Turbulence modulation across the interface of a large deformable drop

Abstract

In this work, momentum transfer across the interface of a single, large deformable droplet in a turbulent channel flow is investigated by direct numerical simulations of turbulence coupled with the diffuse interface approach for interface tracking. A wide range of Weber numbers (ratio between inertia and surface tension) is explored, focusing the analysis to cases of large non-breaking droplets where the two fluids have same density and viscosity and the shear Reynolds number is . Results show the presence of wall-blocking effects where the interface-normal velocity is deflected to the interface-tangential direction. These effects depend on surface tension: more deformable is the droplet, the smaller is velocity deflection. In the proximity of the droplet, at distances comparable to the droplet diameter, the interface-normal velocity shows a decay rate that is similar to that observed in the proximity of flat interfaces. As a result, convective effects at the droplet interface are damped, momentum transport across the interface is limited and turbulence intensity inside of the droplet is dramatically reduced.

Keywords:

• droplet
• turbulence
• diffuse interface

Acknowledgement

The authors acknowledge Partnership for Advanced Computing in Europe (PRACE) for awarding us access to resource Julich Research on Petaflop Architectures (JUROPA) based in Germany at Julich. The EU COST Action MP0806 ‘Particles in turbulence’ is also acknowledged.

Notes

1. At least three mesh points are necessary to fully resolve the interface with the current methodology employed. Larger number of mesh points can be required according to the accuracy of the numerical scheme adopted.

2. The Chebyshev transform, adopted along the wall-normal direction, produces a non-uniform grid spacing highly refined at the walls where large gradients have to be resolved.

3. The extension to larger distances can lead to an increasing statistical non-homogeneity of the sample, thus different analysis framework should be adopted.

4. As in Section 4.2, the result analysis through PDs allow to treat the sample non-homogeneity.