In this article, the limits of the validity of the Furry, Jones and Onsager theory (FJO) has been studied both numerically and experimentally. The commercial Solver Fluent^® (Fluent Inc.) has been used in this study, as a computational fluid dynamics (CFD) tool. It was used in order to simulate the variations of the steady separation that will take place in a thermogravitational column, over a wide range of the product of the Grashof number (Gr) and the Schmidt number (Sc). On the other hand, we have used a thermogravitational column for experimental measurements of the achieved stationary separation in two dilute mixtures of polystyrene in toluene. Separation measures were taken at different values of Grashof and Schmidt numbers.

Both the numerical and experimental results show the validity of the FJO theory whenever the work conditions satisfy the relation:  , were A is the aspect ratio between the height of the thermogravitational column and the work gap (annular space). Outside the validity range the separation strongly decreases with respect to that given by the FJO theory.