RE; Force displacement behaviour and extraction

[STKO Team]

I think the best way for you is to examine the model in detail.
First look at the location of plastic deformations and check they are in the same area as in plaxis.
Also compare the displacement/rotation at bottom and top of the pile.
In this way you can check wether the boundary conditions are correct and the model is displacing the same way.
Then try to inspect the stress-strain behavior at the Gauss points undergoing the maximum plastic deformation, and try to see if there are any visible differences (to understand if the difference comes from the material response)

[jiangtaolei]

Hello, Ryancwl95, I encountered the similar problem for SANISAND model in STKO see attached. So i started to check the soil stress to make sure the plasticity is activated, and I found another serious BUG, if there are more than two brick materials (the pile is brick element as well in my model), the soil stress cannot be recorded and shown in postprocess module. I then have to set the soil as a sperate region in analysis step section and record it. Finally, I found the soil stress is homogenous after lateral loading, this is quite incorrect. so I asked chatgpt, they told me like this: Critical Problem: Shared Material State
In OpenSees:
An nDMaterial is an object, not a template or a copy.
This means:
You have only:
one ManzariDafalias material (tag = 1)
one wrapper (tag = 3)
one wrapper (tag = 4)
But you have hundreds or thousands of elements

Consequence
All integration points in all elements are referencing the same material object
→ Therefore, they share the same internal state variables (e.g., void ratio, fabric, back stress, etc.)

Final Result
Any update at one integration point→ propagates to all others
This leads to: Artificial homogenization of stresses across the entire soil domain

[STKO Team]

Do not trust blindly ChatGPT.
Every element creates its own copy of the one material you define in Tcl.

however, this does not exclude there is a bug.
Can you share your model?

Are you bulding your model in stages?

[jiangtaolei]

Thank you for your response, I noticed you have provided a modified model in my post viewtopic.php?t=3367

But it cannot run properly in my computer, I have left some information on that post

Kind regards