Friday, 14. December 2018

Turbulent thermal and mixed convection with obstacles

Turbulent thermal convection between two horizontal plates with lower heated and upper cooled flat surfaces, has been the subject of numerous experimental and numerical studies. This problem is known as turbulent Rayleigh-Benard convection (RBC). In most applications, like heat exchangers or the climatization of aircraft cabins, turbulent thermal convection appears in much more complicated domains or as mixed convection with additional pressure gradients driving the flow.

In the present project we therefore investigate thermal and mixed convection in box-shaped geometries with obstacles and in parts with additional driving pressure gradients by means of Direct Numerical Simulations (DNS) and Large Eddy Simulations (LES).

On the one hand, four obstacles are attached to each the heated bottom and the cooled top plates to model surface roughness. The influence of  such surface roughness on the  heat transport is important in experimental studies of classical RBC, since with growing Ra the thermal boundary layer thickness decreases monotonically, and any wall roughness becomes influential above a certain Rayleigh number. Industrial investigations of heat transfer aim at the development of more effective heating or cooling devices, which can reduce energy consumption. Consequently, the objective of applied research is often to increase the heat transfer, while keeping constant the difference between the temperatures of the heated and cooled plates. One of the possible ways to increase Nu is to use rough surfaces for the plates instead of smooth ones. Therefore, a comprehensive analysis and advancement of the theory of the heat transfer in convection cells with rough walls are still required. In the present work, the impact on the mean heat transport of periodically distributed roughness elements is studied.

On the other hand, we investigate instantaneous and statistical characteristics of turbulent mixed convection around heated obstacles and compare the obtained results to those of pure forced convection without a driving temperature gradient, for different Grashof numbers and Reynolds numbers. The chosen computational domain, which is a parallelepiped with four parallelepiped obstacles, two inlet ducts and two outlet ducts, represents a generic ventilated room or aircraft cabin.


Instantaneous temperature distributions (Tc<T<Th) with superimposed velocity vectors for Rayleigh number Ra=10e8, Prandtl number Pr=1 and different roughness types.
Sketch of the generic room with inlet and outlet ducts
Instantaneous temperature distribution for Rayleigh number Ra = 10e8, Prandtl number Pr = 0.7


Prof. Dr. Claus Wagner
German Aerospace Center (DLR)
Institute of Aerodynamics and Flow Technology, Department Ground Vehicles
Phone: +49 551 709-2261




German Aerospace Center (DLR), Institute of Aerodynamics and Flow Technology, SCART
Bunsenstraße 10, 37075 Göttingen, Germany