Timoshenko beam

[pavlos]

Hello, I have a quick question is it possible in Timoshenko elastic beam element somehow skip the part regarding coordinate-transformation when i am defining the element. I want to do it because my model is already in correct way but when I run the script I am getting message that it is rquired, so I tried to put 0 in the place but it is not working. Do you have any suggestions how I can do it?
Thanks

This is my code
#element ElasticTimoshenkoBeam $eleTag $iNode $jNode $E $G $A $Jx $Iy $Iz $Avy $Avz $transfTag <-mass $massDens> <-cMass>
element ElasticTimoshenkoBeam 1 1 2 2.1000E11 8.0769E10 1.9801E00 27.221E00 13.610E00 13.610E00 3.76E-01 3.76E-01 0 -mass 15.544E03;

[EricsonEncinaZ]

According to the commands you are using you have a 3D domain, so the vector passed in the geometrical transformation tells the domain how your element is oriented, i.e. the relationship between the element's local reference system and the global reference system. If you do not provide the vector in a 3D domain then OpenSees can not know how you want your element to be oriented.
If your model lived a 2D domain then the vector is not needed.

[pavlos]

So basically I cannot make the model directly in the global reference system but I have to create it in the local and then rotate it as I want to have? Do I understand it correctly?

[EricsonEncinaZ]

I would say elements have their own local reference system, so you can define things like Iy and Iz. As a separate case, the domain has its own reference system. Now both RS do not know each other, so you have to let them know how to interact, this is done by the vector vecXZ and the geometric transformation type.
So the idea of creating everything in the global RS is not possible, that's not the way complex FEA softwares, like OpenSees, work. As an additional example, ETABS has a global RS where the model lives, then every section has its own RF, and then if you assign a pier, for example, then the pier has its own RS as well. The different RSs can match in orientation, but they are not always pointing to the same direction.

[pavlos]

I have one more question is it possible to model hollow circular beam in OpenSees?

[EricsonEncinaZ]

If elastic, you can calculate the properties by hand and assigned them to the elastic element, as you are currently doing.
You can also try a fibre-based section, and then assign that section to a beam-column element.