Please use this identifier to cite or link to this item: https://hdl.handle.net/2440/113426
Type: Theses
Title: The effect of unsteady flow on wind turbine wake development and noise generation
Author: Sedaghatizadeh, Nima
Issue Date: 2018
School/Discipline: School of Mechanical Engineering
Abstract: The wake of a wind turbine is a low velocity region characterised by a complex flow structure and increased turbulence intensity. The interaction of the wake with downstream turbines not only results in power loss, it also increases the emitted noise level and fluctuating loads on the turbines blades. Due to their significant effect on wind farm performance and development, the study of wind turbines wake and the noise emitted from wind farms have been the focus of a great deal of research. However, the structure of the wake, its development and its effect on noise emission and propagation from wind turbines are still not well understood. Aerodynamic noise emitted from wind turbines adversely affects the perception of communities towards wind energy development. Thus, it is vital to find the dominant noise generation mechanism and predict its propagation accurately. Most models developed to predict noise emission from wind turbines are semi-empirical and are based on airfoil noise prediction techniques. These models lack the ability to accurately model the propagation and refraction of the emitted noise, and hence the predicted directivity does not agree well with the observations in the field. In addition to the limitation of these models due to the assumptions they are based on, experimental investigation in this field which are also used to develop wake models, is expensive and difficult in terms of matching the boundary conditions of the experimental setup with the actual conditions in the field. The effect that wakes have on noise emission and performance of wind farms is another concern for the industry, especially during the planning design process. To address this concern, several semi-empirical and analytical models have been developed to predict the wake development, predominantly focused on estimating the velocity deficit in the wake region. Although these models are simple and can provide a reasonable estimation for calculating the power loss in a wind farm, they are not able to accurately determine the effect of incoming instabilities, atmospheric boundary layer, yaw misalignment and wake interaction. It is well known that sound is refracted through the interaction with vortices and shear layer in the flow. The wake of a wind turbine consists of complicated vortical structures accompanied with a shear between the low velocity region in the wake and free stream. Moreover, the wake of a wind turbine is highly affected by different parameters such as, instabilities in the incoming flow, wake interaction and the atmospheric boundary layer. Thus, in addition to the incoming flow condition which affects the noise generated by a turbine, the propagation of the noise signals from wind turbines is also affected by the wake behaviour and its development in different conditions. This thesis commences with a meta-data analysis based on the reported noise perception surveys, combined with the publically available information on the directivity of the airfoil noise at different pitch angles. This approach was used to provide quantitative support for the hypothesised underlying physical mechanisms for noise generation which have been previously reported in literature. The results of this study show that underlying mechanism associated with the perceived noise in the far-field of the turbine blades is amplitude modulation due to partial stall on the blades or interaction of the blades with incoming turbulent structures. This is in contrast with the common belief of trailing edge noise to be the dominant source for the perceived noise and its amplitude modulation due to rotation of the blades. Moreover, it is observed that the majority of the locations in which the noise from whole turbines is perceived by the community, are in close proximity of the wake region. This shows the potential role of the wake on noise propagation patterns and should be accounted for. To develop a detailed understanding about the noise generation mechanisms, especially stall noise, the aeroacoustic behaviour of an airfoil, which is the fundamental element of wind turbine blades, was investigated numerically. A 3D model of a NACA0012 airfoil in a wind tunnel was developed and the flow field was calculated using an embedded large eddy simulation technique. The flow field and aeroacoustic results were validated against experimental data available in published data. Using Ffowcs Williams and Hawkings analogy, the noise signals were calculated from the surface pressure data of the airfoil under different angles of attack. It was shown that the frequency of the noise decreases as the angle of attack increases and the blade experiences deep stall. On the other hand, the integral length scale of the separated vortices increases as the angle of attack increases, as does the noise level. Results show a perpendicular dipole directivity for the peak frequency when the blade is in stall conditions. However, at low angles of attack the peak frequency directivity shifts towards the leading edge of the airfoil. This is in contrast with the results obtained from the trailing edge noise theory and shows the existence of alternative noise sources near the trailing edge of the airfoil which cannot be predicted by the trailing edge noise theory. A first step towards investigating the effect of wind turbine wake on the noise emission and propagation is to study the wake development and its structure. An embedded large eddy simulation model was developed and validated against experimental data. For studies conducted in wind tunnels, the simulated results show that vortical structures in the wake propagate to a distance of up to 20 diameters downstream of the wind turbine when the turbine is confined by wind tunnel walls. The distances that the vortical structures propagates reduces as the walls are eliminated from the computational domain. The length of the turbine wake significantly reduces when the turbine is located in an atmospheric boundary layer, where the wake breakdown occurs at 12 diameters downstream of the turbine with strong downwash due to formation of the longitudinal vortices around the turbine and wake area in the atmospheric boundary layer. Strong downwash in the presence of the atmospheric boundary layer results in a higher velocity magnitude with less turbulence in the far-field region of the wind turbine in this case when compared with the wake formation in uniform flow. This new knowledge may assist wind farm developers in achieving higher turbine densities for future wind farms. The wake development and its structure varies as the terrain on which the turbines are erected varies. It is expected that higher turbulence intensity and shear within the incoming flow results in a reduction in the wake length and stronger downwash due to stronger longitudinal vortices. To calculate the noise signature in far-field, the Ffowcs Williams and Hawkings (FW-H) aeroacoustic analogy was applied to the developed CFD model for wind turbine. Results revealed that, the highest noise level in the vicinity of a wind turbine corresponds to blade pass frequency and is due to amplitude modulation of the trailing edge noise caused by the rotation of the blade, as well as blade tower interaction. Results also showed that the emitted noise is refracted due to the wind shear in atmospheric boundary layer, as well as the wake and associated turbulent structures. The CFD outcomes showed that the wake breakdown occurs at a distance of 12 diameters downstream of the turbine with a strong downwash due to longitudinal vortices. Contours of sound pressure level at the breakdown location of the wake of the wind turbine showed refraction and modulation of the sound at this location. Results also revealed refraction of the noise towards the ground and wider areas due to existence of the longitudinal vortices.
Advisor: Arjomandi, Maziar
Kelso, Richard Malcolm
Cazzolato, Benjamin Seth
Ghayesh, Mergen
Dissertation Note: Thesis (Ph.D.) (Research by Publication) -- University of Adelaide, School of Mechanical Engineering, 2018
Keywords: Research by publication
wind energy
wind turbine wake modelling
atmospheric boundary layer
wake interaction
wind turbine noise
noise directivity
noise propagation
Provenance: This electronic version is made publicly available by the University of Adelaide in accordance with its open access policy for student theses. Copyright in this thesis remains with the author. This thesis may incorporate third party material which has been used by the author pursuant to Fair Dealing exceptions. If you are the owner of any included third party copyright material you wish to be removed from this electronic version, please complete the take down form located at: http://www.adelaide.edu.au/legals
Appears in Collections:Research Theses

Files in This Item:
File Description SizeFormat 
Sedaghatizadeh2018_Phd.pdf12.79 MBAdobe PDFView/Open


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.