Flame Aegis is a fire-fighting drone designed to combat high-rise building fires in urban areas. This product has been requested by the Gezamenlijke Brandweer, and certain stakeholder requirements have been imposed. There are further regulatory constraints at both national and European levels. A 150 kg drone with a capacity of 40 L of water was designed in the end. The water can be delivered by spraying at a rate of 150 L/min through a nozzle or by dropping the contents of the tank within 2 seconds. The drone is able to perform a 15-minute mission with a 5-minute turnaround time.

This thesis presents the development, implementation, and validation of a low-fidelity Aeroacoustic Prediction Framework designed for airborne wind energy systems (AWES), with the Kitepower system as a case study. As AWES technology moves toward commercial viability, understanding and predicting its acoustic emissions becomes critical for regulatory compliance, public acceptance, and design optimization.
The framework integrates established analytical and semi-empirical aeroacoustic models with aerodynamic data based on derived geometry and detailed flight information. It models all major noise sources from the airborne components, such as the Leading Edge Inflatable (LEI) kite, bridle lines, tether, and onboard ram-air turbine. The most significant contributions to the overall noise signature were found to be turbulent boundary layer trailing edge (TBL-TE) noise from airfoils, modeled using the Brooks–Pope–Marcolini (BPM) approach, vortex-shedding noise from cylindrical structures such as the tether and bridle lines, and tonal harmonics produced by the rotating turbine blades, captured through Hanson’s helicoidal surface theory.
To generate aerodynamic input, spanwise airfoil profiles were automatically extracted from 3D CAD models and analyzed through XFOIL. Real-time flight data was provided by an onboard sensor suite and processed through an Extended Kalman Filter (EKF), allowing dynamic simulation of flight conditions. Audio recordings were collected during test flights using GoPro cameras, enabling experimental validation of the acoustic predictions despite the absence of calibrated SPL measurements.
Validation showed strong agreement between predicted and measured spectra up to 5 kHz, particularly for turbine harmonics and general spectral shape. Deviations in the lower tonal harmonics were primarily attributed to acoustic shielding caused by the turbine’s duct structure. Additionally, the use of GoPro cameras introduced limitations due to their lack of calibration data and the presence of internal low-pass filtering above 5 kHz. Despite these constraints, the model successfully predicted tonal peaks, including the blade passing frequency and higher-order harmonics, aligning well with the experimental observations.
Additionally, the framework investigates the influence of the propagation effects, such as atmospheric absorption and geometric spreading, and integrates them to produce realistic observer-based predictions. Despite using non-professional audio hardware, the predictions captured key features including harmonic roll-off and broadband trends, affirming the framework's validity for early-stage design and evaluation.
This work demonstrates that low-order, physics-based models paired with aerodynamic inputs and synchronized flight data can yield meaningful acoustic predictions for AWES. The framework offers modularity, computational efficiency, and adaptability for future upgrades, such as the use of calibrated microphones or high-fidelity CFD data. It serves as a foundation for future extensions in auralization, psychoacoustic testing, and component-level noise reduction strategies.
Ultimately, the thesis bridges theoretical modeling with field-based validation, supporting the responsible integration of AWES technologies into noise-sensitive environments.

Future regional and short to medium-range aircraft are expected to use propellers for sustainable aviation, possibly with (hybrid-)electric propulsion systems. This will enable innovative integration of aero-propulsive systems, potentially increasing the overall aerodynamic efficiency of the airframe and propellers, resulting in a lighter airframe and increased passenger comfort. The presence of electric motors and propellers in such configurations, in addition, offers a unique opportunity to leverage propellers as airbrakes by operating them at negative thrust. This approach offers multiple potential advantages, including shorter landing runs, enhanced landing maneuverability, energy harvesting during braking operations, faster aircraft turnaround times, and reduced community noise exposure.

Despite these potential advantages, the aerodynamic and aeroacoustic characteristics of propellers operating in negative thrust mode remain largely unexplored. This dissertation addresses this knowledge gap through computational analysis, comparing the performance of isolated propeller configurations in negative thrust mode with the well-understood positive thrust mode. The results reveal that the distinct aeroacoustic characteristics of propellers operating in negative thrust conditions, compared to conventional positive thrust conditions, offer a promising avenue for reducing community noise, not only through the possibility of steeper descents but also through changes in the noise emissions from the propeller itself.

This thesis investigates the turbulent jet aeroacoustic phenomena and explores the potential of active flow control using synthetic jet actuators to mitigate noise emissions. The study initially focuses on the development and validation of a comprehensive noise-source prediction model that combines already existing theoretical jet aeroacoustics formulas into an iterative solver that utilises existing jet farfield noise data. The thesis then continues with the goal of designing a synthetic jet actuator to be used as active flow control in the jet that aims to contribute to the emitted noise mitigation. The research begins by analysing the literature revolving around aeroacoustics of jets in general and the efforts to locate and characterise the noise sources in the wake of the jet in specific. The aeroacoustic analogies created for jet noise emissions are reviewed and later on form the basis of the predictive model. The core of this first part of the work involves the integration of the aforementioned analogies in the form of analytical formulas for farfield noise calculation into the modelling framework that incorporates the already existing farfield noise data for the various jets. These noise data are used as the target within a loss function that aims for the convergence of the predicted farfield noise to the experimental one. The characterisation of the noise sources within the jet are then a product of this iterative solving and take the form of energy/frequency requirements. This framework is applied to various jet configurations resulting in a global guide of velocity/frequency requirements for effective actuation of subsonic jets. The major output of the model is regarding a Mach 0.5 jet for which the requirement for effective actuation was to introduce disturbances with an amplitude of 15m/s at the frequency range from 600 to 800Hz. Focus is then given on potential active flow control methods that take the outputs of the noise-source prediction model and are able to provide the specific energy requirement at the required frequency range. Given the Mach 0.5 jet, synthetic jet actuators are chosen as the type of actuation and the research henceforth continuous with the modelling of actuators as a coupled resonator system. Once more, the outcomes of this modelling point at a specific actuator geometry which sparks the design, manufacturing and testing of said actuator. The designed piezoelectric-driven synthetic jet actuator was found to provide the 15m/s requirement at the frequency range of 700 to 900Hz which is deemed sufficient. The report finally discusses the practical implications of these findings in an effort to construct and evaluate an active flow control demonstration using the aforementioned actuators on an actual jet while also putting everything into context for future aeroacoustic modelling and aerospace applications. This work contributes to the ongoing efforts in reducing noise pollution in aerospace environments and provides a foundation for future research in active flow control technologies. The methodologies developed and the results obtained are expected to have broader applications in the design and optimization of quieter, more efficient aircraft.

This UAV system focuses on delivering a 25kg payload to offshore wind turbines. It has a flexible mission profile that can be extended to offshore logistical operations between platforms and ships, as well as on-shore operations. This project aims to provide an innovative, competitive, sustainable, and accessible design capable of fulfilling this mission, all while serving as an educational environment in which a group of 10 students can develop academically.

A decades-long steady rise in air movements has disproportionally impacted airport-neighbouring communities. These communities bear the brunt of the aviation industry's air, and noise pollution, detrimental to their livelihood. Airports near densely populated areas face increasing societal and governmental scrutiny to reduce their noise pollution and improve the quality of life for airport-neighbouring communities. Airframe noise has become a prominent noise source during approach due to the increased usage of quieter high-bypass ratio turbofan engines. Flyover measurements suggest that the leading edge junction between the fuselage, wing, and slat is a high-intensity localised noise source. This airframe region has been sparsely researched. Therefore, this study aims to describe the relation between the aero-structural design of the fuselage-wing-slat junction and its aeroacoustic footprint for an aircraft in approach configuration.

The research set-up was designed for open-jet numerical simulations using the commercial Lattice Boltzmann Method (LBM) based CFD solver PowerFLOW, with future compatibility for experimental open-jet wind tunnel validation studies at the TU Delft's Anechoic wind tunnel. Three variations of a two-sideplate research set-up were created based on a slat-and-main-wing modified 30P30N airfoil cross-section. A `No Gap' (NG) geometry that connects the slat to both sideplates; a `No Horn & Step Stump' (NH) geometry that has a simplified slat side-edge and slat stump, a feature that blends the main wing with the fuselage; and a `Horn & Smooth Stump' (H) geometry, which incorporates a slat horn on the slat side-edge and a slat stump modelled after the Airbus A320. For all three variations, the relative positions of the slat track, slat side-edge and other junction surfaces were modelled after the Airbus A320 as well. All model variants were analysed based on their far-field noise radiation and the near-field behaviour of aerodynamic turbulent structures.

The results of the simulated scaled models were compared to literature. The noise radiation of the `H' model was scaled in power and frequency (Strouhal-based), and compared to beamforming integration flyover data of a junction of an Airbus A320. The model shows good spectral resemblance for higher frequencies, with larger discrepancies at lower frequencies. The low-frequency discrepancies were attributed to beamforming limitations stemming from Rayleigh's criterion, as well as a likely over-prediction of lower frequency noise by the scaled `H' set-up. Comparison to full-scale Reynolds number tests on a model Airbus A320 revealed the slope of the scaled noise spectrum resembled the slope of the slat noise of the Airbus A320 at full Reynolds number both at low and high frequencies.

Contrary to earlier research, the set-up without a sideplate-slat gap, `NG', produced excess noise compared to the `Gap'-models, `NH' and `H'. The excess noise is attributed to a larger spanwise extent of the slat narrow-band peak noise mechanism. The narrow-band peaks stem from a slat-cusp-to-slat-trailing-edge flow-acoustic resonance. The introduction of a side-edge (for the `NH' and `H' models) increases the spanwise velocity in the slat cove and creates a slat cove wake that gets accelerated over the slat track. This study showed that the altered slat cove wake shape prevents the canonical slat-cusp-to-slat-trailing-edge flow impingement from occurring on part of the slat, thus limiting the extent of the resonance mechanism. This phenomenon was portrayed by visualising the slat cove wake through total pressure isosurfaces and visualising the narrow-band peak noise source regions through a Ffowcs-Williams and Hawkings (FWH)-based source visualisation technique. A lack of loading on the slat (which is a property of the non-modified 30P30N cross-section as well) is hypothesised as the reason why the introduction of a slat side-edge does not increase the noise radiation for the models with a slat side-edge (`NH', and `H').

With the ever-increasing issues caused by global warming, the emission of carbon dioxide has to be minimized. With the transport section being responsible for 24% of global emissions, the demand for electric alternatives is on the rise.
The electrification of passenger vehicles comes with benefits like a smoother driving experience, requiring less maintenance, and has led to an overall noise decrease. However, annoying noises that were previously masked by loud engines are now audible. Automotive cooling fan noise is such an example. The tonal character of the fan is annoying to pedestrians and residents, especially during car charging. The batteries reach high temperatures while charging, increasing the cooling demand from the fan module compared to conventional cars. In this situation, no inflow is generated from driving and the fan is operating at high rotational velocities and pressure requirements.

When the (axial) inflow to the fan is non-uniform, (the velocity changes over the circumference), a varying force is generated on the fan blade, when it rotates. When this non-uniformity is stationary (constant in time), these force fluctuations are periodic, and this is the main cause of the tonal noise. The non-uniformities in fan modules arise from inflow distortions, like for example heat exchanger, duct bends, or the shape of the shroud.

In current fan modules, unequal blade spacing is used to reduce the annoyance of these tones. This reduces the noise at the Blade Pass Frequency, but introduces tones at other harmonics of the rotational frequency, overall mainly leading to a reduction in noise annoyance.

Flow obstructions have been shown an effective means of reducing noise in equally spaced axial fans with low Mach number, while minimizing aerodynamic losses. They introduce a secondary non-uniformity in the inflow, which creates a secondary aeroacoustic source on the fan blade. When placed at the correct circumferential angle, cancellation between the primary and secondary source is possible. This research investigates if such flow obstructions can still yield adequate noise reduction on unequally spaced fans, and if (multiple) upstream heat exchangers prohibit the working mechanism of flow obstructions.

It was found that flow obstructions can indeed significantly reduce tonal noise at the fan BPF, by up to 13 dB on the fan axis, and 6dB at an angle 45\degree from the axis. The upstream radiator and condenser did not compromise this reduction, and the obstruction still reduces noise when the fan operates at a lower rotational velocity. Using a superposition of obstructions (creating a multi-modal obstruction) could also control multiple tones. However, with the large number of tones to be controlled in an unequally spaced fan, the overall sound pressure attenuation and annoyance reduction are insignificant.

Furthermore, the low-fidelity analytical tool developed in the thesis was an effective means of rapid obstruction design and optimization. Although information on the fan's primary noise is needed, as well as some parameters to define the obstruction wake (which can be obtained with cheaper, RANS simulations), the model correlated well with simulation data obtained in PowerFLOW.

Turbulent boundary layer trailing edge noise is becoming an increasingly important problem in wind turbine applications. The wind turbine rotor diameter is expected to keep growing bigger in the future and trailing edge noise is known to scale with the relative rotor tip velocity to the power 5-6. One effective and yet relatively novel method to reduce this type of noise is the addition of perforated add-on devices to the rear end of the rotor blades. The perforated add-ons have been extensively tested experimentally in recent years. The experimental studies have been instrumental in developing design rules but currently, the precise mechanisms behind this noise reduction are not yet fully understood. The project objective is to obtain insights into the aero-acoustic flow physics responsible for the far-field noise reduction for perforated add-on trailing edge noise problems. To accomplish this, the first high-fidelity numerical LBM study in which the geometry of the per- forations is fully resolved is performed. The simulation is performed with the commercial Lattice Boltzmann method-based solver, PowerFLOW. A scaled 2.5D airfoil simulation is performed with an inflow of 20 [m/s] in lifting conditions. Three different airfoil add-on configurations are simulated. One baseline, one with a homogenous perforation spacing, and one geometry where this spacing is decreasing in the downstream direction. The simulation results are validated against experiments and a noise reduction of about 3 − 4 [dB] is found for in the lower frequency range Stc < 10. A distribution of sound sources was found over the add-on for all frequencies. The frequencies up to Stc = 10 show signs of destructive acoustic interference contributing to the observed noise reduction. Besides this phenomenon, two additional mechanisms were identified as contributing to the observed noise reduction. The second is the milder ∆p jump at the trailing edge caused by communicating boundary layers on both sides of the add-on. A good indication of this milder pressure jump is the correlation coefficient Rpp between two points on opposite sides of the add-on close to the edge. No development or growth of the zone of the correlation coefficient was found over the add-on except for the vicinity close to the trailing edge. A mean vertical upward flow is observed, flowing out of and extending above the perforations. This is believed to be relevant for the achieved noise reduction. The effect possibly has simultaneous positive and negative conflicting contributions to noise reduction. Lastly, it was concluded that the spectra of the homogeneous and linear perforated geometries showed too much similarity, therefore it was difficult to couple discrepancies in the near-field between the two cases to observations of the far-field spectra.

Airborne wind energy systems offer a promising approach for renewable energy generation. However, the noise emissions associated with these systems should be understood and minimized as well to promote their integration and (social) acceptance. This research aims to identify the noise sources of airborne wind energy systems, systems with leading-edge inflatable (LEI) kites and fixed-wing kites in particular, and establish a foundation for future research in this field. Analytical simulations using the Brooks, Pope, and Marcolini model and the Amiet model were combined with experimental measurements for this research.

The analysis of the fixed-wing kite revealed prominent peaks around 1500 Hz and 2000 Hz in the noise spectra, with the higher frequency peak observed at higher kite velocities. Analytical predictions indicated laminar boundary layer vortex shedding and tether vortex shedding as the main noise sources. The study also investigated the directivity of the turbulent boundary layer trailing edge noise, which revealed dipoles that exhibited slight deformations at higher frequencies.

For the LEI kite, noise analysis identified peaks in the sound pressure level around 300-400 Hz and 1000-2000 Hz. Analytical predictions highlighted turbulent boundary layer trailing edge noise and vortex shedding from the tether and bridle lines as the dominant noise sources.

By considering the implications of these findings, the noise impact of airborne wind energy systems can be minimized, fostering their sustainable deployment and acceptance.

With an unprecedented surge in the number of individuals utilizing electric vehicles for commuting across the globe, it is becoming critical to immediately deliver an adequate and range-friendly degree of thermal comfort in car cabins to help drivers remain focused and attentive. The research presented in this thesis is part of the development work carried out by the Aerothermal team at Tesla. The team aims to better understand high aspect ratio subsonic jets, particularly intersecting twin jets, to ventilate the vehicle's passenger compartment.

The intersecting twin jet has recently captured significant attention from the Aerothermal team for several reasons. Firstly, it provides the ability to actively control the flow angle from the HVAC unit into the vehicle cabin without requiring direct interaction with the vent. Additionally, it facilitates the flushing of the vent outlets with the A-class surfaces in the vehicle interior, helps achieve a simple and minimalist design for the instrument panel and other interior surfaces that align with Tesla's design language

The objective pursued in this thesis is to investigate the physics represented by the bulk flow characteristics of intersecting twin jets, and to scrutinize the installation effect on the merging jet, focusing on the influence of the installation surface offset height. This objective is achieved by characterizing the evolution of large-scale flow structures in High AR intersecting twin jets at moderate Reynolds numbers experimentally using planar Particle Image Velocimetry (PIV) in both installed and uninstalled conditions. Once accomplished, the experimental results are then used to help select, validate, and calibrate the appropriate RANS turbulence model.

The experimental results show that the intersecting jets converge before merging and forming a single jet. This jet behaves like a single jet emanating from a more recessed origin. The variation of the flow ratio between the two nozzle outlets results in varying the jet angle. The study reveals that the placement of the surface impacts the jet angle and velocity and introduces a circulation zone that affects the jet's behavior and direction. The surface's impact depends on the offset distance, length, and profile of the surface. The results also suggest that accurate computational prediction of the jet characteristics requires high mesh resolution and an appropriate selection of turbulence intensity. Among the four turbulence models assessed, the k-omega SST model is found to be adequate for the application, although it exhibits some numerical hysteresis. Overall, the findings provide insights into the dynamics of jet-surface interactions and can guide the design of systems involving such flows.

In recent years, the aviation industry has faced escalating environmental scrutiny, prompted by the global shift towards heightened environmental awareness. This study explores a collaboration framework set up by airlines which envision to partially or totally shoulder the environmental impact costs generated by their airborne emissions. The methodology developed includes the production of detailed emissions data to be employed during the airline’s planning strategy. The model conceived introduces a taxation approach exploited by airlines to include and incentive more environmentally sustainable solutions and, a collaboration mechanism based on the possibility to share passenger demand between collaborating airlines. The environmental incentive and collaboration are implemented in a Linear Programming Network and Frequency Planning model with the goal of maximising airlines profitability. A case study is analysed focusing on two airlines, KLM and TAP, collaborating to achieve improved environmental and economic performance. The airlines engage in a scenario where the environmental impact externalities generated by airborne operations are shouldered by the airlines themselves. Employing the model developed, the airlines explore solutions that either reduce the environmental impact or that solely maximise profit of the two airlines. The results underscore the potential benefits of including collaboration in the airline’s planning strategy, with a total 1.33% reduction in environmental impact when prioritising environmental cost abatement. Collaboration also demonstrated to be a consistent tool to achieve a profit recovery, showing up to a total anticipated profit increase of 2.21% when prioritising profitability or 0.71% when the priority is given to environmental impact reduction. In conclusion, this study highlights the potential efficacy of collaboration in addressing aviation’s environmental and economic challenges.

The research presented in this thesis focuses on the receptivity to surface roughness of swept wing boundary layers dominated by crossflow instabilities (CFI), providing insights into how surface roughness can be used to passively control the developing instabilities. Discrete roughness elements (DRE) arrays and distributed randomized roughness patches (DRP) are employed to investigate the physical phenomena governing receptivity and their impact on CFI onset. The supporting data combine numerical solutions of linear and non-linear stability theory with advanced experimental flow diagnostics.

This booklet is divided into three main parts. The first part investigates the flow mechanisms dominating the receptivity of stationary CFI to the amplitude and location of DRE arrays. The relation between the external forcing configuration and the initial instability amplitude is investigated, along with scaling principles allowing for the up-scaled reproduction of the swept wing leading-edge configurations, which provide experimentally observable configuration.

The second part of this research explores the stationary CFI receptivity to specific up-scaled roughness configurations, including both isolated discrete roughness elements and DRE arrays. These roughness elements are applied at relatively downstream chord locations to enhance the experimental resolution of the near-roughness flow field.

The isolated discrete roughness elements ensure strong boundary layer forcing, which helps to outline the relation between the near-element instability onset and the rapid transitional process. In contrast, the applied DRE arrays configurations provide boundary layers dominated by the development of CFI. In such scenarios, high-magnification tomographic particle tracking velocimetry identifies the dominant near-element stationary instabilities precursor to CFI. Specifically, the presence of transient growth and decay mechanisms in the near-roughness flow region is outlined, exploring their role in the receptivity process and in the CFI onset. This investigation results in the first conceptual map describing the receptivity of swept-wing boundary layers to a wide range of DRE array amplitudes.

Lastly, the acquired knowledge of the near-element flow topology is employed in the final part of this work to develop a passive laminar flow control technique for stationary CFI cancellation. This technique is based on the destructive interference of the velocity disturbances introduced by a streamwise series of optimally arranged DRE arrays. The performed measurements confirm a reduction in the developing CFI amplitude accompanied by a delay of the boundary layer transition. The compatibility of the proposed technique with the control of CFI developing in a realistic free-flight scenario is as well investigated.

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...

Trailing-edge noise can be the dominant noise source of diverse industrial applications, including wind turbines. Noise regulations may limit the power production and installation of new wind farms. There is a need to reduce it. In order to achieve so, we first need to predict it, and to do so we need experimental data to support and validate the new methodologies. A new experimental campaign has been carried out to provide a benchmark of trailing-edge noise applicable to wind turbine noise. It involves a broad range of Reynolds numbers and angles of attack, and covers the noise reduction effect of serrations. This Thesis studies the aforementioned novel data set. The trends of the far-field noise with the forcing of the boundary layer, angle
of attack, and Reynolds number have been studied. Furthermore, scaling laws have been applied to compare different test conditions. The collapse quality has been studied, and the differences have been linked to the aerodynamics and possible post-processing effects. The effect of the serrations has also been studied. Additionally, the noise reduction dependence on the aerodynamic loading is discussed. The influence of the microphone locations on the far-field spectra uncertainty is also studied. The Monte-Carlo method has been applied using the Delay-and-Sum and Clean-SC beamformers. The effect of the input uncertainty correlation, wind tunnel velocity, facility, and algorithm are examined.

The design of electrical machines is complicated due to the number of variables involved and due to competing objectives like efficiency, weight and cost. Another important aspect is the involvement of different physical phenomena such as torque production, electromagnetic fields and thermal heat flow. Thus designs need to satisfy multiple constraints and fulfill competing objectives making the design process tedious. Multi-Objective Optimisation algorithms provide a set of designs that are Pareto optimal i.e. any improvement in one objective comes at the cost of performance in another objective. This gives the designer a set of designs with different trade-offs between objectives and they can choose the design that satisfies the objectives and constraints the best.

There are numerous commercial software available that provide these functionalities. However the methods used to model machines within these packages are not available or cannot be changed. They are also usually expensive. This makes the exploration of new limits and new topologies difficult. Using open source packages allows us to modify these methods according to our application and gives greater control over the process.

This thesis uses PYLEECAN python library that is based on FEMM software to perform the analysis of the machine. A six time-step magnetostatic analysis method to calculate the average torque and iron losses in the stator and rotor core is presented along with methods to calculate the copper losses and windage losses. A steady state Lumped Parameter Thermal Network (LPTN) is developed to calculate the temperatures of various parts of the machine. The LPTN is capable of estimating the temperatures under natural convection and forced air cooling conditions.
Finally a MOO framework was developed using the models developed and the PYMOO python library. This thesis uses NSGA-II to perform the MOO. The MOO framework was used to optimise the design of a machine for a drone application and explore the specific power density limit of the machine. The power density limit was found to 5 − 7 kW/kg based on different slot pole combinations, winding temperature limits and core material used. Further, insights into how different machine parameters affect the specific power density are presented

Aerodynamic noise produced by aircraft, wind turbines, and other objects subjected to airflow contribute to environmental noise pollution, which adversely affects human and animal health. Consequently, governments impose restrictions on aircraft and wind turbine noise levels. These restrictions can have an economic impact by limiting aircraft traffic and reducing wind turbine energy production. Accordingly, improving the design of aerodynamic surfaces to reduce their noise levels benefits health while enabling improved operational efficiency. Therefore, aeroacoustic research focuses on identifying and understanding the physical mechanisms behind aerodynamic noise to improve noise mitigation technologies. This research relies on acoustic wind tunnel measurements to validate simulations, theories, and design improvements.

Closed test section wind tunnels are widely used for aerodynamic testing but are less suitable for acoustic measurements because microphones must be installed in the wall. This location subjects the microphones to pressure fluctuations from the turbulent boundary layer (TBL), which contaminates acoustic measurements and reduces the signal-to-noise ratio (SNR). The impact of the TBL can be mitigated by recessing microphones within cavities and covering them with an acoustically transparent material. Modifying existing wind tunnel walls by installing cavity--mounted microphones is a straightforward and cost-effective improvement that enables combined aerodynamic and acoustic measurement campaigns.

The cavity geometry, i.e., depth, aperture size, wall angle, and presence of a covering determines the amount of TBL attenuation and consequently the improvement to SNR. While several studies have shown empirically that these parameters have an effect, few studies focus on identifying the physical mechanisms that explain the relationship between geometry and the reduction in TBL pressure fluctuations at the microphone. Thus, this thesis aims to identify these physical mechanisms through experiments and different modeling approaches to better explain the relationship between cavity geometry, the amount of TBL attenuation, and the subsequent impact on the measured acoustic signal.

Experimental data were collected to develop an empirical model to quantify how varying cavity geometry affects the measured pressure spectra. Moreover, experiments were also performed to validate simulation results and to quantify the SNR improvement when applying a beamforming algorithm to microphone array data. The modeling and simulation efforts focus on explaining the trends and phenomena identified in the experimental data. Initially, a physical model was developed that assumed acoustic propagation into an axisymmetric cavity with a constant cross-section. This model decomposes a pressure field, resulting from a TBL, into circular duct modes and was used to evaluate the relationship between cavity geometry and the propagation of these acoustic modes into the cavity. This model was followed up with a finite element method (FEM) simulation to study the influence of different cavity geometric parameters and wall materials on the acoustic response of the cavity when subjected to an acoustic wave.

The FEM simulation showed that the cavity's acoustic response is determined by the presence of standing waves in the form of acoustic depth modes. This simulation showed that cavities with angled walls have depth modes with lower amplitude waves and thus distort the acoustic signal less. Furthermore, it is shown that the acoustic responses of cavities formed out of sound-absorbing foam are driven by the shape of the foam holder and not the cavity shapes within the foam. Thus, the holder can be optimized to minimize the acoustic response, while the cavity itself can be optimized to reduce the influence of the TBL. Building upon these simulations, a Lattice Boltzmann based computational fluid dynamics (CFD) method was used to simulate the pressure and flow fields within three uncovered cavities and covered cavities resulting from the presence of a turbulent boundary layer.

The CFD simulations confirmed a significant finding of the physical model, that the amount of TBL attenuation increases as the cavity aperture size increases relative to the TBL streamwise coherence length. This is due to the resulting modal decomposition of the pressure field above larger cavities having more energy distributed across higher-order modes than for smaller cavities. These higher-order modes decay exponentially into the cavity, resulting in increased attenuation of the TBL. Smaller cavities have most of their energy in their first mode, which does not decay with increasing cavity depth. Furthermore, these simulations showed that the pressure field within covered cavities is primarily acoustic and can be decomposed into acoustic circular duct modes. Since the propagation of TBL pressure fluctuations into covered cavities is primarily acoustic, the shape of future cavities can be efficiently optimized using FEM simulations.

Finally, beamforming used with cavities improved the acoustic measurement SNR. Analysis shows that the improvements due to beamforming are independent of those attributed to the cavity geometry. Thus, combining the two approaches improves the SNR of acoustic measurements in closed test section wind tunnels.

This work presents an experimental and analytical investigation of unsteady loading of a low Reynolds number propeller and its relation to aeroacoustics. The propeller is positioned at an angle of attack with respect to the freestream which causes a variation in loading along the azimuth. The propeller is operated at 4000 RPM with an advance ratio of J = 0.4, corresponding to a free stream velocity of 8 m/s. The chord-based Reynolds number of the propeller is of the order of 10⁴ and is investigated at two angles of attack: α_r = 7.5° and α_r = 15°. For reference, a steady case, α_r = 0◦, is presented as well. Loads and noise measurements were carried out in the A-Tunnel of TU-Delft. Two microphone array configurations were used to mimic the observers positions above and underneath the flightpath of the propeller. Underneath the flightpath, an increase in SPL was found with increase of the rotor angle of attack. Above the flightpath, the opposite effect was observed, resulting in an asymmetrical directivity profile when the rotor is at an angle of attack. A quasi-steady loading model, based on the lifting line code Xrotor, is used to determine the azimuthal variation in loading. The variation in tip speed along the azimuth due to the propeller angle of attack is prescribed to determine the variation in loading in one rotation. Far-field loading and thickness noise are computed using an aeroacoustic model based on the Ffowcs-Williams & Hawkings’ equation to predict the harmonic noise in the frequency domain of a rotating dipole and monopole, respectively. The numerical predictions for the steady case show good agreement for both the aerodynamic and aeroacoustic performance when compared with loading and noise measurements. For the α_r≠ 0° cases it was shown that unsteady loading noise increases the SPL for all harmonics, with a larger contribution at higher harmonics.