Learning Forum

[Nev]

Hi all,

I have measured some permittivity and permeability values of my materials which I am hoping to import into Ansys, and I have set up databases with the values in to enable me to calculate it in the program using pwl($e1,Freq)-1i*pwl($e2,Freq), however I am very confused over whether or not to use a minus or plus between the real and imaginary parts! Using basic construction of complex numbers, I would assume I use a plus, but everywhere I look for electromagnetics people use -j, but then say it's just a sign convention and is arbitrary. However, changing the equation in Ansys gives wildly different answers if I use one to another, so it can't be that arbitrary!

If anyone can please help me out here and advise which I should use that would be tremendously helpful! I am very much out of my physics depth here (I'm a chemist by background and stumbled into computationally deriving EM), and was hoping to just get a complex permittivity value out of my kit but alas was not to be!

Thanks so much :)

[Nev]

Hi everyone So, I've been looking a bit more into this and it seems as though it all boils down to the sign convention of the forward propagating wave and whether or not Ansys expects the imaginary part of the permittivity to be a negative value (I think!). My results measured in the lab have a positive e'', which I guess means I need to use e*=e'+ie'', right?
Thanks!

[Nev]

Ah wait! OK, so I appreciate this is turning into more of just me talking to myself, but I always think it's nice when someone posts the answer to their own question for anyone trying to tread the same steps in the future, so here goes! I think the answer to all my issues was found in this paper https://www.imperial.ac.uk/media/imperial-college/faculty-of-engineering/electrical-and-electronic-engineering/public/optical-and-semiconductor-devices/pubs/2012_09_TTST.pdf, which essentially states "Now, it is not unambiguously stated that all input parameters must be real numbers, including relative permittivity and bulk conductivity , in order for the code to calculate e(eff), which is the only parameter needed by the solver". So, for my data (measured on an EpsiMu coaxial cell) which essentially gives me the real and imaginary parts of the relative effective permittivity, I believe all I need to input is the real part of those measurements.
If anyone wants to jump in and say this is all heinously wrong, please feel free! But hopefully this will help other people in the future who have been similarly pulling their hair our in confusion and desperation!

[Praneeth]

Hi Nev,

We apprecitate for your contribution to the learning forum and keep posting such information whenever possible.

Best regards,
Praneeth.