Integration line (waveguide)
I have a rectangular waveguide with dimensions a > b.
Inside I want to have the first mode (TE10 Mode) excited.
_______
|______|<--b
.....^a
Is there a difference in the result between defining the integration line parallel to the a-side, or parallel to the b-side?
Regards!
Yes, the way you define the integration line changes the results.
Since you are using a rectangular waveguide with dimensions a > b. For TE10 mode, the Electric field distribution from the lower broader side wall to the upper broader side wall.
Integration line should be always from E-wall to E-wall. So in this case, the integration line should be parallel to b-side.
Small tip: Even if you don't define the integration line when defining the waveports, that's OK. HFSS will take care of that.
Are you using HFSS for your simulation ?
Thanks. I will remove the integration line and try it again.
Yes I want to simulate the waveguide.
In which way does the way of defining a integration line affect the simulation results?
Because as I have found in the help, it seems that Zpi impendance is not concerning for the integration line but for the area of the wave port. (from: hfss help, "Calculating Characteristic Impedance")
Zpi does not need an integration line as it integrates H around the perimeter of the port. However, if you choose Zpv or Zvi the integration line WILL matter! For example, if in the above waveguide you define your integration line along the long axis of the cross section you will get a very different impedance than if you chose a line along the short axis...why? well, for the TE10 mode (the fundamental) the E fields will be oriented parallel to the b side. If this is the case a line integral of E (for V) perpendicular to this line will NOT be equal to a line integral along the E field. V will actually be zero as the line integral is completely perpendicular to E. So the integration line can matter, but just keep Zpi as it is pretty robust and does not require an integration line unless you want to define an initial field polarization.
have fun :)
Thank you for the help.
I will choose Zpi and do not worry about integration line at this level.