Project: P-TURB - Prandtl number effect on TUrbulent Rayleigh-Bénard convection

Prace Call: 17th
ID: 2018184426, Leader: Enrico Stalio
Affiliation: Università degli Studi di Modena e Reggio Emilia, IT
Research Field: Engineering
Collaborators: Sergio Chibbaro University Pierre et Marie Curie FR , Andrea Cimarelli Università Politecnica delle Marche IT , Andrea Fregni Università degli Studi di Modena e Reggio Emilia IT , Paolo Gualtieri Sapienza Università di Roma IT , Francesco Battista Sapienza Università di Roma IT
Resource Awarded: 65 Mil. core hours on Marconi - KNL


Thermally driven turbulence is a phenomenon of relevance in various areas of science and technology, as it affects geophysical and astrophysical systems and occurs also in industrial applications. Its importance for atmospheric flows involves both small length and timescales for weather predictions and large scales for climate calculations. Rayleigh-Bénard convection, the buoyancy-driven flow in a fluid layer heated from below and cooled from above involves the main thermally driven turbulence features and is a classical problem in fluid dynamics. It is also a rich multi-physics problem with fundamental outcomes in heat transfer and important fluid mechanics implications where the concomitant action of turbulence, plumes and large scale circulation strongly influences the transport properties of the medium. This does not include only heat transfer characteristics but more generally the transport of substances or particles dispersed in the fluid. The aim of this project is to perform three numerical simulation of the Rayleigh-Bénard convection, representing the same physical phenomenon occurring in three very different fluids of interest for applications: liquid metals, air and water. Numerical simulations proposed are as close to reality as possible, because no turbulence model is involved in the calculations and spatial as well as temporal discretisation grids are so fine as to be able to catch all the spatial and temporal scales actually involved. The three simulations will help reserchers to understand important physical processes related to the statistical properties of turbulent fluctuations and buoyancy-induced turbulence, These include the large-scale dynamics of the plumes and its interaction with small-scale fluctuations, the turbulent cascade, and the instantaneous paths of mechanical and thermal energy in the flow.