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 The information provided may be out of date.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.
Circular Polarization Simulation in Time Domain
Posted Jan 24, 2014, 12:50 p.m. EST Version 4.1 8 Replies
Please login with a confirmed email address before reporting spam
Hey,
I am doing a simulation of free space with circular polarization in time doamin.
I am using Comsol 4.1.
I read through some comment of previous discussions.
It looks like people are suggesting to do circular polarization by using two orthogonal electric fields with pi/2 phase delay, but in frequency domain and by using periodic boundary conditions for side boundaries and scattering for front and back..
This is my problem:
In Transient Electromagnetic Waves (temw) , I use periodic boundary conditions for side boundaries, and for front and back I am using scattering boundary conditions. Then I put Ex= cos(2*pi*f*t + pi/2) and Ey= cos(2*pi*f*t) as my source in one of scattering boundaries in 'Incident electric field' option under 'Wave given by' , where f is the frequency.
I am also using these time intervals t_start= 0, t_end= 20*T, time step= T/100. (20*T is large enough such that the wave should be able to travel whole length of free space of my model) where T is I/f. I should say that I also tried Ex= cos(2*pi*f*t + pi/2-k*z) and Ey= cos(2*pi*f*t-k*z) as the source definition.
I see that the electric field vectors are kind of circulating but the frequency does not look like right:
I placed two slices (plane probes) which are one wavelength far away from each other. But the electric filed does not circulate in phase on these slides which means the wavelength or frequency of the source is not right.
I did same simulation with linear polarization source either, but I have the wrong frequency problem still through checking the electric filed vectors one wavelength away from each other.
Now, I do not know what the problem is. Am I defining the circular polarization source wrong or is there any relationship between time steps and frequency and mesh?
To make sure about the frequency. I used this method of simulation for a resonating structure at its resonance frequency, but because the frequency is wrong, I did not see any expected result.
I appreciate if somebody which has this kind of experience could help me ?
I attached my file either in case.
I am doing a simulation of free space with circular polarization in time doamin.
I am using Comsol 4.1.
I read through some comment of previous discussions.
It looks like people are suggesting to do circular polarization by using two orthogonal electric fields with pi/2 phase delay, but in frequency domain and by using periodic boundary conditions for side boundaries and scattering for front and back..
This is my problem:
In Transient Electromagnetic Waves (temw) , I use periodic boundary conditions for side boundaries, and for front and back I am using scattering boundary conditions. Then I put Ex= cos(2*pi*f*t + pi/2) and Ey= cos(2*pi*f*t) as my source in one of scattering boundaries in 'Incident electric field' option under 'Wave given by' , where f is the frequency.
I am also using these time intervals t_start= 0, t_end= 20*T, time step= T/100. (20*T is large enough such that the wave should be able to travel whole length of free space of my model) where T is I/f. I should say that I also tried Ex= cos(2*pi*f*t + pi/2-k*z) and Ey= cos(2*pi*f*t-k*z) as the source definition.
I see that the electric field vectors are kind of circulating but the frequency does not look like right:
I placed two slices (plane probes) which are one wavelength far away from each other. But the electric filed does not circulate in phase on these slides which means the wavelength or frequency of the source is not right.
I did same simulation with linear polarization source either, but I have the wrong frequency problem still through checking the electric filed vectors one wavelength away from each other.
Now, I do not know what the problem is. Am I defining the circular polarization source wrong or is there any relationship between time steps and frequency and mesh?
To make sure about the frequency. I used this method of simulation for a resonating structure at its resonance frequency, but because the frequency is wrong, I did not see any expected result.
I appreciate if somebody which has this kind of experience could help me ?
I attached my file either in case.
Attachments:
8 Replies Last Post Jan 31, 2014, 1:03 p.m. EST
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Mohammad,
your model didn't show the solver configuration. If you used the default solver, then the time stepping in the time dependent solver was set to 'free'. This doesn't resolve the waveform. You must set 'steps taken by solver' to 'strict' and take care that the steps you set in the time dependent study resolve the waveform well enough. Your setting T/100 should be good enough.
Cheers
Edgar
--
Edgar J. Kaiser
emPhys Physical Technology
www.emphys.com
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Thank you Edgar,
Actually I just attached the structure and the file does not have solver now because this file has not been run.
The size of file after running was big to upload.
Anyway, I ran it already and still have the same problem.
Actually I just attached the structure and the file does not have solver now because this file has not been run.
The size of file after running was big to upload.
Anyway, I ran it already and still have the same problem.
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Did you change the time dependent solver setting to strict? Here is the model with strict time stepping and an additional point in the center of the prism.
The result looks ok for me. I did it in 4.4 because I don't have 4.1 available on my machine
Cheers
Edgar
--
Edgar J. Kaiser
emPhys Physical Technology
www.emphys.com
Attachments:
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Thank you Edgar,
I can not open the file you attached because I am using 4.1 version.
I do not know how I can change the time dependent solver settings.
What I do, I start a short simulation. Then when the solver configuration is made under study I go and change from there. I did this and I got good results.
Is there any other option to change it without doing a short simulation first?
And what does it mean "additional point in the center of prism"? where can I add additional points.
I appreciate if you help me to learn these.
Thanks,
Mohammad
I can not open the file you attached because I am using 4.1 version.
I do not know how I can change the time dependent solver settings.
What I do, I start a short simulation. Then when the solver configuration is made under study I go and change from there. I did this and I got good results.
Is there any other option to change it without doing a short simulation first?
And what does it mean "additional point in the center of prism"? where can I add additional points.
I appreciate if you help me to learn these.
Thanks,
Mohammad
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Thank you Edgar,
I can not open the file you attached because I am using 4.1 version.
I do not know how I can change the time dependent solver settings.
What I do, I start a short simulation. Then when the solver configuration is made under study I go and change from there. I did this and I got good results.
Is there any other option to change it without doing a short simulation first?
And what does it mean "additional point in the center of prism"? where can I add additional points.
I appreciate if you help me to learn these.
Thanks,
Mohammad
I can not open the file you attached because I am using 4.1 version.
I do not know how I can change the time dependent solver settings.
What I do, I start a short simulation. Then when the solver configuration is made under study I go and change from there. I did this and I got good results.
Is there any other option to change it without doing a short simulation first?
And what does it mean "additional point in the center of prism"? where can I add additional points.
I appreciate if you help me to learn these.
Thanks,
Mohammad
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
ok, good to hear your model is now doing what you want. You can always get the default solver sequence by right-clicking on the Study node and choosing accordingly.
I sometimes add points into the geometry to have them available as entities in the post processing. There are other options like point probes or cut elements.
Cheers
Edgar
--
Edgar J. Kaiser
emPhys Physical Technology
www.emphys.com
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Thanks Edgar.
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi Edgar,
I have another problem in simulation by using both periodic and port boundary conditions in the structure.
I made another post for this issue. I have this problem for like a year without any solution.
I appreciate if you can help me for this issue.
Here is its link:
www.comsol.com/community/forums/general/thread/41949/
I have another problem in simulation by using both periodic and port boundary conditions in the structure.
I made another post for this issue. I have this problem for like a year without any solution.
I appreciate if you can help me for this issue.
Here is its link:
www.comsol.com/community/forums/general/thread/41949/