Discussion Forum

Hello Felix Vega

Your Discussion has gone 30 days without a reply. If you still need help with COMSOL and have an on-subscription license, please visit our Support Center for help.

If you do not hold an on-subscription license, you may find an answer in another Discussion or in the Knowledge Base.

Has anyone solved this problem for you? I have the same problem.

Hi

you have to check your equations, you are referring to a variable in a circularway and your equations cannot solve, most oftn this happens if you use a COMSOL defined variable for your own, ceck the equations you are using and your variable names, something is wrong in there ;)

--
Good luck
Ivar

I have similar difficulties. Simple section of coax using PEC boundaries and two lumped 50 ohm ports, one on either end (default settings) with one having wave excitation turned on. The model is 2D axis symmetric, basically the domain is a simple skinny rectangle offset from the rotation axis. When I went to solve, The error COMSOL 4 reports is :
Circular variable dependency detected
-Variable mod1.temw.tEphi
-Geometry:1
-Boundary: 2

I have no clue as to what the problem might be, and there are no examples to follow as COMSOL does not include any transient electromagnetic examples...

Does any one know what the problem is?

Hi

Again its probably not something generic, but somethin specific to your model, take a careful look on everything defined on boundary 2, the equations and the COMSOL equations for the BC's interacting with boundary 2. And if you have two distinct ports, are they defined with two different identifiers ?

--
Good luck
Ivar

Ivar,
Thank you for your reply. The two ports had distinct numeric port names, and the following settings:

Boundaries: 2
Port name:1
Type of Port: Coaxial
Terminal Type:Cable
Wave excitation:On
Voltage Vo:1 [V]
Port phase:0 [rad]
Characteristic impedance: 50 [ohm]

Boundaries: 3
Port name:2
Type of Port: Coaxial
Terminal Type:Cable
Wave excitation:Off
Characteristic impedance: 50 [ohm]

The equations used for transient electromagnetics is given on page 102 of the RF Userguide:
muo*sigma*d/dt(A)+muo*sigmao*d/dt(epsilonr*A) + Del X (1/mur Del X A) = 0

I'm assuming Comsol uses the definitions for the lumped ports to come up with suitable boundary conditions, but I don't know how/where to check which variables in the boundary definition might be clashing and or how to change them appropriately? One would assume that for a coaxial port Ephi should be zero...

Hi

I'm not yet really used to the new equations/naming convcention for V4 port physics so I'm not sure what to say, but when I have such problems, I look inside the equation sub-nodes to understand what COMSOL is applying underneath, mostly that gives me the clues for what is wrong.

Also switch to 4.1 as soon as possible, its more stable

--
Good luck
Ivar

Ivar,
Thanks for the reply. It seems there is indeed a lumped port issue for the time dependent simulation addressed an update from COMSOL.

The update 4.0.0.993 list of fixes states:

Radio Frequency

* Lumped port boundary condition now works correctly in time-dependent studies
* Corrected solver sequence for Microwave Heating
* Loss tangent constitutive relation in the Electromagnetic Waves interface now works correctly

So I assume that was the problem. Thanks for the discussion :)

I'm not sure if this thread is still alive or relevant anymore, but I have just encountered the same problem while modeling Reaction Engineering. The circular variable error came because I have used the notation of a parameter the same like the model uses (as Ivar already said).

Just to be clear, I took a screenshot.