15 th International Conference on Fluid Control, Measurements and Visualization 27-30 May, 2019, Naples, Italy Full-field and time-resolved tomographic PIV of turbulent thermal convection inside a cylinder Gerardo Paolillo 1,* , Carlo Salvatore Greco 1 , Tommaso Astarita 1 , Gennaro Cardone 1 1 Department of Industrial Engineering, University of Naples “Federico II”, Naples, Italy * corresponding author: [email protected] Abstract Rayleigh-Bénard convection, the buoyancy-driven flow induced by temperature gradients par- allel to the gravity, is relevant to a variety of processes spontaneously occurring in nature or used in technological applications. Understanding of its inherent chaotic evolution is possible only through mea- surement techniques that allow to access the unsteady, three-dimensional behavior of all the thermo-fluid dynamic quantities of the flow. This work presents full-field and time-resolved tomographic particle im- age velocimetry measurements of non-rotating and rotating Rayleigh-Bénard convection inside a slender cylinder with aspect ratio equal to 1/2 at Rayleigh and Prandtl numbers equal to 1.86 × 10 8 and 7.6. Keywords: thermal convection, rotating convection, particle image velocimetry 1 Introduction Rayleigh-Bénard (RB) convection, the buoyancy-driven flow induced by temperature gradients parallel to the gravity, plays a key role in a large number of natural processes (wind generation, oceanic circulation, convection in the stars’ interior, etc.) and technological applications (such as turbomachinery flows, melting of pure metals and gas separation based on centrifugation). In several contexts, a background rotation occurs to significantly change the dynamics of the thermal convection. For instance, the interplay between the buoyancy and the Coriolis forces is at the basis of the formation of tropical hurricanes and it might even explain the genesis of the Earth’s magnetic field according to the geodynamo theory. Since the study of such phenomena is very complicated, the basic mechanisms of RB convection are typically investigated in confined, basic geometries, like rectangular or cylindrical cells [1, 2, 5]. This allows, on one side, to simplify the theoretical analysis, on the other, to obtain comparable results from the experimental and numerical sides, since the flow boundary conditions are easily reproducible in both cases. In particular, the cylindrical geometry is of much interest, because it has one direction of statistical homogeneity (the azimuthal one, provided that the cylinder is leveled), which prevents the existence of preferential orientations in the development of the large-scale structures of the turbulent flow. The inherent turbulent nature of RB convection makes three-dimensional measurements mandatory for the analysis of such a phenomenon. In this work, the time-resolved tomographic particle image velocimetry (PIV) is used to investigate the whole domain of RB convection inside a cylinder with aspect ratio equal to 1/2 at Rayleigh and Prandtl numbers equal to 1.86 × 10 8 and 7.6, respectively. The effects of rotation on the flow dynamics and evolution are also investigated in similar operating conditions by varying the Rossby number. 2 Experimental setup The convection cell is a Plexiglas cylinder filled with water, heated from below by a flat heater coupled with a copper plate and cooled from above by means of a transparent water exchanger. Both plates are maintained at constant temperature by means of a thermoelectric controller performing a PID control with a stability up to 0.01 ◦ C. The flow is seeded with fluorescent orange polyethylene seeding particles illuminated by a Nd:YAG laser shaped into a cylindrical beam, which covers the whole cylinder interior. Image particles are recorded by four sCMOS cameras (Andor Zyla, 5.5 megapixels), arranged in an planar configuration with a constant angular spacing of 40 ◦ . The digital resolution of the cameras is about 14 voxel/mm, while the sampling frequency is 7.5 Hz. This is sufficient to obtain time-resolved measurements by virtue of the slow velocities of the thermal convection. The duration of each experiment is 4 hours, in such a way that a reliable statistical analysis can Extended Abstract ID:25 1