A minimal FDTD proof of concept (FDTDPoC). The aim of the project is to implement a simple FDTD solver in python3, and use it to help people familiriarize with the workings of the algorithm. To this aim, the solver should work in 2D (for TE and/or TM polarisations) and allow for easy configuration of a TFSF plane wave source as well as simple geometries. A real time animation could also have a good teaching impact.

It has the following features:

• Additive dipolar sources can be added at arbitrary positions
  • An independent pulse component representing the time dependence of the source
• Reflective boundaries through perfect electric conductor (PEC)
• Absorptive (i.e. open system) boundaries through absorbing boundary conditions (ABC)

The code has been written on Linux (Ubuntu 17.10) using python3. Packages required:

• numpy

Packages required to run the test driver:

• matplotlib

List of desired features for the FDTD solver

1. Working 2D solver (TE/TM)
2. Boundaries
   • Absorbing boundaries (ABC/PML)
   • Reflecting boundaries (PEC)
3. Sources
   • Dipolar additive source
   • TFSF box
4. Materials
   • Dielectrics (arbitrary n)
   • Metals (Drude model)
   • Two level system (?)
5. ...

and the GUI and data analysis/presentation

1. Real time field animation
2. Easy geometry definition
3. Real time (?) DFT/FFT
4. Real time envelope extraction (Hilbert transform)
5. ...

This looks ambitious!

Look here