Example Pluto run for the Whole Sun convection benchmark

This repository holds an example run for the Whole Sun convection benchmark. This example is meant to be compiled and run with the PLUTO code. This example corresponds to runs #3 of the benchmark with theta=10, sigma=0.1.

Version : This repo has been tested and is working with Pluto 4.4

Files description

The files in the repository are the following :

• init.c: the initial conditions, the boundary conditions and the body-force definitions
• definitions.h: definitions for the run, including user parameters
• pluto.ini: the input configuration file for the run
• tc_kappa.c: definition of the Thermal conduction parameter (kappa)
• visc_nu.c: definition of the viscosity parameter (nu)
• update_stage.c: the update algorithm. The file is modified to force boundary mass flux to zero.
• plot_run.py: the plotting and info extraction script (see below)

How to compile and run

Provided you are working with a UNIX-compliant system, that PLUTO is installed and the $PLUTO_DIR variable is set up, the run is setup using the pluto script :

python $PLUTO_DIR/setup.py

The file definitions.h should already have every parameter setup for the run, so press enter until reaching the makefile configuration. Pick a makefile adapted to your environment and exit the configuration tool. The code is then built using make

make

and finally run, either in serial mode:

./pluto

or in parallel using mpi (if you picked a mpi-compilation makefile in the configuration tool):

mpirun -np 8 ./pluto

Remarks

Since the last version of this repo (feb. 2021) substantial changes have been made. In particular in the treatment of boundary conditions. Before the update, we used to define the boundary conditions in the first cell of the domain (at top and bottom boundaries) to avoid mass loss due to the lack of control in the results of the Riemann solver. Since we have adopted a different strategy where the flux is set manually in update_stage.c. The values in the ghost layers are still set for the diffusive kernels (thermal conduction and viscosity) to take place. This allows us finer control over what is happening at the boundary and hence perfect mass conservation.

Important note : Running this repo as is will generate ~250Gb of data. This is due to the simultaneous generation of Pluto's binary files (data.xxxx.dbl) AND of vtk files (data.xxxx.vtk) for visualization. Feel free to deactivate vtk outputs in pluto.ini to reduce the generated data to ~150Gb.

Plotting and data extraction

We have also provided in the repo a python (3!) script plot_run.py which extracts all the information necessary for the plotting of the runs. The script undergoes the following steps :

1. Rendering the simulation. The results are stored as a series of png files in the folder render.
2. Extracting the time evolution of the simulation. This will result in the creation of the pluto_time.csv file in the format required for the benchmark as well as the creation of a time_evolution.png image with the evolution in time of the mass as well as the kinetic, internal and total energies.
3. Extrating the time evolution of temperature profiles. This will result in the creation of a temperatures.png file where each line corresponds to the temperature profile of a snapshot. By default we draw one line every 50 snapshots (hence 20 lines for 1000 snapshots). All the lines should be starting at z=0 from the top temperature (Ttop=1.0) and should end at z=1 with the same gradient.
4. Finally, the averaged fluxes and profiles for the analysis. These are horizontally averaged and time averaged between the times t=0.895 and t=0.905. This step writes a pluto_prof.csv file with the fluxes formatted correctly for the benchmark analysis.

Each step can be deactivated by using the following command line arguments after calling the python script : --no-render (step 1); --no-time-evolution (step 2); --no-temperatures (step 3) and --no-profiles (step 4).

Example run

The default run assumes theta=10.0, and sigma=1.0. Here are some results obtained with the script and with paraview :

time_evolution.png

temperatures.png

Render plot (at t=30.0)

3D visualization in Paraview at t=30.0