For convective mixing, store the saturated dissolution factor. #5774
Conversation
|
jenkins build this please |
|
new file is missing from CMakeLists_files.cmake. EDIT: never mind, it's not a new file. |
| const auto& liquidPhaseIdx = (FluidSystem::phaseIsActive(FluidSystem::waterPhaseIdx)) ? | ||
| FluidSystem::waterPhaseIdx : | ||
| FluidSystem::oilPhaseIdx; |
There was a problem hiding this comment.
I don't think you need references here–i.e., auto&. The phase indices are integers so you can just store them by value.
|
Make sense to put rssat in the intensiveQuantities rather than the fluidstate. So I will close my PRs. I my PRs I changed default enableConvectiveMixing to false and only activated it for gas+water+dissolution and gas+oil and it's thermal/diffusion variants. Does it make sense to do something like this in the PR as well? |
| */ | ||
| void updateSaturatedDissolutionFactor_() | ||
| { | ||
| auto& fs = asImp_().fluidState(); |
There was a problem hiding this comment.
Another tiny nit: The fs object is used in just a single location–the call to saturatedDissolutionFactor()–so I guess you could just as well use asImp_().fluidState() directly in that function call.
Gives a significant speedup with DRSDTCON.
Same idea as OPM/opm-common#4364 and #5771, but with some differences. I managed to avoid noticing that one, or I would not have made this...