Random-vortex-particle methods applied to broadband fan interaction noise

The general aim of this thesis is to investigate the suitability of a stochastic method for computational aeroacoustics, the specific objective being to devise a stochastic method to generate synthetic turbulence and combine it with the linearised Euler equations to predict broadband fan interaction noise. In modern turbofan designs broadband fan noise is a dominant source of aircraft noise, the most efficient source being the interaction between upstream turbulence and the stator vanes. The stochastic method developed to generate synthetic turbulence reproduces twodimensional isotropic turbulent flows by filtering a random field. The fillter is expressed in terms of the energy spectrum and controls the spatial properties of the synthetic turbulence. In contrast with previous work, non-Gaussian filters are developed to model more realistic energy spectra such as Liepmann and von Karman spectra. The temporal decorrelation present in turbulent flows is modelled using Langevin Equations. A standard Langevin equation and a second-order Langevin model are derived in details and validated for fan interaction noise. In contrast with classical methods to generate synthetic turbulence, random-vortex-particle methods can be extended to cope with inhomogeneous non-stationary turbulence with little modification from the formulation for homogeneous turbulence. The stochastic method is applied for first time to broadband fan interaction noise. The method is firstly validated for frozen turbulence interacting with an airfoil. The temporal decorrelation is then included in the method to assess the influence of the integral time scale on the radiated acoustic sound field. The method is also combined with an existing wake model to represent the inhomogeneous non-stationary turbulent flow found downstream of a fan. Finally, comparison with existing experimental data for an isolated airfoil in a turbulent jet demonstrates the benefits of using more realistic energy spectr