Bug using Bloch boundary conditions for a 1-dimensional model of a PEC terminated dielectric layer using set_boundary

[andrew-ueltd]

If you model a PEC-terminated dielectric layer with a skew-incident plane wave source propagating in the xz plane, one way to do this is to use Bloch boundary condition with a one dimensional computational domain (along the z-direction with the incident plane wave propagating with wave vector in the xz plane). By default the Bloch condition is applied in both x and z directions so a PEC termination can be represented using sim.set_boundary(mp.Z,mp.High,.p.Metallic) just prior to the simulation with the layer which overrides the Bloch condition in the +z direction. This method works with a 2D computational domain. However, in 1D this results in a segmentation fault-core dumped message and incorrect results (reflection coefficient not unity as expected). Any ideas anyone? File enclosed.
refl-angular_1D_4.zip

[oskooi]

The issue may be fixed by #3157.