Discussion Forum

Please note, I tried setting the following in v4.0a:

Electric conductivity = 6e-5/1e-9
Thickness (found right under "Electric Conductivity" setting) = 1e-9

i.e. I thought by dividing the reciprocal of the surface impedance (reciprocal of surface impedance is, for me, 6e-5 Ohm) by the thickness (some number plucked out of the air, here it is 1 nm) I should return to the correct units for conductivity. By which I mean, I imagined that to get surface impedance it'd multiply the Elec conductivity by the Thickness and take the reciprocal. So I tried to trick it by giving it what it might need in order that in the end, the surface impedance comes out to 1/6e-5.

However, this gives me a strange result and not the same (correct) result as I got in comsol v3.5 when setting Surface Impedance = 1/6e-5, so I presume this is incorrect.

Hi

mostly if you have problems to understand the units, check the 2D - 3D dimensions and say to yourself that it applys to the mesh element of infinitesimal size dx*dy*dz. In 2D dz is replaced by 1[m] hence a division by [1/m] of your variable

--
Good luck
Ivar

Thanks Ivar. I thought about it a lot and contacted COMSOL. I appreciate your thoughts though.

Here is the correct answer, from COMSOL support email address:

============
The Transition Boundary Condition (TBC) in version 4 is not the same as in
version 3.5a. The TBC in 3.5a is valid for electrically thin layers, i.e.,
d << delta, for thickness d and skin depth delta. The TBC in version 4 is
extended to a more general form that holds also for layers that are not
electrically thin.

If you look at the equations for the TBC in the documentation, you see
that they reduce to the ones used in 3.5a when kd << 1. Unfortunately,
there is an error in the TBC which will not be fixed until version 4.2.
This error makes the TBC wrong for d << delta.

If you just want to use the same TBC as in 3.5a, you can add a surface
current feature on your layer boundary and define the surface current as Js
= 1/Z*tE, where Z is the surface impedance and tE is the tangential
electric field.

If you need more help correcting the TBC equations or setting up a surface
current boundary condition, please get back to me.
============

Hi

thanks good to know, luckily v4.2 is supposed to be just "around the corner" ;)

--
Good luck
Ivar

Is this limitation/bug also valid for the impedance boundary condition? I'm getting some strange results doing mode analysis on a microstrip using IBC, and this could explain what I'm seeing.

Could you ask Comsol support, as it is linked to your ongoing discussion with them?

Thanks,

Maarten.

could you please show how to set surface impedance using surface current boundary in detail?
I tried to modify its equation, but failed.
my modified equation in 2D simulation is:

emw.Jsx=sig1*emw.tEx
emw.Jsy=sig1*emw.tEy
emw.Jsz=sig1*emw.Ez
emw.Js0x=0
emw.Js0y=0
emw.Js0z=0


quote:
" If you just want to use the same TBC as in 3.5a, you can add a surface
current feature on your layer boundary and define the surface current as Js
= 1/Z*tE, where Z is the surface impedance and tE is the tangential
electric field."