Learning Forum

[zhang yukang]

Dear teachers:

When using FDTD to simulate a finite number of arrays, for example, a photonic crystal plate with a 30*30 array, observe the transmission spectrum. There are the following questions:

1. Can the transmission spectrum be observed at this time (I see that other posts say that the transmission spectrum is generally not observed for finite structures)? My simulations resulted in very low transmission power peaks.
2.  If the transmission spectrum can be observed, what settings should be paid attention to to achieve a good transmission spectrum? 
3.  If the transmission spectrum is not suitable for observation, how do I calibrate the impact of the number of arrays?

 

 

 

 

 

[Guilin Sun]

transmission can be well calculated for periodic structure+plane wave, and waveguide, since the incident power is finite and does not change with the size of the injection source.

When the array has finite size, I guess you use PML BCs for all the sides. In such case, if you use a plane wave source, the simulation or source area will determine the injection power. However the true transmitted power from the device is fixed but the extra dimension larger than the device will directly pass it. Thus the transmission will be variable and dependent on the source/simulation area. If you use TFSA source it will have similar effect.

The viable way is to use a known waist Gaussian beam. You will make sure in experiment it uses the same Gaussian beam. Otherwise the beam waist will affect the transmission again.

So please consider what exactly you desire and how to control in experiment. 

Maybe you can get "tranmsission" with arbitrary unit, eg, normalized to the peak of the "transmission" to show the result qualitatively. It is not uncommon in practice.