A PIC simulation code moves particles (electrons and ions) on a given domain. It couple the motion of the particles with the generated electric and magnetic field.
- Only the Poisson equation is solved, generating a self-consistent electric field
- The particles are moved in one direction, even is there velocity is defined in 3D
We can use it to study interesting 1D kinetic effects such as plasma instabilities, to improve existing models, or to illustrate plasma lectures.
The advantage of a PIC code is that the equations are simple to solve. The disadvantage of a PIC code is that it as to deal with LOTS of particles. The performance is a big issue
This code only uses Python and standard Python modules:
Numpy, Numba, tkinter, matplotlib, Pytest
Because performance is a really big issue, I compared different functions from different modules, but most of the time the fastest solution was to use Numba.jit.
Is is true by example for interpolation and density estimation.
You just need to clone the repo, and you should be able to execute both main files main.py and Main.ipynb
Feel free to fork and contribute to the code. However, I should specify that this project's main objective is to test solutions for the PlasmaPy project. Hence, contributing to PlasmaPy could be more interesting.
- Physic and Solvers
- Electrostatic particle pusher (Leap-frog scheme)
- Electromagnetic particle pusher (Boris scheme)
- Poisson solver (Thomas's algorithm)
- Wall boundary conditions
- periodic boundary conditions
- interface and user experience
- GUI for initialisation
- GUI of the evolution of the simulation
- diagnostics (data output)
- Restars
- Software part
- Tests
- 95 % Coverage
- Performance profiling
- Continuous integration
- Parallelistation
This project is licensed under the MIT License - see the LICENSE.md file for details
I have to think about it ! The list of my references is long, even though this project is only my own... I'd say mostly Vivien Croes and Trevor Lafleur, for now.
I'll improve the list later.