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 The information provided may be out of date.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.

Need help with Meshing

Sanjay Mukuntha Madhava

Please login with a confirmed email address before reporting spam

Hi Everyone,
I need to simulate the deflection of a Membrane with Internal Stress in it. I have attached the file here and the problem is it take too long to execute and i believe the problem is with meshing. Can someone please check it and let me know if there is a better way to mesh the model.

Thanks in Advance
Sanjay


7 Replies Last Post 21 févr. 2012, 11:06 UTC−5
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 21 févr. 2012, 08:05 UTC−5
Hi

Indeed your aspect ratio is extreme. I would suggest that you mesh the top in "triangles", then use a sweep mesh with 3 or more elements in the tickness.

Now your short beams will take all the stress at their sharp corners, ideally you should have at least 3 elements acros the width of the beams. One way is to split them with a layer, or tu set up the "Size custom" Resolution of narow regions = 3.5 or so, and check that the min, max sizeas are coherent

--
Good luck
Ivar
Hi Indeed your aspect ratio is extreme. I would suggest that you mesh the top in "triangles", then use a sweep mesh with 3 or more elements in the tickness. Now your short beams will take all the stress at their sharp corners, ideally you should have at least 3 elements acros the width of the beams. One way is to split them with a layer, or tu set up the "Size custom" Resolution of narow regions = 3.5 or so, and check that the min, max sizeas are coherent -- Good luck Ivar

Sanjay Mukuntha Madhava

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 21 févr. 2012, 08:18 UTC−5
Thanks for your reply but Can you please elaborate your solution further?
I get this error when i try to follow your instructions.

Regards
Sanjay
Thanks for your reply but Can you please elaborate your solution further? I get this error when i try to follow your instructions. Regards Sanjay


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 21 févr. 2012, 08:30 UTC−5
Hi

are you sure you got all the top surfaces, also the small ones of the beams ?

--
Good luck
Ivar
Hi are you sure you got all the top surfaces, also the small ones of the beams ? -- Good luck Ivar

Sanjay Mukuntha Madhava

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 21 févr. 2012, 08:31 UTC−5
Yes i guess and also am new to this (was using the Quad for the entire geometry till now) and its not helping either .
Also can i run upto 4 Parametric sweeps at the sametime (width, length, stress and number of beams in this case)

Thanks and Regards
Sanjay
Yes i guess and also am new to this (was using the Quad for the entire geometry till now) and its not helping either . Also can i run upto 4 Parametric sweeps at the sametime (width, length, stress and number of beams in this case) Thanks and Regards Sanjay

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 21 févr. 2012, 09:39 UTC−5
Hi
Another point I did not catch was your initial stress value, how are you sure about the stress tensor ?

Have you though about modelling only a section using symmetry ?

Anyhow, snall beams like that have high stress concetration at the sharp corners => needs small mesh => require time to solve, so spending a little time to reduce the model complexity can globally gain a lot of time ;)

--
Good luck
Ivar
Hi Another point I did not catch was your initial stress value, how are you sure about the stress tensor ? Have you though about modelling only a section using symmetry ? Anyhow, snall beams like that have high stress concetration at the sharp corners => needs small mesh => require time to solve, so spending a little time to reduce the model complexity can globally gain a lot of time ;) -- Good luck Ivar

Sanjay Mukuntha Madhava

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 21 févr. 2012, 09:44 UTC−5
The stress is between -100 MPa to + 100 MPa, later i'd have to expand it to 1 GPa .
I wasnt aware that only a section can be modelled. And are there any tutorials for the same.
And finally are there any study material for Meshing that you would recommend :)

Thanks and Regards
Sanjay
The stress is between -100 MPa to + 100 MPa, later i'd have to expand it to 1 GPa . I wasnt aware that only a section can be modelled. And are there any tutorials for the same. And finally are there any study material for Meshing that you would recommend :) Thanks and Regards Sanjay

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 21 févr. 2012, 11:06 UTC−5
Hi

I'm not aware of anything special apart testing out the model library examples and learn from there. Otherwise COMSOL organises regularly some excellent courses, also on advanced meshing

But check out your initial stress description, and test out if there is not a more relaistic way to introduce your stress than "just" play with a few diagonal tensor elements, specially since Si is normally crystal rahter anisotropic !

--
Good luck
Ivar
Hi I'm not aware of anything special apart testing out the model library examples and learn from there. Otherwise COMSOL organises regularly some excellent courses, also on advanced meshing But check out your initial stress description, and test out if there is not a more relaistic way to introduce your stress than "just" play with a few diagonal tensor elements, specially since Si is normally crystal rahter anisotropic ! -- Good luck Ivar

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.