Learning Forum

[Muhammad Danial Haziq Azizan]

My question is how to plot the graph for resonant (a.u. vs wavelength), not the (a.u. vs frequency). Which part I need to change. Urgent..Tqs

##############################################

# Q analysis

# This script calculates the quality factor from

# the slope of the decaying envelope.

#

# Input properties

# make plots: make plots of signal, envelope, slope, spectrum

# 1 for yes, 0 for no

# number resonances: number of resonances to look for

#

# Output properties

# f0: vector of resonant frequencies

# Q: vector of Q factors of resonances

# delta_Q: error in Q factor

#

# Tags: resonator high q analysis quality factor

#

# Copyright 2012 Lumerical Solutions Inc

##############################################

 

# simplify input variable names by removing spaces

number_resonances = %number resonances%;

make_plots = %make plots%;

 

min_filter_width = 1; # min width of filter in units of resonance FWHM

max_filter_width = 6; # min width of filter in units of resonance FWHM

filter_width_test_points = 20;

zero_pad = 2^16; # fft zero padding

# Note fft zero pad should be a power of 2,

# and larger gives more resolution in the

#frequency domain.

 

for(N=0; (N+1) <= nx*ny*nz; 0) {

N=N+1;} # up to the sum of # of monitors = (nx*ny*nz)

 

#################################################

# get the monitor data for the first monitor

#################################################

t = getdata("t1","t");

field0_t_Ex = pinch(getdata("t1","Ex"));

field0_t_Ey = pinch(getdata("t1","Ey"));

field0_t_Ez = pinch(getdata("t1","Ez"));

 

#################################################

# do fft to frequency domain for all monitors

#################################################

w = fftw(t,1,zero_pad);

field_w = matrix(length(w),6*N);

for(i=1:N) {

mname = "t" + num2str(i);

for(j=1:6) {

if(almostequal(j,1)) { component = "Ex"; }

if(almostequal(j,2)) { component = "Ey"; }

if(almostequal(j,3)) { component = "Ez"; }

if(almostequal(j,4)) { component = "Hx"; }

if(almostequal(j,5)) { component = "Hy"; }

if(almostequal(j,6)) { component = "Hz"; }

if(j > 3.5) { extra_factor = sqrt(mu0/eps0); }

else { extra_factor = 1; }

if(havedata(mname,component)) {

 

field_w(1:length(w),6*(i-1)+j) = 2*extra_factor*( (1:length(w)) <= (length(w)/2+0.1)) *

fft(pinch(getdata(mname,component)),1,zero_pad);

}

}

}

 

 

#################################################

# find resonant peaks, including all monitors

#################################################

f_spectrum = sum(abs(field_w)^2,2);

 

p = findpeaks(f_spectrum,number_resonances);

f0 = w(p)/(2*pi);

 

#################################################

# find quality factors

#################################################

 

# reserve memory for results

peak_spectra = matrix(length(w),number_resonances);

peak_filters2 = matrix(length(w),number_resonances);

 

# calculate slope of decay using 40-60% of time signal

tp1 = round(0.4*length(t));

tp2 = round(0.6*length(t));

t2 = t(tp1:tp2);

log_field_all = matrix(tp2-tp1+1,number_resonances);

 

Q = matrix(number_resonances);

delta_Q = matrix(number_resonances)+1e300;

 

# loop over each peak

for(i=1:number_resonances) {

# find FWHM of peak

peak_val = f_spectrum(p(i));

continue_search = 1;

for(p1=p(i)-1; (p1>1) & continue_search ; 1) {

if(f_spectrum(p1)<=peak_val/2) {

continue_search = 0;

} else {

p1 = p1-1;

}

}

continue_search = 1;

for(p2=p(i)+1; (p2

if(f_spectrum(p2)<=peak_val/2) {

continue_search = 0;

} else {

p2 = p2+1;

}

}

if(p1 < 1) { p1 = 1; }

if(p2 > length(w)) { p2 = length(w); }

FWHM = w(p2)-w(p1);

 

for(filter_width=linspace(min_filter_width,max_filter_width,filter_width_test_points)) {

# calculate the filter for the peak

peak_filter = exp( -0.5*(w-w(p(i)))^2/(filter_width*FWHM)^2 );

 

# inverse fft to get data in time domain

field2_t = 0;

for(mcount=1:6*N) {

field2_t = field2_t + abs(invfft(pinch(field_w,2,mcount)*peak_filter))^2;

}

field2_t = field2_t(tp1:tp2);

log_field = log10(abs(field2_t));

 

# calculate slope and Q from the slope of the decay

# estimate error from the slope

slope = (log_field(2:length(t2))-log_field(1:length(t2)-1))/

(t(2:length(t2))-t(1:length(t2)-1));

slope_mean = sum(slope)/length(slope);

slope_delta = sqrt( sum((slope-slope_mean)^2)/length(slope) );

Q_test = -w(p(i))*log10(exp(1))/(slope_mean);

delta_Q_test = abs(slope_delta/slope_mean*Q_test);

if(delta_Q_test < delta_Q(i)) {

Q(i) = Q_test;

delta_Q(i) = delta_Q_test;

 

# collect data for final plot

peak_spectra(1:length(w),i) = f_spectrum * peak_filter^2;

peak_filters2(1:length(w),i) = peak_filter^2;

log_field_all(1:length(t2),i) = log_field;

 

}

}

# output summary of peak results to script window

?"Resonance " + num2str(i) + ":";

?" frequency = " + num2str(w(p(i))/(2*pi)*1e-12) + "THz, or "+num2str(2*pi*c/w(p(i))*1e9)+" nm";

?" Q = " + num2str(Q(i)) +" +/- " + num2str(delta_Q(i));

}

 

Q_matrix=Q;

 

Q = matrixdataset("Q");

Q.addparameter("lambda",c/f0,"f",f0);

Q.addattribute("Q",Q_matrix);

Q.addattribute("dQ",delta_Q);

 

spectrum = matrixdataset("spectrum");

spectrum.addparameter("lambda",c/(w/2/pi),"f",w/2/pi);

spectrum.addattribute("spectrum",f_spectrum);

 

#################################################

# plot the results

#################################################

if (make_plots) {

# plot signal and envelope for first monitor

field_t_Ex = invfft(pinch(field_w,2,1));

field_t_Ex = field_t_Ex(1:length(t));

field_t_Ey = invfft(pinch(field_w,2,2));

field_t_Ey = field_t_Ey(1:length(t));

field_t_Ez = invfft(pinch(field_w,2,3));

field_t_Ez = field_t_Ez(1:length(t));

plot(t*1e15,field0_t_Ex,abs(field_t_Ex),field0_t_Ey,abs(field_t_Ey),field0_t_Ez,abs(field_t_Ez),"time (fs)","field envelope");

legend("field (Ex)","envelope (Ex)","field (Ey)","envelope (Ey)","field (Ez)","envelope (Ez)");

 

# plot the slopes of the decaying fields

plot(t2*1e15,log_field_all,"time (fs)","log10(|field(t)|)","Decay for each resonance");

 

# plot spectra

p1 = find(w,0.8*min(w(p)));

p2 = find(w,1.2*max(w(p)));

f = w/(2*pi);

plot(f(p1:p2)*1e-12,peak_spectra(p1:p2,1:number_resonances)/max(f_spectrum)

,"frequency (THz)","Arbitrary units","Spectrum of resonances");

plot(f(p1:p2)*1e-12,f_spectrum(p1:p2)/max(f_spectrum),peak_filters2(p1:p2,1:number_resonances)

,"frequency (THz)","Arbitrary units","Spectrum and filters");

}

 

[Muhammad Danial Haziq Azizan]

Then, what means by these codes to plot the resonant? why need to multiply 0.8 and 1.2?

# plot spectra

p1 = find(w,0.8*min(w(p)));

p2 = find(w,1.2*max(w(p)));

[Dev]

Hello, this is simply to set the axis limit. For eg if you change value from 1.2 to 1.8 the X-axis limit will incease to the right side. This wont affect your result. 

Regards

Devika

[Dev]

Hello, Can you share mode details? device under simulation and whee you accure this script? 

 

[Dev]

Hello, 

Add the following lines at end of the script. You can plot the result in wavelength.

 

plot(c/f(p1:p2)*1e9,peak_spectra(p1:p2,1:number_resonances)/max(f_spectrum)

,"wavelength (nm)","Arbitrary units","Spectrum of resonances");

plot(c/f(p1:p2)*1e9,f_spectrum(p1:p2)/max(f_spectrum),peak_filters2(p1:p2,1:number_resonances)

,"wavelength (nm)","Arbitrary units","Spectrum and filters");

Note: When you use Forum next time please try to share more details. The more details you write the faster you get a suggested solution.

 Ansys Insight: Ansys Innovation Space —- Forum and Feed

[Muhammad Danial Haziq Azizan]

Can I know who is an expert in this topic? To find the Q factor with the existence of resonant which means calculate the Q factor based on resonant wavelength divided with FHWM. I really stuck using this software

[Dev]

Hello Muhammad, 

If you are using an academic license, through this forum you can ask questions and reach out to Ansys. Our support team replies most of the questions. 

You can find equation and more details here. Quality factor calculations for a resonant cavity – Ansys Optics.

If you are new to using FDTD siftware, 

You can refer this manual to learn more about the software: FDTD product reference manual – Ansys Optics

I recomend to take free course on FDTD: Ansys Lumerical FDTD – Learning Track | Ansys Courses

Thanks and reagrds

Devika