[Gfs-devel] patches for conservation validation in MHD module

[zhangbruno <zha...@gm...>]

Hi,Stephane

   As you've mentioned some days before, I modify the module and the
test in order to check  whether the current density is conservative  in
a control volume, which is adhered in attachment(and also I push them
into the repository).

   The results showed the maximum divergence of current density is
smaller than 1.0e-3 . 

   As I calculated, the divergence of current at a center of a control
volume is obtained by using current density fluxes at the cell face.

  But there is another way to calculate the divergence by using
gfs_divergence(cell,magn->j),it's obtained by using the value of j in
neighbor cells.However, because the current density flux is located on
cell face , so I think it's more convincing using the first method.How
do you think of that?


Cheers

Aaron