Note: This discussion is about an older version of the COMSOL Multiphysics® software. The information provided may be out of date.
Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.
Comsol 4.2a, RF MODULE, Scatering field and Meshing
Posted 29 sept. 2012, 16:22 UTC−4 RF & Microwave Engineering Version 4.2a 2 Replies
Please login with a confirmed email address before reporting spam
Dear All,
I have module with one half sphere glass, one half sphere air.
I use scattering field. Now I have some doubt.
I found for version 3.5 that some people used for first half of sphere equation E0*exp(-j*k0_rfw*z*nglass), and for second half of sphere (air) they used E0*exp(-j*k0_rfw*z). k0_rfw wave number in vacuum, nglass refractive index of glass. They combined these two equations in one equation with "sign" function. They have do all these things because of total internal reflection between two surfaces with different refractive indices.
Now I have two questions.
1. What is the best yo use in 4.2a? Is it necessary to dothis in 4.2a version, or is it just enough to put E0*exp(-j*emw.k0*z), and everything is ok (becuase I allready defined glass and air like dielectric functions with relative dielectric constants 2.3 and 1). So my question does this problem with total internal reflection exist in 4.2a.
2. When I mesh this geometry, if I put for the air meshing max element size for example S, am I need to put max element size for glass S/nglass or S*nglass, or something like that, or is it not necessary?
Did any of you compare this cases??
Thank you very much on every response.
I have module with one half sphere glass, one half sphere air.
I use scattering field. Now I have some doubt.
I found for version 3.5 that some people used for first half of sphere equation E0*exp(-j*k0_rfw*z*nglass), and for second half of sphere (air) they used E0*exp(-j*k0_rfw*z). k0_rfw wave number in vacuum, nglass refractive index of glass. They combined these two equations in one equation with "sign" function. They have do all these things because of total internal reflection between two surfaces with different refractive indices.
Now I have two questions.
1. What is the best yo use in 4.2a? Is it necessary to dothis in 4.2a version, or is it just enough to put E0*exp(-j*emw.k0*z), and everything is ok (becuase I allready defined glass and air like dielectric functions with relative dielectric constants 2.3 and 1). So my question does this problem with total internal reflection exist in 4.2a.
2. When I mesh this geometry, if I put for the air meshing max element size for example S, am I need to put max element size for glass S/nglass or S*nglass, or something like that, or is it not necessary?
Did any of you compare this cases??
Thank you very much on every response.
2 Replies Last Post 30 sept. 2012, 11:17 UTC−4
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi
you are right one can easily get confused, I mostly check my variables and intermediate reasults by plotting them out and verifying separately. It is easy to define the index in the material AND in the formulas, while here there should be an exclusive OR else, you double up.
So make a simple model and check if the beahviour is correct the way you have defined your materials, if not you might nee to correct the formulas.
For the mesh, indeed one should have a max mesh size of at most <1/3 (better <1/5) of a wavelength IN THE MATTER and since c0 = n_index * lambda_wavelength * frequency, and that c0 and frequency remains invariant, the higher index materials have a shorter lambda, hence needs a denser mesh (I hope I got this right ;)
--
Good luck
Ivar
you are right one can easily get confused, I mostly check my variables and intermediate reasults by plotting them out and verifying separately. It is easy to define the index in the material AND in the formulas, while here there should be an exclusive OR else, you double up.
So make a simple model and check if the beahviour is correct the way you have defined your materials, if not you might nee to correct the formulas.
For the mesh, indeed one should have a max mesh size of at most <1/3 (better <1/5) of a wavelength IN THE MATTER and since c0 = n_index * lambda_wavelength * frequency, and that c0 and frequency remains invariant, the higher index materials have a shorter lambda, hence needs a denser mesh (I hope I got this right ;)
--
Good luck
Ivar
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hello Ivar,
Thank you very much on your answer, especially because doubts concerning meshing I had.
I will use your approach, apply this on some simple module, and go step by step.
Just help me if you can with something.
Probably you are familiar with modules with TIP above substrate, antennas on the substrate, etc...
I have two questions.
1. Do you think there is enough good to use Johnson and Christi values for dielectric functions of material,
or is it better to use Drude model or Lorentz expansion model, which are maybe better adapted
for FDTD simulations.
2. What do you think about situation when is antenna placed on the substrate. Is it better put antenna directly on the substrate with fine meshing, or is it maybe better to place it for example 2-3nm above substrate?
In any case thank you very much for first answer.
Best regards,
Nikola
Thank you very much on your answer, especially because doubts concerning meshing I had.
I will use your approach, apply this on some simple module, and go step by step.
Just help me if you can with something.
Probably you are familiar with modules with TIP above substrate, antennas on the substrate, etc...
I have two questions.
1. Do you think there is enough good to use Johnson and Christi values for dielectric functions of material,
or is it better to use Drude model or Lorentz expansion model, which are maybe better adapted
for FDTD simulations.
2. What do you think about situation when is antenna placed on the substrate. Is it better put antenna directly on the substrate with fine meshing, or is it maybe better to place it for example 2-3nm above substrate?
In any case thank you very much for first answer.
Best regards,
Nikola