Appel à candidatures | Recherche, Emploi

PhD: Towards rare events simulations of a bistable turbulent jet

Du 1 septembre 2025 au 31 août 2028

Laboratoire de Mécanique des Fluides de Lille
Contacts : joran.rolland@centralelille.fr

Un sujet de thèse (ci joint) sur la simulation d'évènements rare en turbulence est ouvert au Laboratoire de Mécanique des Fluide. Il s'adresse à des étudiants de master 2 de mécanique des fluides, de physique ou de mathématique appliqué avec de bonnes bases en dynamique des fluides et en simulation numérique. A PhD position is open in the Laboratoire de Mécanique des fluides de Lille. It concerns the use of rare event simulation methods for the study of multistable jet. Interested student can contact the supervisor with a CV and the email address of lecturers/supervisors that can write recommendations

Turbulent flows are not altogether irregular and are often organised in large scale coherent circulations. In some cases, one finds several distinct such circulations for a given value of the dimensionless control parameters of the flow. In that case, the flow is multistable and can transit from one circulation to another. This phenomenon is important
for aerodynamics of transportation (like an Ahmed body) and ubiquitous in atmospheric and oceanic flows with a high impact on climate.

We study multistability in in LMFL in the case of the bistable jet in the wake of two parallel bars. In that flow, the jet can point either leftor right of the centerline, provided the gap between the bars is narrow enough. So far we have mainly worked on data sampled during experimental campaigns. This gave the occasion to study the effect of the distance between the bars on bistability, display the specific bifurcation toward bistability and train analytic stochastic models that highlighted the bifurcation and rationalised the transition rates. However, many questions as to the physical mechanisms driving this bistable flow remain open. This PhD topic proposes to study the configuration numerically and in particular take the successive steps towards the use of rare events simulation methods to efficiently sample many realisations of change of 
direction of the jet at a very reasonnable cost. Such a numerical study will help understanding key physical mechanisms related to the creation of these bistable states and the transition from one to the other. Moreover, it will address the methodological challenge of performing rare event simulations of three dimensional high Reynolds numbers turbulent flows. This project will build upon the fact that Large Eddy Simulations of these flows have recently shown a good compromise between affordability and reliable rendering of transitions between turbulent bistbale states.
Moreover it will take advantage of the supervisor’s experience on the topic of rare events and bistability modelling and simulation, as well as his dedicated suite of code.