[frubicle93]

Hi all,

I am trying to solve an advection-diffusion problem in OpenFOAM for a passive scalar (say temperature) where I have a Dirichlet boundary condition on the flow outlet (velocity field that advects the flow is correctly computed using SIMPLE).

Usually, we would place a fixedValue temperature at the flow inlet and a zeroGradient at the outlet, but I am considering a problem where I have a Robin/mixed condition at the inlet and fixedValue at the outlet.

The solver I am using for temperature is from scalarTransportFoam.C.

Depending on the divscheme (linear or upwind), I either get a constant solution (value dictated by Robin condition with gradient term zeroed) with oscillations at the outlet to get to the fixedValue, or I get a solution that is constant near the inlet and outlet with a smooth transition between them at some location in the domain.

I am starting to think that this might be a bit ill-posed or maybe difficult for OpenFOAM to solve? Has anyone managed to get this type of problem working well?