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 The information provided may be out of date.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.
internal varialbe(rfwh.epsilonrxx)missing for silver (defined with polarization) in RF 3D in transient analysis in version4.0
Posted 18.01.2011, 10:41 GMT-5 RF & Microwave Engineering, Materials, Modeling Tools & Definitions, Parameters, Variables, & Functions Version 3.5a, Version 4.0 3 Replies
Please login with a confirmed email address before reporting spam
Since I have a silver metal film, I use multiphysics coefficient PDE combined with transient to include the drude model of silver.
Everything goes fine and there is no error in version3.5a. The strange thing is that if I do the same process in version4.0. It always give an error:
Failed to evaluate variable Jacobian.
- Variable: rfwh.epsilonrxx
- Geometry: 1
- Domain: 2 4 6 10
I checked the variables in the equation term of each domain and boudnary.
In the subdomain of silver, where I choose polarization Px, Py and Pz for the material, there is no definition of rfwh.epsilonrxx, rfwh.epsilonryy etc. If I choose refractive index instead of polarization, there is definition of rfwh.epsilonrxx again.
The thing iswhenscattering boundary condition is used for the outer boundary, the variable of rfwh.epsilonrxx... are needed . If the outer boudnary include the boudnary of silver, definition of rfwh.epsilonrxx is missing. This is why the error complains.
However, if I choose PEC, there is no error, because, no variable like rfwh.epsilonrxx are needed for PEC boundary.
In version3.5a, for the subdomain of silver, the definition of rfwh.epsilonrxx is still there although it seems COMSOL do not use it for silver. So everything goes well. I just guess in version4.0, they think the varialbe is not used and they simply delete it. But this will cause an error when the outer boundary of silver is set as scattering boundary condition.
Does anyone have the same problem?
Please login with a confirmed email address before reporting spam
Have you checked with 4.1?
It might be a conversion issue (if you imported the 3.5 file) or some internal weakness, the v4.0 version was rather "early" I did not use it for "serious" work, only for learning COMSOL à la v4 ;)
That was also what I understood from the COMSOL presentation of the Conference in 2009, announcing v4.0 in Milano, that serious work would start with 4.1, and since I have 4.1 I have no major issues (apart that everything is new, but I find easier and more powerfull.
--
Good luck
Ivar
Please login with a confirmed email address before reporting spam
We do not update our license from version4.0 to 4.1. We wondered whether we need the newest version.
What does the new version4.1 improve?
It would be very nice if there are any free test version4.1. Since there is no need for us to update the license if we find out the version4.1 still does not solve the above problem.
Can you help me how I can find out?
Please login with a confirmed email address before reporting spam
it solves a few bugs, including some errors in the ACDC physics formulas, and it improves the CFD meshing, certainly a few other things I havent even noticed :) It's installed in a few minutes
--
Good luck
Ivar