Re: [Meep-discuss] Meep Units and custom sources

[Hannes Kohlmann]

Actually this also solved 2), because now I made a loop, that creates a bunch of sources that are stored in a list. When initializing the source I take into account the time retardation as an argument and simply subtract it from the variable t in heidler_fct!

 

Concerning 1), I think my thoughts were not so wrong. Either I decide for the spacial unit (e.g. 1000m), then one meep time unit t equals some value in microseconds or a fix the meep time unit t (e.g. 1 microsecond) then 1a = 300m, for example.

 

Thanks! Now my lightning return stroke already radiates nicely something that was expected ;)

 

Cheers

 

Gesendet: Mittwoch, 27. Juni 2018 um 22:06 Uhr
Von: "Hannes KOHLMANN" <hannes.kohlm...@gmx.at>
An: "Christopher Hogan" <ho...@simpetus.com>
Betreff: Re: [Meep-discuss] Meep Units and custom sources

Her Chris,

thanks, the wrapper solution does the job ;)

Hannes
 

Am 27. Juni 2018 17:30:56 MESZ schrieb Christopher Hogan <ho...@simpetus.com>:

Hi Hannes,

 

Regarding question 3, you could try something like this:

 

def heidler_wrapper(T1, T2, n):

    def heidler_fct(t):

        eta = m.exp((T1/T2)*(n*T2/T1)**(1/n))

        if t > 0 and t < 1 :

           return 1/eta*(t/T1)**n/((t/T1)**n+1)*m.exp(-t/T2)

        return 0

    return heidler_fct

 

Then, create your CustomSource like this:

 

mp.CustomSource(heidler_wrapper(x, y, z), start_time = ...)

 

This evaluates heidler_wrapper and returns heidler_fct with T1=x, T2=y, and n=z.

 

Unfortunately I can't help with the other 2 questions.

 

Chris

  

 

On Tue, Jun 26, 2018 at 12:08 PM, Hannes Kohlmann <hannes.kohlm...@gmx.at> wrote:

Hello everyone,

 

I have three questions, where I got stuck so far with MEEP:

 

I want to simulate a lightning return stroke (a channel with many CustomSources at different heights, that become active one after the other to mimique a current propagation with speed of about 1/3*c_o to 1/2*c_0).

 

1) The first thing is that I am getting a little confused with the meep units, since it is written in the docs, that the time unit is chosen such that c=1 for speed of light in vacuum.

I struggle with the interpretation of this. Why? Because I keep looking at the sim.meep_time, which is used for printing the fields to h5 or calling step functions etc, and after the following interpretation c=1 doesn't make much sense to me.

 

So I defined a grid, lets say cell = Vector3(20,20,0) (for simplicity I will stay 2 dimensional first).

Then I said: meep length unit = 1km, so my grid is 20km long and 20km high.

This means that if I have no material specified, I have vacuum propagation speed v=c_0=3e8. Based on that, it would make sense to me that one meep_time unit equals 1000m/(3e8m/s)=3.3 microseconds ... is that correct? Or the other way round, I could define 1 meep_time equals 1microsecond, so my gris is 1us*3e8 = (20 x 20) * 300m = (6km x 6km)

 

I guess the c=1 is a more useful interpretation when I want to relate the frequency to the wavelength...

 

As already mentioned, I want to activate sources one after another, having something like:

 

def heidler_fct(t,T1=1, T2=0.1, n=2):

     eta = m.exp((T1/T2)*(n*T2/T1)**(1/n))

     if t > 0 and t < 1 :

         return 1/eta*(t/T1)**n/((t/T1)**n+1)*m.exp(-t/T2)

     return 0

 

sources = [mp.Source(mp.CustomSource(heidler_fct, start_time = i/3), component=mp.Ez, center=mp.Vector3(-7,src_y_start+i/10), size=mp.Vector3(0,0))]

for i in range(100): 

   sources.append(mp.Source(mp.CustomSource(heidler_fct, start_time = (i+1)/3), component=mp.Ez, center=mp.Vector3(-7,src_y_start+(i+1)/10), size=mp.Vector3(0,0)))

 

2) Now the problem I figured out was that Meep uses always the current sim.meep_time, which means that the later I start, the more my function has already declined. I could solve it by subtracting sim.meep_time() from t in the heidler_fct ... but how could I access the meep time in the CustomSource function just like in a step function, where this is the first parameter is the simulation instance sim?

 

3) I havent figured out yet how I could set the arguments T1, T2, and n in heidler_fct when defining the source... using something like lambda: heidler_fct(T1=x, T2=y, n=z) didn't work but this was just guessing, trying out and hoping that it would work.

 

I would be very glad, if you can help me with those three points.

Thank you very much in advance!

 

Cheers,

Hannes

 

(PS: should I - in the future - split up those questions into contextually separated mails (e.g. CustomSources and Meep Units, instead of squeezing all questions into one mail?)

_______________________________________________
meep-discuss mailing list
meep-discuss@ab-initio.mit.edu
http://ab-initio.mit.edu/cgi-bin/mailman/listinfo/meep-discuss

--
Diese Nachricht wurde von meinem Android-Gerät mit K-9 Mail gesendet.

_______________________________________________
meep-discuss mailing list
meep-discuss@ab-initio.mit.edu
http://ab-initio.mit.edu/cgi-bin/mailman/listinfo/meep-discuss