Guilin Sun
Ansys Employee

For question 1, I guess you used the default Metal boundary condition. So it is normal to have more number of modes in larger box (think it as a metal box partially filled with waveguide materials). But they do not affect the mode you desire. Most often, customers are only interested in the guided modes, or the fundamental mode, which needs to be identified, regardless all other higher "modes".  The FDE size should be ok, as long as the mode is not truncated to certain degree. You can view the mode amplitude in log scale and check the magnitude at the boundaries.

for question 2, it all depends on what you want to do. If you want to get the best coupling from actual beam to the waveguide, you will need to know the experimental beam property, say the beam waist. Most often it is the laser beam passing through a microscope objective, with certain numerical aperture NA. Once you know the NA, or beam waist on the waveguide to be coupled, you can create a Gaussian beam from the GUI.

 

NA beam

Scalar beam: