“and one more thing if I run the anlso solve mode it would give me mode!!!
is there anything behind of this?”
Yes, there is something behind it. Both the aniso mode solver and the FD mode solver search for modes as eigenvectors in a Krylov subspace using the Implicitly Restarted Arnoldi Method (IRAM). The ADI mode solver, on the other hand, uses BPM with the propagation going in an imaginary direction. The imaginary direction is to promote exponential growth of the fundamental mode (a circular function in the real direction). Imaginary direction BPM would be described, in mathematical terms, as a power method. IRAM is a more sophisticated method than the power method, so it is not surprising that FD and aniso mode solvers can sometimes get it right when ADI gets it wrong.
In fact, this property of the aniso mode solver was noted before we developed the FD mode solver. I guessed the reason for aniso’s success was IRAM, so I developed an isotropic mode solver that uses IRAM instead of imaginary BPM. The result: FD mode solver! Introduced in OptiBPM 12 and found in recent releases of OptiFDTD too!
.
