Correct handling of diffusive fluxes

In the seminar we came up with the following strategy:

• When useMoles = false we will have to add the diffusive fluxes to the total mass balance equation (eqIdx = replaceCompIdx)
• Fick's law will always be mole based (as in its derivation). However, the fluxes are transformed into mass fluxes if useMoles = false
• Update documentation
• Return vector of fluxes from fick's law and think about maxwell-stefan diffusion when implementing it

An idea for tracer (doesn't need to be conservative in the sense that it could react) was that

• Fluidsystem specify numTracerComponents
• sum x_i = 1 (i = main components, excluding tracer components)
• x_tracer = mass_tracer / mass_phase(excluding tracer mass)
• Tracer compoments get their own advection diffusion equation that is not considered by replaceCompIdx / total mass balance
• Tracer have no influence on molar mass, density, viscosity, ... any other fluid phase property

Would be nice: mechanism to attach a list of tracer to the fluid system at compile time and access it at runtime.

Opened branch fix/fs364-massFormulationDiffusiveFluxes

Move the tracer discussion to issue #365 (closed)

changed milestone to %2.11

marked the task Update documentation as completed

Move "UseComplexRelations" discussion to #367 (closed)

mentioned in issue #366 (closed)

removed milestone

added ~113 label

marked the task Update documentation as incomplete

marked the task Implement static_assert(useMoles || numComp < 3) as incomplete

marked the task Fick's law will always be mole based (as in its derivation). However, the fluxes are transformed into mass fluxes if useMoles = false as incomplete

marked the task When useMoles = false we will have to add the diffusive fluxes to the total mass balance equation (eqIdx = replaceCompIdx) as incomplete

removed ~112 ~97 labels

changed milestone to %6

added ~97 label

assigned to @heck

marked the checklist item Return vector of fluxes from fick's law and think about maxwell-stefan diffusion when implementing it as completed

changed the description

marked the checklist item Fick's law will always be mole based (as in its derivation). However, the fluxes are transformed into mass fluxes if useMoles = false as completed

Return vector of fluxes from fick's law is not implemented? There seems to be an open, yet un-merged branch concerning this issue. This would also fix the bug of the forgotten density on the box Fick's law.

changed the description

marked the checklist item Return vector of fluxes from fick's law and think about maxwell-stefan diffusion when implementing it as incomplete

marked the checklist item Return vector of fluxes from fick's law and think about maxwell-stefan diffusion when implementing it as completed

changed the description

marked the checklist item When useMoles = false we will have to add the diffusive fluxes to the total mass balance equation (eqIdx = replaceCompIdx) as completed

What is the status of this?

I do not think we need the static_assert and I cannot remember why we wanted it? All calculations are now one molar basis so I guess this is obsolete now? As for the documentation, we might need to update that a bit but it should not be too much. We should be able to finish that on Wednesday

changed the description

mentioned in merge request !1132 (merged)

marked the checklist item Update documentation as completed

closed via commit 93713f20

closed via merge request !1132 (merged)

mentioned in commit 93713f20