[perao]

Hi,

to make it short I am going straight to the point. My apologies for the lack of accuracy with the language and/or technicalities with regards to either the programming language or the OpenFOAM's code terminologies.

For the solver in the subject we find under the multiphaseSystem.C all the aliases for the multiphase system types (the ones we must specify in constant/phaseProperties when setting up a case). For example:

    typedef
        ThermalPhaseChangePhaseSystem
        <
            PhaseTransferPhaseSystem
            <
                TwoResistanceHeatTransferPhaseSystem
                <
                    MomentumTransferPhaseSystem<phaseSystem>
                >
            >
        >
        thermalPhaseChangeMultiphaseSystem;

    addNamedToRunTimeSelectionTable
    (
        phaseSystem,
        thermalPhaseChangeMultiphaseSystem,
        dictionary,
        thermalPhaseChangeMultiphaseSystem
    );

Questions:

1.The parameter that is past to these PhaseSystems is the phaseSystem class. Is it correct to say that all these typedefs are also a phaseSystem?

2.Most of member functions used by these PhaseSystems come from the phaseSystem class. For example, the pure virtual member function:

virtual autoPtr<momentumTransferTable> momentumTransfer() = 0

. Now, looking at the innermost class within the above typedef, i.e., the MomentumTransferPhaseSystem, we find this such member function declaration and definition.

In .H:

virtual autoPtr<phaseSystem::momentumTransferTable> momentumTransfer();

In .C:

template<class BasePhaseSystem>
Foam::autoPtr<Foam::phaseSystem::momentumTransferTable>
Foam::MomentumTransferPhaseSystem<BasePhaseSystem>::momentumTransfer()
{ <body's function> }

Moving outwards in the typedef, we pass through the TwoResistanceHeatTransferPhaseSystem class, then this is past to PhaseTransferPhaseSystem class. However, in the latter, we find once more the declaration and the definition of

momentumTransfer()

.

In .H:

virtual autoPtr<phaseSystem::momentumTransferTable> momentumTransfer();

In .C:

template<class BasePhaseSystem>
Foam::autoPtr<Foam::phaseSystem::momentumTransferTable>
Foam::PhaseTransferPhaseSystem<BasePhaseSystem>::momentumTransferf()
{
    autoPtr<phaseSystem::momentumTransferTable> eqnsPtr =
        BasePhaseSystem::momentumTransferf();

    phaseSystem::momentumTransferTable& eqns = eqnsPtr();

    this->addDmdtUfs(totalDmdtfs(), eqns);

    return eqnsPtr;
}

a.Since the PhaseTransferPhaseSystem is a class that deals with mass transfer, is it correct to assume that the momentum appears here again in order to be taken into account as a consequence of the mass transfer process?


b.If so, this is why the MomentumTransferPhaseSystem is an inner class in this typedef, so that by using

BasePhaseSystem::momentumTransferf()

the PhaseTransferPhaseSystem class is actually using the momentum modeling available under the BasePhaseSystem instead of providing a new one?



There are other member functions that could be used in place of the

momentumTransfer()

, but I guess that the same rationale applies to the others as well.



Extra question: If someone would be able to explain better these templated classes under PhaseSystem which are of type BasePhaseSystem, but at the same time inherits from BasePhaseSystem itself, it would be a great help.

Thank you very much for your time.