ETD

Archivio digitale delle tesi discusse presso l'Università di Pisa

Tesi etd-06092006-094508


Tipo di tesi
Tesi di laurea specialistica
Autore
Giannotti, Emanuele
Indirizzo email
Emanuele.Giannotti@poste.it
URN
etd-06092006-094508
Titolo
Development and evaluation of an integrated method for liner wall temperature prediction in gas turbine combustors
Dipartimento
INGEGNERIA
Corso di studi
INGEGNERIA AEROSPAZIALE
Relatori
relatore Paganucci, Fabrizio
relatore D'Agostino, Luca
relatore Lazzeretti, Renzo
Parole chiave
  • Trent 500
  • Thermal Analysis
  • Rolls-Royce
  • Combustors
  • CFD
  • Aircraft Engines
  • Wall temperature
  • Gas Turbines
  • Jet Engine
  • Finite Elements
Data inizio appello
03/07/2006
Consultabilità
Non consultabile
Data di rilascio
03/07/2046
Riassunto
The effort to reach higher gas turbine efficiencies and reduce the specific fuel consumption has led modern aircraft engines towards a steady increase of the temperatures within the combustor. At the same time, the number of tests carried out on this component during the development phase is decreasing, due to the will to shorten the overall design time. It is then apparent how important would be the availability of an accurate method to predict combustor wall temperatures, and thus the thermal-fatigue phenomena. During the six months spent at the Rolls-Royce Combustion Systems Department in Derby (UK), an integrated method for the prediction of combustor tile wall temperatures (using Computational Fluid Dynamics and Finite Elements Analysis tools) has been devised, implemented and evaluated. The method consists in using the network-based software Flownet to predict the annulus temperatures and mass flows, which are imposed as boundary conditions for a CFD analysis of the combustor liner – carried out using PACE – and for Sc03, which is the standard FEA tool used for temperature and stress prediction in Rolls-Royce engines. The results have been compared with thermal paint data obtained in tests on the Rolls-Royce Trent 500 combustors. It has been found that currently the method cannot be considered entirely predictive, due to the necessity of an extensive matching on the thermal model. The necessary improvements to the CFD code and to the empirical correlations implemented in Sc03 are also suggested.
File