Patch: make Refluxing infrastructure independent of GRHydro

Issue #1862 closed
David Radice created an issue

GRHydro can optionally run with conservative AMR using the "Refluxing" thorn. The infrastructure for refluxing is very generic, but the "Refluxing" thorn currently contains GRHydro-specific code that makes it incompatible with any other solver.

To overcome this limitation, I prepared:

  1. A patched version of the Refluxing thorn, with a generic MoL-like registration mechanism for the fluxes
  2. A new thorn containing all of the old GRHydro-specific code from Refluxing.

The new code passes all existing tests and allow other (M)HD thorns to register variables for refluxing in a very simple way. The code is in two git repositories on bitbucket:

https://bitbucket.org/dradice/refluxing https://bitbucket.org/dradice/grhydro_refluxing

Notes:

  1. GRHydro parfiles that use conservative AMR should be modified to include GRHydro_Refluxing
  2. The new Refluxing thorn no longer supports delayed refluxing, which was never used for any production run, and/or special GRHydro-specific atmosphere hacks.

Keyword:

Comments (10)

  1. Roland Haas
    • changed status to open
    • removed comment

    I added minor comments on grhydro_refluxing to https://bitbucket.org/dradice/grhydro_refluxing/commits/8f972d5dd86c72cb9fe6ac473dc3640d8be152d4 . Nothing big, mostly just some unused routines (are yout ''supposed'' to be unused ?) and a request to not use sprintf and to document a purpose for the thorn. Otherwise grhydro_refluxing is fine with me.

    I added a single minor comment to refluxing here https://bitbucket.org/dradice/refluxing/commits/faad5bc55245e102022a958dde6918f30d2d3b43 .

    Note that the new refluxing removes various experimental code that never quite worked as desired. This is with approval of the author (Erik) and user (Christian Reisswig) for whom these were added (as far as I know). While this technically removes features from an ET thorn, the features were highly experimental.

  2. David Radice reporter
    • removed comment

    Roland: the latest version of the GRHydro_Refluxing thorn should address your questions. I have not changed the parameter name in the Refluxing thorn: do you consider it necessary for inclusion in the ET?

  3. Log in to comment