I`ve considered flow of water in complex geometry with orthotrophic porous media blocks. I use constant density model and f(T) (or constant) model for viscosity. I`ve got temperature diminishing in some cells at the boundary: porous media - fluid - wall, but I use fixed temperature (or adiabatic) boundary conditions at the wall and so, i think that this result is unphysical.

If somebody knows the nature of this unphysical result or, maybe, comes across such problem, i will be very thankful for any information.

Actually I have been wondering about this as well. I did a room vent problem some time ago and modelled the human skin as walls of fixed temp (36'C). Loading vertex data gives near-freezing temperatures! There were no heat sinks. Averaged cell data gives the correct value. Or was it the other way round? as CAVE-ed post data will not preserve boundary values.