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.
How to simulate the propagation of particular mode in a waveguide
Posted 14.10.2013, 01:56 GMT-4 RF & Microwave Engineering Version 5.1 7 Replies
Please login with a confirmed email address before reporting spam
Hello.
I have rectangular metallic stripe embedded in homogeneous dielectric as a plasmonic waveguide. I have performed a mode analysis for this waveguide at particular frequency (the wavelength is 632nm) and found the modes which are supported by this waveguide. Now I want to simulate how one of these modes propagate through the waveguide.
To do this I have created a simple 2-port structure - the metallic stripe embedded in homogeneous dielectric of 1um thickness from each side (see the attached picture, blue is the waveguide, red is PLM layer). The waveguide is terminated by 2 ports at each transverse cross cuts.
Now I want to simulate the propagation of one the the modes. I have specified the propagation constant from the previous simulation at ports settings and run the simulation.
Then I have noticed that no matter which propagation constant is specified the results are the same. What can be wrong in the model? Maybe I should specify something else to excite the particular mode in this waveguide? Thank you
I have rectangular metallic stripe embedded in homogeneous dielectric as a plasmonic waveguide. I have performed a mode analysis for this waveguide at particular frequency (the wavelength is 632nm) and found the modes which are supported by this waveguide. Now I want to simulate how one of these modes propagate through the waveguide.
To do this I have created a simple 2-port structure - the metallic stripe embedded in homogeneous dielectric of 1um thickness from each side (see the attached picture, blue is the waveguide, red is PLM layer). The waveguide is terminated by 2 ports at each transverse cross cuts.
Now I want to simulate the propagation of one the the modes. I have specified the propagation constant from the previous simulation at ports settings and run the simulation.
Then I have noticed that no matter which propagation constant is specified the results are the same. What can be wrong in the model? Maybe I should specify something else to excite the particular mode in this waveguide? Thank you
Attachments:
7 Replies Last Post 29.06.2016, 17:56 GMT-4
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Evgeniy,
I would use two numeric ports (excitation and passive) and set two Boundary Mode Analysis steps (for each port) followed by Frequency Domain step. You might not need PMLs if dielectric domain is sufficiently large compared to width of metal strip.
Regards,
Sergei
I would use two numeric ports (excitation and passive) and set two Boundary Mode Analysis steps (for each port) followed by Frequency Domain step. You might not need PMLs if dielectric domain is sufficiently large compared to width of metal strip.
Regards,
Sergei
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Thank you for reply.
Yes, I've done exactly the same things for mode analysis. For example the picture A.png shows one of the modes of this waveguide (for this mode the top and bottom dielectrics are different). This is plot of emw.normEbm_1 for excitation port 1. Now I want to study the propagation of this mode not in transverse but also in longitudinal directions. To do this I am using the same model. I have specified port 1 as 'user defined', wave excitation 'on', Pin '1W', electric field Eo(0,0,1) V/m, propagation constant as was found from the mode analysis (picture A.png). After that I ran the simulation (Frequency Domain) at the same frequency freq=c_const/lam0 where lam0=632[nm].
The results are shown on figures waveguide2 and waveguide3. Yes, my mode was excited as could be seen from transverse cross cut, but the wave also propagates through the top and bottom dielectric. But they shouldn't because I have specified the propagation constant for only one mode. What could cause such a result?
Yes, I've done exactly the same things for mode analysis. For example the picture A.png shows one of the modes of this waveguide (for this mode the top and bottom dielectrics are different). This is plot of emw.normEbm_1 for excitation port 1. Now I want to study the propagation of this mode not in transverse but also in longitudinal directions. To do this I am using the same model. I have specified port 1 as 'user defined', wave excitation 'on', Pin '1W', electric field Eo(0,0,1) V/m, propagation constant as was found from the mode analysis (picture A.png). After that I ran the simulation (Frequency Domain) at the same frequency freq=c_const/lam0 where lam0=632[nm].
The results are shown on figures waveguide2 and waveguide3. Yes, my mode was excited as could be seen from transverse cross cut, but the wave also propagates through the top and bottom dielectric. But they shouldn't because I have specified the propagation constant for only one mode. What could cause such a result?
Attachments:
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Evgeniy,
Not sure I understand how you can excite correct mode using “User defined” port. In my mind, “Numeric” type of port should be used to excite proper mode found by “Boundary Mode Analysis” step, as shown in the attached image. Some trial and error is usually required to set up correct transformation shift in the Boundary Mode Analysis step in order to excite desired mode.
Regards,
Sergei
Not sure I understand how you can excite correct mode using “User defined” port. In my mind, “Numeric” type of port should be used to excite proper mode found by “Boundary Mode Analysis” step, as shown in the attached image. Some trial and error is usually required to set up correct transformation shift in the Boundary Mode Analysis step in order to excite desired mode.
Regards,
Sergei
Attachments:
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Thank you very much for your response.
I have a last question. When I perform a mode analysis I receive a good mode electric field distribution for the waveguide embedded in homogeneous dielectric. But when I try to do the same with inhomogeneous claddings (the waveguide on silicon substrate with silica on the top) I still receive the mode profile but with "noise" (see the pictures) in dielectric with higher permittivity (silicon). How can I avoid such a "noise"??
I have a last question. When I perform a mode analysis I receive a good mode electric field distribution for the waveguide embedded in homogeneous dielectric. But when I try to do the same with inhomogeneous claddings (the waveguide on silicon substrate with silica on the top) I still receive the mode profile but with "noise" (see the pictures) in dielectric with higher permittivity (silicon). How can I avoid such a "noise"??
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Evgeniy,
I can see two possibilities:
1.Somehow you hit the radiating mode and solution is not a noise but radiating field (similar to microstrip antenna).
2. Solution is artifact caused by the presence of infinite silicon/silica interface. It is a challenging task to model infinite interface. If you are still using PMLs then solution is quite questionable, since PMLs are designed to perform properly if they enclose domain with homogeneous material (but not different materials). In this situation, I would enclose silicon and silica layers by air domain and study the effect of silicon and silica layer dimensions on the propagating mode. This might give insight on what is going on.
Regards,
Sergei
I can see two possibilities:
1.Somehow you hit the radiating mode and solution is not a noise but radiating field (similar to microstrip antenna).
2. Solution is artifact caused by the presence of infinite silicon/silica interface. It is a challenging task to model infinite interface. If you are still using PMLs then solution is quite questionable, since PMLs are designed to perform properly if they enclose domain with homogeneous material (but not different materials). In this situation, I would enclose silicon and silica layers by air domain and study the effect of silicon and silica layer dimensions on the propagating mode. This might give insight on what is going on.
Regards,
Sergei
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
I have tried to simulate the structure with and without PML layers. The strange thing is the difference between the results. While the real part of effective mode index is the same the imaginary part differs around 7 times - the effective mode index is 1.547-0.0036i and 1.549-0.0252i for the model with and without PML layer respectively. Moreover the mode analysis depends on the size of surrounding dielectric. I am not talking about the sizes less than the wavelength, but about the difference between surrounding area of 1 and 3 microns. I thought that it shouldn't have any effect on the results.
Taking into account all above I began to doubt about all results. What is the right model in this case?
Taking into account all above I began to doubt about all results. What is the right model in this case?
Please login with a confirmed email address before reporting spam
Posted:
9 years ago
Sergei
Can I do the similar process with a 2D waveguide structure?
Can I get high order modes at boundary mode analysis (at port), and simulate how the high order modes propagate?
Thanks
Can I do the similar process with a 2D waveguide structure?
Can I get high order modes at boundary mode analysis (at port), and simulate how the high order modes propagate?
Thanks
Evgeniy,
I would use two numeric ports (excitation and passive) and set two Boundary Mode Analysis steps (for each port) followed by Frequency Domain step. You might not need PMLs if dielectric domain is sufficiently large compared to width of metal strip.
Regards,
Sergei