Computational aeroacoustics based on sound sources obtained by a fast random particle-mesh method : assessment and application to aircraft wing slat

Despite considerable improvements in high-performance computing technologies enhancing the capabilities of CFD methods in the last decades, aeroacoustic simulations remain highly expensive in terms of computational time, making it less suitable for industrial purposes. This is mainly due to the numerical resolution required to solve the turbulence generating aeroacoustic sources. This master thesis focuses on an alternative method to numerically simulate aerodynamically generated noise at a reasonable cost. The Random Particle-Meshtechnique (RPM) is a stochastic method that consists in reconstructing a noise source by synthesizing turbulent velocities that accurately represent the turbulence properties provided by a steady simulation of the flow. The present work is divided in three parts. In a first step, a theoretical background of the RPM method is presented, detailing the synthetic turbulence generation. The second part presents a validation of the RPM methodthrough a first implementation on an elementary Matlab code. The statistical noise results will be computed on a synthetic benchmark. The final part consists in simulating a slat-wing configuration using an in-house RPM code of SIEMENS. This constitutes an unprecedented study case for the code on a slat-wing geometry. The application problem was selected to reproduce the radiated sound field and experimental setup of Marc Terracol et al. from the Investigation of the Unsteady Flow and Noise Generation in a Slat Cove study (2016). The turbulence statistics reconstruction is analyzed and improved through a parametric analysis before comparing the computed acoustic predictions to results found in the literature and to experimental anechoic wind tunnel data.