With a symmetric rotor, tip vortex helices develop regularly before interacting, following the leapfrogging instability. The latter can occur earlier when the helices are placed at an initial radial offset, which is realized by considering blades of different lengths. This study investigates the spatio-temporal development of near-wake behavior for rotors with a significant blade length difference. Large eddy simulations and actuator line method on a modified NREL 5MW turbine under various inflow conditions were conducted to evaluate the impact of blade length differences ranging from 5 to 30%. The development of tip vortex helices, the onset of leapfrogging, vortex merging, and ultimately their three-dimensional breakdown were monitored. The analysis is corroborated by using a simplified (two-dimensional vortex-filament model).
The results show that the leapfrogging process begins immediately downstream of the vortex release when blades of different lengths are considered. The instability growth rate obtained from the 2D vortex model agrees with the LES data. Although the rotor asymmetry accelerates the leapfrogging and, in some conditions, the vortex merging process alone, it proves insufficient to cause a large-scale breakdown of the helix system and enhance wake recovery. Under laminar and 5% turbulence inflow conditions, the rotor asymmetry has minimal effect on momentum deficit recovery. In other conditions, no appreciable variations of wake recovery can be observed.

This thesis proposes techniques to produce wall-shear stress estimates from three-dimensional Lagrangian particle tracking (LPT). Several works have already faced the problem of determining near-wall velocity and in particular skin friction for the case of a flat surface. Here, the problem is translated to generic three-dimensional objects. The work makes use of an experimental database, recently produced, with LPT (Hendriksen et al. 2024), that provides in-situ registration of the object, for three shapes of increasing complexity: a cube, an airfoil and a cyclist. Four techniques are examined and compared, including interpolation method as well as local data regression. The uncertainty is evaluated a-posteriori, comparing the results with a local coin-stacking technique, where applicable. All techniques yield accurate and robust representations of the skin friction lines around three-dimensional objects, allowing for an insightful inspection of the near-surface flow topology. Instead, distinct differences are found when the skin-friction magnitude is estimated.

The increasing population density in metropolitan areas, coupled with the faster pace of everyday life, calls for the development of Inter-Urban Air Mobility (UAM) solutions. Currently, there is a gap in the market since there is no competitor that can facilitate all market segments and also provide “last-mile transport” by air. This report aims to identify a promising Inter-UAM design, the Swing, focused on minimising the aircraft’s ground and environmental footprint. The small ground footprint will enable the vehicle to take off from smaller vertiports, helipads or even driveways, allowing it to penetrate urban areas and land where other UAM vehicles cannot.

Particle Tracking Velocimetry (PTV) is an effective and non-intrusive flow measurement technique that is able to provide quantitative information of the full velocity field at a certain instant in time, by tracking each particle individually with high-speed cameras. This permits to have a very high precision in measuring the flow evolutions. The abundant time information can then be leveraged by reconstruction algorithms not only to bring the information into a Cartesian format, but also to increase the space resolution of the original measurements. This thesis aims to explore the enhancement of the Navier-Stokes based reconstruction method called VIC-TSA (Vortex-in-Cell with Time-Segment-Assimilation) through the implementation of Radial Basis Functions (RBF). Moreover, it investigates whether the incorporation of RBFs can enable VIC TSA to match or exceed the accuracy of the VIC+ method. The improvements are evaluated numerically with a synthetic flow field (Taylor-Green sine wave vortex lattice) and experimentally (flow over a bluff body) with a real Lagrangian Particle Tracking (LPT) experiment in the TU Delft W-Tunnel.

Understanding and predicting vortex behavior is crucial to maximize performance using aerodynamic design. Vortex breakdown can be induced to benefit the flow field or has to be prevented to avoid losing performance. It is therefore important to study the dynamics of vortex breakdown.

The main objective of this thesis was to assess the feasibility of measuring a free, streamwise vortex and its breakdown from data acquired with robotic volumetric Particle Image Velocimetry (PIV), whilst analyzing the relation between the local flow properties and the location of the breakdown. Experiments were conducted at the W-Tunnel at the TU Delft High-Speed Lab, where vortex breakdown was induced by the adverse pressure gradient field in front of a cylinder. The robotic Coaxial Volumetric Velocimeter (CVV) was used to assess the feasibility of using this system for vortex breakdown identification, in combination with data processing using a Lagrangian Particle Tracking (LPT) technique (Shake-The-Box), to detect and analyze vortices and their breakdown. The results of this experimental campaign were analyzed to identify vortices and their breakdown from the CVV data by characterizing the vortex’s internal structure. The analysis included establishing the relations of the adverse pressure gradient and the Reynolds number with the breakdown location.

The results showed that vortex breakdown is induced by the adverse pressure gradient. However, the breakdown location is not solely dependent on it. Furthermore, the effect of the Reynolds number at the breakdown location is barely measurable with the robotic volumetric PIV system.

This exploratory and experimental study showed that the CVV in combination with LPT can be used to recognize the structure of the vortex in both its pre- and post-breakdown regimes. However, for the analysis of vortex breakdown, it was found that the data was affected by the low spatial resolution. Thus, robotic volumetric PIV cannot provide conclusive results in terms of the relation between the location breakdown and local flow properties. Several recommendations regarding the experimental set-up are given for future research.

Wind-turbine noise can restrict the growing implementation of renewable energy sources and their application close to urban environments. The largest contributor to the noise of modern turbines is the scattering of the turbulent fluctuations at the blade trailing edge. This source of noise is directly correlated with the turbine’s extracted power. Therefore, operating in noise-restricted environments and at night times entails lower energy production. An extensively applied solution for reducing the noise of wind turbines is the use of trailing-edge serrations, i.e. imposing periodic variations in the geometry of the blade trailing edge. Serrations reduce the effectiveness of the scattering at the trailing edge as the turbulent fluctuations reach the trailing edge at different times along the blade span, consequently reducing the wind-turbine noise. Although extensive literature and knowledge exist on serrations, their measured performance does not compare with the predicted one. Even more problematic, the trends predicted for the geometric alterations of the serrations are not observed in reality. Notably, two things are worth mentioning: first, geometries shown as optimal by theory perform worse than other concepts, and second, the noise from serrations is affected by the angle between the insert and the flow. As a result, the design of trailing-edge serrations still requires dedicated experiments and numerical simulations, hampering the assessment of several geometries necessary for complete optimization of the serration design. This work seeks a physical interpretation of the noise generation mechanisms of trailing edges with serrated add-ons. This interpretation is focused on understanding the underlying physical principles of the flow surrounding a serrated trailing edge. This is carried out in this work with three studies, respectively on the observation, modelling, and control of the flow and acoustic properties of serrated trailing edges...@en

The available energy density supplied by conventional light sources prohibits the use of time-resolved volumetric Particle Image Velocimetry (PIV) with submillimetre helium-filled soap bubbles (HFSB) on a cubic metre scale. In connection with attempts to expand the maximum attainable measurement domain of PIV, a research project is carried out into the upscale of the seeding particles, capable of scattering more light. The production flexibility and scalability of HFSB is investigated by designing and improving the conventional HFSB nozzle generator used by Delft University of Technology. Experiments are performed on various nozzle designs, focusing on the average bubble diameter. Through geometrical scaling and further developments, the bubble diameter is scaled from 0.5 mm, produced by conventional nozzles, to 2.5 mm. The obtained upscale in bubble diameter corresponds to a 3-fold increase along all dimensions compared to the current operational domain for HFSB experiments. Furthermore, a novel physics-based model is proposed relating particle size to the nozzle orifice and flow rates of air and helium and is benchmarked with the acquired experimental data.

Turbulent boundary layers are responsible for up to 89 % of the skin friction drag of civil aircraft. This shows that in regard to the current social sensitivity towards climate friendly aviation a reduction of tur- bulent skin friction could have a large impact. On the one side greenhouse gas emissions by aviation could be reduced and on the other hand fuel costs could be saved, an important consideration for air- lines. The research on turbulent drag reduction distinguishes between passive and active techniques. Passive techniques have the advantage that no energy input is necessary. Prominent examples are riblets and dimples which have shown turbulent drag reductions up to 10 %. Active techniques reach turbulent drag reduction percentages of up to 45 %, however it is in doubt if the energy savings are larger than the energy required to operate the control technique. One promising active technique which has proven to partially generate a positive energy effect is spanwise wall oscillation. However, researchers have opposing opinions of the mechanism which leads to the drag reduction. As turbulent boundary layers are strongly three-dimensional, the work of this thesis investigates three-dimensional flow fields of turbulent boundary layers subjected to spanwise wall oscillation by means of tomographic PIV. These are used to study the changes in pointwise statistics and coherent structures in order to derive a descriptive model of the drag reducing mechanism.

Reduction of skin-friction drag over a fully developed canonical zero-pressure gradient turbulent boundary layer (ZPGTBL) subjected to spanwise oscillation is measured using planar particle image velocimetry (PIV). The experiments are conducted at Re_θ of 1000 and 1800, the chosen range of spanwise oscillations amplitude and frequency are around the optimum reported in literature studies (T⁺_osc = [100-700], A⁺_osc= [50, 150]). A high-resolution planar PIV apparatus is employed to measure the drag reduction directly by wall shear stress estimates on the oscillating wall. The measurement uncertainty of the drag estimates from PIV measurements is examined. The results show drag reduction in the order of 15% after 6 boundary layer thicknesses from the beginning of the oscillating section. Variations of the drag reduction follow the trends reported in literature. The PIV measurements enable the analysis in terms of Reynolds shear stresses, turbulence production and allow visualising vortex packets by vorticity. A pronounced drop of turbulence production in the range y⁺ = [5, 30] is observed. The vorticity analysis indicates a distortion from the well-known hairpin-packet arrangement, suggesting that the mechanism of hairpin auto-generation may be inhibited by spanwise wall oscillations

Car industry and motorsport development nowadays strongly rely on wind tunnel testing. As the literature survey shows, in order to fully understand the characteristics of the flow around a car, large-scale and volumetric flow field measurements are required. For its features, PIV can be pointed as a useful technique for automotive wind tunnels. However, its use in an automotive wind tunnel is strongly conditioned by the industrial environment itself, which requires particular efforts for what concerns safety and economic aspects.

The advent of Robotic Volumetric PIV has permitted the measurement of time-averaged properties of large-scale complex aerodynamic flows by the combination of coaxial volumetric velocimetry and robotics. Firstly, the use of Helium Filled Soap Bubbles as seeding permits to enlarge the measurement volume due to their higher scattering capability. Furthermore, the compactness of the coaxial velocimetry probe together with the robotic actuation allows to measure large-scale volumes, partitioning the entire volume and reconstructing the time-averaged flow filed from multiple PIV acquisitions.
Since its presentation, Robotic Volumetric PIV has been used to investigate low velocities flows, with free-stream values ranging between 2 m/s and 14 m/s. More specifically, 15 m/s can be considered the largest velocity that can be analysed with the state-of-the-art of the system due to its hardware limitations in the maximum acquisition frequency.
For this reason, to meet the requirements in terms of flow velocities proper of industrial environments, the range of analysable velocities has to be extended.

To address the identified requirement of high-speed volumetric measurements, a new acquisition technique has been designed and proposed. Even if a standard double-pulse strategy can measure high-speed flows due to the short time that can elapse between the two subsequent pulses, in order to increase the accuracy of the final results, a two-$\Delta t$ method is proposed. Firstly, a predictor is built using a Double-Pulse, Double-Frame acquisition with a short pulse separation time $\Delta t_1$. Afterwards, the predictor is used to allow the stretching of the second pulse separation time $\Delta t_2$ without encountering errors given by the false pairing that would be present due to the longer displacement of the particles.

To test the performance of the new method, the near wake of a $50\%$ replica of the Ahmed body with a 25$^{\circ}$ slant angle has been studied through an experimental campaign in the Open Jet Facility (OJF) of TU Delft. Firstly, a measurement at 12 m/s has been performed in order to be able to assess the performances of the proposed method w.r.t. the time-resolved acquisition strategy and the DP standard -DF approach.
Afterwards, an airspeed of 20 m/s has been considered, in order to demonstrate the capability of the new technique of extending the velocity range of Robotic Volumetric PIV.

The conducted measurements demonstrate the possibility of extending the velocity range of the Robotic Voluemtric PIV system with the proposed method. However, this is achieved at the expense of a lower accuracy, due to the lack of temporal information, and a longer measurement time, given by the necessity of multiple acquisitions.

Aerodynamics research and development often involves large-scale and turbulent flows; may this be a rotor flow, the wake of a car, or the flow passing a human athlete to name a few examples. Volumetric particle image velocimetry (PIV) is well suited for the study of such flows. The introduction of Helium Filled Soap Bubbles (HFSB) for wind tunnel measurements has pushed the achievable measurement domains using state-of-the-art tomographic PIV systems to the order of 50 liters. Despite the dramatic increase in the measurement volume size by the introduction of HFSB, the observation volumes in tomographic PIV are still relatively limited. Further, the system complexity is high, constraining optical access on complex geometries and prohibiting a rapid repositioning of the measurement domain.

To address the identified need of large-scale volumetric flow field measurements around complex geometries, a novel robotic PIV approach is presented. The proposed method relies on a coaxial volumetric velocimetry probe comprising four high speed cameras at low tomographic aperture, aligned with the illumination. The compact system greatly enhances optical access and it gives rise to a universal calibration, meaning that it can take full advantage of robotic manipulations. Therefore, large domains can rapidly be partitioned into a series of adjacent sub-volume measurements, controlling the velocimeter position by means of a robotic arm.

The feasibility of the proposed system is demonstrated in a study of the near flow field around a full-scale replica of a time-trialling cyclist at 14 m/s. 450 time-resolved volumetric PIV measurements are acquired to compound the time-averaged flow field on a 2 m³ domain. The analysis of the flow topology developing around the cyclist shows good agreement with available literature and it indicates the potential of the proposed robotic PIV approach in other engineering applications.

The fundamental properties of computed flow fields using particle imaging velocimetry (PIV) have been investigated, viewing PIV processing as a black box without going in detail into algorithmic details. PIV processing can be analyzed using a linear filter model, i.e. assuming that the computed displacement field is the result of some spatial filtering of the underlying true flow field given a particular shape of the filter function. From such a mathematical framework, relationships are derived between the underlying filter function, wavelength response function (MTF) and response to a step function, power spectral density, and spatial autocorrelation of filter function and noise.
A definition of a spatial resolution is provided independent of some arbitrary threshold e.g of the wavelength response function and provides the user with a single number to appropriately set the parameters of the PIV algorithm required for detecting small velocity fluctuations.
The most important error sources in PIV are discussed and an uncertainty quantification method based on correlation statistics is derived, which has been compared to other available UQ-methods in two recent publications (Sciacchitano et al. 2015; Boomsma et al. 2016) showing good sensitivity to a variety of error sources. Instantaneous local velocity uncertainties are propagated for derived instantaneous and statistical quantities like vorticity, averages, Reynolds stresses and others. For Stereo-PIV the uncertainties of the 2C-velocity fields of the two cameras are propagated into uncertainties of the computed final 3C-velocity field.
A new anisotropic denoising scheme as a post-processing step is presented which uses the uncertainties comparing to the local flow gradients in order to devise an optimal filter kernel for reducing the noise without suppressing true small-scale flow fluctuations.
For Stereo-PIV and volumetric PIV/PTV, an accurate perspective calibration is mandatory. A Stereo-PIV self-calibration technique is described to correct misalignment between the actual position of the light sheet and where it is supposed to be according to the initial calibration procedure. For volumetric PIV/PTV, a volumetric self-calibration (VSC) procedure is presented to correct local calibration errors everywhere in the measurement volume.
Finally, an iterative method for reconstructing particles (IPR) in a volume is developed, which is the basis for the recently introduced Shake-the-Box (STB) technique (Schanz et al. 2016).@en

Wings operated at low and moderate Reynolds number such as the ones of UAVs or the blades of small turbine and of compressors, can be the source of an aero-acoustic phenomenon called laminar boundary layer instability noise. The narrow band noise can be attenuated using trailing edge (TE) serrations (Chong et al. 2013, Moreau et al. 2012) but the mechanism behind tonal attenuation is yet to be investigated in detail. An experimental study (1.32X105 < Re < 5.30X105) is conducted in an open-jet low turbulence wind tunnel using NACA 0018 airfoil (c = 0.20m) with modified TE. The acoustic emission in the far field is recorded using far-field microphones whereas the developing field near the TE is studied using Time resolved planar particle image velocimetry and oil flow visualization techniques. Study showed the serration-3 (2h = 20mm, λ = 10mm) has maximum tonal noise attenuation and effect the point of separation of laminar separation bubble. The span-wise change in flow characteristics in case of serrations can be attributed to the effective chord-length of the wing at each span-wise position. Further span-wise correlation of chord-wise flow fluctuations show the flow being turbulent upstream of serration-3 TE modification.

The behaviour of an inviscid, constant density fluid on which no body forces act, may be modelled by the two-dimensional incompressible Euler equations, a non-linear system of partial differential equations. If a fluid whose behaviour is described by these equations, is confined to a space where no fluid flows in or out, the kinetic energy, vorticity integral and enstrophy integral within that space remain constant in time. Solving the Euler equations accompanied by appropriate boundary and initial conditions may be done analytically, but more often than not, no analytical solution is available. The solution, however, can always be approximated and usually a computer is used to provide a numerical approximation.

The aerodynamic noise generated by trailing edge of aircraft wings at take-o_ and approach con_guration continues to be a critical factor in the future development of aviation. Trailing edge noise is also a serious concern for wind turbines, marine propellers, helicopter rotors, axial and centrifugal turbines and compressors. A fundamental understanding of how the energy contained in turbulent motions around trailing edges is converted to sound will be bene_cial for designing silent airfoils and control surfaces. Current research aims at applying new tools of ow _eld interrogation i.e. high-speed planar and tomographic PIV to the canonical problem of trailing edge noise. In analytical models, the far _eld noise can be described using turbulent quantities like the spanwise coherence. The use of high-speed tomographic PIV creates a link between the phenomenological description of the ow and acoustic _eld in terms of coherence of structures and noise. The geometric con_guration used involves an airfoil with an elliptic leading edge and an asymmetric beveled trailing edge with 25 deg tip angle (_) and three di_erent radii of curvature (R). The measurements are taken at four di_erent Reynolds numbers.

This study examines the effects of slot suction on the mean characteristics of a two-dimensional turbulent boundary layer. Previous studies on this subject have mainly focused on low Reynolds number boundary layers (i.e. Re_ B 2,000). Furthermore, numerical tools are often not able to accurately predict the effects of slot suction. To grasp the effects of slot suction at conditions more relevant to engineering problems, the present research focuses on a turbulent boundary layer with a Reynolds number of Re 10,000.

One of the main limitations of microphone measurements in an open-jet wind tunnel is spectral broadening. The phenomenon of spectral broadening is observed when a single frequency sound wave passes through a free shear layer. Inside the turbulent shear layer the wave interacts with the turbulence, which leads to broadening of the tone in the frequency domain. The spectrum of a single frequency tone measured with a microphone will not only include the peak but also two shoulders containing the broadened frequencies, one on each side of the peak. Spectral broadening does not only a_ect single microphone measurements but also has a detrimental e_ect on acoustic array measurements. The distortion caused by the shear layer will reduce the coherence between the microphones in the array. This limits the size and thus resolution of the array. This spectral broadening is particularly problematic when examining objects which produce a tonal noise, e.g. open rotors, sharp edges.