Setting about FDTD Region and mesh



What is the most efficient way to set FDTD Region or mesh?

Adjusting Mie Scattering 3D sample files
Waveguide filter with nanodisk resonators
Mesh problem in Graphene silulation
Defining FDTD Region while taking PBC into account
Not getting the correct reflection-transmission spectrum
Oscillation in frequency domain E field amplitude
Wired FDTD simulation result on racetrack micro-resontor?
How to calculate the transmittance of a nanotube array?
Mode in optical microocavity

Here is the process I use when setting up the mesh for an FDTD simulation:

  • Setup the size of the FDTD simulation region. For PML boundaries, it’s best to leave about 1/2 wavelength of space between the edge of the critical structures and the PML boundary.

  • Set the ‘mesh accuracy’ to a value of 1 or 2. This will give a mesh size of about 6 or 10 mesh points per wavelength, which is a good mesh size for initial simulations. Later, when I want to get higher accuracy results, I will increase the mesh accuracy setting, often to around 3-5.

  • Add mesh override regions around the critical regions of the simulation volume. For example, if there is a small metal particle in the simulation, I would probably add an override region around the particle. Initially, I will set the mesh size to be reasonably large, such as using the same mesh size that is generated by the automatic mesh. However, during convergence testing, the mesh override region can be used to force the mesh to be much smaller. Typically, it is more important to make the mesh small around the critical structures of the device, rather than making the mesh small throughout the bulk regions of the simulation (ie. with the mesh accuracy setting).

Why my FDTD solutions do not converge?
Absorption vs wavelength
Mesh problem in Graphene silulation
2D TE mode for tracing the Poynting vectors
How can i calculate Transmission in graphene ribbon?
Result is not matching even if simulation environment is the same
Trying to calculate power absorbed
Mesh order's effect on regions that don't overlap
Ridge waveguide to slab waveguide modal reflection