Drops dynamics and heat transfer in turbulence

Abstract

In this thesis, the dynamics and heat transfer of large and deformable drops in turbulence are numerically investigated. Direct Numerical Simulation (DNS) is used to resolve the turbulent flow and energy field, while Phase Field Method (PFM) is used to describe the interface behaviour. The first purpose of this work is to study the interaction between drops and turbulence, and in particular to understand how this interaction is affected by density and viscosity differences between the fluids. To this aim, a campaign of simulations of a drop-laden turbulent channel flow is performed, exploring several density and viscosity contrasts and two different surface tensions, and analysing the drops dynamics in terms of deformation, breakage and coalescence. We observe that the effects of density and viscosity are negligible when surface tension is sufficiently high, while they become more significant as surface tension is reduced. In the latter case, they visibly affect drops deformability in opposite ways: while increasing drops viscosity reduces drops deformability, increasing drops density enhances drops deformability. Viscosity, in particular, shows a stronger impact, affecting deformation as well as breakage and coalescence rates. On the contrary, density induces visible deformations of the interface, but does not show any influence on breakage and coalescence events. The flow modifications inside the drops are then evaluated in terms of turbulent kinetic energy (TKE). As expected, an increase of drops density produces a higher internal TKE, while an increase of viscosity causes a strong suppression of turbulence and a reduction of TKE. The drops internal flow can influence the mixing of the heat contained by the drops, and how fast it is transferred to the carrier fluid. The second purpose of this work is to provide a better understanding of this phenomenon, within the passive scalar approximation. The heat transfer in a drop-laden turbulent channel flow is studied by performing numerical simulations, where warm drops are released in a cold carrier fluid. Different Prandtl numbers are considered, which are changed by varying the thermal diffusivity, while keeping a constant momentum diffusivity. Computing the time behaviour of the drops and carrier fluid average temperatures, we find that an increase of Prandtl slows down the heat transfer process. These results are explained by deriving a simplified phenomenological model, showing that the time evolution of the drops average temperature is self-similar, and a universal behaviour can be found upon rescaling by t/(Pr^(2/3)). This scaling can be explained via the boundary layer theory and is consistent with previous theoretical/numerical predictions. In the last part of this work, the study of heat transfer in drop-laden turbulence is extended to evaluate the influence of viscosity differences between the fluids on the heat transfer process. While keeping a constant density and thermal diffusivity, two viscosity contrasts are considered, in order to mimic respectively the cases of warm oil drops in cold water and warm water drops in cold oil. It is observed that in the oil phase, which is more viscous, the heat transfer is slowed down, while in the water phase, which is less viscous, the heat transfer is accelerated. These effects balance each other in the two cases, which therefore result in having an equal heat transfer rate.