Session: 7.1 - Heat and Mass Transfer in multiphase flows

Paper Number: 158101

158101 - Mass Transfer in Bubble-Laden Flows 

Abstract: 

We use Direct Numerical Simulation (DNS), coupled with a Volume of Fluid (VoF) method, to study the mass transfer process of a chemical species in a wall-bounded turbulent bubbly flow. Bubbles, which are initially fully-saturated with a given chemical species, are injected into the turbulent flow, in which the chemical species is initially not present. This induces the mass transfer (flux of species) from the bubbles (whose volume fraction is about 5%) to the carrier turbulent flow. The main physical parameters that govern the problem are the Schmidt number Sc (momentum to mass diffusivity ratio) inside the bubbles and in the carrier flow, the bulk Reynolds number Re (inertia to viscous forces ratio) and the Weber number We (inertia to surface tension forces ratio). In fact, we keep the Reynolds and Weber number constant and equal to Re = 10000 and We = 3100, respectively, and we vary the Schmidt numbers considering two different cases. Case A, where the Schmidt number in the carrier flow, Sc_c, is varied from 0.5 to 4, and is equal to the Schmidt number inside the bubbles, Sc_b, so that the diffusivity ratio between the carrier flow and the bubbles is unitary, D_r = Sc_c/Sc_b = 1; and Case B, where Sc_c is varied from 0.5 to 4, while the Schmidt number inside the bubbles is kept constant and equal to Sc_b = 0.01, so that D_r is systematically increased from 50 to 400. Following this strategy, we can analyze the interaction between turbulent transport and molecular diffusivity on the mass transfer rate. Our results show peculiar scaling behaviors of the Sherwood number Sh (i.e. the dimensionless mass transfer rate), with Sc. For Case A, Sh scales as Sh~Sc^1/2, consistently with classical theoretical predictions. In contrast, for Case B, Sh scales as Sh~Sc^2/3, highlighting the important role of the diffusivity ratio D_r on mass transfer efficiency. We expect these findings to help the development of improved models and parametrizations of the process.

Presenting Author: Francesco Zonta Newcastle University

Presenting Author Biography: