Ansys Products

Ansys Products

Discuss installation & licensing of our Ansys Teaching and Research products.

Lumerical 3D FDTD simulation for Chi2 photonic crystal not guided

    • Jordi Lopez Leyva
      Subscriber

      I'm trying to simulate a 1D photonic crystal nanobeam of GaAs with pockels Chi2 effect. I'm doing so by implementing 2 dipole sources at the center. However, the movie shows no light is guided for high powers of the dipoles. For low dipole power I don't have this issue and light is perfectly guided.

      I'm using 2 electric dipole source in FDTD, with total amplitude ~ 2e-24 and power ~2mW. I obtain power via dipolepower command. My material is GaAs nanobeam with Chi2, with air holes, and surrounded by SiO2 and a polymer of n = 1.6. I'm using standard PML boundaries and high mesh density.

      Questions I have

      1. Am I calculating the power correctly?
      2. Do I need to change boundary conditions?
      3. How can I fix the simulation instability?

      For reference, here's a snapshot of the movie:

    • Dev
      Ansys Employee

      Hello,

      Please try not to post duplicated post in another Forum.

      You can refer the optical forum post response: 3D FDTD simulation with 1mW dipole and nonlinear chi2 material not stable (ansys.com)

      Thanks and regards

      Devika

       

Viewing 1 reply thread
  • The topic ‘Lumerical 3D FDTD simulation for Chi2 photonic crystal not guided’ is closed to new replies.