eigenmode source unable to launch modes in negative direction

[oskooi]

The EigenModeSource can be used to launch oblique-angle, forward-propagating planewaves as demonstrated by the following script.

import meep as mp
import numpy as np
import matplotlib.pyplot as plt

resolution = 50 # pixels/μm                                                                                                                                              

cell_size = mp.Vector3(14,10,0)

pml_layers = [mp.PML(thickness=2,direction=mp.X)]

# rotation angle (in degrees) of planewave, counter clockwise (CCW) around z-axis                                                                                        
rot_angle = np.radians(0)

fsrc = 1.0 # frequency of planewave (wavelength = 1/fsrc)                                                                                                                

n = 1.5 # refractive index of homogeneous material                                                                                                                       
default_material = mp.Medium(index=n)

k_point = mp.Vector3(fsrc*n).rotate(mp.Vector3(z=1), rot_angle)

sources = [mp.EigenModeSource(src=mp.ContinuousSource(fsrc),
                              center=mp.Vector3(),
                              size=mp.Vector3(y=10),
                              direction=mp.AUTOMATIC if rot_angle == 0 else mp.NO_DIRECTION,
                              eig_kpoint=k_point,
                              eig_band=1,
                              eig_parity=mp.EVEN_Y+mp.ODD_Z if rot_angle == 0 else mp.ODD_Z,
                              eig_match_freq=True)]

sim = mp.Simulation(cell_size=cell_size,
                    resolution=resolution,
                    boundary_layers=pml_layers,
                    sources=sources,
                    k_point=k_point,
                    default_material=default_material,
                    symmetries=[mp.Mirror(mp.Y)] if rot_angle == 0 else [])

sim.run(until=100)

nonpml_vol = mp.Volume(center=mp.Vector3(), size=mp.Vector3(10,10,0))
ez_data = sim.get_array(vol=nonpml_vol, component=mp.Ez)

plt.figure()
plt.imshow(np.flipud(np.transpose(np.real(ez_data))), interpolation='spline36', cmap='RdBu')
plt.axis('off')
plt.show()

The three examples shown above for rotation angles (rot_angle) of 0°, 20°, and 40° involve a k_point in which all vector components have positive values. However, when trying to launch a planewave propagating in the negative direction at 0° via a negative k_point (shown below), the fields blow up:

k_point = mp.Vector3(-fsrc*n).rotate(mp.Vector3(z=1), rot_angle)

The fields also blow up if the rotation angle is >90° which involves negative components for k_point.

[stevengj]

This should be a simple bug somewhere — nothing fundamental prevents this, so there is probably a line or two that is incorrectly assuming the sign of k

[stevengj]

Should be fixed now, e.g. I tried your script for an angle of 110° and got the expected result.