Discussion Forum

Physically, a GHz wave will be very strongly attenuated in water. So you have nothing to calculate.

Another thing: you need to use a circular port, not rectangular.

DWGreve
DWGreve Consulting

Thank you very much for your reply. when i used circular port i had an error. I want just to see the transmission loss.
Please, what do you mean by strongly attenuated. any ideas about value of insertion loss?

> when i used circular port i had an error.

Comsol doesn't make it obvious how to do this. AFTER you select circular port, you need to right-click the port node and select Points > Circular port reference axis. THEN you can define the polarization direction of the waveguide mode- because, after all, any direction gives a valid waveguide mode for a circular waveguide.

By the way, there are interesting things that happen if you choose different reference axes on the input and output.

> what do you mean by strongly attenuated

Water is somewhat conductive. In the GHz range the skin depth in water is of the order of cm. (This is why your baked potato gets cooked on the inside). So a water-filled waveguide will have essentially no transmission if many cm in length.

Further, water (at least Comsol water) has a permittivity of 80, thus a rather short wavelength. So unless you have many elements per wavelength you are guaranteed to have an erroneous result.

Another issue, incidentally, is the skin depth of iron. This is far less than a pipe thickness. It is probably better to use the impedance boundary condition for the walls if you want to understand the attenuation due to wall loss.

DWGreve

Thank you very much for your reply and explications, I will try all this options and recommendations.