basement interaction with soil

[spater]

Dear STKO team

Thank you for your reply.

This is my new model, I have finished meshing, but when I “partition”, STKO will exit automatically.

Could you help me to check what is the problem? If it is because of the low configuration of the computer that it cannot be used normally, if not, please give your opinion on modification.

When I uploaded the file, it showed error "Sorry, the board attachment quota has been reached." So I attached the mesh information of the model.

屏幕截图 2022-01-21 152819.png (7.92 KiB) Viewed 3363 times

And I attached the initial model without meshing,

geometry.rar

(2.29 MiB) Downloaded 170 times

Can you give me some suggestions about the mesh size whether my elements exceeded the ultimate number?

Thank you very much!

[STKO Team]

The model you sent me does not have the same mesh seed (mesh size) and mesh controls you used.
In order to replicate your issue, I need the exact model you used. If the size is too large for the forum, you can use something like WeTransfer and obtain a link (post it here).

Otherwise you can send me the model without the mesh, but with the same mesh seed and controls you used

[spater]

Dear STKO team
I uploaded it to the Google Drive
https://drive.google.com/file/d/1AVg9LL ... sp=sharing
Thank you very much

[STKO Team]

I ran the partition command on your model and when the METIS library is called (the library we use for partitioning) there is a huge peak in memory.

However it works even for larger meshes when the mesh size is not that different between the outer box and the inner box.

It looks like there is an issue in partitioning the 2 solids when their mesh size is so different. It will take a while for us to find the source of the issue.

In the meantime I would suggest you to make the inner box mesh size larger. Keep in mind that a model with more than 2 million nodes, will take more than 300 GB or RAM during factorization...

[spater]

Dear STKO team

Thank you for suggestions!

I remeshed the model, but I still can't partition, So I planned to use the openseessp to calculate, but it could not converge and it took too much time to calculate for one increment.

Can you help me check what's the problem?

THANK YOU VERY MUCH!
The figure is the information of the mesh and the error of the calculating procedure.

屏幕截图 2022-01-21 200634.png (10.89 KiB) Viewed 3366 times

屏幕截图 2022-01-21 200418.png (49.51 KiB) Viewed 3366 times

And I also attached the remeshed model, this is mesh side is larger than the last one.

case1.rar

(14.99 MiB) Downloaded 177 times

It looks like there is an issue in partitioning the 2 solids when their mesh size is so different. It will take a while for us to find the source of the issue.

When can you solve this problem? If you finished it, could you please inform me in time? Thank you!

Based on the openseesmp and openseessp, I want to ask three questions:

• 1. what is the difference between openseesmp and openseessp? (openseessp-one processor, openseesmp-multiple processros)

• 2. Now that both of them can calculate the large model and and apply paralell analysis, what is the different efficiency and speed when calculating a same model? which one is better for my large model?

• 3. what is the different settings in the analyze command when I ues them seperately? Except ParellelRCM.

Thank you

[Clarabella]

Regarding performance you can read this paper:
https://www.researchgate.net/publicatio ... e_analysis

The authors used STKO for partitioning the model.

Regards

[STKO Team]

I remeshed the model, but I still can't partition

It's because you decreased the number of elements in the outer box, which was already coarse.
I mean, try to increase the mesh size in the inner box.
The mesh size you used in the XY plane is very dense and make the elements very distorted because the mesh size is larger in the Z direction.

Try to increase the mesh size in the inner box and let's see what happens

[spater]

Dear STKO team and Clarabella

Thank you for replying!

I remeshed the model, it can be partitioned now.
The outer-soil: grid size is 10m in X-Y plane, and the inner-soil: grid size is 1m in x-y plane, and the basement box: grid side is 1m in X-Y plane. The whole model gird size in Z direction is 5m. Based on this mesh size, the node number is 192155, the element number is 533181.

But it failed in analyzing. What is the problem? Can you help me modify it.
I uploaded the model to the Google Drive. https://drive.google.com/file/d/1n9MZ6Y ... sp=sharing

I need your help urgently.
Thank you very much.

屏幕截图 2022-01-22 101941.png (112.07 KiB) Viewed 3366 times

[Lucy2012]

Dear Spater,

You have problems of convergence please check our webinar on convergence issues, nonlinear analysis is a complicated task and requires time, experience and effort.

Best regards

[spater]

Dear Lucy and STKO team
I have studied the webinar carefully but I couldn't fix it.

Can you help me fix it?

Thank you very much

https://drive.google.com/file/d/1n9MZ6Y ... sp=sharing

Beijing time is 7 hours ahead of Italian time, and I really cherish the time to talk with you.

Thank you so much for helping me.