Wrong sign on divu(ueint) and pdivu ?

[marchdf]

I think that the line here: https://github.com/AMReX-Fluids/CAMR/blob/development/Source/Hydro/Hydro_utils_eb_K.H#L107
should be: divu(i,j,k,UEINT) -= pdivu;. And similarly for this line: https://github.com/AMReX-Fluids/CAMR/blob/development/Source/Hydro/Hydro_utils_eb_K.H#L156, it should be divu(i,j,k,UEINT) -= ... . And similarly for the 3D version in that function (4 spots total).

This would make things consistent with the non-EB version:

CAMR/Source/Hydro/Hydro_utils_K.H

Line 256 in 6062c91

update(i, j, k, UEINT) -= pdivu(i, j, k);

The reason for an issue vs a opening a PR is that I am not 100% sure because it seems to make no difference when I run the shock-cylinder case. Thoughts @asalmgren ?

[jbbel]

Marc, i have looked over the code and i believe you are correct. go ahead and submit a PR. this only does something when equation for total energy gives bad results. It may well never be used in a shock-cylinder case. Usually associated with hypersonic stuff from astrophysics or cosmology

[marchdf]

Thanks, John. I fixed it (and a couple others) in #66