Electric - magnetic field due to a dipole source

[psabbagh]

Hi.
I am trying to get the electric and magnetic field in a location “r” due to the excitation of a dipole source at location “r_prime”. The desired wavelength range is from 10 um to 13.5 um. In fact, there are two thin film SiC layers with 100 nm and 10 nm thicknesses. They are separated by a vacuum gap of 100 nm. I want to put a dipole somewhere in SiC layer with thickness of 100 nm and get the electric and magnetic field in the middle of the vacuum gap between layers by setting a time monitor. In my opinion, electric and magnetic field should go to zero as simulation time increases. However, I cannot see such a thing in my case. I am curious why electric and magnetic field are not converging to zero in time domain?
If you open the file attached here, you can see that there is a time monitor in the middle of vacuum gap where I like to see electric and magnetic field as a function of time. Any idea on this is really appreciated. Thanks

Dipole.fsp (307.2 KB)

[bkhanaliloo]

Dear @psabbagh

You arr right and electromagnetic fields should go to zero over time as they get absorbed by PML layer.
I made a few changes in your simulation file:

• I used finer mesh along x-axis. Originally you had a mesh aspect ration of 30, which might result in poor PML absorption. This is discussed in the links below:
  https://kb.lumerical.com/en/index.html?sp_sers.html
  https://kb.lumerical.com/en/index.html?rf_methodology.html

• I increased simulation time to 10,000 to make sure that simulation trigger auto shut off level.

• I used standard PML layers with 16 layers. I am not an expert in custom PML, and 16 layers of standard PML was giving sufficient results anyways

Here is the screenshot of the results for your review:

and here is the modified simulation file:
Dipole_modified.fsp (248.2 KB)

Please let me know of your thoughts and I will be happy to be of a help.

Thanks

[psabbagh]

Thanks a lot dear bkhanaliloo,
It helped me a lot. I have one more question, though. While I could converge the electric field at the target point, I could not converge it when the dipole is along the y axis. Is there any hint that I did not consider for dipole in y axis?! Thanks so much again.

[bkhanaliloo]

Dear @psabbagh

Thanks for the updates.

Can you please clarify what were the results that you were trying to converge? One thing to note is that, in 2D simulations, for a dipole with y-polarization (x-polarization), majority of its radiation is along x axis (y axis). You can test this by simulating a dipole in a free space and adding monitors far from dipole along x and y axis to monitor transmission.

In your case when dipole is oriented along y-axis, you may need a larger simulation region for the light to escape the emitter region before it gets absorbed by PML. Thus, your results might vary by a lot as you increase simulation region as majority of radiation is along x-axis. Since your simulation is in 2D, you can increase the FDTD region until your results converge without waiting long for simulation to finish. For the x-polarized dipole, since most of the radiation is in y-axis, you will need smaller simulation region.

I hope this solves the convergence problem but please let me know if you had further questions.

Thanks

[psabbagh]

Thank you so much Behzad. It was so helpful. Thanks