Master thesis and internship[BR]- Master's Thesis : Optimal criterion to define convergence in unsteady CFD simulations for applications in Formula 1 car aerodynamics[BR]- Internship (linked to master's thesis)
Promotor(s) : Terrapon, Vincent
Date of defense : 25-Jun-2020/26-Jun-2020 • Permalink :
|Title :||Master thesis and internship[BR]- Master's Thesis : Optimal criterion to define convergence in unsteady CFD simulations for applications in Formula 1 car aerodynamics[BR]- Internship (linked to master's thesis)|
|Translated title :||[fr] Critère optimal de convergence pour les simulations numériques d'écoulements instationnaires appliqué à l'aérodynamique d'une Formule 1|
|Author :||Cotteleer, Léo|
|Date of defense :||25-Jun-2020/26-Jun-2020|
|Advisor(s) :||Terrapon, Vincent|
|Committee's member(s) :||Andrianne, Thomas
|Number of pages :||115|
|Keywords :||[fr] CFD, aerodynamics, convergence, statistics|
|Discipline(s) :||Engineering, computing & technology > Aerospace & aeronautics engineering|
|Target public :||Researchers|
Professionals of domain
|Institution(s) :||Université de Liège, Liège, Belgique|
|Degree:||Master en ingénieur civil en aérospatiale, à finalité spécialisée en "aerospace engineering"|
|Faculty:||Master thesis of the Faculté des Sciences appliquées|
[fr] Computational Fluid Dynamics (CFD) analyses have become a tool of choice to deepen phys- ical understanding and a precious help to solve practical problems. In the light of the continuous enhancement in computer power, the use of this new numerical tool is not planned to stop. This growth in computing capacity is leading to an increasing interest in scale-resolving turbulence such as detached-eddy simulation (DES). The unsteady pattern of the flow in Formula 1 racing therefore results to simulations that are very computational resources consuming. The resolution of numer- ous turbulent scales requires small time steps. In addition, time-dependent simulations take time to develop from arbitrary initial conditions to a statistically stationary state such that long samples are necessary for reliable statistical estimates. The initial transient duration and the required simulation time are highly case-specific, user burden and a priori unknown.
This work introduces a methodology that automates the initial transient estimation during a numerical CFD simulation and the prediction of the required simulation time to obtain reliable statistics. The techniques rely on statistical considerations for aerodynamic quantities of inter- est and have shown satisfactory results, in line with the a posteriori intuitions. In the two first parts, a rigorous mathematical background is provided and the methodology is described as well as its previous versions. In particular, a chapter is dedicated to the estimation of the statistical error by means of a second order model. The discussions and results provide a wide range of im- plementation issues and test cases. Especially, Delayed Detached Eddy Simulations (DDES) are performed for common bluff bodies (high angle-of-attack airfoil, triangular cylinder, rotating cir- cular cylinder). Finally, the methodology has been implemented as a User-Defined Function (UDF) in ANSYS Fluent and in the last chapter are exposed the different interesting quantifications that can be extrapolated from the methodology. A concise summary of the main outcomes is provided in the conclusion. The latter is written in such a way that each principle finding is linked to the corresponding section.
Cite this master thesis
The University of Liège does not guarantee the scientific quality of these students' works or the accuracy of all the information they contain.