Fluid-particle interactions: from the simple pendulum to collective effects in turbulence
Interaction particule-fluide : du pendule simple aux effets collectifs en turbulence
Résumé
Flows transporting particles are ubiquitous in natural and human–made systems. The study of material inclusions interacting with a carrier flow, that can be laminar or turbulent, is therefore not only an activity of scientific interest by itself, but it is also related with lots of natural processes and it has many industrial applications. We can focus in the study of a lonely particle in a laminar flow, where it is still a challenge to find and understand the forces resulting from the coupling to the flow (specially drag and lift forces), with direct applications to aerospace and automotive industries. On the other hand, the focus can be directed to the dynamics of many particles in a turbulent flow, modeling collective phenomena present in nature, like in clouds, astrophysical flows, etc... Although this phenomena have been known for centuries, it is only in the last two decades when proper time and space resolution has been possible to achieve in experiments and numerical works. This is due to the development of new experimental techniques such as high–speed imaging and particle image velocimetry and numerical tools like large-eddy and direct numerical simulations of high Reynolds number flows. In many cases, this new data has been in contradiction with many theoretical models of particles in turbulence, showing that we have to reconsider part of our previous knowledge of fluid– particle interactions.
This PhD work is therefore based in a systematic study of these interactions, in an effort to explore the problems and limitations of current models. These phenomena includes an enormous family of different systems, taking into account the richness of geometries and flows that can be considered. The purpose of this work is to contribute to the understanding of many different scenarios, using similar experimental and analytical tools. We show that what we can consider a priori as different effects (instabilities, hysteresis, clustering) are caused by very deep and simple principles: the richness that emerges when Navier–Stokes equations are coupled with different mechanical systems. This work is of great importance to the understanding of social phenomena, as each part has many natural, industrial and environmental applications.
The manuscript is organized in two parts. The first part focuses on the interaction of a single particle with a mean flow at different turbulent intensities, when the particle motion is constrained to one and two degrees of freedom. From a fundamental point of view, this situation is motivated by the necessity to better understand the drag (and lift) effects for particles with inertia in a turbulent flow. Besides, this configuration has direct applications in the context of towed systems, which are important for many practical situations. For instance acoustic streamers, where a sonar array is towed at the tip of a long cable attached to a ship or a submarine are commonly used to detect and analyze sonic signals in the ocean. Aerial systems towed by aircrafts have also been used for express mail delivery in the first half of the twentieth century, and applications are still considered for precision payload delivery or snatch pick-up, aerial refueling and low-altitude atmospheric, research among others.
The second part of the work studies turbulent flows with many inertial particles. An inertial particle is a material inclusion that does not follow the flow. A striking feature of such turbulent flows laden with inertial particles is the so-called preferential concentration or clustering that leads to very strong inhomogeneities in the concentration field at any scale. This effect has been proved to be ubiquitous in many environmental, natural and industrial situations: sedimentation in rivers, rain formation in warm clouds, plankton dispersion in the ocean, optimization of chemical reactors and of various industrial processes including combustion of liquid fuel. This phenomenon is also essential to the understanding of pollutant dispersion, particularly important for emergency situations such as radioactive leaks and chemical or biological attacks.
Cette thèse est organisée en deux parties. Après une brève discussion présentant la soufflerie du LEGI et des techniques expérimentales utilisées (chapitre 2), une première partie étudie les effets individuels des particules dans les écoulements tantôt laminaires et turbulents. Les propriétés d’un pendule immergé dans une écoulement moyen sont analysées dans le chapitre 3. Les chapitres 4, 5 et 6 abordent ensuite la question de la dynamique d’objets tractés. Le Chapitre 4 concerne les instabilités de sillage d’une sphère tractée dans un réservoir d’eau tandis que les autres chapitres portent sur la dynamique d’objets tractés dans la soufflerie sous un écoulement laminaire (chapitre 5) et turbulent (chapitre 6), pour lesquels les instabilités aéroélastiques jouent un rôle prépondérant.
Dans une seconde partie je me suis intéressé aux effets collectifs d’une population dense d’inclusions en interaction avec un champ turbulent. Trois chapitres traitent la question de la concentration préférentielle : des bulles d’air dans un canal à eau (chapitre 7), des particules solides dans un écoulement en eau de von Kármán (chapitre 8) et des gouttelettes d’eau en turbulence de grille active dans la soufflerie (chapitre 9). Le premier chapitre de cette partie (chapitre 7) propose une introduction générale de la problématique de la concentration préférentielle de particules en turbulence, alors que le dernier chapitre (chapitre 9), présente quelques conclusions générales.
Loading...