Combustion noise prediction using linearized navier-stokes equations and large-eddy simulation sources

Wolfram Christoph Ullrich, Christoph Hirsch, Thomas Sattelmayer, Kilian Lackhove, Amsini Sadiki, André Fischer, Max Staufer

Research output: Contribution to journalArticlepeer-review

11 Scopus citations

Abstract

In modern aeroengines, combustion noise has become a significant source to the overall noise, particularly at approach conditions. This requires further advances in understanding and predicting combustion noise of turbulent flames. This is the aim of the present study, where a hybrid computational fluid dynamics/computational aeroacoustics approach is applied on a generic premixed and pressurized combustor to assess its accuracy for combustion noise predictions. The hybrid approach consists of Reynolds-averaged Navier-Stokes (RANS) or largeeddy simulations (LES) mean flow and frequency-domain simulations based on linearized Navier-Stokes equations that are fed by combustion noise source terms. The latter are obtained from both the application of a statistical noise model on the RANS simulations and a postprocessing of incompressible LES, in a first step. The acoustic simulation results are compared with experimental pressure measurements conducted by the Centre National de la Recherche Scientifique. Very good agreement is found over the entire frequency range if the LES source model is applied. Sensitivity studies with respect to feeding lines, mean flowfield, and mesh were performed. The resulting comparisons of the linearized Navier-Stokes equation simulations based on the RANS and LES flowfields revealed that the combustion noise spectrum is mainly governed by the heat release spectrum but not by the aerodynamic combustor flowfield. However, this issue needs further investigation.

Original languageEnglish
Pages (from-to)198-212
Number of pages15
JournalJournal of Propulsion and Power
Volume34
Issue number1
DOIs
StatePublished - Jan 2018

Fingerprint

Dive into the research topics of 'Combustion noise prediction using linearized navier-stokes equations and large-eddy simulation sources'. Together they form a unique fingerprint.

Cite this