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 The information provided may be out of date.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.

Can ALE deal with such kind of moving and deformation of some interior boundary?

Posted 27 févr. 2011, 11:12 UTC−5 Version 4.0 13 Replies

Xiaoyi Yu
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Details please see the attachment, thank you!

Attachments:

13 Replies Last Post 17 juin 2011, 12:30 UTC−4
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 28 févr. 2011, 01:11 UTC−5
Hi

I would say that is what ALE is for, at least from what I understand of your image

--
Good luck
Ivar
Hi I would say that is what ALE is for, at least from what I understand of your image -- Good luck Ivar
Xiaoyi Yu
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 1 mars 2011, 09:05 UTC−5

Hi

I would say that is what ALE is for, at least from what I understand of your image

--
Good luck
Ivar


Thank you, Ivar!
As time going on, the deformation maybe become too large (a limitation of ALE), are there any good method to deal with this situation?
[QUOTE] Hi I would say that is what ALE is for, at least from what I understand of your image -- Good luck Ivar [/QUOTE] Thank you, Ivar! As time going on, the deformation maybe become too large (a limitation of ALE), are there any good method to deal with this situation?
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 1 mars 2011, 09:39 UTC−5
Hi

Yes: stop, remesh and restart. That feature was there in 3.5a not sure how its done in v4, havent used it so far, or perhaps it only comes in next release I do not know, check the doc

--
Good luck
Ivar
Hi Yes: stop, remesh and restart. That feature was there in 3.5a not sure how its done in v4, havent used it so far, or perhaps it only comes in next release I do not know, check the doc -- Good luck Ivar
Danial Tz
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 3 mars 2011, 12:20 UTC−5

Details please see the attachment, thank you!


I would assume it will become messy with large deformations. Can't you use Level-set methods?


--
Comsol 4.1
Ubuntu 10.04.1
[QUOTE] Details please see the attachment, thank you! [/QUOTE] I would assume it will become messy with large deformations. Can't you use Level-set methods? -- Comsol 4.1 Ubuntu 10.04.1
Xiaoyi Yu
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 4 mars 2011, 09:24 UTC−5


Details please see the attachment, thank you!


I would assume it will become messy with large deformations. Can't you use Level-set methods?


--
Comsol 4.1
Ubuntu 10.04.1


Thank you for you answer!
Yes, I have considered to use the Level -set method. But the Level set method in Comsol is designed for the multiphase flow, my model is solidification (phase change), and the moving interface is the phase change region. If I want to use the level set method (coupled with heat transfer), I need to know the heat flux difference across the moving interface and the temperature of the interface should be fixed at the melting temperature, I don't know how to realize these two points (or maybe comsol cannot do these). If you have any good suggestions, it would be appreciated greatly!
[QUOTE] [QUOTE] Details please see the attachment, thank you! [/QUOTE] I would assume it will become messy with large deformations. Can't you use Level-set methods? -- Comsol 4.1 Ubuntu 10.04.1 [/QUOTE] Thank you for you answer! Yes, I have considered to use the Level -set method. But the Level set method in Comsol is designed for the multiphase flow, my model is solidification (phase change), and the moving interface is the phase change region. If I want to use the level set method (coupled with heat transfer), I need to know the heat flux difference across the moving interface and the temperature of the interface should be fixed at the melting temperature, I don't know how to realize these two points (or maybe comsol cannot do these). If you have any good suggestions, it would be appreciated greatly!
Danial Tz
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 4 mars 2011, 10:55 UTC−5



Details please see the attachment, thank you!


I would assume it will become messy with large deformations. Can't you use Level-set methods?


--
Comsol 4.1
Ubuntu 10.04.1


Thank you for you answer!
Yes, I have considered to use the Level -set method. But the Level set method in Comsol is designed for the multiphase flow, my model is solidification (phase change), and the moving interface is the phase change region. If I want to use the level set method (coupled with heat transfer), I need to know the heat flux difference across the moving interface and the temperature of the interface should be fixed at the melting temperature, I don't know how to realize these two points (or maybe comsol cannot do these). If you have any good suggestions, it would be appreciated greatly!


I see. I am not sure that mass/energy is conserved along level-set boundaries in Comsol, but I can guess can be tough. In that sense, ALE is much better choice. If you have large deformations then you will need to stop the simulation and do a remesh (you can do this automatically via Matlab). It is possible and not a very big hassle. I remember there was such an example on comsol website (user models), but it works for old comsol v3.5a. At least you will get the idea.

Danial


--
Comsol 4.1
Ubuntu 10.04.1
[QUOTE] [QUOTE] [QUOTE] Details please see the attachment, thank you! [/QUOTE] I would assume it will become messy with large deformations. Can't you use Level-set methods? -- Comsol 4.1 Ubuntu 10.04.1 [/QUOTE] Thank you for you answer! Yes, I have considered to use the Level -set method. But the Level set method in Comsol is designed for the multiphase flow, my model is solidification (phase change), and the moving interface is the phase change region. If I want to use the level set method (coupled with heat transfer), I need to know the heat flux difference across the moving interface and the temperature of the interface should be fixed at the melting temperature, I don't know how to realize these two points (or maybe comsol cannot do these). If you have any good suggestions, it would be appreciated greatly! [/QUOTE] I see. I am not sure that mass/energy is conserved along level-set boundaries in Comsol, but I can guess can be tough. In that sense, ALE is much better choice. If you have large deformations then you will need to stop the simulation and do a remesh (you can do this automatically via Matlab). It is possible and not a very big hassle. I remember there was such an example on comsol website (user models), but it works for old comsol v3.5a. At least you will get the idea. Danial -- Comsol 4.1 Ubuntu 10.04.1
Xiaoyi Yu
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 4 mars 2011, 15:26 UTC−5




Details please see the attachment, thank you!


I would assume it will become messy with large deformations. Can't you use Level-set methods?


--
Comsol 4.1
Ubuntu 10.04.1


Thank you for you answer!
Yes, I have considered to use the Level -set method. But the Level set method in Comsol is designed for the multiphase flow, my model is solidification (phase change), and the moving interface is the phase change region. If I want to use the level set method (coupled with heat transfer), I need to know the heat flux difference across the moving interface and the temperature of the interface should be fixed at the melting temperature, I don't know how to realize these two points (or maybe comsol cannot do these). If you have any good suggestions, it would be appreciated greatly!


I see. I am not sure that mass/energy is conserved along level-set boundaries in Comsol, but I can guess can be tough. In that sense, ALE is much better choice. If you have large deformations then you will need to stop the simulation and do a remesh (you can do this automatically via Matlab). It is possible and not a very big hassle. I remember there was such an example on comsol website (user models), but it works for old comsol v3.5a. At least you will get the idea.

Danial


--
Comsol 4.1
Ubuntu 10.04.1


Thank you, Danial!
I searched the remesh example you said on comsol website but failed to find it. Do you still remember where it is ? Thank you!
[QUOTE] [QUOTE] [QUOTE] [QUOTE] Details please see the attachment, thank you! [/QUOTE] I would assume it will become messy with large deformations. Can't you use Level-set methods? -- Comsol 4.1 Ubuntu 10.04.1 [/QUOTE] Thank you for you answer! Yes, I have considered to use the Level -set method. But the Level set method in Comsol is designed for the multiphase flow, my model is solidification (phase change), and the moving interface is the phase change region. If I want to use the level set method (coupled with heat transfer), I need to know the heat flux difference across the moving interface and the temperature of the interface should be fixed at the melting temperature, I don't know how to realize these two points (or maybe comsol cannot do these). If you have any good suggestions, it would be appreciated greatly! [/QUOTE] I see. I am not sure that mass/energy is conserved along level-set boundaries in Comsol, but I can guess can be tough. In that sense, ALE is much better choice. If you have large deformations then you will need to stop the simulation and do a remesh (you can do this automatically via Matlab). It is possible and not a very big hassle. I remember there was such an example on comsol website (user models), but it works for old comsol v3.5a. At least you will get the idea. Danial -- Comsol 4.1 Ubuntu 10.04.1 [/QUOTE] Thank you, Danial! I searched the remesh example you said on comsol website but failed to find it. Do you still remember where it is ? Thank you!
Danial Tz
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 4 mars 2011, 15:48 UTC−5
You're right. It is not there anymore.

Danial

--
Comsol 4.1
Ubuntu 10.04.1
You're right. It is not there anymore. Danial -- Comsol 4.1 Ubuntu 10.04.1
Erwin Bielert
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 16 juin 2011, 04:49 UTC−4
What might be even more interesting is whether you can simulate a moving phase front which is "going around a corner" or passes through a T-junction or bifurcation. In that case the original boundary has to split up in two separate ones. Does anyone have an idea if it is possible to use the ALE method in those cases? (See similar picture as before in the attachment).

Thanks!

- Erwin
What might be even more interesting is whether you can simulate a moving phase front which is "going around a corner" or passes through a T-junction or bifurcation. In that case the original boundary has to split up in two separate ones. Does anyone have an idea if it is possible to use the ALE method in those cases? (See similar picture as before in the attachment). Thanks! - Erwin

Attachments:

Erwin Bielert
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 16 juin 2011, 04:49 UTC−4
What might be even more interesting is whether you can simulate a moving phase front which is "going around a corner" or passes through a T-junction or bifurcation. In that case the original boundary has to split up in two separate ones. Does anyone have an idea if it is possible to use the ALE method in those cases? (See similar picture as before in the attachment).

Thanks!

- Erwin
What might be even more interesting is whether you can simulate a moving phase front which is "going around a corner" or passes through a T-junction or bifurcation. In that case the original boundary has to split up in two separate ones. Does anyone have an idea if it is possible to use the ALE method in those cases? (See similar picture as before in the attachment). Thanks! - Erwin
Xiaoyi Yu
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 17 juin 2011, 09:35 UTC−4

What might be even more interesting is whether you can simulate a moving phase front which is "going around a corner" or passes through a T-junction or bifurcation. In that case the original boundary has to split up in two separate ones. Does anyone have an idea if it is possible to use the ALE method in those cases? (See similar picture as before in the attachment).

Thanks!

- Erwin


To my knowledge, ALE cannot deal with such kind of topological change problem.
[QUOTE] What might be even more interesting is whether you can simulate a moving phase front which is "going around a corner" or passes through a T-junction or bifurcation. In that case the original boundary has to split up in two separate ones. Does anyone have an idea if it is possible to use the ALE method in those cases? (See similar picture as before in the attachment). Thanks! - Erwin [/QUOTE] To my knowledge, ALE cannot deal with such kind of topological change problem.
Erwin Bielert
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 17 juin 2011, 09:41 UTC−4
I also got a reply from COMSOL support and they prefer to use level-set or phase-field in such a case. However, than you run into the difficulties you mentioned before. Have you solved the issues for the unknown heat flux difference across the moving interface and the fixed temperature of the interface? If you know how to implement these, it would also be of great help for me!

Thanks,

- Erwin
I also got a reply from COMSOL support and they prefer to use level-set or phase-field in such a case. However, than you run into the difficulties you mentioned before. Have you solved the issues for the unknown heat flux difference across the moving interface and the fixed temperature of the interface? If you know how to implement these, it would also be of great help for me! Thanks, - Erwin
Xiaoyi Yu
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 17 juin 2011, 12:30 UTC−4

I also got a reply from COMSOL support and they prefer to use level-set or phase-field in such a case. However, than you run into the difficulties you mentioned before. Have you solved the issues for the unknown heat flux difference across the moving interface and the fixed temperature of the interface? If you know how to implement these, it would also be of great help for me!

Thanks,

- Erwin


If you use the ale method, you can:
For the heat flux difference across the moving interface, you can use the down() or up() operators to calculate it.
For the fixed temperature of the interface, you can treat it as a boundary condition.

But for the level set or phase field, I still don't know how to make it.

By the way, are you simulating phase change model?
[QUOTE] I also got a reply from COMSOL support and they prefer to use level-set or phase-field in such a case. However, than you run into the difficulties you mentioned before. Have you solved the issues for the unknown heat flux difference across the moving interface and the fixed temperature of the interface? If you know how to implement these, it would also be of great help for me! Thanks, - Erwin [/QUOTE] If you use the ale method, you can: For the heat flux difference across the moving interface, you can use the down() or up() operators to calculate it. For the fixed temperature of the interface, you can treat it as a boundary condition. But for the level set or phase field, I still don't know how to make it. By the way, are you simulating phase change model?

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