Makes it clear that UseDivergenceCleaning is and additional bit of code
Updating docs for PoissonDivergenceCleaning
And put it in the docs
update note about comoving Phi source term for Dedner.
There was a note about future implementation of the Phi expansion terms. It has been determined that this is not necessary.
@dcollins4096 , the documentation you’ve created conflicts with PR #412 (which rearranges everything in the doc/manual/source/parameters directory). We should hold off on accepting this PR until #412 gets accepted, and then update your docs. Also, please note that there’s a merge conflict in src/enzo/Grid_PoissonSolverTestInitializeGrid.C that also needs to be fixed!
I fixed the comments @bwoshea made. I left the recommendation about UsePoissonDivergenceCleaning purposefully cryptic; it probably works and does something useful, but the cost/benefit hasn’t been established.
@dcollins4096 Are there methods papers you can link to that describe the implementation of methods 3 and 4 of UsePoissonDivergenceCleaning?
Are you looking for an Enzo related method paper or for more general information regarding conjugate gradient methods?
When we discussed this at the workshop no one was aware of an Enzo related method paper describing the actual implementation/testing, which is why we decided to add a warning in the documentation.
I see! I was just looking for something that described either the implementation in Enzo or a paper that describes the general approach the authors were following. But if there isn’t one, then I think the warning is enough. Thanks!
I’ll run the test suite to ensure that things work. Assuming they do, I think this one is ready to merge!
Don’t worry about it @bwoshea , I can run the test suite!
Not a problem, I know you’re on vacation. I just ran it, and it passes. I’ll go ahead and accept this PR!