Menu

Discussion Forum

New Discussion
Filtrer

Note: This discussion is about an older version of the COMSOL Multiphysics® software. The information provided may be out of date.

Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.

Swept Mesh Vs Traingular Mesh

Posted 29 nov. 2011, 13:50 UTC−5 RF & Microwave Engineering, Studies & Solvers Version 4.2a 3 Replies

Rui Silva
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Hi folks!

I am having a bit of a problem trying to get a simulation to converge.

Initially I used a triangular mesh with a boundary layer on a 3D pipe.
Later I realized that there were some numerical variations (some triangular elements were reversed)
on the boundary layer and after getting in touch with SUPPORT I was advised to use a SWEPT mesh.

The problem is that using the swept mesh for low flow velocities the simulation converges, but for higher is a no go.
And using the triangular mesh I get convergence in both low and high velocities.

I really need to get a convergence using the swept mesh. Anyone got any thoughts on this?

I have attached the files so its more clear.
Due to the files size I have cleared the mesh and solution.
Please mesh it again.

Best Regards,

Rui
3 Replies Last Post 29 mars 2012, 02:47 UTC−4
Rui Silva
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 30 nov. 2011, 09:39 UTC−5
Anyone?
Anyone?
Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 30 nov. 2011, 15:08 UTC−5
Hi

Another way is to mesh the input area with triangles, add a boundary layer on this face, and then sweep the "remaining" tube. your mesh elements might become qite "long" and not very square, hence have a poor quality, up to you to define a valid "dimension" node
But as you are studying turbulence, you must ensure your mesh is small enough. I'm no CFD expert so I'm not sure what final criteria to use for checking the appropriate mesh dimensions for such cases

--
Good luck
Ivar
Hi Another way is to mesh the input area with triangles, add a boundary layer on this face, and then sweep the "remaining" tube. your mesh elements might become qite "long" and not very square, hence have a poor quality, up to you to define a valid "dimension" node But as you are studying turbulence, you must ensure your mesh is small enough. I'm no CFD expert so I'm not sure what final criteria to use for checking the appropriate mesh dimensions for such cases -- Good luck Ivar
Glenston Miranda
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 29 mars 2012, 02:47 UTC−4

Hi

Another way is to mesh the input area with triangles, add a boundary layer on this face, and then sweep the "remaining" tube. your mesh elements might become qite "long" and not very square, hence have a poor quality, up to you to define a valid "dimension" node
But as you are studying turbulence, you must ensure your mesh is small enough. I'm no CFD expert so I'm not sure what final criteria to use for checking the appropriate mesh dimensions for such cases

--
Good luck
Ivar


Hi

I'm building this geometry and im trying to build the mesh for this.
The model's big so ive used swept meshes to reduce the no of elements.

I want to keep the no of elements to within 1 million to solve on a 16 GB system.

The default size is 1250 micrometer based on the minumum wavelength/6.
Been struggling with this for quite some time.

Any kind of help would be greatly appreciated.
Ive attached the model in this post.

Mapped1 and Swept 1 are for domain 8 which is one arm of the spiral. This mesh builds successfully.
Mapped2 and Swept 2 for domain 7 which is another arm of the spiral shows an error.
This is strange as both the arms are identical in geometry and in the mesh settings.

Regards
Glenston
[QUOTE] Hi Another way is to mesh the input area with triangles, add a boundary layer on this face, and then sweep the "remaining" tube. your mesh elements might become qite "long" and not very square, hence have a poor quality, up to you to define a valid "dimension" node But as you are studying turbulence, you must ensure your mesh is small enough. I'm no CFD expert so I'm not sure what final criteria to use for checking the appropriate mesh dimensions for such cases -- Good luck Ivar [/QUOTE] Hi I'm building this geometry and im trying to build the mesh for this. The model's big so ive used swept meshes to reduce the no of elements. I want to keep the no of elements to within 1 million to solve on a 16 GB system. The default size is 1250 micrometer based on the minumum wavelength/6. Been struggling with this for quite some time. Any kind of help would be greatly appreciated. Ive attached the model in this post. Mapped1 and Swept 1 are for domain 8 which is one arm of the spiral. This mesh builds successfully. Mapped2 and Swept 2 for domain 7 which is another arm of the spiral shows an error. This is strange as both the arms are identical in geometry and in the mesh settings. Regards Glenston

Attachments:

Note that while COMSOL employees may participate in the discussion forum, COMSOL® software users who are on-subscription should submit their questions via the Support Center for a more comprehensive response from the Technical Support team.

Suggested Content