[yosuu]

Hello Foamers!


I have discovered a serious bug (in my opinion) in the buoyantBoussinesq solvers (version 2.1.1). I checked how the kappa is implemented, and i have my doubts that it is correctly done.

In the original code is:

kappat = turbulence->nut()/Prt;
kappat.correctBoundaryConditions();

volScalarField kappaEff("kappaEff", turbulence->nu()/Pr + kappat);

but i think that there should be a Cp multiplying the kappa terms, so the final code

kappat = Cp*turbulence->nut()/Prt;
kappat.correctBoundaryConditions();

volScalarField kappaEff("kappaEff", Cp*turbulence->nu()/Pr + kappat);

because from definition of Prt=nut/alphat=nut*Cp*rho/kappa. Since the equations are incompressible, that ends up as Prt=nut*Cp/kappa.

Note: to make it work, Cp must be added in readTransportProperties.H from buoyantBoussinesqSimpleFoam.

What do you think about it? When I run laminar cases, it takes forever to see the temperature field to change from the initial condition when using the original code.

[yosuu]

ok, i might have been wrong all the time. So, kinematic thermal conductivity is not the same as normal thermal conductivity coefficient...