Plane source

fdtd

#1

I would like to model an nanosecond laser pulse in the UV region. However when I try to change the PlaneSource the set frequency/wavelength are dependent on set the time domain. Is there a way to have a 4 ns laser pulse at 337 with a repetition rate of 60 Hz?


#2

Hi @stopka

You can set the frequency of the plane source as is shown in the screenshot below:

To model the laser that creates pulses with repetition rate of 60 Hz, you will need to have a few sources and set the offset (fs) accordingly. For example if the offest of first pulse is set to τ, second pulse offset needs to be set to τ+16.7ms.

When I was trying to set the offset of the pulse, I noticed that sometime the shape of injected pulse is weird (see below):

I examined whether this is the problem with pulse injection or some sampling problem (effecting visualising the injected source) by placing a time domain monitor to capture the injected light. It looks like this is a sampling problem bug (which our team will look into that to fix it), but should not effect your simulations. Here is the time domain electric field for the pulse is shown in the second figure:

and my simulations for you review:
4nsPulse.fsp (271.1 KB)

Please let me know if you have any other quesitons.
Thanks


#3

Thank you for looking into, however the laser has a rep rate of 60 Hz, if I enter this value this does not work. Do you have an idea to fix this?

Best,
Stopka


#4

Hi @stopka

Where did you enter the value? Can you clarify it more or provide me with screenshot?

Thanks


#5


#6

You can modify the offset (fs) to be, for example, 1e7 and FDTD simulation time (fs) to 4e7 (from General tab of Edit FDTD simulation windows).


#7

You can also import your custom time signal as is explained here.

Thanks


#8

I am still having trouble creating this plane source. Even when following the custom source example. Could you help me, i need a pane wave source, with 60 Hz, 337 nm, 3 nm pulse width


#9

Hi @stopka

Here how I set it and it is working fine:

What version of the software are you using? I am running on 2017a 8.17.1102. Maybe you need to upgrade to the newest version from here.


#10

I have input these parameters but the frequency needs to be set to 60 Hz. Pulse length= 3 nm, 337 nm wavelength. I have 2017a 8.17.1072


#12

yes I need 60 hz not 60 THz, but if i change this value the wavelength changes to the incorrect one. Does that make sense?


#13

Hi @stopka

Why you are setting frequency to 60Hz? Repetition rate is the rate which your pulses at 337nm occur. For a pulse which is 5ns and is injected at 337, above screenshot demonstrates how you can do it.

I explained above how you can set the repetition rate. For this you need to set multiple source.


#14

I have this screen shot


#15

for the multiple source I can have one as the defined time domain and the other wavelength?


#16

Here is the file, is this correct, thanks for helping me.

bowties based off Zhou test.fsp (269.6 KB)


#17

Dear @stopka

I will try to explain it again, hopefully I can clarify misunderstanding.

You have two ways to setup your source: in frequency/wavelength domain by setting your wavelength/frequency and bandwidth OR in time domain by setting central frequency and its length. Once you choose time domain, wavelength/frequency settings gets deactivated. Thus, you have to put 890 THz in the frequency panel as in the screenshot I provided.

This source will inject a pulse at time τ that you select for offset. Now for the second pulse, you need to use the exact parameter for frequency and pulse length, but set offset to be τ+(1/601e15)[fs] and for the third pulse to be τ+2(1/60*1e15)[fs].


#18

So the offset for a 3 ns pulse width, wv 336.8 would be 5e19, is that correct


#19

One more question, the run time is hundreds of hours, is there a way to reduce that.


#20

Hi @stopka

Can you please tell me what you are trying to simulate?

For FDTD simulations, we use simulation times in the order of fs or few nano seconds. This means that if you want to simulate a long pulse or pulse that are separated by few ms, your simulation time will be quite long (as you noticed). We can always come up with ideas to simplify the simulation so that they require less time, but give precise results.

If you want to replicate your experimental setup exactly, simulation time will be quite long in this case.

Thanks