PHARE
Parallel hybrid particle-in-cell code with adaptive mesh refinement
PHARE is an open-source hybrid particle-in-cell code for collisionless plasmas. Ions are followed as macro-particles; electrons are a massless, isothermal fluid whose momentum equation gives the electric field. Patch-based adaptive mesh refinement (SAMRAI) refines the grid in user-defined boxes or where the magnetic field varies sharply. PHARE runs in 1D, 2D and 3D with MPI and is driven from Python.
1D · 2D · 3Dinterpolation orders 1 to 3
AMRpatch-based (SAMRAI), refinement ratio 2
MPIoptional dynamic load balancing
Pythoninput scripts and post-processing
GPLv3licence
News
- MHD open, reflecting and inflow boundaries
- Particle splitting corrected to Smets et al. (2021); refined 2D results change
- Second-order mesh refinement in space and time, divergence-free B
- MHD adaptive time stepping
- Finite-volume MHD solver, 1D, 2D and 3D
- 3D hybrid runs with mesh refinement
- VTK-HDF output, readable by ParaView
- Dynamic load balancing across MPI ranks
- PHARE method paper, Computer Physics Communications
- 2D automatic refinement
- Checkpoint and restart
- 2D hybrid runs with mesh refinement
- 1D automatic refinement on magnetic field gradients
- First 1D run with self-consistent refinement
- Particle splitting paper (Smets et al.), Computer Physics Communications
- Python interface to set up and run simulations
- First two-level 1D run: Alfvén wave through a refined patch
- HDF5 output: fields, ion moments, particles
Work in progress
to validate- in review External magnetic field
- in review Ion-ion collisions
- in review Choice of refinement criterion, including Löhner's
- in review Installation with pip
- draft MHD and hybrid levels in one run
- draft Multithreading
Week-by-week development activity: activity reports.