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.

Urgent help in Meshing required...

Amey Revandkar
Hi,
Im working on building a rf MEMS switch since past months. The big problem i'm facing is in user control meshing. i'm still not clear what to select under ''size of minimal element, Size of maximum element, Resolution of narrow regions''.
I'm finding extremely difficult to mesh structure as the minimum size of my part of the structure is as less as 0.1 micron and maximum as 80 micron.
Regards

10 Replies Last Post Oct 1, 2013, 5:29 PM EDT
Josh Thomas Certified Consultant
Posted: 4 years ago Aug 1, 2013, 1:17 PM EDT
Amey-

For thin structures, it is recommended to use mapped and/or swept meshing for these features. The free operations will give a less efficient mesh (if you happen to get no errors) as it seems you've noticed.

Best regards,
Josh Thomas
AltaSim Technologies
Amey- For thin structures, it is recommended to use mapped and/or swept meshing for these features. The free operations will give a less efficient mesh (if you happen to get no errors) as it seems you've noticed. Best regards, Josh Thomas AltaSim Technologies

Amey Revandkar
Posted: 4 years ago Sep 28, 2013, 4:41 AM EDT
Hi,
i am trying to use a physics induced mesh for the entire structure. But,after meshing i'm getting the following error and i'm struck badly in overcoming this.
The error is; 'Edge is much shorter than specified minimum element size and Face is much shorter than minimum specified element size'.
Please help me.
Hi, i am trying to use a physics induced mesh for the entire structure. But,after meshing i'm getting the following error and i'm struck badly in overcoming this. The error is; 'Edge is much shorter than specified minimum element size and Face is much shorter than minimum specified element size'. Please help me.

Josh Thomas Certified Consultant
Posted: 4 years ago Sep 30, 2013, 9:15 AM EDT
Amey-

The message you described is not actually a fatal error, just a warning. However, if you'd like to clear out the warning, simply define a user-controlled mesh and reduce the minimum element size parameter.

Best regards,
Josh Thomas
AltaSim Technologies
Amey- The message you described is not actually a fatal error, just a warning. However, if you'd like to clear out the warning, simply define a user-controlled mesh and reduce the minimum element size parameter. Best regards, Josh Thomas AltaSim Technologies

Fnu Mohammed Taqiuddin
Posted: 4 years ago Sep 30, 2013, 3:44 PM EDT
Hi Josh,

what if i cant use swept or mapped mesh?(thin layers curved and tapered) because of their limitations?
i have thin layer(thickness=0.101mm).
Hi Josh, what if i cant use swept or mapped mesh?(thin layers curved and tapered) because of their limitations? i have thin layer(thickness=0.101mm).

Josh Thomas Certified Consultant
Posted: 4 years ago Sep 30, 2013, 3:56 PM EDT
Fnu-

Well, apart from just using the free tetrahedral mesher:

1) You can try using the "Form Assembly" option when you finalize your Geometry. This will enable you to mesh your domains independently which may cut down on total degrees of freedom as well as enable you to mesh without errors. The cost here is that you have to define identity pairs where the boundaries touch between the different domains. This may cause other solving difficulties as well.

2) Depending on the physics of your problem, you can consider approximating the behavior of your thin 3D layers using boundary (2D) equations. This is only if important variables don't change much across the thickness of your problem.

Best regards,
Josh Thomas
AltaSim Technologies
Fnu- Well, apart from just using the free tetrahedral mesher: 1) You can try using the "Form Assembly" option when you finalize your Geometry. This will enable you to mesh your domains independently which may cut down on total degrees of freedom as well as enable you to mesh without errors. The cost here is that you have to define identity pairs where the boundaries touch between the different domains. This may cause other solving difficulties as well. 2) Depending on the physics of your problem, you can consider approximating the behavior of your thin 3D layers using boundary (2D) equations. This is only if important variables don't change much across the thickness of your problem. Best regards, Josh Thomas AltaSim Technologies

Amey Revandkar
Posted: 4 years ago Oct 1, 2013, 5:43 AM EDT
hi,
how much element size parameter i need to keep? that's the big problem,i have build my structure by rectangular domains on top of each other as if it would like front view of an fabricated switch model. I can do Modal analysis but fail in stationary analysis..i don't get beam displacement properly.
I used mapping meshing but error is, 'failed to create discretization of edge 6'. moreover when i do physics induced mesh i get warning,'face and edges are much smaller that specified minimum element size'.
kindly help,
Regards
hi, how much element size parameter i need to keep? that's the big problem,i have build my structure by rectangular domains on top of each other as if it would like front view of an fabricated switch model. I can do Modal analysis but fail in stationary analysis..i don't get beam displacement properly. I used mapping meshing but error is, 'failed to create discretization of edge 6'. moreover when i do physics induced mesh i get warning,'face and edges are much smaller that specified minimum element size'. kindly help, Regards

Fnu Mohammed Taqiuddin
Posted: 4 years ago Oct 1, 2013, 2:59 PM EDT
Hi Josh,
thanks for the reply

my physics is piezoelectric devices, i guess i cant use identical pair assembly approach.
I have attached my model file(v4.3b), if its possible for you to have a look and give some suggestions(meshing),it will be very helpful.

many thanks.
mohammed.
Hi Josh, thanks for the reply my physics is piezoelectric devices, i guess i cant use identical pair assembly approach. I have attached my model file(v4.3b), if its possible for you to have a look and give some suggestions(meshing),it will be very helpful. many thanks. mohammed.


Josh Thomas Certified Consultant
Posted: 4 years ago Oct 1, 2013, 3:38 PM EDT
Fnu-

It looks like you need to first consider that your geometry is not being created as you might anticipate. Your 2 imported files are not being unioned properly on the finalize node. I'd recommend performing the geometry Boolean operation in your native CAD package.

Also, have you considered by suggestion 2) above? You may consider not meshing in the thickness and instead using 2D boundary equations defined on the surfaces of the airfoil. I'm not an expert in the physics but it is worth considering.

Best regards,
Josh
Fnu- It looks like you need to first consider that your geometry is not being created as you might anticipate. Your 2 imported files are not being unioned properly on the finalize node. I'd recommend performing the geometry Boolean operation in your native CAD package. Also, have you considered by suggestion 2) above? You may consider not meshing in the thickness and instead using 2D boundary equations defined on the surfaces of the airfoil. I'm not an expert in the physics but it is worth considering. Best regards, Josh

Fnu Mohammed Taqiuddin
Posted: 4 years ago Oct 1, 2013, 5:15 PM EDT
Hi Josh,
thanks for reply.

i will try to re-import with Boolean geometry.

Also when you said '2D boundary equation defined on airfoil surface' can you name a example of that kind?
so that i can go through and try to implement on my model. coz if i can use that approach i can save lot of computation time and space.

many thanks.
mohammed.
Hi Josh, thanks for reply. i will try to re-import with Boolean geometry. Also when you said '2D boundary equation defined on airfoil surface' can you name a example of that kind? so that i can go through and try to implement on my model. coz if i can use that approach i can save lot of computation time and space. many thanks. mohammed.

Josh Thomas Certified Consultant
Posted: 4 years ago Oct 1, 2013, 5:29 PM EDT
Fnu-

This video gives an overview of how a film boiling layer was approximated using 2D equations.

www.comsol.com/video/heat-transfer-oil-quench-simulation

Regards,
Josh Thomas
AltaSim Technologies
Fnu- This video gives an overview of how a film boiling layer was approximated using 2D equations. http://www.comsol.com/video/heat-transfer-oil-quench-simulation Regards, Josh Thomas AltaSim Technologies

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.