EME: Sweep
In practice, a taper length needs to be optimized to ensure proper modal transmission and low loss. This example shows how the sweep function can be used to visualize the transmission of an EME taper simulation.
This code example is licensed under the BSD 3-Clause License.
import emodeconnection as emc
import numpy as np
## Set simulation parameters
wavelength = 1550 # [nm] wavelength
dx, dy = 20, 20 # [nm] resolution
h_core = 220 # [nm] waveguide core height
h_clad = 1000 # [nm] waveguide top and bottom clad
window_width = 4000
window_height = h_core + h_clad*2
num_modes = 10 # [-] number of modes
BC = 'TE'
## Connect and initialize EMode
em = emc.EMode(simulation_name = 'taper', verbose = True)
## Settings
em.settings(
wavelength = wavelength, x_resolution = dx, y_resolution = dy,
window_width = window_width, window_height = window_height,
num_modes = num_modes, boundary_condition = BC,
background_material = 'SiO2')
## Draw shapes
em.shape(name = 'BOX', material = 'SiO2',
height = h_clad)
em.shape(name = 'core', material = 'Si',
height = h_core, etch_depth = h_core)
## Launch FDM solver and label profiles
em.shape(name = 'core', mask = 1000)
em.label_profile(name = 'a') # skips solving the modes here
em.shape(name = 'core', mask = 2000)
em.label_profile(name = 'b') # skips solving the modes here
## Draw EME sections
em.straight_section(
profile = 'a',
length = 2000)
em.taper_section(name = 'taper',
profile = 'a', profile_end = 'b',
length = 8000)
em.straight_section(
profile = 'b',
length = 2000)
## Run EME sweep
data = em.sweep(key = 'section, taper, length',
values = np.arange(0, 8001, 250),
result = ['S_matrix'])
## Close EMode
em.close()
Console output:
EMode3D 1.0.6 - email
Sweeping section parameter 'length'...
Solving EME: length = 8000...
Solving slice at 6666.7 nm... completed in 1.4 sec
Solving slice at 4000.0 nm... completed in 1.4 sec
Solving slice at 1333.3 nm... completed in 1.2 sec
Solving slice at 7333.3 nm... completed in 1.2 sec
Solving slice at 6000.0 nm... completed in 1.4 sec
Solving slice at 4666.7 nm... completed in 1.5 sec
Solving slice at 3333.3 nm... completed in 1.4 sec
Solving slice at 2000.0 nm... completed in 1.2 sec
Solving slice at 666.7 nm... completed in 1.3 sec
Solving slice at 7666.7 nm... completed in 1.2 sec
Solving slice at 6333.3 nm... completed in 1.4 sec
Solving slice at 5666.7 nm... completed in 1.4 sec
Solving slice at 5000.0 nm... completed in 1.4 sec
Solving slice at 4333.3 nm... completed in 1.4 sec
Solving slice at 3666.7 nm... completed in 1.4 sec
Solving slice at 3000.0 nm... completed in 1.4 sec
Solving slice at 2333.3 nm... completed in 1.4 sec
Solving slice at 1666.7 nm... completed in 1.2 sec
Solving slice at 333.3 nm... completed in 1.4 sec
Solving slice at 6166.7 nm... completed in 1.4 sec
Solving slice at 5833.3 nm... completed in 1.4 sec
Solving slice at 5166.7 nm... completed in 1.4 sec
Solving slice at 4833.3 nm... completed in 1.4 sec
Solving slice at 4500.0 nm... completed in 1.2 sec
Solving slice at 4166.7 nm... completed in 1.4 sec
Solving slice at 3833.3 nm... completed in 1.4 sec
Solving slice at 3500.0 nm... completed in 1.4 sec
Solving slice at 3166.7 nm... completed in 1.4 sec
Solving slice at 2833.3 nm... completed in 1.4 sec
Solving slice at 1833.3 nm... completed in 1.2 sec
Solving slice at 6083.3 nm... completed in 1.4 sec
Solving slice at 5916.7 nm... completed in 1.4 sec
Solving slice at 3416.7 nm... completed in 1.4 sec
Solving slice at 3250.0 nm... completed in 1.4 sec
Solving slice at 3083.3 nm... completed in 1.4 sec
Solving slice at 1916.7 nm... completed in 1.2 sec
Solving slice at 1750.0 nm... completed in 1.2 sec
Solving slice at 6125.0 nm... completed in 1.4 sec
Solving slice at 1958.3 nm... completed in 1.3 sec
Solving slice at 6104.2 nm... completed in 1.4 sec
Solving slice at 1937.5 nm... completed in 1.2 sec
Solving slice at 1927.1 nm... completed in 1.2 sec
Solving slice at 1921.9 nm... completed in 1.2 sec
completed in 1 min 32.9 sec
Solving EME: length = 7750... completed in 0.5 sec
Solving EME: length = 7500... completed in 0.5 sec
Solving EME: length = 7250... completed in 0.5 sec
Solving EME: length = 7000... completed in 0.5 sec
Solving EME: length = 6750... completed in 0.5 sec
Solving EME: length = 6500... completed in 0.5 sec
Solving EME: length = 6250... completed in 0.5 sec
Solving EME: length = 6000... completed in 0.5 sec
Solving EME: length = 5750... completed in 0.5 sec
Solving EME: length = 5500... completed in 0.5 sec
Solving EME: length = 5250... completed in 0.5 sec
Solving EME: length = 5000... completed in 0.5 sec
Solving EME: length = 4750... completed in 0.5 sec
Solving EME: length = 4500... completed in 0.5 sec
Solving EME: length = 4250... completed in 0.5 sec
Solving EME: length = 4000... completed in 0.5 sec
Solving EME: length = 3750... completed in 0.5 sec
Solving EME: length = 3500... completed in 0.5 sec
Solving EME: length = 3250... completed in 0.5 sec
Solving EME: length = 3000... completed in 0.5 sec
Solving EME: length = 2750... completed in 0.5 sec
Solving EME: length = 2500... completed in 0.5 sec
Solving EME: length = 2250... completed in 0.5 sec
Solving EME: length = 2000... completed in 0.5 sec
Solving EME: length = 1750... completed in 0.5 sec
Solving EME: length = 1500... completed in 0.5 sec
Solving EME: length = 1250... completed in 0.5 sec
Solving EME: length = 1000... completed in 0.5 sec
Solving EME: length = 750... completed in 0.5 sec
Solving EME: length = 500... completed in 0.5 sec
Solving EME: length = 250... completed in 0.5 sec
Solving EME: length = 0... completed in 0.3 sec
completed in 1 min 48.2 sec
Exited EMode
While no figures are generated in the EMode script, a separate script can be used to plot the results.
import emodeconnection as emc
import numpy as np
## Extract sweep results without an EMode license
data = emc.get(variable = 'sweep_data', simulation_name = 'taper')
## Plot
import matplotlib.pyplot as plt
from matplotlib import rc as mplrc
fw, LW = 8/2.54, 0.5
mplrc('font',**{'family':'sans-serif','size':7})
mplrc('axes', linewidth=LW, axisbelow=True)
mplrc('xtick', bottom=True, top=True, direction='in')
mplrc('ytick', left=True, right=True, direction='in')
mplrc('xtick.major', size=3, width=LW)
mplrc('ytick.major', size=3, width=LW)
mplrc('figure',figsize=[fw, fw/2**0.5])
fig, ax = plt.subplots(1, 1)
ax.set_xlabel('Taper length (nm)')
ax.set_ylabel('Transmission (%)')
ax.grid(visible=True, which='major', axis='both', linewidth=LW/2, color='gray', alpha=0.25)
transmission = np.array([s.transmission() for s in data['S_matrix']])
lines = []
for kk in range(transmission.shape[-1]):
line, = ax.plot(data['values'],
100*np.abs(transmission[:,kk,0])**2,
marker = '', linestyle = '-', lw = LW*1.5,
label = r'Mode 0$\rightarrow$%d' % kk)
lines.append(line)
ax.set_xlim([
np.min(data['values']),
np.max(data['values'])])
lg = ax.legend(handles = lines[0:1], loc = 'center right')
lg.get_frame().set_linewidth(LW/2)
lg.get_frame().set_edgecolor('k')
ax.set_ylim([70, 101])
fig.savefig('taper_1.png', dpi=600, bbox_inches='tight')
lg = ax.legend(handles = lines[1:10], loc = 'upper right')
lg.get_frame().set_linewidth(LW/2)
lg.get_frame().set_edgecolor('k')
ax.set_ylim([-1, 25])
fig.savefig('taper_2.png', dpi=600, bbox_inches='tight')
Figures:
