T equation

Dear all,

I am looking into the scalar transport models in order to implement them in another solver. I've found that all of the employed transport equations are of the form below (taken from Teqn in cfdemSolverPisoScalar):


fvScalarMatrix TEqn
(
fvm::ddt(voidfraction,T) - fvm::Sp(fvc::ddt(voidfraction),T)
+ fvm::div(phi, T) - fvm::Sp(fvc::div(phi),T)
- fvm::laplacian(alphaEff*voidfraction, T)
==
Tsource
+ radiation->ST(rhoCpRef, T)
);
<\code>

I'm confused by the 2nd and 4th terms. It seems they are there to remove that special and temporal variation of the void fraction from the equation. Is this because this is a transport equation for temperature, not for thermal energy (intensive vs extensive property)?

I'm trying to implement non-constant k, rho and Cp, which are being represented by volScalarFields. I've pasted the changes I made below, but these result in unphysical heat-fluxed resulting from gradients in k/rho/Cp. If someone has experience with this, could you give me some feedback?

Many thanks!


fvScalarMatrix TEqn
(
fvm::ddt(voidfraction*rho*Cp,T) - fvm::Sp(fvc::ddt(voidfraction)*rho*Cp,T)
+ fvm::div(phi*fvc::interpolate(rho*Cp), T) - fvm::Sp(fvc::div(phi)*rho*Cp,T)
- fvc::div(k*voidfraction, fvc:grad(T))
==
Qsource
);
<\code>