06 octobre 2022

Webinaire Alberto Vela-Martín

ZARM, University of Bremen
Causality-based modelling of drop breakup in turbulence

Abstract: The dynamics of inmiscible binary mixtures in the presence of turbulence are largely controlled by the breakup of fluid particles, which needs to be accurately modelled to predict the evolution of these flow. This, however, remains a challenge due to our poor understating of the mechanisms that cause turbulent breakup. In this talk we will derive a causality-based model of drop breakup that accurately reproduces the probability of drop breakup in homogeneous isotropic turbulence. We used a novel GPU code to perform thousands of independent direct numerical simulations of single drops at different Weber numbers, and gathered statistics of the breakup process for drops in mixtures of equal density. This extensive database reveals that, for Weber numbers close to the Hinze diameter [1], the breakup of single drops resembles a memoryless statistical process characterised by a single parameter, the breakup rate. This quantity depends exponentially on the inverse of the Weber number, but not on the Reynolds number of the drop. To explain these results, we consider breakup from an energetic perspective, and leverage our recent findings on the mechanisms that cause the increments of the surface energy [2]. Causal analysis is applied to the deformation of drops to show that breakup is driven by the stretching of the drop surface by eddies far from the drop, whereas the interactions between the surface of the drop and eddies close to it provide only a ‘passive’ dissipation mechanism. This causal picture is used to construct a model based on extreme-value theory which reproduce the memoryless statistics of breakup and the scaling of the breakup rate observed in our simulations. This model can be easily extended to different material properties of the fluid-fluid mixture (bubbles or viscous drops), and suggests that the breakup of drops at low Weber numbers is caused by the interaction of the drop with the most extreme events in the flow. An important implication of these results is that, although the breakup probability becomes exponentially less likely with decreasing Weber, it is non-zero below the maximum stable diameter given by Hinze [1]. It is argued that this is a consequence of intermittency effects in the flow, which should be explicitly considered by breakup models.

[1] Hinze, J.O., AIChE 3, 289–295 (1955).
[2] Vela-Martin, A, Avila, M, J. Fluid Mech. 929 (2021).

06 octobre 2022, 16h3017h30
Webinaire (veuillez contacter J.-P Laval ou F. Romano pour le lien)

Prochains évènements

Voir l'agenda
01 décembre 2022

Wébinaire Esteban Ferrer

New avenues in high order fluid dynamics

Esteban Ferrer est professeur de mathématiques appliquées à l'école d'aéronautique (ETSIAE-UPM). Il a obtenu son doctorat à l'université d'Oxford (Royaume-Uni) et possède 20 ans d'expérience industrielle et universitaire dans le développement de techniques numériques pour les problèmes de fluides. Il travaille activement avec l'industrie et coordonne deux projets Européens. Ses principaux intérêts sont les méthodes d'ordre élevé pour la dynamique des fluides, la modélisation de la turbulence, l'apprentissage automatique, l'aéroacoustique pour l'aéronautique et l'énergie éolienne. Il a rédigé plus de 90 articles de journaux et de conférences sur ces sujets.