Running a simulation

Running an ibis simulation is done in three parts:

1. Pre-processing
2. Execution
3. Post-processing

Pre-processing

To run your first simulation, change directory into one of the examples.

cd examples/wedge

The first thing is to prepare the simulation. For this, we need the grid. ibis has no grid generation capabilities, so you need to generate the grid using some other software. For this example, grid.py uses gmsh to generate the grid. First install gmsh through pip, then execute the script:

python grid.py

This will write a su2 file.

Now you can prepare the ibis job. This is done with the following command:

ibis prep

This will look for job.py, execute it, and write detailed configuration, grid, and flow files.

Execution

Now it is time to run the simulation. This is done with the following command:

ibis run

This command looks for the files generated in the pre-processing stage, and runs the simulation. It will periodically write flow solutions out (as often as you asked it to).

NOTE: This stage doesn’t look at job.py. If you make changes to that, you have to re-run the pre-processing stage.

Post-processing

The flow solutions written during the simulation are in native ibis format, designed to write as little information as possible for efficiency. Post-processing takes this information and recreates the full flow state, and writes more interesting information in a more widely used format. Currently, only vtk format is implemented.

To generate vtk files, run:

ibis post plot

This will generate vtk files in a directory called plot. You can open plot/plot.pvd using paraview to see the results of the simulation.