Vortex migration in a disk with a cavity
@alainm @astr-mle @morby @crida @dkloster @hmeheut @crobert
We have recently compared results from fargOCA and Nirvana ( Nirvana testing done by R. Nelson). We have discussed the results on a meeting of the group on monday, so that everybody is on this issue.
We have found discrepancies between fargOCA and Nirvana codes on 3D test with a Jupiter massive planet in zero viscosity disks.
The main difference concerns vortex migration: vortex produced at outer gap edge appear to migrate in our code, they do not in Nirvana.
Clement is studying vortices at the edge of a cavity in 2D disks with the AMRvac code (and finds no vortex migration)
So I have implemented the cavity prescription (Clement, Meheut 2020 submitted paper) in our last code version and in a old fargo version (2D, not in the git repository)
I have done 2 test for each code: using FARGO algorithm for the transport and using standard transport
The result is the following: new code and old code match when using standard transport they do not when using FARGO
Concerning vortex migration: I do not see the vortex migrating (on a time = 800x6.28 code units) in the old code (both FARGO and standard transport) nor in the new version with STANDARD transport I do see the vortex migrating in the new version with FARGO
So we have a problem ( an annoying one not seen on the series of test that we use to check the code behaviour) on the FARGO algorithm
I am working to set up a short test so that we can use git bisect to detect what is the problem,
@alainm then I will probably need your help to do the git bisect I keep you updated
any other idea is welcome