Discussion Forum

Is this an RF or electrostatics problem? Are you talking about a point RF source or a point charge? If it is a stationary point charge, then this is an electrostatics problem. If so, you should specify a small circle (instead of a point) and put a constant potential on it. Then you'll get a radial E distribution.

Thanks for your reply. Its an RF problem. I have attached an image below how I want it to be.

I've attached an example.

Hi Dear sir its very useful file. But once i simulated then its field is inverted as compared to yours plot. Then i needed to put -1A to get same like plot of yours, whereas you put 1A. Where am i making mistake?
Thanks & Regards

Hi Sir I found correction. You used study at 180 degree and by default its at 0 degree. Whats reason behind out of phase plotting the data?
Thanks & Regards

Hi Sir Robert,
I found correction. You used study at 180 degree and by default its at 0 degree. Whats reason behind out of phase plotting the data?
Thanks & Regards

I plotted it that way because the resulting plot was prettier. The plot is just a snapshot at a particular phase. Feel free to pick and plot any other phase snapshot of the solution that you prefer.

Dear Robert,
This took me a day to decode this phase change plot :-)
1. What source do you prefer to use for E-field incidence in RF problem in 2-D. Its either line point source, point dipole, electric field at some boundary or background E-field?
2. Does COMSOL caters E-Filed attenuation and phase changes automatically as wave travels into different multi-layers medias?
3. For how long time span the simulation is performed, once frequency domain study is selected? Actually, i need to figure out the travelling and return time of RF wave from scattering multi-layer dielectrics and link it to analytical solutions?

Thanks & regards

OK, I'll attempt to help....

Your Question 1. What source do you prefer to use for E-field incidence in RF problem in 2-D. Its either line point source, point dipole, electric field at some boundary or background E-field?

My Answer 1. It depends on the problem. If I'm doing a scattering problem, I typically choose the "scattered field" formulation. (You might want to look that up in the help system.) That lets you set up a clean plane wave as a background wave. But if I want to launch something else, and I often do, then I may use some other source or antenna configuration. Your goal should be to create a model that accurately represents your physical problem and which works well. There actually may be several different ways to do that. And by the way, don't forget about specifying the correct polarization.

Your Question 2. Does COMSOL caters [hmm, did you mean "calculate"?] E-Filed [field] attenuation and phase changes automatically as wave travels into different multi-layers medias?

My Answer 2. If I've understood your question correctly, then yes. But there is an exception. Don't use the "scattered field" formulation if you can't find a way to model your situation as an isolated scattering target surrounded by a single-type of propagation media. Otherwise, you will have to launch your own wave with your own source. In 2D, that would be likely via a point or edge boundary condition, and there are a variety to choose from.

Your Question 3. For how long time span the simulation is performed, once frequency domain study is selected? Actually, i need to figure out the travelling and return time of RF wave from scattering multi-layer dielectrics and link it to analytical solutions?

My Answer 3. Your question suggests to me that you may be confused about the difference between "time domain" and "frequency-domain." A frequency-domain problem does not even have time as a parameter. Any/all of the waves involved in the frequency-domain problem are sinusoidal, and all of them exist for all time, in the problem. Now, for a time-domain problem, but one in which the waves have a sinusoidal dependence, then the wave phases are well-defined and advance linearly with time. But in a time-domain problem, there is no requirement that waves always be sinusoidal. And if the waves of interest to you are truly sinusoidal, then you are usually better off analyzing them in the frequency domain. If you need to compute time differences in a frequency-domain problem, you'll need to extract them from the phase differences and the wave velocity, which (of course) may itself vary as the wave passes through different media within your problem. If you can't find an unambiguous way to analyze that, then maybe you should set it all up as a time-domain problem instead, right from the start. But bear in mind that setting up time-domain problems properly in Comsol Multiphysics is usually a little trickier than setting up frequency-domain problems.

So good luck. I hope that helps.

Dear Robert,
Thanks a lot for the clarifications. Actually, these are weaknesses of EM theory at my part so i needs further assistance from you on subject concepts.

1. How a Dipole antenna is approximated in 2-D problem? Is it by line current source or electric point dipole? And more importantly how to link up magnitudes e.g 1 Vp-p at 50 Ohm Dipole to Current of line source or Dipole moment of electric point dipole?

2. In frequency domain analysis, after how long time (steady-state time) the simulation results are displayed on graphics. I mean to ask that plots that we see are calculated after how much time-stamp, by default?

3. Moreover, for S-parameter calculation ratio of Erec/Einc is calculated after how long time reflected back signal?

Thanks & Best Regards
Awais Qureshi

Well, I think you may need more assistance than I'm prepared to offer, so I encourage you to contact Comsol support if you need more help with the software.

In regard to better understanding antennas, I suggest you take a look at:
www.st-andrews.ac.uk/~www_pa/Scots_Guide/RadCom/part6/page1.html
and
www.st-andrews.ac.uk/~www_pa/Scots_Guide/RadCom/part7/page1.html
These are part of a more general series of RF-educational webpages, found at:
www.st-andrews.ac.uk/~www_pa/Scots_Guide/RadCom/intro.html
and for electronics in general, at:
www.st-andrews.ac.uk/~www_pa/Scots_Guide/intro/electron.htm

Hi Dear Robert,
Thanks a lot for the useful links.
Best Regards
Awais Qureshi

Dear Sir,

I hope every thing is fine with you.

I want to add 2 electric field sources around sphere. I used 2 electric point dipole in 3D around it on the same axis(z).
1- I determine 2 physical points and then linking between them and dipoles.
2- After that I choose the dipole moment option and I get 3 data entries for x, y and z, I know that these values determine the polarization of dipole moment and according to these values the power will be perpendicular to the polarization direction.

My first question is how can I determine the polarization direction and the physical points play a role in that.

The second: how could I convert from [A.m] unit to electron volt.

The attached pic is my structure and it is radio frequency model

Thanks in advance

Dear Yasmien ,
I am a new user and I have to solve the RF scattering problem in a scenario which is similar to yours.
Could you please tell me how you set the point source in a 3D scenario?
I have drawn a point and then I tried to assing an electric point source as you do in a 2D scenario but it does not work.
Thank you in advance for your suggestions.
Regards
Domenica

Hi Robert

I am working on SPP wave excitation on graphene 2D in COMSOL. I modeled graphene in COMSOL to excite SPP using a point source, I used a circle above graphene surface as a point source. Well, I am interested in TM mode so I used electric field on circle with only x and y components and z is zero. Let me say, if I use Electric field on circle with E0 having components as x=1 [V/m], y=1[V/m] and z=0. After running simulation I saw SPP excited.
I also know that SPP on graphene can also be excited using a port connected to the graphene sheet where we can give port input power=1[W] etc. But in my case I am using excitation using a point source (circle) above graphene which radiate electric field on x and y axis and it excite SPP on graphene sheet which is 20nm below point source (circle). But now I am looking to know the power level I have used for excitation using this method.

Can you guide me how to know the input power I used with given x=1 [V/m], y=1[V/m] and z=0 of Electric field components of the excitation source (ie circle above graphene)?. Do we need to convert x[1 V/m], and y in [1 V/m] to power in watts or current in ampere ? Any suggestions please ?

Many thanks