self-centering controlled rocking RC shearwall

[martin0429]

Dear STKO Support Team,

I would like to develope a self centering, controlled rocking RC-shearwall modell. Currently I have several problems. Could you review my model and reveal my misstakes?

The dimensions are N and mm.

- RC shearwall: Layered shell; ASDShellQ4 element
B500 rebar: equivalent steel plates (0 and 90 angle) PlateRebar; Material Steel02;
C30/37 concrete: PlateFromPlaneStress; Material ASDConcrete3D

- unbonded tendons (posttension) in the RC-wall (Tensile Strength: 1770 MPa - nominal diameter: 15.7 mm)
steel04 material; truss property, truss element

I have right now a conventional, but posttensioned shear wall (fixed base). I will develope the gap opening (for the rocking behaviour) between the foundation and the bottom of the wall later, when the model run succesfully in the current stage.

At the support of the tendons I have several equalDOF and zero length element to perform a "tension only" connection. The analysis steps are the following: self-weight only, post tensioning phase (calculeted support displacement for 400 MPa stress), pushover analysis. (and later time history analysis)

I have attached the current version of my model.

*error while creating input files: invalid type of element or number of nodes

Thank you in advance!

[STKO Team]

The error is due to the use of ASDShellQ4 on a mesh that contains triangles:

mesh_01.png (48.7 KiB) Viewed 2339 times

Reason:
The top surface is not a 4-sided surface, thus cannot be mesh with a structures quad algorithm

mesh_02.png (7.91 KiB) Viewed 2339 times

Solution:
Add lines to cut the surface into all 4-sides surfaces, like this and use the merge command:

mesh_03.png (5.4 KiB) Viewed 2339 times

Other issues:
Now your model will not run due to singularities. They come from the way you are using a truss and how you are linking it to the ground and to the wall. Please review the connections

[martin0429]

Dear STKO Support Team,

thank you for your quick reply. I tried different configurations this week for the unbonded post-stressed tendons, but the model did not converge. Without the tendons, the model run perfectly (the model is like a conventional RC shearwall).

I'm relatively new to the opensees community, could you help me find the cause of the convergence? I have read a lot about the elements on the berkeley's and openseespy's websites, but i can't find the issue.

I drew a shematic overview about the configuration of the tendon:

tendon.jpg (133.13 KiB) Viewed 2268 times

I have also attached the current version of my model.

Please suggest if you think another solution would be more suitable for modelling the post tensioned tendons

Thank you in advance!

[STKO Team]

I think you can simplify the way you modelled you tendons...
Anyways, pay attention when you model with trusses and translational-only springs: you have to fix with rigid body rotations:

• Trusses SHOULD NOT be meshed, otherwise your internal nodes will be free to move in the X direction (rigid body translation in X + rigid body rotation about Y)

  

  PT_01.png (57.86 KiB) Viewed 2253 times

• Same thing here. Don't know why you put these in series like this, but if you really need to do it, pay attention to UX, RY rigid body modes

  

  PT_02.png (32.39 KiB) Viewed 2253 times