Title:	Noise generation by turbulent boundary over an airfoil
DNr:	Berzelius-2021-39
Project Type:	LiU Berzelius
Principal Investigator:	Ardeshir Hanifi <hanifi@kth.se>
Affiliation:	Kungliga Tekniska högskolan
Duration:	2021-10-04 – 2022-05-01
Classification:	20306
Homepage:	https://www.kth.se/mech/flow-stability-turbulence
Keywords:

Abstract

Noise emitted by airfoil, wings and blades occurs due to the interaction of turbulence with the neighbouring solid surfaces. At low angles of attack and high Reynolds numbers, boundary layers over an airfoil surface are often turbulent. The interaction of advecting turbulence with the trailing edge of the airfoil cause noise. The generated noise is a result of the trailing edge inducing an abrupt change of boundary conditions for pressure and/or velocity. Trailing-edge noise is the dominant sound-generation mechanism for an airfoil at low angle of attack. Despite long-time research on the topic, still adequate prediction of the generated noise is a very difficult task. The difference between the predicted and the actual level of sound can be several decibels, mostly because the noise generation mechanisms are not fully understood. The goal of the proposed project is to further investigate the mechanism of noise generation in compressible turbulent boundary layers using advanced methods of analysis (e.g. SPOD and resolvent analysis). Especially, we aim at understanding the role of wavepackets inside the turbulent boundary layers, which recent works have confirmed their existence and correlation with the farfield noise, in this process. It must be mention that, due to computational costs, there are only few works on the subject in the literature and they are mostly based on simulations of incompressible flows. This type of studies requires highly resolved simulations requiring HPC resources. In the present work we perform wall-resolved large-eddy simulations using the PyFR code which has shown good performance on GPUs. Preliminary tests have been performed at PDC (Tegner system). In our first simulations we consider an airfoil with a chord Reynolds number of 100,000. The gained insight in turn will allow us to design/optimize more efficient noise- control devices. Further, the investigation will help us to improve prediction of generated noise by applying relatively simple numerical tools to the turbulent flow field given by ordinary Raynolds-Avergaed Navier-Stokes computations.