Currently working on verifying my CHT with a richard bénard simulation (the solver is not as in a boussinesq approximation, but fully compressible) although the pressure variations are 0. This is because my solver later will need large temperature variations. When verifying I am going for a Gr = 10^3, Ra = 7.1e-2 Using Pr = 0.71, beta = 0.01 (as I simulate ideal gas, from 100 to 99 K), Cp = 1000 (stabilization), P = 1e5, M = 8.314 g/mol, rho = 1 (solution of the paramters). dynamic viscosity = 0.001. For the wanted Gr I find g = 0.1 This leads to alpha = 1.40845e-3 kappa_f = 1.40845 I use Kr = kappa_s/kappa_F = 10 kappa_s = 14.0845 I initialize the temperature field already with T(t=0) = 100 -5/6 x Then I run my simulation, I calculate the heatFlux on the CHT wall with. Finally finding a Nu = 1.5, although it should be 1.04