PORTALE DELLA DIDATTICA

PORTALE DELLA DIDATTICA

PORTALE DELLA DIDATTICA

Elenco notifiche



Innovative approaches to the simulation of turbulent flows in aerospace propulsion systems

01UMKIW

A.A. 2020/21

Course Language

Inglese

Degree programme(s)

Doctorate Research in Ingegneria Aerospaziale - Torino

Course structure
Teaching Hours
Lezioni 15
Lecturers
Teacher Status SSD h.Les h.Ex h.Lab h.Tut Years teaching
Ferrero Andrea   Professore Associato IIND-01/G 15 0 0 0 3
Co-lectures
Espandi

Context
SSD CFU Activities Area context
*** N/A ***    
PERIOD: APRIL-MAY The course is aimed to the presentation of some recent developments related to the numerical simulation of turbulent flows in aerospace propulsion systems. The study of turbulence in these systems is still challenging due to the multiscale nature of the problem and to the complexity of the involved geometries. Large benefits can be obtained by improving the prediction capability on this kind of flows since this would allow to improve the performances and reduce the emissions of next generation propulsion systems. The techniques discussed during the course can be applied to several fields: turbomachinery for aeronautical or civil applications, internal and external aerodynamics, rocket engines for space propulsion, automotive and naval applications. The course starts from a presentation of the state of the art with particular emphasis on Reynolds averaged (RANS) model and Large Eddy Simulations (LES). Some recent developments in the field of machine learning have paved the way to the automatic inclusion of high fidelity data (from experiments or LES simulations) in RANS models. These data-driven approaches can significantly improve the prediction capability of existing RANS models while keeping the computational cost to an acceptable level. These techniques can have several industrial applications for problems in which the cost of LES simulations is prohibitive due to the large Reynolds number which usually characterises this kind of flow fields.
Basic fluid dynamics concepts.
* Introduction and description of the different approaches to the simulation of turbulence (1.5 hours) * RANS methods (3 hours) - State of the art: models with 2, 3 and 4 equations - Examples of flow fields in propulsion systems where RANS methods fail * LES and hybrid RANS/LES methods (3 hours) - Numerical schemes for LES simulations - Problems related to parallelisation, data management and high Reynolds number flows - Example: low pressure aeronautic gas turbine * Improving RANS models with high fidelity data (6 hours) - Importance of RANS models in the design of industrial components - A brief review on machine learning - The field inversion method - The adjoint method: automatic differentiation tools with exercise - Machine-learnt algebraic Reynolds stress models - Physics-informed machine learning - Examples of the previous approaches * Immersed/embedded boundary methods for the simulation of turbulent flows in complex geometries (1.5 hours)
On line synchronous mode
Written report presentation
P.D.2-2 - April