Less6_3dB_coupler.bpd as you have attached it has 2 issues:
- The example spaces the linear waveguides making the coupler at X = +/-3 um. You have set yours to +/- 4 um. If you make that correction you will see the coupling.
- In a couple of places at around the 470um and 530 points your have a waveguide coming from just before it (468) and right at it (470). They should both have the same value. The same is true at around the 530 um point.
As for EvCouplingTwo.bpd the results look to me like they are going non-physical as you go around the bend. Based on your design this could be a couple of things that would require investigation:
- You have the core and cladding separate due to the way we layout the 2D. You are using only the core as the propagation constant in your Global simulation settings. I would determine the mode of the whole waveguide and use that manually.
- It is possible that the bend is too steep with the refractive index contrast. This would need investigation.
- Possible convergence issue which would need to be investigated though this one is less likely versus the other two based on the field distribution.