Turbulent Evaporation and Condensation – TEC2

To study experimentally the behavior of droplets in a turbulent flow, it is essential to follow their trajectory in real time. This is a real challenge, since it is necessary to determine at each moment the position in space and the diameter of fast-moving droplets ranging from a few tens to hundreds of microns, in a space of several tens of centimeter-cube. In this project, we have developed several new techniques to solve this problem: digital holography, which involves making holograms using a laser and a fast CCD camera and using a reverse method to reconstruct the images and determine the size of the drops all along their movement; 3D tracking, using four high-speed cameras, combined with new enlighting methods and data processing. The experiments were carried out in a specific high turbulence wind tunnel, in an original «turbulence box« using loudspeakers to produce turbulence and a setup where turbulence is produced by the rotation of two counter-rotating disks !Von Karman). The experimental approach has been complemented by massive numerical simulations, in which the exact equations of fluid mechanics are solved, both to study the phase change at the scale of a single bubble or a network of bubbles, and for very large sets of point particles (of the order of a billion) in a homogeneous and isotropic turbulent flow.

Among the significant results, we would highlight the non-uniformity of the rate of evaporation of the drops during their trajectory, which depends closely on the history of the drops in the turbulence. Indeed, they tend to explore particular regions of the flow, which significantly modifies the convective exchanges. From these results we can hope to obtain more accurate evaporation laws than the empirical laws used up to now. This result is to be associated with the discovery by numerical simulation of the role of the temperature and concentration fronts on the phase change, the particles accumulating preferentially in these regions. The simulation also revealed an unexpected mechanism of multiple collisions between the same particles, which could have significant consequences on coalescence, and in particular the growth of raindrops. The experiments carried out show the role of zones of zero acceleration on segregation, and confirm, at least for the parameter range discussed here, one of the mechanisms proposed to explain the phenomenon.

This project made it possible to answer several open questions about the effect of turbulence on phase change in dispersed diphasic medium. The perspectives offered relate to several points:
- Improve existing models for the evaporation or condensation laws in the presence of turbulence. This work is ongoing.
- Introducing the finite size effects of particles in macroscopic models: the present study has shown the limits of the point particle approach. Very interesting results have already been obtained on the scale of some objects, the passage to the macroscopic scale remains to be done.
- Develop Lagrangian models to describe the phenomena of segregation and preferential concentration and their effect on evaporation / condensation. TEC2 has shown that the number of Stokes is not sufficient to describe the process, and stochastic models have already been proposed.
- Continue investigations on interparticle turbulence collisions, which show the importance of multiple collisions even in a priori diluted environments.

The results of this fundamental study were published in 25 peer-reviewed journals and 50 national and international conferences.

Dispersed two-phase flows occur in many industrial applications and natural phenomena. In the past decade, significant progress has been achieved in this field, due both to a revolution in experimental techniques (PIV, PLIF, PTV…) and to the extraordinary development of numerical simulations. Interestingly, this domain has been tackled by different scientific communities, in Fluid Mechanics, Physics, Meteorology, or Geophysics, with different focuses and different methods. The case of inertial particles is an emblematic example, with the development of efficient Eulerian or Lagrangian models in Engineering, associated with deep insights in the dynamics of particle-turbulence interaction by physicists studying turbulence, which helped to improve current models. The participants to this project were involved in this global effort and mixing of scientific cultures in a previous ANR program DSPET, devoted to inertial particles. The case of dispersed flows involving mass transfer, due to evaporation, condensation, or dissolution or coalescence is even more intricate, due to the dependence in time of the size of the particulates (drops, bubbles, and particles), which precludes some convenient approximations such as point-particles and one-way coupling. Among the many fundamental difficulties of the problem, the present project objective is to focus on the interaction between turbulence and phase change in particulate flows. Although the specific features of flows with either drops, bubbles or particles make it unlikely that all situations can be described by a single general framework, we believe that common physical principles can be identified in all cases, and that much can be understood with the newly developed experimental and numerical tools. Since the problem is clearly multiscale, the project involves three steps corresponding (i) to a local study at the scale of the particles (Bubble, drop or solid particle), (ii) to the intermediate level (mesoscale) of a swarm of interacting inclusions, basically through turbulent induced collisions, (iii) to the macroscopic level of mean field equations used in applications. Although several aspects of the problem have been investigated, other important aspects, such ascrossing trajectories effects and preferential concentration, or finite size effects on mass transfer and forces, have received little attention until recently. The goal of this proposal is to take advantage of recent advances in the investigation of this long-standing problem, by using the latest available techniques, and a strong coupling between experiments, simulations and modeling. The challenge is to determine both the evaporation/condensation rate along the trajectories, and the local characteristics of the continuous phase in the vicinity of the particle. Experimental techniques developed in the previous ANR program DSPET by the same group (high-speed PTV, high-speed holography, PIV+PLIF) appear to be well adapted for this purpose. Numerical methods accounting for severe evaporation conditions will also been developed. In a second methodological step, we shall consider situations involving many condensing/evaporating bubbles or drops. In such situations, phase changes can be greatly modified by two classes of mechanisms: collisions and coalescence processes, and collective effects. The first one involves essentially binary collisions between particulates, which are affected in a still incompletely understood way by turbulence. The four teams involved in the project, at the LMFA, the Physics Laboratory at ENS Lyon, the LEGI and the OCA, gather some unique competences in advanced measuring techniques and theoretical and numerical modeling. They have a strong experience of common projects, and use to share experimental equipments and numerical codes.