Afivo-streamer is a tool for the simulation of streamer discharges in 1D, 2D and 3D, using plasma fluid models. Afivo-streamer makes uses of the afivo framework, which for example provides adaptive mesh refinement and parallelization.
Recommended reading for new users
Git repository links
Features
- Adaptive mesh refinement
- OpenMP parallelization
- Multigrid field solvers
- Plasma fluid model for streamer simulations
- Photoionization using a Helmholtz or Monte Carlo approach
- Support for tabulated transport data
- Support for user-defined chemical reactions
- Flexible usage through configuration files and command line arguments
More details about the implementation are given in the publications describing afivo-streamer.
Source code structure
| Folder | Contents |
| src | The source code of afivo-streamer |
| src/config_fortran | Module for configuration files |
| src/lookup_table_fortran | Module for lookup tables |
| src/rng_fortran | Module for random numbers |
| afivo | The afivo library |
| documentation | The documentation (mostly Markdown files) |
| lib_1d | The afivo-streamer library (1D version) |
| lib_2d | The afivo-streamer library (2D version) |
| lib_3d | The afivo-streamer library (3D version) |
| makefiles | Makefiles for building the library and programs |
| programs | Streamer simulation programs |
| tools | Python scripts for e.g. converting data |
| transport_data | Input data files for simulations |
Opiniated comparison against typical alternatives (advantages and disadvantages)
Strong points
- Computational efficiency: relative to typical other codes, afivo-streamer has a high computational efficiency, see e.g. this paper. This usually makes it possible to do 2D simulations on a normal computer rather interactively.
- Possibility of doing 3D simulations: due to the computational efficiency it is possible to do relevant 3D simulations in a reasonable time on typical hardware
- Usage: the code has frequently been used for streamer simulations in various publications, which also means that things like plasma chemisty and different photoionization models are included
- Open source: anyone can in principle change the code (although doing so might not always be easy)
Neutral points
- Linux/unix: basic familiarity with such systems (and the command line) is required to compile and run the code
- Fortran: experience with Fortran is required in order to change the code
Weak points
- Complexity: it can be rather difficult to change the code or to add new physics, in particular due to the use of adaptive mesh refinement and parallelization
- Research type code: which means there is not that much documentation and there is no graphical user interface
- Complex geometries: although curved electrodes can be included, curved dielectrics and complex geometries are generally not supported
Links and related software
- Lxcat website, where input data can be obtained
- Bolsig+, which can be used to compute transport data from electron-neutral cross sections
- Particle_swarm, which can also be used to compute transport data using a Monte Carlo approach
- Chombo discharge, “A multiphysics code which uses Chombo for discharge simulations with adaptive mesh refinement (AMR) on embedded boundary grids”, developed by Robert Marskar
- PASSKey, "Parallel Streamer Solver with Kinetics" by Yifei Zhu and others at LPP
