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 The information provided may be out of date.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.

Transmission Coefficientof Electromagnetic Wave in Frequency Domain

Mohammad Parvinnezhad Hokmabadi

Please login with a confirmed email address before reporting spam

Hi,
I am doing a simulation of free space. My Comsol version is 4.1
My goal is to find transmission and reflection coefficients in order to determine phase of reflection or transmission.
I know how to find them by using 'Lumped Port' at front and back boundaries of the space, since I can access to scattering parameters S11, and S12, ... . But in this method of simulation, it looks like I have to use PEC and PMC at side boundaries in order to have a definition of polarization of incident filed.
Now, this is the problem:
I am going to use periodic boundary at side boundaries with linear polarization and later I will apply circular polarization for my next steps. Therefore 'Lumped Port' is not useful.
I saw that some people do simulation with 'Port' at front and back boundaries. The good point of this is that in 'Port' boundaries I can define circular polarization or any kind of polarizations. I am trying to do same simulation as they do, but my simulation makes error. It does not matter whether I use PEC, PMC boundaries or Periodic boundaries or even circular polarization or linear. In any case I am not able to simulate. It makes the same error always. I am not even sure if I can access to scattering parameters in this 'Port' method of simulation in order to calculate phase of transmitted and reflected waves.
This the error:
"Failed to find a solution.
Singular matrix.
For mesh-case 0 there are 1 void equations (empty rows in matrix) for the variable mod1.emw.S2x
at coordinates: (0,0,0), ...
and similarly for the degrees of freedom (empty columns in matrix)
Returned solution has not converged."

I appreciate if somebody can help me. I attached my file either in case if you would like to look at that and let me know what is wrong with my simulation.


5 Replies Last Post 2014年2月14日 GMT-5 17:54
Luke Gritter Certified Consultant

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 2014年2月3日 GMT-5 11:12
Mohammad,

You have not specified the port mode settings for Port 2, and this is causing the error. If you enter an appropriate eigenmode and propagation constant, you should get a solution. You could also use a scattering boundary condition instead of Port 2, in which case the software would still calculate the value of the S11 parameter.

--
Luke Gritter
AltaSim Technologies
Mohammad, You have not specified the port mode settings for Port 2, and this is causing the error. If you enter an appropriate eigenmode and propagation constant, you should get a solution. You could also use a scattering boundary condition instead of Port 2, in which case the software would still calculate the value of the S11 parameter. -- Luke Gritter AltaSim Technologies

Mohammad Parvinnezhad Hokmabadi

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 2014年2月3日 GMT-5 12:12
Thank you Luke,
Actually you are right. I tried it on the weekend and it works now.
I have one more question. I am using Ey = E0*exp(-j*emw.k0*z) with beta=emw.k0, Ex = 0, and Ez = 0 with beta =0 in E field options of port1 such that incident wave flows into free space in my model. I am using exactly the same E field for port2. So, it looks like the incident field comes from just port 1. Is this right?
And if this is right and then I embed a structure in my model, will S11 and S21 parameters be reflection and transmission coefficients of the structure with these settings of ports?

Thank you so much,
Mohammad
Thank you Luke, Actually you are right. I tried it on the weekend and it works now. I have one more question. I am using Ey = E0*exp(-j*emw.k0*z) with beta=emw.k0, Ex = 0, and Ez = 0 with beta =0 in E field options of port1 such that incident wave flows into free space in my model. I am using exactly the same E field for port2. So, it looks like the incident field comes from just port 1. Is this right? And if this is right and then I embed a structure in my model, will S11 and S21 parameters be reflection and transmission coefficients of the structure with these settings of ports? Thank you so much, Mohammad

Mohammad Parvinnezhad Hokmabadi

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 2014年2月4日 GMT-5 11:38

Mohammad,

You have not specified the port mode settings for Port 2, and this is causing the error. If you enter an appropriate eigenmode and propagation constant, you should get a solution. You could also use a scattering boundary condition instead of Port 2, in which case the software would still calculate the value of the S11 parameter.

--
Luke Gritter
AltaSim Technologies


Thank you Luke,
Actually you are right. I tried it on the weekend and it works now.
I have one more question if you do not mind. I am using Ey = E0*exp(-j*emw.k0*z) with beta=emw.k0, Ex = 0, and Ez = 0 with beta =0 in E field options of port1 such that incident wave flows into free space in my model. I am using exactly the same E field for port2. So, it looks like the incident field comes from just port 1. Is this right?
And if this is right and then I embed a structure in my model, will S11 and S21 parameters be reflection and transmission coefficients of the structure with these settings of ports?

Thank you so much,
Mohammad
[QUOTE] Mohammad, You have not specified the port mode settings for Port 2, and this is causing the error. If you enter an appropriate eigenmode and propagation constant, you should get a solution. You could also use a scattering boundary condition instead of Port 2, in which case the software would still calculate the value of the S11 parameter. -- Luke Gritter AltaSim Technologies [/QUOTE] Thank you Luke, Actually you are right. I tried it on the weekend and it works now. I have one more question if you do not mind. I am using Ey = E0*exp(-j*emw.k0*z) with beta=emw.k0, Ex = 0, and Ez = 0 with beta =0 in E field options of port1 such that incident wave flows into free space in my model. I am using exactly the same E field for port2. So, it looks like the incident field comes from just port 1. Is this right? And if this is right and then I embed a structure in my model, will S11 and S21 parameters be reflection and transmission coefficients of the structure with these settings of ports? Thank you so much, Mohammad

Luke Gritter Certified Consultant

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 2014年2月4日 GMT-5 14:00
Mohammad,

Yes, there is only an incident field entering through the port if "Wave excitation at this port" is set to "On", regardless of what is entered under the Port Mode Settings.

If you have two ports in a model with excitation at Port 1, the software will calculate S21 and S11. For a transmission line, this would correspond to transmission and reflection, respectively. Whether these properly correspond to transmission and reflection of your structure may depend on the details and implementation of your model.

--
Luke Gritter
AltaSim Technologies
Mohammad, Yes, there is only an incident field entering through the port if "Wave excitation at this port" is set to "On", regardless of what is entered under the Port Mode Settings. If you have two ports in a model with excitation at Port 1, the software will calculate S21 and S11. For a transmission line, this would correspond to transmission and reflection, respectively. Whether these properly correspond to transmission and reflection of your structure may depend on the details and implementation of your model. -- Luke Gritter AltaSim Technologies

Mohammad Parvinnezhad Hokmabadi

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 2014年2月14日 GMT-5 17:54

Mohammad,

Yes, there is only an incident field entering through the port if "Wave excitation at this port" is set to "On", regardless of what is entered under the Port Mode Settings.

If you have two ports in a model with excitation at Port 1, the software will calculate S21 and S11. For a transmission line, this would correspond to transmission and reflection, respectively. Whether these properly correspond to transmission and reflection of your structure may depend on the details and implementation of your model.

--
Luke Gritter
AltaSim Technologies


Hey Luke,

Thank you so much for your answer.
I have another question if you do not mind.
I am going to find S21 and S11 parameters of a Split Ring Resonator in order to calculate phase of reflection and transmission. I used Lumped Port with PEC-PMC boundaries and also Port with either PBC or PEC_PMC. For lumped port I am using
Type of Port = Uniform
Terminal Type: Cable
In All methods one of the terminal ports is on and the other is off.
I calculated and plotted the phase and absolute values of scattering parameters.
All look so similar, except the phase of S21 parameter of Lumped Port method which has almost 3 units offset than the others. Therefore I am not sure which one is right.
I included a ppt file including all simulations required information and the results.
I appreciate if you take a look at that ppt file and inform me what is the issue and which method is right for finding refelction and transmission coefficients. If other information is required or I am doing wrong simulation please let me know.

Thanks
Mohammad
[QUOTE] Mohammad, Yes, there is only an incident field entering through the port if "Wave excitation at this port" is set to "On", regardless of what is entered under the Port Mode Settings. If you have two ports in a model with excitation at Port 1, the software will calculate S21 and S11. For a transmission line, this would correspond to transmission and reflection, respectively. Whether these properly correspond to transmission and reflection of your structure may depend on the details and implementation of your model. -- Luke Gritter AltaSim Technologies [/QUOTE] Hey Luke, Thank you so much for your answer. I have another question if you do not mind. I am going to find S21 and S11 parameters of a Split Ring Resonator in order to calculate phase of reflection and transmission. I used Lumped Port with PEC-PMC boundaries and also Port with either PBC or PEC_PMC. For lumped port I am using Type of Port = Uniform Terminal Type: Cable In All methods one of the terminal ports is on and the other is off. I calculated and plotted the phase and absolute values of scattering parameters. All look so similar, except the phase of S21 parameter of Lumped Port method which has almost 3 units offset than the others. Therefore I am not sure which one is right. I included a ppt file including all simulations required information and the results. I appreciate if you take a look at that ppt file and inform me what is the issue and which method is right for finding refelction and transmission coefficients. If other information is required or I am doing wrong simulation please let me know. Thanks Mohammad

Note that while COMSOL employees may participate in the discussion forum, COMSOL® software users who are on-subscription should submit their questions via the Support Center for a more comprehensive response from the Technical Support team.