Sorry for my late reply. We had really hectic support days.
I do not think that the problem is due to type of the source i.e. using pulse or CW mode source. Rather it can be due to mesh, boundary conditions, and injecting a proper mode source.
Regarding your first post: I noticed that your bandwidth is small. Do you expect to see a resonance in this spectrum? Maybe you can use a larger bandwidth? Also, the plot for the CW source in your earlier post is very strange. It looks to be a problem with PML boundary conditions.
Any how, I started working on your simulations from yesterday. Since this is a big file, it takes a long time for each simulation to finish (~1-2 hours) and it makes troubleshooting very hard. I was thinking of some alternative approaches to start with and gain some knowledge before starting 3D FDTD:
2D FDTD: This can be a very good test to start with. I haven’t done this simulation, but I am very hopeful that we can learn a lot from it like resonances and coupling efficiencies.
varFDTD: This can be a good next step. Please note that the results will not be as accurate as 3D FDTD, but are computationally cheap and troubleshooting will be much easier.
This link might be also very useful to locate the resonances of the microdisk:
I also attached the modified simulation file for your review and here is some initial results (coarse mesh) with a broad band source obtained from transmission monitor:
microdisk_Modified.fsp (678.0 KB)
Can you please update me with the results of 2D FDTD, varFDTD? Also please keep working on 3D simulations by running them with a finer mesh. I will be glad to help you to troubleshoot the problem and provide you with my idea, but I will need your updates to make decision on the next steps.