Here we have a 2D simulation with a cylindrical structure which acts as a cavity.
This cavity supports whispering gallery modes which have strong fields around the outer
edge of the cylinder.
The mode that we want to excite in this simulation is the whispering gallery mode at a wavelength
of 418 nm.
Add a dipole source from the sources drop down menu.
You can click and drag the position of the source to place it near the edge of the structure
where it should couple light into the desired mode.
It's also possible to edit the source and type in the source position manually.
Open the edit window for the source by clicking the edit button or using the E keyboard shortcut.
In the General tab, you can set the dipole type to electric or magnetic, and you can
set the dipole orientation by specifying theta and phi angles.
Either type of dipole, electric or magnetic will work to excite the mode of the cavity.
In this case I happen to know that the mode that I want to excite will have magnetic fields
in the z-direction, so select the magnetic dipole and set angle theta and angle phi to
Click OK and check the orientation of the dipole in the CAD view ports.
The direction of the green arrow shows that the dipole is oriented along the z-direction.
Edit the source again and under the Frequency/Wavelength tab, set the wavelength to 418 nm to match
the expected resonance wavelength.
To ensure that light from will couple into the desired mode, you can add additional dipoles
at different locations.
Select the dipole and use the duplicate button in the edit toolbar to create copies of the
dipole and drag the duplicated dipoles to different positions around the edge of the
Now run the simulation
and after running, right-click on the profile monitor and visualize the H result to see
the magnetic field profile of the resonant mode.
Here are some tips for setting up the dipole source.
Avoid placing the dipole source right next to Bloch boundaries.
A distance of at least 1 mesh cell should be used between the source injection region
and Bloch boundaries to avoid source injection errors.
As mentioned in the previous unit, if you want to simulate uniform isotropic radiation
from a point source rather than the dipole radiation, this can be done by running 3 simulations
using 3 orthogonal dipole orientations, then averaging the fields from the three simulations.
Since the actual power emitted by the dipole is modified by surrounding structures, you
can re-normalize power transmission results by multiplying the transmission by the Purcell
You can get the Purcell factor as a result from the source or measure it using the transmission
box analysis group.