Hello STKO team,
I hope this message finds you well.
My apologies for posting a new message here. In our previous discussion thread (link:viewtopic.php?t=3283), although there were 14 messages in total, only 13 are currently visible. I'm concerned that my earlier request for assistance might have been overlooked, so I'm reposting my questions here.
Following your previous suggestions, I have carefully studied the webinar on ZeroLengthContactASDimplex and implemented corresponding modifications to my model. Specifically, I changed the connection method from node-to-node to node-to-element and made adjustments to the mass and load applications on the structure. However, I continue to encounter convergence issues during execution. The model hasn't even successfully completed the gravity analysis. I have reattached the recently revised model file and would greatly appreciate it if you could help analyze the problems.
Additionally, my model involves groundwater conditions, but I haven't applied any specific treatment to the soil-structure interface surfaces. I'm uncertain whether my current approach is correct and would welcome your guidance on this matter.
The second model file is intended for geostatic stress balance verification. While checking the post-analysis results, I noticed potential errors and would be grateful for your assistance. Could you also provide some guidance on how to determine whether geostatic stress balance has been successfully achieved?
During this process, I observed that applying volume forces directly to elements yields better convergence compared to applying loads separately. The latter loading method fails to execute successfully, and I haven't been able to identify the specific cause. I would appreciate your insights into this issue as well.
Based on this model, I also attempted to modify the soil-structure connection method, replacing the previous rigid link constraints with ZeroLengthContactASDimplex elements. However, I encountered similar convergence problems with this approach.
Finally, I have one additional question regarding result data processing: in a 3D model where a single node is shared by eight elements, how should the stress value at this node be accurately determined?
Thank you once again for your time and support. I look forward to your reply.
Persistent Issues: Model Convergence, Geostatic Balance Verification, and Nodal Stress Calculation
Persistent Issues: Model Convergence, Geostatic Balance Verification, and Nodal Stress Calculation
Last edited by liushun on Sat Nov 22, 2025 8:02 am, edited 1 time in total.
Re: Persistent Issues: Model Convergence, Geostatic Balance Verification, and Nodal Stress Calculation
Dear user, let's tackle one issue at a time, otherwise it's difficult for us to look into a file for ALL possible issue.
Post 1 file, and describe 1 issue.
Post 1 file, and describe 1 issue.
Re: Persistent Issues: Model Convergence, Geostatic Balance Verification, and Nodal Stress Calculation
Hello STKO team,
I hope this message finds you well.
Indeed, focused discussions often yield better results, and I fully agree with this approach. Therefore, I have selected what I consider the most crucial issue from my previous questions - the analysis of geostatic stress balance. I would like to begin this discussion with a case study, which can be found at the following URL:https://opensees.berkeley.edu/wiki/inde ... _Analysis)
This is a one-dimensional soil column model where I used the single-layer soil model code. I only modified the recorder in the original code to extract the vertical displacement at a depth of 5 meters. As shown in the attached figure, the displacement value appears excessively large. Based on my understanding, the displacement generated during the gravity analysis phase should be very small (approximately on the order of E-10).
This is the code I used to extract the data:
recorder Node -file dash/dash-Gdisplacement.out -time -node 281 -dof 2 disp
In this model, a viscoelastic boundary condition was initially applied at the base. Since rigid boundaries are more commonly used in my work, I modified the viscoelastic boundary to a rigid boundary using the command fix $nodes-bot 1 1. Using the acceleration time history as input, I compared the results of the two configurations and observed differences even during the gravity analysis phase—an outcome I had not anticipated.
I would greatly appreciate your assistance in addressing this issue. Specifically, I would be grateful if you could provide guidance on how to determine whether geostatic stress balance has been successfully achieved, as well as insights into the displacement differences resulting from different bottom boundary settings.
Thank you for your time and help. I look forward to your reply.
I hope this message finds you well.
Indeed, focused discussions often yield better results, and I fully agree with this approach. Therefore, I have selected what I consider the most crucial issue from my previous questions - the analysis of geostatic stress balance. I would like to begin this discussion with a case study, which can be found at the following URL:https://opensees.berkeley.edu/wiki/inde ... _Analysis)
This is a one-dimensional soil column model where I used the single-layer soil model code. I only modified the recorder in the original code to extract the vertical displacement at a depth of 5 meters. As shown in the attached figure, the displacement value appears excessively large. Based on my understanding, the displacement generated during the gravity analysis phase should be very small (approximately on the order of E-10).
This is the code I used to extract the data:
recorder Node -file dash/dash-Gdisplacement.out -time -node 281 -dof 2 disp
In this model, a viscoelastic boundary condition was initially applied at the base. Since rigid boundaries are more commonly used in my work, I modified the viscoelastic boundary to a rigid boundary using the command fix $nodes-bot 1 1. Using the acceleration time history as input, I compared the results of the two configurations and observed differences even during the gravity analysis phase—an outcome I had not anticipated.
I would greatly appreciate your assistance in addressing this issue. Specifically, I would be grateful if you could provide guidance on how to determine whether geostatic stress balance has been successfully achieved, as well as insights into the displacement differences resulting from different bottom boundary settings.
Thank you for your time and help. I look forward to your reply.
Last edited by liushun on Sat Nov 22, 2025 8:03 am, edited 1 time in total.
Re: Persistent Issues: Model Convergence, Geostatic Balance Verification, and Nodal Stress Calculation
Hello STKO team,
I hope this message finds you well.
During my recent studies, I discovered a method to address the large deformation issues in geostatic stress balance analysis. By implementing the initialstateanalysiswrapper material, the deformations during the gravity analysis phase can be forced to zero.
Based on this approach, I have made further revisions to my previous 3D model. However, the analysis still fails to run successfully. I have reattached the revised model file and would greatly appreciate it if you could find time to analyze the potential causes. Here is the link to the discussion thread: viewtopic.php?t=3283
Thank you very much for your assistance. I look forward to receiving your reply in the thread I provided and remain truly grateful for your support.
I hope this message finds you well.
During my recent studies, I discovered a method to address the large deformation issues in geostatic stress balance analysis. By implementing the initialstateanalysiswrapper material, the deformations during the gravity analysis phase can be forced to zero.
Based on this approach, I have made further revisions to my previous 3D model. However, the analysis still fails to run successfully. I have reattached the revised model file and would greatly appreciate it if you could find time to analyze the potential causes. Here is the link to the discussion thread: viewtopic.php?t=3283
Thank you very much for your assistance. I look forward to receiving your reply in the thread I provided and remain truly grateful for your support.
Discussion on Nodal Stress Calculation: Reasonable Computational Methods for Stresses at Shared Nodes in 2D/3D Models
Hello STKO team,
I hope this message finds you well.
I have a question regarding nodal stress calculation in finite element analysis:
In a 2D model using quad elements, a single node is typically shared by four surrounding elements. Similarly, in a 3D model using stdbrick elements, one node is shared by eight adjacent elements.
What would be the most appropriate method to calculate the stress at such shared nodes? Is taking the average value of stresses from all connected elements considered a reasonable approach?
Thank you very much for your assistance. I look forward to your professional insight.
I hope this message finds you well.
I have a question regarding nodal stress calculation in finite element analysis:
In a 2D model using quad elements, a single node is typically shared by four surrounding elements. Similarly, in a 3D model using stdbrick elements, one node is shared by eight adjacent elements.
What would be the most appropriate method to calculate the stress at such shared nodes? Is taking the average value of stresses from all connected elements considered a reasonable approach?
Thank you very much for your assistance. I look forward to your professional insight.
Re: Persistent Issues: Model Convergence, Geostatic Balance Verification, and Nodal Stress Calculation
STKO already had this feature.What would be the most appropriate method to calculate the stress at such shared nodes? Is taking the average value of stresses from all connected elements considered a reasonable approach?
Stress is computed at gauss points, extrapolated at nodes and smoothed if requested