mesh problem
-
shijun@hit.edu.cn
- Posts: 37
- Joined: Thu Mar 10, 2022 12:02 pm
mesh problem
There is a mesh problem in my model . And this is the model .
Because the line element was used to establish the boundary column, the actual size of line element need to extend the cetain distance and there is the repeat area . I want to separate the boundary column and web wall in the shear wall to garantee the actual length . The interaction of General link is defined between boundary column and web wall . However, I dont know if the set is correct and the model is abovementioned.
Because the line element was used to establish the boundary column, the actual size of line element need to extend the cetain distance and there is the repeat area . I want to separate the boundary column and web wall in the shear wall to garantee the actual length . The interaction of General link is defined between boundary column and web wall . However, I dont know if the set is correct and the model is abovementioned.
Re: mesh problem
There were some issues in your model:
The first error in the mesh is because you used quadratic line elements (3-node line elements) with a beam element that wants only 2 nodes.
Keep in mind that the "quadratic" order means quadratic order of interpolation for displacement and generates element with mid-edge nodes.
Another error is the mesh type. You need:
Finally the interaction you used are wrong. You need a node-to-node interaction, where the beams are the masters and the shell-sides are the slaves.
Then you need to use a RigidLink boundary condition on them, otherwise the interaction alone is just a geometric feature.
The first error in the mesh is because you used quadratic line elements (3-node line elements) with a beam element that wants only 2 nodes.
Keep in mind that the "quadratic" order means quadratic order of interpolation for displacement and generates element with mid-edge nodes.
Another error is the mesh type. You need:
- Triangular shape with Linear order on the top face
- Quadrilateral shape with Linear order on the other faces
Finally the interaction you used are wrong. You need a node-to-node interaction, where the beams are the masters and the shell-sides are the slaves.
Then you need to use a RigidLink boundary condition on them, otherwise the interaction alone is just a geometric feature.
- Attachments
-
- actual size-480-2.zip
- (187.86 KiB) Downloaded 303 times
-
shijun@hit.edu.cn
- Posts: 37
- Joined: Thu Mar 10, 2022 12:02 pm
Re: mesh problem
Thank you for your reply.
Re: mesh problem
Dear STKO Team,
I also encountered this problem, and after making many changes the error was still not solved, I don't know what went wrong. I have attached the file below.
I would be grateful if you could help me with this problem.
Regards,
Charon
I also encountered this problem, and after making many changes the error was still not solved, I don't know what went wrong. I have attached the file below.
I would be grateful if you could help me with this problem.
Regards,
Charon
Last edited by Charon on Wed Mar 08, 2023 10:53 am, edited 1 time in total.
Re: mesh problem
The error given in red is pretty clear:
You assigned a 2-node Element Property (dispBeamColumn) to a linear mesh with 3 nodes. This means that somewhere in your mode you use a line mesh with quadratic interpolation (i.e. a 3-node line element).
Please, select all those edges and assign Linear interpolation and re-mesh
After this, you will see another error regarding the pile-cap coupling condition. I don't know what you are trying to do there, but the interaction you use is wrong
You assigned a 2-node Element Property (dispBeamColumn) to a linear mesh with 3 nodes. This means that somewhere in your mode you use a line mesh with quadratic interpolation (i.e. a 3-node line element).
Please, select all those edges and assign Linear interpolation and re-mesh
After this, you will see another error regarding the pile-cap coupling condition. I don't know what you are trying to do there, but the interaction you use is wrong
Re: mesh problem
Thank you for your reply! I will check my model.
Re: mesh problem
Dear STKO Team:
Thanks for the quick reply! I set it up as you suggested and re-mesh it, now it works fine and I can enter the analysis program.
1. As for the interaction issue you mentioned, I am using beamSolidCoupling for my pile-cap connection because I want to use a rigid connection between the pile and the bearing. What is wrong with this?
2. I want to simulate a cluster pile centrifuge shaker test, I set my analysis step according to the webinar section on SSI, but I get an error when running and it shows no convergence, can you help me to check that this is the correct way to set the analysis step? What should I do next?
Sorry I have no experience with other FEM software, so I will make mistakes in many of the most basic questions, thank you for your patience and answers!
Regards,
Charon
- Attachments
-
- 20230311114412.jpg (101.42 KiB) Viewed 5401 times
Last edited by Charon on Tue Mar 14, 2023 1:10 am, edited 1 time in total.
Re: mesh problem
There's a problem with the interaction you used for the beam-solid-coupling.
The master should be the beam-end-node (pile-top node)
The slave should be the face of the cap solid
However, since your cap is made of a single cube, there is only one face at the bottom. Doing this, the beam-solid-coupling constraint will make the whole face rigid.
Instead, you should divide it in such a way that you have sub-faces that correspond to the pile-cross section. That is the part that is rigid with the pile
The master should be the beam-end-node (pile-top node)
The slave should be the face of the cap solid
However, since your cap is made of a single cube, there is only one face at the bottom. Doing this, the beam-solid-coupling constraint will make the whole face rigid.
Instead, you should divide it in such a way that you have sub-faces that correspond to the pile-cross section. That is the part that is rigid with the pile