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.

Moving Mesh

Please login with a confirmed email address before reporting spam

I am trying to simulate a Electromagnetic Generator 3D version but I cant do the geometry move. In the 2D example, i saw how the mesh and the geometry move together. I defined a assembly and the pair between stator and rotor and put all the parameters on the Physics but I just get movement on the mesh while the geometry still freeze...What is the problem here?

Thanks in advance


Leo

11 Replies Last Post Apr 19, 2012, 3:27 p.m. EDT

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Jan 28, 2010, 11:20 a.m. EST
Hi Leonardo,
You should wait for an expert to answer, but I think it is normal. It was the same for me and if you look at some Comsol Model from the library, you'll see that the geometry doesn't move.

Ronan
Hi Leonardo, You should wait for an expert to answer, but I think it is normal. It was the same for me and if you look at some Comsol Model from the library, you'll see that the geometry doesn't move. Ronan

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Jan 28, 2010, 12:46 p.m. EST
Thank you Roman for your answer...I got your point and I agree with you that the logic says to move the mesh and not the geometry...but my doubt appears when i was studying the 2D generator example because (as i can see or in my oppinion)...the geometry is rotating..!!!!!!

Leo
Thank you Roman for your answer...I got your point and I agree with you that the logic says to move the mesh and not the geometry...but my doubt appears when i was studying the 2D generator example because (as i can see or in my oppinion)...the geometry is rotating..!!!!!! Leo

Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Jan 29, 2010, 2:26 a.m. EST
Hi

I can only advise to go carefully through the examples, step for step and ensure you understand what is happening. If you model does not behave as you expect, I suppose you have missed something in the treatmant of the "Frames", this is delicate and confusing, at least it was for for me, in the beginning. Moving parts means also using the "assembly" option correctly and managing the "identity pairs" fluently.

The examples and several models are excellent teachers, but I agree there are many windows to check and it demands some excercice, but it's absolutely worth it

There are also a couple of discussion on the forum around this subject with working and non-working examples, try a "search"

Good luck
Ivar
Hi I can only advise to go carefully through the examples, step for step and ensure you understand what is happening. If you model does not behave as you expect, I suppose you have missed something in the treatmant of the "Frames", this is delicate and confusing, at least it was for for me, in the beginning. Moving parts means also using the "assembly" option correctly and managing the "identity pairs" fluently. The examples and several models are excellent teachers, but I agree there are many windows to check and it demands some excercice, but it's absolutely worth it There are also a couple of discussion on the forum around this subject with working and non-working examples, try a "search" Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Jan 29, 2010, 9:33 a.m. EST
Thank u Ivar...I will follow your recommendations, I noticed that there is an option on Postprocessing Menus to avoid the geometry while your are plotting any result....this could be a good solution for my case...anyway...on any electromagnetic problem the geometry is not really important...

Leo
Thank u Ivar...I will follow your recommendations, I noticed that there is an option on Postprocessing Menus to avoid the geometry while your are plotting any result....this could be a good solution for my case...anyway...on any electromagnetic problem the geometry is not really important... Leo

Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Jan 29, 2010, 10:17 a.m. EST
Hi

Yes there are many options there, among other to select from which "frame" you are postprocessing, this is important with ALE or with moving parts.

In anycase I forgot to say that as user of COMSOL I highly appreciate the COMSOL training courses, as they gave me a kick-start for each new subject I have followed, I have even take one twice, with a year and a half interval, with two different professors, each equally good, but I had evolved too, and my questions and understanding allowed me each time to gain new knowledge on how to best use COMSOL. So if you have the chance to get on to one of these, or just the shorter ones at the COMSOL yearly conference it is worth it.

Good luck
Ivar
Hi Yes there are many options there, among other to select from which "frame" you are postprocessing, this is important with ALE or with moving parts. In anycase I forgot to say that as user of COMSOL I highly appreciate the COMSOL training courses, as they gave me a kick-start for each new subject I have followed, I have even take one twice, with a year and a half interval, with two different professors, each equally good, but I had evolved too, and my questions and understanding allowed me each time to gain new knowledge on how to best use COMSOL. So if you have the chance to get on to one of these, or just the shorter ones at the COMSOL yearly conference it is worth it. Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Feb 1, 2010, 7:27 a.m. EST
Hi,

If you Provide more details about your problem it might be helpful for us to understand which kind of problem you are facing.
Also, I could not understand the point of "the mesh is moving but the geometry no"!!!
Your mesh is embedded in your geometry and delimited by the boundaries of your geometry. If your Boundaries are moving => everything should move!!!

Hope it helps

Cheers
Hi, If you Provide more details about your problem it might be helpful for us to understand which kind of problem you are facing. Also, I could not understand the point of "the mesh is moving but the geometry no"!!! Your mesh is embedded in your geometry and delimited by the boundaries of your geometry. If your Boundaries are moving => everything should move!!! Hope it helps Cheers

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago May 20, 2010, 8:14 p.m. EDT
I had the same problem. All you have to do is go to Results/Data Sets/Solution "x", where "x" is the solution set that you are plotting from. In the Frame dropdown box, make sure that you are using the "Spatial" frame. The default is the "Material" frame, which means that the fields, when plotted, are not using the rotated coordinate values. Thus, they appear not to move. Hopefully this helps.

Cheers!
I had the same problem. All you have to do is go to Results/Data Sets/Solution "x", where "x" is the solution set that you are plotting from. In the Frame dropdown box, make sure that you are using the "Spatial" frame. The default is the "Material" frame, which means that the fields, when plotted, are not using the rotated coordinate values. Thus, they appear not to move. Hopefully this helps. Cheers!

Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago May 21, 2010, 2:30 a.m. EDT
Hi

I can only recommend: read through the doc (it's long but there are a lot oof material), in V4 ther is the emphasis on the different "frames": spatial, material, mesh ... (naming has slightly changed too from 3.5) this is related to the Eulerian or Lagrangian representations (see en.wikipedia.org/wiki/Lagrangian_and_Eulerian_coordinates), and to the use or not of ALE or moving meshes.

Do not forget that COMSOL is attempting to be universal for all physics and all of its representations, the developpers are doing a tremendous job to unify this, as by habits we used different approaches for historical reasons, what is really the difference of solid mechanics and fluidics ? It's the same basic underlying equations.

There are complements in the sme_ug.pdf, mems_ug.pdf, cfd_ug.pdf ans specially the mph_ug.pdf ... of V4

... the Navier-Stokes equations are solved in the spatial (deformed) frame while the solid mechanics physics is defined in the material (undeformed) frame, a transformation ...

Read also the section on "Frames" p549 mph_ug.pdf V4

its also about the subtilities of "d/dt" and "D/dt" common to the fluidics experts

1st important rule: understand clearly the differences of the different frames, and when to use which !
it's not innocent on your results. This expalins also the change of names from small "x" to uppar case "X" on "ux" and "uX" between V3.5 and V4 (plus "m" for mesh)

Have fun Comsoling
Ivar
Hi I can only recommend: read through the doc (it's long but there are a lot oof material), in V4 ther is the emphasis on the different "frames": spatial, material, mesh ... (naming has slightly changed too from 3.5) this is related to the Eulerian or Lagrangian representations (see http://en.wikipedia.org/wiki/Lagrangian_and_Eulerian_coordinates), and to the use or not of ALE or moving meshes. Do not forget that COMSOL is attempting to be universal for all physics and all of its representations, the developpers are doing a tremendous job to unify this, as by habits we used different approaches for historical reasons, what is really the difference of solid mechanics and fluidics ? It's the same basic underlying equations. There are complements in the sme_ug.pdf, mems_ug.pdf, cfd_ug.pdf ans specially the mph_ug.pdf ... of V4 ... the Navier-Stokes equations are solved in the spatial (deformed) frame while the solid mechanics physics is defined in the material (undeformed) frame, a transformation ... Read also the section on "Frames" p549 mph_ug.pdf V4 its also about the subtilities of "d/dt" and "D/dt" common to the fluidics experts 1st important rule: understand clearly the differences of the different frames, and when to use which ! it's not innocent on your results. This expalins also the change of names from small "x" to uppar case "X" on "ux" and "uX" between V3.5 and V4 (plus "m" for mesh) Have fun Comsoling Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago May 25, 2010, 2:33 p.m. EDT
Hi,

I just want to add some additional comments on what was already mentioned:

First of all, you have to avoid this confusing understanding between the "geometry" and the "mesh": This means that your object your or geometry is only one part, one block. You can even retrieve the geometry from the mesh!! Your geometry is delimited by edges or boundaries which are at the same time part of your mesh at a certain level.
Moving the mesh is directly related to the collective motion of your boundaries.
This is just to clarify the point!!

You can differentiate between the "geometry edges" : fixed boundaries which represent the initial geometry, and the "edges", which represents the moving one. You can find this in >plot parameters > plot type. This is can be found for 2D or 3D also.

The second point is, if you said that your mesh is moving but not your geometry. This is a bit confusing!! following what was said above this means that your boundaries are not moving but your bulk mesh does!!?? in that case you will get inverted mesh as warning!!?

I think you have to check again the conditions on your boundaries first. I am 100% sure that something is wrong with your boundaries.
Another thing is that you have to be sure that you are using the ALE frame not the reference. This can be found also on >plot parameters > plot type.


Keep us informed about the progress and I hope it helps,

Cheers

Hi, I just want to add some additional comments on what was already mentioned: First of all, you have to avoid this confusing understanding between the "geometry" and the "mesh": This means that your object your or geometry is only one part, one block. You can even retrieve the geometry from the mesh!! Your geometry is delimited by edges or boundaries which are at the same time part of your mesh at a certain level. Moving the mesh is directly related to the collective motion of your boundaries. This is just to clarify the point!! You can differentiate between the "geometry edges" : fixed boundaries which represent the initial geometry, and the "edges", which represents the moving one. You can find this in >plot parameters > plot type. This is can be found for 2D or 3D also. The second point is, if you said that your mesh is moving but not your geometry. This is a bit confusing!! following what was said above this means that your boundaries are not moving but your bulk mesh does!!?? in that case you will get inverted mesh as warning!!? I think you have to check again the conditions on your boundaries first. I am 100% sure that something is wrong with your boundaries. Another thing is that you have to be sure that you are using the ALE frame not the reference. This can be found also on >plot parameters > plot type. Keep us informed about the progress and I hope it helps, Cheers

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Oct 29, 2011, 10:40 a.m. EDT
I 'd like to inquire from users who are experienced with mesh movement whether 'mesh movement' will help me in meeting my purpose, which is as follows:

I'm modelling a bunch of cilia(hair like structures....(like those cilia that line the inside of you nose)...these cilia are attached to the wall and stay submersed in a gel-like fluid......this ENTIRE stucture, consisting of the cilia and the gel-like fluid that it is inside form the POROUS LAYER.....HOWEVER, THE MAIN CATCH IS THAT each cilia has its own independent SWAYING movement....so the oscillation of the entire porous layer depends on the SUM-TOTAL of movements of the cilia.
Note also that the cilia do not translate.....their bottom remain fixed to the wall, only top portions sway.

As far as I understand, mesh movement as done on comsol will give movement of the mesh as an ENTIRETY...like a blob of mass, which is deformed....how will I be able to account for the movement of INDIVIDUAL cilia within the porous medium?
I 'd like to inquire from users who are experienced with mesh movement whether 'mesh movement' will help me in meeting my purpose, which is as follows: I'm modelling a bunch of cilia(hair like structures....(like those cilia that line the inside of you nose)...these cilia are attached to the wall and stay submersed in a gel-like fluid......this ENTIRE stucture, consisting of the cilia and the gel-like fluid that it is inside form the POROUS LAYER.....HOWEVER, THE MAIN CATCH IS THAT each cilia has its own independent SWAYING movement....so the oscillation of the entire porous layer depends on the SUM-TOTAL of movements of the cilia. Note also that the cilia do not translate.....their bottom remain fixed to the wall, only top portions sway. As far as I understand, mesh movement as done on comsol will give movement of the mesh as an ENTIRETY...like a blob of mass, which is deformed....how will I be able to account for the movement of INDIVIDUAL cilia within the porous medium?

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Apr 19, 2012, 3:27 p.m. EDT
In your response you mention examples. Where can I find the examples? It would be preferable if the examples came with the PDF files. That way we can unnderstand what is going on in the examples.
In your response you mention examples. Where can I find the examples? It would be preferable if the examples came with the PDF files. That way we can unnderstand what is going on in the examples.

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.