How to run the EU hydrocode comparison test problems with FARGO


A recent project coordinated by Stockholm Observatory consists in  comparing the behavior of different hydrocodes on simple disk-planet problems. Namely, there are four such problems, involving respectively a Jupiter mass planet and a planet of one tenth of a Jupiter mass (called for short 'Neptune') embedded either in a viscous or in a inviscid disk. FARGO has been run on these test problems, and the four corresponding parameters files (stockholm*.par) are provided in the distribution. FARGO was however written before the file format specifications of the EU test problem were defined, so a module which outputs the torques and inner/outer disk masses, in the format required by the Stockholm team, was recently added to FARGO (see src/stockholm.c). In order to activate this module at runtime, FARGO must be launched with the '-e' flag. As a result, an additional file is produced in the output directory, named 'torque0.dat' (for the first planet; for the second one it would be 'torque1.dat', etc. The EU test problems however only involve one planet at a time). Therefore, launching the EU test problem involving a Jupiter mass planet embedded in a viscous disk is done by issuing the following command in the fargo directory:

./fargo -e in/stockholm_MJ_visc.par

In addition to the fine grain torque sampling every 1/20th of orbit, a snapshot of hydrodynamics variables is output every orbit, in the standard FARGO format. In order to convert these raw format files to the ASCII files required by the Stockholm group, one can use the following simple IDL routine. Just copy it to the output directory, and there, from the IDL prompt, issue something like:

IDL> convert,nb=10,radix='1MJ-visc'

if, for instance, you wish to convert files output at t=10 orbits.

Note: the runs performed by the above commands correspond exactly to the Stockhom group specifications (potential smoothing length, resolution, mesh size, time sampling, etc.) except for the boundary conditions. A desirable property of the boundary conditions is that they avoid reflection of the wake triggered by the planet. For this purpose, the Stockholm group recommends to impose a damping of disk perturbations in narrow annuli near the boundaries, whereas FARGO uses a built-in non-reflecting boundary condition which gets rid of the reflected wake without any damping (see figures here).