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 The information provided may be out of date.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.

ERROR: FAILED TO EVALUATE VARIABLE JACOBIAN, Variable: Freq

Please login with a confirmed email address before reporting spam

Hi,

I am getting the following error:

Failed to evaluate variable Jacobian.
- Variable: freq
- Geometry: 1
- Domain: 1 2 3
- Feature: Stationary Solver 2 (sol2/s2);
- Error: Failed to evaluate variable Jacobian.

I perform a frequency domain study (Study 1: Frequency Domain) and I calculate the electric field inside a solar cell (due to the 'light') with an 'Electromagnetic Waves, Frequency Domain (emw)' node.

The divergence of the Poynting vector:

div(S) = (d(emw.Poavx, x)+d(emw.Poavy, y)+d(emw.Poavz, z))

is then used for a second study (Study 2: Step 2: Stationary 2), via a function G = div(S)/hf, where h and f are constants. In the second study the drift-diffusion model for semiconductors is implemented and G is the generation term used in the 'Transport of Diluted Species' reaction term.

WHEN I try to calculate Study 2 I get the previous error. There must be some problem with the expression of div(S). However I can plot div(S) without any problems.

When instead of G = div(S)/hf I use another expression that does not include div(S) the simulation converges.

I have already tried to obtain a smooth electric field (Study 1: Frequency Domain) by selecting a Cubic discretization and very fine mesh (not in the attached model), however I cannot improve the situation.

Do you know how to fix this problem?

Thanks in advance for the comments


6 Replies Last Post 16.01.2013, 15:33 GMT-5

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 14.01.2013, 07:54 GMT-5
just a small correction, not included in the model:

div(S) = abs(d(emw.Poavx, x)+d(emw.Poavy, y)+d(emw.Poavz, z))
just a small correction, not included in the model: div(S) = abs(d(emw.Poavx, x)+d(emw.Poavy, y)+d(emw.Poavz, z))

Sergei Yushanov Certified Consultant

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 14.01.2013, 08:31 GMT-5
Samuele,

Define global parameter “freq=c_const/lambda_1” and re-run Study 2.

Regards,
Sergei
Samuele, Define global parameter “freq=c_const/lambda_1” and re-run Study 2. Regards, Sergei

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 14.01.2013, 15:33 GMT-5
Thanks a lot. That works! :-) Do you know why? Is that a bug?
Thanks a lot. That works! :-) Do you know why? Is that a bug?

Sergei Yushanov Certified Consultant

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 14.01.2013, 16:09 GMT-5
Kind of bug.
See Ivar's explanation
www.comsol.com/community/forums/general/thread/33767/
Kind of bug. See Ivar's explanation http://www.comsol.com/community/forums/general/thread/33767/

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 15.01.2013, 00:53 GMT-5
Hi

I find it rather logical, that COMSOL defines only required variables when needed. But OK when you combine several solvers, what is needed by the first is not necessarily required by the second one. COMSOL is a GUI upona programming language so some rules must be learned ;)

--
Good luck
Ivar
Hi I find it rather logical, that COMSOL defines only required variables when needed. But OK when you combine several solvers, what is needed by the first is not necessarily required by the second one. COMSOL is a GUI upona programming language so some rules must be learned ;) -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 16.01.2013, 15:33 GMT-5
Thanks both :-)
Thanks both :-)

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.