Heat transfer through natural turbulent thermal convection is an important mechanism in geophysics and many engineering problems. A canonical well-known experiment is the Rayleigh-Bénard (RB) cell, which consists of an enclosure with adiabatic side walls uniformly heated from below and cooled from above. The direct correlation between the local momentum and heat transfer is currently under investigation, especially for large aspect ratios. In this situation the dominant large-scale circulation roll, as typically found for low aspect ratios, is substituted by a whole pattern of long-living rolls. Nowadays, commonly advanced optical methods are applied to investigate velocity fields, which makes a transparent heating or cooling plate necessary. For the current study glass plates were coated with a thin metal oxide layer enabling heating via the Joule effect. However, since the temperature homogeneity is crucial for the setup, two heating plates had to be combined to enable a homogeneous temperature distribution at the bottom of the convection cell. Additional difficulties arise from the large aspect ratio and thus the large observation angle. This may cause optical aberrations and systematic errors due to the perspective bias. Furthermore, the observation duration has to be in the order of minutes to hours to resolve the dynamics of the slowly evolving large-scale flow. This results in additional challenges for the seeding generation. To prove the reliability of the experimental approach, a convection cell with an aspect ratio Γ=l/h=10 was placed in the SCALEX (Scaled Convective Airflow Laboratory Experiment) facility. This facility enables experiments with sulfur-hexafluoride (SF6) and air under pressures of up to 10 bar to achieve very high Rayleigh numbers. Here, results of measurements for a Rayleigh number of 5×105 in SF6 at ambient pressure and in air at a pressure of 5.7 bar will be discussed and show the applicability of this approach. Stereoscopic particle image velocimetry (PIV) in horizontal planes covering the whole cross section of the cell (300×300mm2) was used for the estimation of all three components of the velocity vector. The optical access for the laser light sheet was provided by transparent sidewalls. The aim of the current study was to prove the possibility of reliable PIV measurements with reproducible homogenous temperature boundary conditions in the SCALEX facility. Issues of the temperature distribution at the heating plate, tracer particles, illumination and data evaluation will therefore be addressed in greater detail.
Read full abstract