FDTD frequency domain problem

[dmacload]

Hi all
I want to write a FDTD code for analyzing the light after it passes throw a dispersive medium I know that I must send an impulse into the medium and take fft of the transmited field I know that my medium interacts with light at 200 THz and I must see a peak at 100 THz in the transmited wave my question is that how can I do this?(what is the incident pulse ?)

 

[iyami]

Basically anything you like, but to keep errors low you should take a pulse whose spectrum peaks in the frequency region you are interested in say 40-400THz.

 

[dmacload]

Basically anything you like, but to keep errors low you should take a pulse whose spectrum peaks in the frequency region you are interested in say 40-400THz.

How can I show this pulse?

 

[iyami]

How can I show this pulse?

Do you mean generate? If so look at the definition of Gaussian pulses e.g.

Gaussian function - Wikipedia, the free encyclopedia

If you mean show (which I doubt) then how about plotting field values against time?

 

[ark5230]

As I understand, the impulse should be a sum of several frequencies of interest (may be continuous function) with number of waves consistent with the Nyquist criterion.

Nyquist