i am simulating a periodic nanoantenna sensor, the sensor consists of the gold disk on top of a silica disk on a substrate. i measure the reflection by the upper monitor at background index 1.33 and another simulation at backgroundindex 1.34 to calculate the sensor sensitivity = change in resonance wavelength / change in Refractive index (background index) . the simulation attached seems to resonate correctly but the graph has ripples compared to same simulation on CST
I think we discussed your simulations already here, however I tried to provide you how you can deal with this case here again.
These simulations are taking a long time which is the nature of your simulation. This is because you have a gold disk that acts like a high-Q cavity and stores the field for a long time.
Please see the edges of your gold disk below, they tend not to decay and simulation takes a long time:
I did some modification and ran it for a relatively long time and here is the plot that I got:
I have my simulation file attached:
Mushroom_Modified.fsp (253.7 KB)
Here are a few things to do:
- Use a finer mesh and make sure that results converge.
- Use a higher number of PML layers
- Use a longer simulation time
- Use a longer pulse in time domain and run simulation for wavelengths span of 3150-3250.