Soil-Pile Interaction Modeling

[kolozvari]

Hello,

I would like to model soil-pile interaction, and I found this video (E-Learning Course Soil-Pile Interaction - Modeling with STKO's Python API on YouTube; https://www.youtube.com/watch?v=iDgFM-ZjHVo). However, the video doesn't show step-by-step process to define geometry, interactions, etc. I was wondering if there is any similar video that you would recommend to start modeling of the soil and pile geometry from scratch. If not, can you please walk me through step-by-step process to do this.

Thank you.

[kesavapraba]

Hi. Since the pile-soil interaction covered in the webinar is an advanced subject, Dr. Massimo did not dedicate time to creating the geometry in order to focus more time on explaining the specific things related to the topic. Nevertheless, he explains in the video (in the initial 10 mins) the process of creating the geometry and interactions, which is akin to developing any other model. Kindly view the video linked below to gain insight into the development of geometry, interactions, and local axes. When you encounter any specific challenges with geometry or interactions in your model, feel free to reach out for additional assistance. Thank you.

https://www.youtube.com/watch?v=w8FCrWkeSyI

[kolozvari]

Thank you, but I am still having an issue with creating my geometry and meshing the model. As I said, I am trying to redo the example (E-Learning Course Soil-Pile Interaction - Modeling with STKO's Python API on YouTube; https://www.youtube.com/watch?v=iDgFM-ZjHVo), but with a slightly different geometry.

1) There is a line that was generated by extruding a circle (see yellow highlight in the screenshot below) when I was trying to replicate the model from the example. This line is creating problems as STKO reports that there is a mismatch with DOFs at nodes where this line intersects with solid elements.

Capture1.PNG (134.37 KiB) Viewed 5848 times

2) For some reason, even though I selected the structured Hexa/Quad mesh, my mesh is still triangular (see second screenshot below). I am not sure how to change this.

Capture2.PNG (256.23 KiB) Viewed 5848 times

Please see the attached model. I could really use a step-by-step explanation about how to create a geometry from the example.

Please let me know. Thank you.

[STKO Team]

That line is created by the extrusion of the circle. More specifically, by the vertex used as the second point in the construction of the circle.

The reason why it does not allow a structured hexa mesh is because that extra line creates faces with more than 4 vertices and 4 edges, thus not a candidate for structured meshes.

In the webinar example, I avoided this by creating a circle using a direction (second_point - first_point) of 45° from the global X-axis, so that the vertex will coincide with the diagonal lines created at 45°.

DIAG.png (16.44 KiB) Viewed 5834 times

Now, since you already did it this way, I used a simple trick to make all the faces 4-sided.
I simply created 2 orthogonal planes along the global X and Y directions passing through the center.
I used them to cut the solid so that all sub-geometries are candidates for structured mesh:

DIAG CORR.png (143.56 KiB) Viewed 5834 times

Here's a modified file:

Pile_v1.zip

(285.21 KiB) Downloaded 407 times

[kolozvari]

Thank you very much for the quick response. Makes sense. The mesh now is indeed what I wanted. However, the model doesn't run because there is an error that NDOFs are not consistent (3 vs 6 DFOs at the node). I did some investigation by reducing the number of elements to locate the node where this happens, and it seems that it is a node at the base of the face of the opening in the soil (see screenshot below).

Capture.PNG (198.11 KiB) Viewed 5826 times

When I delete that line, the face disappears and the problem moves to the next one. If I delete the entire face of the opening, there is no problem. I don't understand why would a line have any DOFs and I am not sure how to solve this problem. This problem doesnt occur in your model. Can you please advise or fix the problem?

Thank you.

[STKO Team]

Here's a fixed version of your file:

Pile_v1.zip

(295.5 KiB) Downloaded 432 times

Both interactions using the BeamContact3dp or BeamEndContact3dp must be node-to-element.
One was missing the material property
The side interaction was missing the assignment of the cylindrical local axes
Pay attantion to convergence issues given by contact!


Here's the same model made with the new contact elements we developed aming at improving convergence of contact problems:

Pile_v2.zip

(297.02 KiB) Downloaded 427 times

If you activate the IMPLEX algorithm the convergence will be further improved

[kolozvari]

Thank you very much!

[kolozvari]

Hello, I have a few questions regarding some parameters of ContactMaterial3D material and zeroLengthContactASDImplex element used to model the contact between the pile and the soil.

1) In your example shown in the YouTube video, you are using BeamContact3dp element, while you suggested for me to use zeroLengthContactASDImplex. Can you please explain what is the difference between the two and why you selected to use BeamContact3dp in your example instead of zeroLengthContactASDImplex? Which one is better and which one would you recommend for the soil pile interaction as modeled in the example and my file that I shared earlier?

2) I have noticed that the friction coefficient (mu) must be defined in ContactMaterial3D material and zeroLengthContactASDImplex element. I am assuming that the value of mu should be the same in both element and material, but this is unclear from the user documentation. Can you please explain what is the relationship between the two values of the friction coefficient and how they need to be defined?

3) Similarly, ContactMaterial3D material requires the definition of G, while zeroLengthContactASDImplex element requires the definition of Kt (and Kn). Is there any relationship between the G in the contact material dn Kt in the zero length element? Can you provide some guidance on how to select these parameters?

4) Regarding the Rigid Gap flag. In my model (the one you have downloaded and reviewed before - see thread above), I have a non-zero distance between the nodes and the master node because the pile is modeled with a beam-column (line element). Also, the pile is my master geometry, which has rotational DOFS. Since both of these criteria are per the user manual explanation for the Rigid Gap, then I should be using the Rigid Gap flag. Can you please confirm?

Thank you,

[STKO Team]

1) In your example shown in the YouTube video, you are using BeamContact3dp element, while you suggested for me to use zeroLengthContactASDImplex. Can you please explain what is the difference between the two and why you selected to use BeamContact3dp in your example instead of zeroLengthContactASDImplex? Which one is better and which one would you recommend for the soil pile interaction as modeled in the example and my file that I shared earlier?

Because at the time we did webinar #22 the ASD element was not yet developed.
Check webinar #24 to see a dedicated webinar about the new ASD contact element
#024 Implicit-Explicit Contact Element Applied Examples: soil-foundation interaction and RC structures

2) I have noticed that the friction coefficient (mu) must be defined in ContactMaterial3D material and zeroLengthContactASDImplex element. I am assuming that the value of mu should be the same in both element and material, but this is unclear from the user documentation. Can you please explain what is the relationship between the two values of the friction coefficient and how they need to be defined?

It's the same. the friction coefficient = tan(friction angle)

3) Similarly, ContactMaterial3D material requires the definition of G, while zeroLengthContactASDImplex element requires the definition of Kt (and Kn). Is there any relationship between the G in the contact material dn Kt in the zero length element? Can you provide some guidance on how to select these parameters?

It's the same G = Kt = tangential interface stiffness. While Kn is the normal interface stiffness (should be infinite in theory, but in a penalty based method it just a large value)

4) Regarding the Rigid Gap flag. In my model (the one you have downloaded and reviewed before - see thread above), I have a non-zero distance between the nodes and the master node because the pile is modeled with a beam-column (line element). Also, the pile is my master geometry, which has rotational DOFS. Since both of these criteria are per the user manual explanation for the Rigid Gap, then I should be using the Rigid Gap flag. Can you please confirm?

Yes.
Watch webinar #24, There you can find all use-cases for this element. Including the soil-pile one

[kolozvari]

Hello, again. I am still working with this model and I decided to simplify it by trying to connect the pile (modeled as beam-column element) and the surrounding soil in a way so they move together, without modeling contact using contact elements. I am using beamtoSolid multiple constraints to connect the pile to the soil domain at the sides and Equal DOF constrain at the base. When I plot the deformed shape, it appears that the pile is penetrating the soil only at the very top (see screenshot below).

Capture.JPG (75.21 KiB) Viewed 5103 times

Also, I am doing a force-controlled analysis using 10,000 kN, but the overall total reaction at the end of the analysis is 120,000 kN. Not sure why is this. When I use exactly the same analysis settings in models with explicit contact modeling, the two numbers match,

Can you please let me know what could be causing these two issues? I attached the model for your convenience.

Thank you,