes without element property

[gercekonur]

Dear STKO Team,

I have created my model. However, in the TCL file it is shown that there are some nodes without element property. I used planar surface for the shell element. I have added some extra nodes on the planar surface in order to assign zero length elements. Also, I have another error which says "no local y vector for twonodelink element'.

I would like to ask you could help me with the error.

Thank you in advance.

[STKO Team]

It was writing nodes with no element assigned because some faces of one wall did not have any element assigned.

Then I slightly changed the tie connections, to have:

• rigid link in the middle
• zero length with pinching on the 2 sides with only the Uy DOFs
• all other DOFs of the zero-length node are connected to shell nodes with EqualDOFs in all DOFs except Uy (better than using stiff materials)

Now it runs. However, I do not fully understand your boundary conditions. You are using imposed displacement in Uy in 2 nodes, but then you use a uniform excitation. Furthermore, you did not assign any mass and damping.

[gercekonur]

Dear STKO Teams,

Thank you so much for your help.

I have corrected the parts that you mentioned (removing the nodal load for time history and adding damping etc). However, I couldn't perform the modal analysis with the script that you posted on the forum (viewtopic.php?f=40&t=1513&p=2642&hilit= ... 6cfe#p2642). In order to perform the modal analysis, I have assigned masses for the leaves using FaceMass. I didn't use any distributed mass on the elements. Is there any other thing that we need to take into consideration?
I also want to ask you that Can we perform the modal analysis while we are using openseesmp and can we use the monitor for openseesmp as well?

Ps I am not sure that these questions are in the correct place. Otherwise, I would like to post them again in the correct section.

[STKO Team]

I also want to ask you that Can we perform the modal analysis while we are using openseesmp and can we use the monitor for openseesmp as well?

Modal analysis can be done in OpenSeesMP, but only if you use 1 processor. Unfortunately the Eigen command in OpenSees does not work in parallel.

The monitor can be used in parallel analyses as well.


What problem do you have with the modal analysis? what error does it show?

[gercekonur]

Dear STKO Team,

Thank you so much for your support.

Sorry, I forgot to attach the error that I received. I also used -fullGenLapack as eigen solver in the analysis.

Best regards,

[STKO Team]

The error says that the lambda in sqrt(lambda) is not in the valid range for the sqrt funciton. It means that you have a negative eigenvalue. It can happen when your model is not well constrained, or when you run an eigen command after your structure goes into a softening stage.

Make sure your model is well constrained. If I recall properly, your model was not constrained in Y.