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.

Boundary condition in 3D transient analysis of RF module

Please login with a confirmed email address before reporting spam

I am simulating a 3D waveguide geometry for observing nonlinear effects using 3D transient analysis of RF module. The problem is defining the electric field distribution at the the input boundary face of the waveguide.

In the RF model library, there is a case of gaussian beam propagation in such 3D waveguides where the input field is defined by the help of analytical expression of gaussian beams. But I have no such analytical expression for the particular guided mode that I want to simulate.

Nevertheless, I do know how the input field precisely looks like based on the results I obtained using "2D Perpendicular waves" analysis method. I am wondering if there is any way of exporting this result to "3D transient analysis" method in order to define the input field configuration? I will greatly appreciate if anyone could inform me if that is even possible or not. And in case it is, how.

Thanking in anticipation.

6 Replies Last Post 31 mar 2011, 15:31 GMT-4

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 18 feb 2010, 04:40 GMT-5
Ok, I might modify the question somewhat as I have moved to Comsol's scripting environment:

I have the numerically computed field distribution which I want to inject into one of my boundaries. How can I do this?

Again, if I had an analytical formula, I could have easily done this with the following two commands:

bnd.E0 = {{0;0;0},{0;0;0},{'Ex';'Ey';'Ez'}};
bnd.ind = [1,1,3,1,1,2,3,1,1,2,2,2,2,2,1,1];

Structure bnd above is a member of xfem structure containing information about the whole problem. In the first command, strings 'Ex', 'Ey' and 'Ez' denote analytical expressions already defined in the field of boundary expressions. The second command just injects this analytical expression into the boundary parts referred to by the number 3. Comsol's solver then evaluates these values in runtime on mesh points and pump in for transient analysis.

Now, I want to replace 'Ex', 'Ey' and 'Ez' with, say, arrays whose values correspond to the numerical value of these electric field components. Can anyone give me an idea on how to do this?

Ok, I might modify the question somewhat as I have moved to Comsol's scripting environment: I have the numerically computed field distribution which I want to inject into one of my boundaries. How can I do this? Again, if I had an analytical formula, I could have easily done this with the following two commands: bnd.E0 = {{0;0;0},{0;0;0},{'Ex';'Ey';'Ez'}}; bnd.ind = [1,1,3,1,1,2,3,1,1,2,2,2,2,2,1,1]; Structure bnd above is a member of xfem structure containing information about the whole problem. In the first command, strings 'Ex', 'Ey' and 'Ez' denote analytical expressions already defined in the field of boundary expressions. The second command just injects this analytical expression into the boundary parts referred to by the number 3. Comsol's solver then evaluates these values in runtime on mesh points and pump in for transient analysis. Now, I want to replace 'Ex', 'Ey' and 'Ez' with, say, arrays whose values correspond to the numerical value of these electric field components. Can anyone give me an idea on how to do this?

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 18 ago 2010, 07:55 GMT-4
Hi, maybe you can go to Options>Functions, there you can attach a file, by choosing analytic, piecewise analytic or interpolation, you can plug your value either analytically or numerically.

Hi, maybe you can go to Options>Functions, there you can attach a file, by choosing analytic, piecewise analytic or interpolation, you can plug your value either analytically or numerically.

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 18 ago 2010, 08:05 GMT-4
You can combine transient RF mode with boundary mode analyis by choosing RF Module>Boundary Mode analysis.
There is a similar tutorial model called 'waveguide adapter', maybe you can have a look.
I tried this previously, but I found that it is difficult to plug the numerical values of different boundaries at the same time. It seems it can only plug value from one boundary with homogeneous medium easily like what is done in the tutorial 'waveguide adapter'.


You can combine transient RF mode with boundary mode analyis by choosing RF Module>Boundary Mode analysis. There is a similar tutorial model called 'waveguide adapter', maybe you can have a look. I tried this previously, but I found that it is difficult to plug the numerical values of different boundaries at the same time. It seems it can only plug value from one boundary with homogeneous medium easily like what is done in the tutorial 'waveguide adapter'.

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 18 ago 2010, 09:03 GMT-4
Hi Jing,

Thank you for the response. This is an old question I had posted when I was beginning to use Comsol. Now I know how it is done and it is just the way you described.

I have not tried plugging values simultaneously from multiple boundaries but I shall be surprised if it is not possible.

Hi Jing, Thank you for the response. This is an old question I had posted when I was beginning to use Comsol. Now I know how it is done and it is just the way you described. I have not tried plugging values simultaneously from multiple boundaries but I shall be surprised if it is not possible.

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 29 mar 2011, 18:21 GMT-4
Hi Shakeeb and others,

I want to do time domain analysis of antenna which has to radiate gaussian short pulses (UWB).
Could you anyone advise about the boundary conditions for time domain analysis and to generate UWB pulses?

Thanks,
Kirthika,

Hi Shakeeb and others, I want to do time domain analysis of antenna which has to radiate gaussian short pulses (UWB). Could you anyone advise about the boundary conditions for time domain analysis and to generate UWB pulses? Thanks, Kirthika,

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 31 mar 2011, 15:31 GMT-4
Hi,

You need scattering boundaries to excite source. Remember that scattering boundaries are not able to perfectly absorb back reflections like PMLs. And until 3.5a, I know for a fact that PMLs are not available in transient module. May be it has changed in 4.x but I didn't see it mentioned in change logs.

All the best
Hi, You need scattering boundaries to excite source. Remember that scattering boundaries are not able to perfectly absorb back reflections like PMLs. And until 3.5a, I know for a fact that PMLs are not available in transient module. May be it has changed in 4.x but I didn't see it mentioned in change logs. All the best

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.