asdeasoft.net

Dear STKO Team,
I am writing you as I need help on a Soil-Structure Interaction problem, I hope you can help me find a solution. I have some questions on the Recorders and Post processing.
Based on your example on Absorbing boundaries: https://www.youtube.com/watch?v=OGII3a0 ... Technology
and on the Excavation examples developed by Arduino et al.: https://opensees.berkeley.edu/wiki/inde ... _Pile_Wall
I have developed two models of a retaining wall supported by piles. The main difference is is the type of interaction between the wall and the soil:
1) In the first model, the abutment and the soil share the same nodes and therefore they are constrained to move together (i.e. equal DOF interaction).
2) In the second one, there are contact 2D (penalty) elements between the wall and soil, so the geometry and mesh were slightly modified in order to connect the nodes of the wall and soil.
I have run the transient analysis with an earthquake record, so I managed to extract correctly the results on the nodes (like accelerations, displacements).
Then, I would like to extract the bending moment and shear on the wall. In the first model, I can visualize the bending moments on the wall (although they are in the Z direction, but this is a minor issue) but I cannot extract them.
Please note that I used the Recorder E5 LocalForce and then selected M in the drop-down menu. So, the first question is:
A) Is there a way to extract the bending moments (and also strain and stresses) on the BeamColumn elements in a similar way as with the accelerations on the nodes?
The second problem is related to model 2, the one with Contact 2D elements. In this model, using the same approach as in the first one, I tried to visualize E5 Localforce results but I obtain a result which is not consistent with the results obtained with model 1. So, the questions are:
B) First of all, do you think that I model correctly the interaction between wall and soil in model 2?
C) How can I visualize and extract the bending moments in a model with contact 2D? Do I have to create specific recorders? Or is there a way using Python to extract the results?

I am attaching the two different models (SCD files).

Thank you in advance for your response

I tried the penalty-based contact in 2D but it seems to have some issues (the 3D was working fine).
So using the lagrange-multiplier-based model it seems fine.
Keep in mind that with Lagrange-multipliers you need a convergence test based on residual vector.
Also, the TRBF2 integrator works better for this kind of contact conditions, otherwise you will see weird oscillations with standard Newmark, leading to unrealistically high accelerations

Now the model seems to work and you can correctly visualize the beam diagram.
If you want to extract data using the STKO extract tool, you need a result on gauss points, so section.forces, but for this you need a DispBeamColumn or a ForceBeamColumn (the elastic beam has no gauss output)

Thank you very much for the explanation, I can see that now the Contact 2D elements are giving correct results.

Regarding my question:

"C)How can I visualize and extract the bending moments in a model with contact 2D? Do I have to create specific recorders? Or is there a way using Python to extract the results?"

I have tried to do what you suggest to extract the bending moments on the wall (they are DispBeamColumn). I have set the view to see only the retaining wall, then I have extracted the chart data, but I have got an error (please see attached figure).
Please note that when I open the Gauss point plot, I can only select E4 (strain) or E5 (stresses), while the bending moments are only visible in the Beam Diagram plot.

Could you please help me to solve also this problem?

Thank you very much

Dear gabrifio,


"C)How can I visualize and extract the bending moments in a model with contact 2D? Do I have to create specific recorders? Or is there a way using Python to extract the results?"

You should use the Python API. With the Python API you can create all the engineering parameters you want including integrating the forces to obtain the bending moments. Dr. Petracca gave some webinars (https://asdeasoft.net/?stkowebinars-for-opensees from 20 to 24) and also in the future he is planning to show the potential of the Python interface in particular for researchers.

Regards

this is because the localForce output is not a result on gauss points (it is on the element's end-nodes).

If you need beam forces along the element, in each gauss point, you need the section.force output.
But to do so, you first need to use either a forceBeamColumn or a dispBeamColumn, i.e. those elements where you can specify an integration rule.

Dear STKO team,

Thank you for your responses, I managed to solve the issue thanks to your advice.

Regards