Natural convection - Theoretical aspects

Heat (or mass) transfer by convection occurs when heat (or the component) is transferred to, or through, a flowing fluid. As opposed to forced convection, where a flow is created by an external force (pressure gradient), convection is said to be free or natural if fluid movement is due solely to variations in density. These variations are generally caused by a non-uniform distribution of temperature and/or species concentration in a mixture.

The difference between these two convection modes is essential in that, in the case of forced convection, temperature or concentration can be passive scalars, advected by an otherwise imposed flow. In natural convection, on the other hand, density gradients are the driving force behind the flow of the fluid, which also transfers heat or solute to its surroundings. There is therefore a fundamental coupling between the velocity field and the various scalar fields.

Understanding the mechanisms of natural convection is a major scientific challenge, enabling us to model them and predict their effects in the environment or in equipment. Understandably, engineers have to deal with a wide variety of situations, depending on the size and geometry of the field of study, the nature of the fluids involved and their thermophysical properties, the boundary conditions and the magnitude of the density gradients involved, and so on. The aim of this article is not, of course, to cover every possible configuration, but to set out the physical principles governing this type of flow and show how they affect heat or mass transfer. We shall therefore confine ourselves to the most fundamental aspects, considering only simple, sometimes academic, configurations. We will try to refer readers to the main references if they wish to delve deeper into more specific points or more complex situations, or if they are looking for specific correlations to solve practical problems (see on this subject the "Thermal and mass convection" articles [BE 8 205][BE 8 206]