Rectangular Waveguide Bend simulation in HFSS

[Emanuel Ricardo]

Hi

I am doing simulations for a rectangular waveguide bend using HFSS 15, in the band 20-100 GHz. In these frequencies, several modes can be created, when TE01 mode, the input,goes through the bend. I am using HFSS for the first time and i cannot find anything about my question:

I have the bend geometry and 2 waveports, one in the begining of the bend and one in the end. The boundary of the bend is made with copper. Te01 mode (the second mode of the rectangle) is introduced in one waveport using "edit sources", choosing only the mode i want, (am i do it right?). Here is the circular bend:



Because new modes can appear due the bend, when i will define the waveport do i need to insert the maximum number of modes ? (25). Or, because I just want to see what TE01 power i have in the exit of the bend and i dont want to know how this power was lost, i just need to define 2 modes in each waveport? (have to define 2 because TE01 is the second)

In another way, the number of modes that i define in the waveports have influence in the field solution inside the bend?




Thank you

 

[eraste]

Hi,

In fact, you can define the number of mode you want and after excite (with the edit sources) your bend (on one wave port : eg 1 or 2) with one mode. After simulation, you can plot the S parameters. Ex : S(1:1,2:1 ) who means transfer S parameter from port 1 to port 2 only for mode 1 (ie : TE10)
other examples : S(1:2,2:1) transfer from port 1 to port 2 for mode 2 (TE01) toward mode 1 (TE10).
or S(1:2,1:1) reflexion coefficient on port 1 for mode 2 toward mode 1.

Bye.

eraste

 

[Emanuel Ricardo]

Hi,

In fact, you can define the number of mode you want and after excite (with the edit sources) your bend (on one wave port : eg 1 or 2) with one mode. After simulation, you can plot the S parameters. Ex : S(1:1,2:1 ) who means transfer S parameter from port 1 to port 2 only for mode 1 (ie : TE10)
other examples : S(1:2,2:1) transfer from port 1 to port 2 for mode 2 (TE01) toward mode 1 (TE10).
or S(1:2,1:1) reflexion coefficient on port 1 for mode 2 toward mode 1.

Bye.

eraste

I want to know about TE01 mode, my point is keep its power after the bend. So i have to plot S(1:2,2:2). My problem is that when i define more modes in the ports, result are different. This makes sense? When i set the number of modes this changes the solution inside the bend? The source is the same, always TE01 mode.

 

[eraste]

Ok, it seems to be strange. But the reason is, perhaps, that HFSS consider only the mode you define in each wave port and consquently distribute the energy between them. So if you define only 2 modes at each ends of your waveguide the energy is distibuted only into this two modes if you increase the number of mode, energy is distribute for this new modes ...What is your frequency ? i suppose that you work between the cutoff frequencies of the second and third mode...So you will have only two excited fields.

Bye.

 

[Emanuel Ricardo]

Ok, it seems to be strange. But the reason is, perhaps, that HFSS consider only the mode you define in each wave port and consquently distribute the energy between them. So if you define only 2 modes at each ends of your waveguide the energy is distibuted only into this two modes if you increase the number of mode, energy is distribute for this new modes ...What is your frequency ? i suppose that you work between the cutoff frequencies of the second and third mode...So you will have only two excited fields.

Bye.

Thats the problem. As i said, the waveguide is oversized. In sources i always use te01 as input. And the result is different when i set the number of modes in the waveports.