/test_particle

Test particle code to study particle transport or energization

Primary LanguageC

Test particle simulation

Test particle code to study particle transport or energization. The code is parallelized using MPI and OpenMP, and it use HDF5 for output.

Compile

  1. You need a MPI library and parallel HDF5.
  2. You may need to change some variables in Makfile, including CC, LD, and HDF5_ROOT.

Then, simply make in the root directory.

Run the code

  1. Copy config_files/init.dat and config_files/test_particle.sh into the directory that you are going to run the code. And you need to change init.dat and test_particle.sh. The variables in the files should be self-explained.
    • The included test_particle.sh is designed for using the fields generated by 3D VPIC simulations. You may have to write some code and scripts to load the fields from your own simulations.
  2. You can run the code by ./test_particle.sh.
  3. In default, the code generates four files.
    • diff_coeffs.dat (ASCII): diffusion coefficients. Please check diagnostics.c for details.
    • espectrum.dat (ASCII): energy spectrum. The first row is particle energy bins. The rest rows are spectrum at a series of time frames.
    • particle_diagnostics.h5 (HDF5): particle trajectories.
    • particles_fields_init.h5: initial particle information, including particle position, velocity, time, charge/mass, electric field, and magnetic field.

Plot

You can use the provided Python scripts to plot the data.

  1. Diffusion coefficients
python reconnection_analysis.py --run_name $run_name --run_dir $run_dir \
       --tinterval $tinterval --tframe $tframe --diff_coeff --ptl_vel 0.7

where you have to define the simulation's parameters. For multiple injection velocity, add --multi_vel flag.

  1. Energy spectrum: change --diff_coeff to plot_spectrum in the above script.
  2. Particle trajectory:
    • particle_diagnostics.h5 saves each particle information in a dataset. In order to read the data using ParaView + H5Part, we need to transfer the HDF5 file to a H5Part file using
    python reconnection_analysis.py --run_name $run_name --run_dir $run_dir \
       --tinterval $tinterval --tframe $tframe --tran_hdf5 --ptl_vel 0.7
    Then, we can load the generated file into ParaView and plot all particles at a single time step. You can generate time series be selecting one specific particles using ParaView's selection tools, but the process could be tedious. You's better write a python script to do that.
    • We can also generate piecewise trajectories for a single particle, so have a rough idea of the trajectory curve. It uses pyetvk to generate a single .vtu file for each time step.
    python reconnection_analysis.py --run_name $run_name --run_dir $run_dir \
       --tinterval $tinterval --tframe $tframe --piecewise_traj --ptl_vel 0.7
    You may also need to specify --iptl and --tint to get a specific trajectory and the number of steps before and after current step. Then, you can load these .vtu files into ParaView to show piecewise trajectories. You can use Paraview's ProgrammableFilter to generate vtkPolyData to make the trajectories curves. The ProgrammableFilter is given in vtu_vtkPolyData.py. Then, you can use the tube filter to make the trajectory look even better.