Lumerical Forum -

Why? Because photonics is hard. Unlike circuit simulation, where "ground" is a safe assumption, in FDTD (Finite-Difference Time-Domain) solutions, everything is boundary conditions and mesh order.

It is chaotic. It is occasionally pedantic. But it is arguably the single greatest repository of applied nanophotonics troubleshooting on the internet.

Whether you are designing a sub-wavelength metasurface or trying to suppress crosstalk in a silicon photonic modulator, the moment you hit "Run" in Lumerical, you enter a gray zone between pure mathematics, material science, and hope. When the simulation crashes—or worse, runs successfully but produces physically impossible results—where do you turn?