I think you may have to do it using 2 ports, each operating with the TE01 mode. You may need to introduce a short septum between them, but you may not. You can excite them both at once, out of phase, I think. By TE02 I presume you mean the electric field across the long waveguide dimension, with one period across the short dimension.

You may need to excite each TE01 with a discrete source in its own half-width (height) waveguide section.

Actually, I think you can excite it by asking for 3 modes - the TE01 may be the third mode - and putting a grid of wires just after the port to short out all the TE1n modes.

Hello Arka,

Yes, it is possible. In the screenshot you showed you are selecting how many modes are calculated at the port, which is the right first step.

2 - You will need to perform a simulation with "calculate port modes only" activated so that CST calculates all possible modes (up to 12 or more in your case) including the one you are interested in.

3 - After the calculation of the port modes is finished, you can see how the modes look like at each of the ports in "2D/3D results - Port modes" to check that the mode you are looking for is exactly the number 12.

4 - Once you find it, remember the mode number, go to the solver configuration (time or frequency domain), and look for the section "Excitation".

There you can select which ports you want to excite and also which modes you would like to excite. If you select only one mode (for example, number 12), then CST will only excite that mode in the ports.

5 - Do not forget to uncheck "calculate port modes only" before simulating again so that CST now simulates in the whole volume.

I hope this answer helps you. Otherwise, please ask.

Regards,

Alejandro

Hello,

Yes, it is possible as you mentioned in your table. For any particular mode, use parametric analysis for any port. There are lot of option, in any solver.

One file is attached.

Thanks,

You can choose it from the solver setup.

Alejandro Jiménez Sáez: Many thanks! That helps a lot, it was really efficient.

Thanks to Ronis Maximidis too for depicting the same clearly!

PS: I noted that the excitation was possible only on one port though, not on both.

Jinfeng Li: Totally agree with your view, I was just wondering from reducing a problem complexity and simulation time point of view. Alejandro's idea helped in doing so.

In Frequency Domain if you choose Source Type -> List than press Selection, there you can actually tick/untick the modes/ports that you want to excite

Malcolm White Many thanks for your suggestions, as always. Yes, the TE02 field distribution is as you mentioned. I have not tried using a discrete port yet (shall try the same) however I did not clearly understand your suggestion: is it to align 2 discrete ports parallel to the longer dimension a at of heights b/4 and 3b/4?

Ronis Maximidis That's really great! Thank you very much!

That might work, with them in opposite directions but you might also need a conducting plate halfway between them for a short distance, but the conducting plate might not be needed because this is an equipotential surface when the discrete ports are in opposite phase. However, discrete ports will also excite higher order modes, and if you want these to be much smaller than the mode you want then a conducting plate between the ports will attenuate the higher order modes more. Choose the length of the plate so that there is still enough of the mode you want (it is also evanescent) but the higher order ones are attenuated even more, so that they are small enough..

May this page help to understand Malcolm's idea.