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.

Which solver for time dependent : Electromagnetic Waves + Solid Mechanics

Jean-Christophe Lavocat

Please login with a confirmed email address before reporting spam

Dear all,

I am faced with a tricky problem for me, on which I cannot find a solution in the documentation. My model is representing a polymer that, subjected to some laser beam, expands in illuminanted area. So the geometry is changing with time : at t=0, the light is more intense in some places. t+1 the polymer deforms, and so induce a different light absorption pattern.

I thus have RF + Solid mechanics + Moving mesh modules. If I run a frequency analysis on the RF, I see the instensity of light being absorbed through the polymer. Then, if I run a stationnary study on the object, it deforms accordingly.

But both studies are not coupled, and I need to run 10 or more (who knows) studies like that (frenquency then stationnary) to get an idea of the real stationnary state.

My question is : is there a way to couple EMW and a moving structure with some study type? I discovered the "Frequency Stationnary" one, but it's only used with Joule heating, which is not the case here.

[Edit] : I read that : www.comsol.com/community/forums/general/thread/20275/ but couldn't find any positive answers.

6 Replies Last Post 2013年5月8日 GMT-4 18:19
Jean-Christophe Lavocat

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 2012年1月20日 GMT-5 13:06
When I try to run a time-dependent study only, the E field is not among the variable solved for ('Dependent variables).
If I try a time-discrete study only or a frequency+time dependent study, I can have the field E solved. But when I run these two, I get an error message like the following :

"Nonlinear solver did not converge.
Singular matrix.

There are 19958 void equations (empty rows in matrix) for the variable mod1.E10
at coordinates: (2.5e-06,3.4e-06) ...."

The coordinates correspond to the moving structure.

So my feeling is that the problem is not really coupled, and the EM module doesn't see the structure deformation.

Any idea?
When I try to run a time-dependent study only, the E field is not among the variable solved for ('Dependent variables). If I try a time-discrete study only or a frequency+time dependent study, I can have the field E solved. But when I run these two, I get an error message like the following : "Nonlinear solver did not converge. Singular matrix. There are 19958 void equations (empty rows in matrix) for the variable mod1.E10 at coordinates: (2.5e-06,3.4e-06) ...." The coordinates correspond to the moving structure. So my feeling is that the problem is not really coupled, and the EM module doesn't see the structure deformation. Any idea?

Andrew Prudil Nuclear Materials

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 2012年1月20日 GMT-5 15:10
I don't know how your problem is formulated but if you have equations for the strain caused by the EM field then you might want to try the following. Use a solid-mechanics model and select the option to have it deform the spatial frame. Use the initial stress and strain option on the linearly elastic material node to add the strains caused by the EM field.
Now, use another physics module to calculate the field. Make sure that it is set to use the spatial frame (which is the frame deformed by the solid-mechanics). Now solve with either an iterative solver or fully coupled solver.

Good luck. I Hope that helps.
I don't know how your problem is formulated but if you have equations for the strain caused by the EM field then you might want to try the following. Use a solid-mechanics model and select the option to have it deform the spatial frame. Use the initial stress and strain option on the linearly elastic material node to add the strains caused by the EM field. Now, use another physics module to calculate the field. Make sure that it is set to use the spatial frame (which is the frame deformed by the solid-mechanics). Now solve with either an iterative solver or fully coupled solver. Good luck. I Hope that helps.

Jean-Christophe Lavocat

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 2012年1月24日 GMT-5 04:26
Thanks Andrew. Your answer seems very interesting, but I could not find the options you are talking about.

Where can I find the "use the spatial frame ". I can see that in the post-processing options, but not in the Emw module. And how do I deform the spatial frame in the solid mechanics module?
Thanks Andrew. Your answer seems very interesting, but I could not find the options you are talking about. Where can I find the "use the spatial frame ". I can see that in the post-processing options, but not in the Emw module. And how do I deform the spatial frame in the solid mechanics module?

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 2012年1月24日 GMT-5 06:10
Hi

the full spatial frame x=X+u etc is turned on with the Solver Include Geometric non-linearities, in v4.2a
In previous versions of v4 (<42a) it was in the physics, at the end of linear, hyperleastic material tab

--
Good luck
Ivar
Hi the full spatial frame x=X+u etc is turned on with the Solver Include Geometric non-linearities, in v4.2a In previous versions of v4 (

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 2012年11月29日 GMT-5 11:01
Hi!

I have the 4.2a, but cannot find the option "Solver Include Geometric non-linearities". Where exactly is it?

tnx!
Hi! I have the 4.2a, but cannot find the option "Solver Include Geometric non-linearities". Where exactly is it? tnx!

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 2013年5月8日 GMT-4 18:19
Hi,

I have the same exact problem. The suggestion proposed by Andrew seems interesting but I can't figure how to do it. Can any of you guys post an example or elaborate on the solution?

Thank you.
Hossein
Hi, I have the same exact problem. The suggestion proposed by Andrew seems interesting but I can't figure how to do it. Can any of you guys post an example or elaborate on the solution? Thank you. Hossein

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.