Politecnico di Torino (logo)

Study of turbomachinery flow using Large-Eddy Simulation

Kristi Shtembari

Study of turbomachinery flow using Large-Eddy Simulation.

Rel. Francesco Larocca, Andrea Ferrero. Politecnico di Torino, Corso di laurea magistrale in Ingegneria Aerospaziale, 2020

PDF (Tesi_di_laurea) - Tesi
Licenza: Creative Commons Attribution Non-commercial No Derivatives.

Download (8MB) | Preview

The present thesis is dedicated to the study of turbulent flows inside an aeronautical engine turbine by means of Large-Eddy Simulation (LES). For many decades, the turbomachinery flow has been one of the main focuses of research in the Computational Fluid Dynamics (CFD). The significant developments achieved in aero-engines performances have brought to different complex geometries and flows being employed. Thus, there is a greater need for high fidelity simulations to be performed in order to reach a certain level of accuracy and quality in the design and optimization processes of such flows. Due to its reduced computational cost with reference to Direct Numerical Simulations (DNS) and much higher accuracy if compared to RANS, the LES approach has become a very powerful analysis tool in many engineering fields nowadays. Even though many issues haven’t been addressed yet, the LES approach has proven its effectiveness in complex flow geometries such as Low Pressure Turbine (LPT) or gas turbine combustors. In the present work, a first insight on the LES formalism and Sub-Grid Scale closure models has been presented. An existing code based on a Discontinuous-Galerkin finite element method was used to perform an LES on the mildly loaded LPT cascade blade in its T106A configuration. The impact of the adopted discretization schemes and physical models on some of the variables of interest was analyzed and the achieved numerical results were compared to experimental data and other numerical work present in literature. The motivation of this work was to retrieve high accuracy data of the flow field in the whole computational domain, in order to improve existing RANS models by techniques of Machine Learning.

Relators: Francesco Larocca, Andrea Ferrero
Academic year: 2020/21
Publication type: Electronic
Number of Pages: 69
Corso di laurea: Corso di laurea magistrale in Ingegneria Aerospaziale
Classe di laurea: New organization > Master science > LM-20 - AEROSPATIAL AND ASTRONAUTIC ENGINEERING
Aziende collaboratrici: UNSPECIFIED
URI: http://webthesis.biblio.polito.it/id/eprint/17048
Modify record (reserved for operators) Modify record (reserved for operators)