FDTD modelling of electrically thin frequency dependent layers in large-scale electromagnetic problems

The computational cost is one of the major concerns of the standard finite-difference time-domain (FDTD) method based on a uniformly spaced orthogonal Cartesian lattice when it is employed to solve large-scale electromagnetic problems. To ensure numerical accuracy, the FDTD cell size has to be determined based on the size of the smallest object within the FDTD space. This spatial constraint causes very fine meshing of the entire FDTD space unnecessarily and also leads unreasonably small time step usage under CFL (Courant-Fiedrichs-Lewy) stability condition. Hence, any practical engineering problem containing an electrically small object relatively to a large physical space, especially in the presence of a thin layer in the environment would demand prohibitive computational resources in terms of memory and CPU time. An approach is proposed to permit arbitrarily thin layers to be placed in the FDTD space, without need of reducing the FDTD cell size in proportion. This allows significant reductions of the computational memory and execution time to be achieved, and thus enables us to simulate large-scale electromagnetic problems with a reasonable computational cost.