Turbulence is a commonly encountered state of fluid dynamics. Unsteady turbulent flows in pipes are present in many engineering applications and also in biological flows. However, the various processes active in such flows are not well understood. The present work employs stereo-PIV to investigate the effects of a sinusoidal pressure gradient on the various turbulence parameters, including the terms of the turbulent kinetic energy budget equation. The bulk flow rate was oscillated with a frequency of 0.5 Hz with a mean Re 26,000 and an amplitude of modulation, 0.23 times the mean value. It is seen that there is a delay in the response of turbulence to the oscillations of the bulk flow and the delay increases with increasing distance from the wall. The axial and the in-plane turbulence parameters show a difference in the delay of their responses. This delay extends into the small scales responsible for the dissipation of turbulent kinetic energy. Changes are also observed in azimuthal length scales when the flow oscillates.The effects of oscillation on the streaks of low momentum are also discussed and the structural organization in unsteady pipe flows are found to be different from that in steady pipe flows

Recently the interest in hydrofoil vessels using a fully submerged hydrofoil increased. Hydrofoil vessels experience a resistance hump and possible instabilities during their takeoff of which both are influence by waves. In order to determine the required power during the takeoff to overcome the resistance hump and to investigate whether the hydrofoil vessel is stable during the takeoff numerical simulations or experiments have to be performed. Experiments covering the full takeoff require a long towing tank and high towing velocities and therefore these experiments are costly. The alternative, performing numerical simulations, is favorable but requires a suitable numerical code. As no validated code is available in which a takeoff can be performed in waves, in this research Panship, a potential flow code suitable for high speed craft and adapted for use with hydrofoil craft, is validated for the takeoff of a hydrofoil vessel. In this research the results of Panship are compared with results of experiments and other numerical codes in four different stage as little validation data is present for the takeoff of a hydrofoil vessel. The four different stages are: a foil without the influence of the free surface, a foil with the influence of the free surface, two foils with the influence of the free surface and the hull and foil system near during the takeoff. In these four stages the lift and drag determined in Panship is compared with available experimental results and numerical codes. As the simulations in Panship require a vast amount of input parameters, the influence of these parameters is studied first to ensure the reliability of the results. Comparing the results of Panship for a deeply submerged foil with another potential flow solver shows that the lift is predicted correctly, but that the induced drag is underpredicted. If viscous effects are required in the solution, the quality of the results of Panship is dependent on the type of hydrofoil used. As some foils have a large viscous effect on lift, the lack of viscosity on Panship, leads to a low quality of the results for lift and induced drag as they are overpredicted. The trends for interaction of a foil with the free surface are predicted correctly but the underprediction of induced drag influences the quality of the results. Panship is able to predict the effects of foil interaction for lift correctly, but the results for foil interaction for drag are poorer. During the takeoff Panship is able to predict the lift fairly correct but the total drag is underpredicted severely. Recommendations are given to improve the results of Panship and for future work.

When drilling for oil or gas, wells are typically constructed using a conventional or telescopic well design. After a section of the well has been drilled, a steel casing (or liner) is inserted into the hole preventing the collapse of the hole. The next well section has a smaller diameter as it’s casing must pass through the previously installed casing. This process leads to a stepwise reduction of the well diameter which, especially in deep or complicated wells, restricts the maximum production rate of oil and gas. Mono-diameter technology utilizes casings which are plastically deformed after they have been placed in the well. This plastic deformation is done by pulling a conically shaped tool, known as the cone, through the newly installed casing. The plastic deformation increases the diameter of the new casing to (nearly) the same as that of the previous casing, therefore the well diameter no longer decreases with every new section. The expansion requires a force in the order of 1000-2500[kN]. The work done by this force leads to high local heat generation due to friction and plastic deformation. This heat, poses a risk for the expansion process as the lubricant (RPSLF) decomposes above 250[C]. Decomposition of the lubricant leads to an increase of the friction, inducing more heat release and further decomposition of the lubricant. The expansion force may then exceed the pulling capacity of the drill rig or strength of the drill string, leaving the cone stuck in the well. Previous research provides the distribution of pressure on the contact surface of the cone, the strain distribution inside the liner and the temperature dependent friction coefficient of the lubricant, by combining these the frictional heat is calculated. Heat generation from plastic deformation is a complex phenomenon especially at low strain (-rates). The Taylor-Quinney coefficient defines the ratio between the work done on a material and the heat generated. A strain-dependent model for the Taylor-Quinney coefficient, developed by Zehnder, is used to determine the heat release from plastic deformation. A numerical heat transfer model is built in Matlab, based on the finite volume method. The model uses a mesh based on skewed quadrilateral cells. Numerical stability of the model is verified for the cell size, time step and time integration method. Movement of the cone and the plastic deformation of the casing are modeled by taking a reference frame fixed relative to the motion of the cone and adding a ”convective” term inside the liner. Heat transfer phenomena at all boundaries are investigated using thermal resistance models. A series of experiments is executed to verify the numerical model. The experiments consist of full-scale expansion test using a cone equipped with 30 thermocouples placed just below the contact surface. Additionally also the temperature of the liner, expansion speed and expansion force are measured and recorded. Comparison of the experimental and numerical results leads to the observation that the pressure profile determined in previous work is likely valid at the first onset of expansion. Over time, the forming of a lubricant film however leads to the re-distribution of roughly 20% of the force from the rear of the cone to the front. When the pressure profile is corrected for this effect; the temperatures inside the cone, temperature of the liner, expansion force and overall heat balance show a good match between the numerical and experimental results with an error of around 5%. From the heat balance based on the experimental results, it follows that between 54% - 60% of the work done on the system is converted into heat. As all work done in the form of friction is converted into heat the remaining energy must be lost in the work done to plastically deform the casing, this provides indirect evidence for the validity of the Zehndermodel. The validated numerical model is used to simulate a reference case field expansion, the maximum contact temperature here is 164.4[C]. This value is well below the lubricant decomposition temperature of 250[C], which leads to the conclusion that the process is safe and the expansion speed could even be increased from a thermal point of view.

The total electrical energy consumption by all the operational data centers (located all over the world) is enormous (approx. 1% of the global electricity demand 20000TWh). This electrical energy is required 24x7 to operate and cool all the IT equipment present in the data center. The electrical energy required to cool all the servers present in the white space can range from as low as 10% to as high as 40% of the total data center electrical consumption. The server's inlet temperature has to be within the ASHRAE recommended range (18o C – 27o C), so that they can function correctly.   A simplified design of a raised floor white space with hot aisle / cold aisle configuration is considered. The tile flow rate through the floor tiles influences the server's inlet temperature. To control the tile flow rate, 11 design variables of the data center white space are identified. These are the position of the 4 perforated plates, the amount of perforations of each perforated plates, the floor tiles perforation, the raised floor height, and the CRAH distance to the cabinet. 600 design samples of the white space are generated by applying the Latin Hypercube Sampling (LHS) technique on these 11 design variables. A well-validated CFD software 6SigmaRoom by Future Facilities is used to generate the CFD results. The standard k-ε model is used to model the turbulence in the CFD simulations, in a steady-state condition. A database of 600 samples is generated by recording the cabinet's inlet temperature, flow rate of floor tiles from the CFD simulations, and the corresponding changeable design parameters generated by the LHS technique. The error due to CFD simulation is estimated at less than 4% for the tile flow rate and 1.7o C for the server inlet temperature.  Four Artificial Neural Networks (ANN) are trained on the data from the database to predict the floor tile's tile flow rate and the cabinet's inlet temperature, respectively. The average R2 prediction (testing) accuracy is 0.97 for the tile flow rate predictions and 0.92 for the cabinet's inlet temperature predictions. Their average prediction error is less than 5% for the tile flow rate and less than 20 C for the cabinet's inlet temperature. The Non-dominated Sorting Genetic Algorithm-II (NSGA-II), a variant of the genetic algorithm, is used to find the optimum values of the 11 design parameters of the white space. These optimum values are going to ensure the server's inlet temperature to be within the ASHRAE recommended range. The genetic algorithm optimizes the design variables based on the predictions made by the neural network predicting the tile flow rate. The values of the optimized design parameters are verified using the 6SigmaRoom software by comparing the server's (mean) inlet temperature of the optimized case with the non-optimized case. More number of servers have their (mean) inlet temperature below 270 C in the optimized case as compared to the non-optimized case. The electrical power required by the CRAHs to cool the white space is reduced by 10% in the optimized case as compared to the power which is required by the CRAHs in the non-optimized case. In this study, a CFD simulation of the white space took 40 minutes. The neural networks took less than a minute to make the predictions, and the NSGA-II algorithm took less than 10 minutes to find the optimized design parameters of the white space.  Thus, in this thesis, it is shown that using an artificial neural network and a genetic algorithm, in combination with computational fluid dynamics gives satisfying results in optimizing the white space design, required to keep the server's inlet temperature within the ASHRAE recommended range. The computational time required to find the optimum white space design is also reduced by using a neural network and a genetic algorithm. The prediction by the neural network and the optimization performed by the genetic algorithm can be improved further with the availability of more training data and in-depth knowledge of applying these techniques (the neural network and the genetic algorithm) in predicting and optimizing the solutions respectively.  

In this thesis a method of combined PIV and LIF experiments on the Turbulent/Non-Turbulent interface (TNTI) of a round jet is presented. By varying the distance from the nozzle in the measurements, the Kolmogorov scale of a large range of Reynolds number (2×103 ≤ Re ≤ 48×103) is kept constant, allowing for a constant resolution. The large range in Reynolds number is needed to identify a difference in scaling of the TNTI, which is difficult to capture due to the weak dependency of the separation of Kolmogorov and Taylor scales on the Reynolds number (λ/η ∼ Re1/4). A factor of over 2.5 is achieved in the current work. An experimental setup for these measurements is presented, as well as the design of experiments for seven Reynolds numbers based on the boundary and initial conditions of the setup. A design for measurements with a camera moving with the flow is presented as well. Conditional statistics are used on the jet to look at the change of span-wise vorticity over the interface. For the detection of the interface scalar images from LIF are used, where the interface is detected using a threshold of dye concentration (Rhodamine B). All Reynolds numbers show self-similarity within the measured range, comparable to literature and with a spatial resolution better than 4η. Some small differences between the low and high Reynolds number measurements can be seen. They are not expected to have significant impact on the results and can likely be solved by slight changes in the experimental setup. The influence of the threshold on the shape and width measurements of the mean conditional vorticity profile is examined, indicating a significant dependency on chosen threshold. Current results indicate a scaling of the interface thickness with the Taylor length scale, around a value of 0.5-0.6λ, or 10-15η This is in accordance to the literature for low Reynolds numbers. Due to this significant influence of the threshold on the outcome, these results cannot be seen conclusive. For this, further investigation is needed on the dependency of the width on the threshold, and validation of the jet contour. Universal ways to compute the threshold, jet contour and mean conditional vorticity profiles are currently missing, while they have shown to be potentially of significance on the outcome. Another method of studying TNTI scaling is via the local velocity scales, which can be found via current data, which could provide a comparison of methods. The measurements with the moving camera could help identifying the link between physical processes and the scaling of the interface.

Aerodynamics has played a significant role in the industry of motorsports in improving the performance and handling of the race car. Rob Smedley, the former head of vehicle performance at Williams Racing stated that - "Where teams have problems is when their development or simulation environment – so CFD [Computational Fluid Dynamics] or wind tunnel – doesn’t describe well what happens in reality (although in truth, no-one’s wind tunnel correlates absolutely 100%)". One of the reasons for this poor correlation could be arising from the fact that, in real life on track scenarios, the race cars undergo accelerating, decelerating, or cornering motion which can have a different influence on the aerodynamics of a race car which is not accounted for in the wind tunnel and simulation environment where a steady constant flow is employed. This research aims to numerically investigate the flow past the front wing of a Formula One car in ground effect subjected to accelerating and decelerating flows to understand the trends in the aerodynamic performance. A scaling analysis is performed to determine the relevant non-dimensional numbers that influence the flow for translationally accelerating airfoils and two dimensionless numbers are arrived at, namely, the Reynolds number and the Froude number. Numerical investigations are carried out for translationally accelerating wings in ground effect to determine the influence of these dimensionless numbers on the aerodynamic forces. Transient simulations were performed on a two-dimensional airfoil and a three-dimensional wing in ground effect subjected to translational acceleration and deceleration. The Shear Stress Transport (SST) based on k-ω was employed to model the turbulent flow. The results from the numerical investigations revealed a temporary change in the downforce and the drag force coefficients, as the airfoil (or wing) in ground effect is subjected to translational acceleration (or deceleration). In this study, the mechanisms that contribute to this temporary change in the aerodynamic force coefficients are discussed.

Accurately predicting the ship motions, is essential knowledge when operating in water. To predict the ship motions, multiple ship characteristics have to be known. For the roll motion, one of these characteristics is the roll damping coefficient. Nowadays the industry standard to determine the roll damping coefficient of a ship is by performing model tests or using empirical formulas created in the late ’70s. In recent years computational power has increased significantly. This increasemakes it possible to start using computational fluid dynamics (CFD) for practical purposes instead of solely scientific research. The roll damping coefficient is a very interesting parameter to examine with viscous flow solvers as the roll motion is heavily influenced by the viscosity of the water. Multiple studies have been performed with 3D and 2D CFD simulations. By simplifying the simulation to 2D, calculation time is reduced drastically. This simplification is performed by only simulating one cross-section of the midship. Only simulating one cross-section causes a loss of accuracy in how well the roll damping coefficients can be determined as the damping effects of the bow and stern are not taken into account, or are roughly estimated. During this research, a method to accurately determine the roll damping coefficients of a ship, while maintaining low computation time, is investigated. This method is based on performing multiple 2D CFD simulations of different cross-sections and combining the results to capture the behaviour of the entire ship. With this approach, the high accuracy of 3D simulations and the low computation time of 2D simulations are combined. The accuracy of this 2D section method is determined by comparing the results with experimental data. The experimental data is obtained through free roll decay tests of a ship from Van Oord on model scale. A 3D CFD model, which simulates a free roll decay test, is created using OpenFOAM. The results of this simulation are compared to the experimental data to determine how well a 3D simulation can predict the roll damping coefficients. Next, a 2D CFD model is created using OpenFOAM which is validated using experimental data of a structure which has a uniform cross-section along the entire length. Lastly, the validated 2D CFD model is used for the 2D section method. Multiple cross-sections of the same ship from which the experimental data is obtained and with which the 3D CFD simulation is performed, are simulated. The results of the 2D section method are compared to the experimental results in order to determine how well this method can determine the roll damping coefficient of a ship. The simulation of the 3D free roll decay test underestimated the linear damping. The non-linear damping was within the margins of the experimental data. The 2D forced oscillation simulations showed an excellent agreement in linear and non-linear damping with both the experimental data as with a comparable CFD simulation. The 2D section method slightly overestimates both the linear and non-linear damping of an entire ship but overall this method shows promising results. However, more research should be done to optimise the method further.

Turbulent scalar mixing is prominent in all sorts of situations, being it in air pollution or making tea. However, this does not often lead to a homogeneously mixed fluid, but rather leads to regions of higher concentrations next to regions with significantly lower concentrations. This can be characterized by dividing the concentration field in a turbulent flow into a few regions that have almost the same concentration: Uniform Concentration Zones. How these Uniform Concentration Zones are organized is yet unclear, but a possible candidate in the velocity field is the use of Lagrangian Coherent Structures. These structures are, in principle, transfer barriers in the flow and defined as local maxima of the Finite Time Lyapunov Exponent (FTLE), which is a measure of exponential fluid-separation over time. In order to successfully apply the concept of Lagrangian Coherent Structures in an experiment, turbulent structures should remain in focus long enough to measure the exponential separation of fluid parcels. As almost all fully developed turbulence comes with a mean flow, the observation time in an Eulerian field-of-view is too short. Therefore an experiment was designed to move the observation equipment with the mean flow. With this experimental setup, the concentration field and the velocity field are measured simultaneously via respectively LIF and PIV. Comparing the Uniform Concentration Zones with both the forward and backward time FTLE-field, shows that there is not much correlation between the concentration field and the forward time FTLE. However, local maxima of the backward time FTLE seem to overlap often with boundaries between Uniform Concentration Zones.

The research in resolving the smaller scales of turbulence has gained significant attention in recent years, owing to the advancements in the measurement techniques. High resolution experimental studies are required to investigate high Reynolds number flows where the scales of turbulence are very small. This is of particular importance in addressing the theoretical issues in studying high Reynolds number wall turbulence which finds applications in transport, energy, and aerospace industries. Hot-Wire Anemometry (HWA) is often employed to reach very high spatial resolution. Early research in Particle Image Velocimetry (PIV), a qualitative flow visualization technique, was not able to achieve a higher resolution as comparable with that of HWA. However, recent developments in PIV allow high resolution measurements, using information on the time averaged ensemble correlation. The ensemble correlation can be used to reduce the size of the interrogation windows employed in a PIV analysis, without compromising for the particle image density in an interrogation window. This is achieved by using a large number of image pairs. In this thesis, ensemble correlation is used to study the turbulence in two specific cases: a turbulent jet and high Reynolds number pipe flow. The ensemble correlation, though being a time averaged quantity, contains information on the velocity Joint Probability Distribution Functions (JPDFs) which can be used to estimate the turbulence statistics. This information can be retrieved from the shape of the ensemble correlation. It is assumed that the ensemble correlation is the convolution of the auto-correlation and the velocity JPDFs. The velocity JPDFs are retrieved by estimating the second moments of the ensemble correlation by fitting a Gaussian profile, and further using a correction for auto-correlation. The second moments can be used to estimate the Reynolds stresses in the flow. The proposed method is validated by implementing it to study the turbulent jet. For this purpose, 9000 image pairs are acquired. The results show that the retrieved moments predict the Reynolds stresses accurately. Further, high Reynolds number pipe flow experiments are performed in the Alpha Loop facility at Deltares. The bulk Reynolds number is varied from 0.3 to 0.6 million, acquiring 20,000 image pairs per measurement series. Also, the pressure drop is measured across the pipe along with the PIV measurements to estimate the roughness of the pipe. The results from the PIV measurements show that the mean velocity profile follows a similar trend as observed in the literature. A small bias is observed between the mean velocity obtained from the ensemble correlation and that obtained by averaging the instantaneous velocity vectors, in regions of high velocity gradients. The Reynolds shear stress estimated from the shape of the ensemble correlation is underestimated. However, the streamwise turbulent fluctuations estimated follow a trend, similar to that observed in the literature.

At the turbulent/non-turbulent interface (TNTI) of a jet flow, momentum is transferred from the turbulent jet fluid to the fluid at rest. This transfer is governed by two mechanisms. One of them is related to the large scales of the flow and the other relates to the small scales of the flow. Which of the two is dominant is still a point of discussion. To analyse the TNTI the instantaneous information of the flow and the location of the interface is of great importance. However, the interface simultaneously develops and travels downstream from the nozzle. With stationary measurement techniques this limits the number of frames the interface can be seen developing as it travels in and out of the FOV. In this thesis the TNTI is measured using a camera system that moves along with the TNTI to get high resolution instantaneous measurements of the same part of the evolving interface. The measurement techniques that are used for this experiment are Particle Image Velocimetry (PIV) and Laser Induced Fluorescence (LIF). The cameras for these measurements are mounted to a motorised frame to keep the same part of the interface in view of the cameras as it develops. Three Reynolds numbers are measured with this setup and the cameras move approximately 50Dn at a velocity close to the velocity of the interface along a diagonal path to follow the evolution of the TNTI. One measurement with a Reynolds number of approximately 1.2 × 104 has been processed to show the quality of the results that can be obtained from such a measurement. The velocities are computed by an in-house interrogation analysis program in MATLAB to overcome the wide range of particle displacement found in this experiment, due to the presence of both the centreline and the TNTI of the jet. The TNTI is detected using the LIF data and a threshold detection method from literature that determines a threshold value. The results from the PIV and LIF processing is combined to compute the average conditional vorticity over the interface. The results show that the PIV analysis is able to compute the velocities of almost the entire jet. Only showing a lot of spurious vectors close to the nozzle in the core of the jet. The LIF edge detection algorithm, on the other hand, does not perform as expected. In multiple instances, an internal interface is detected instead of the TNTI. The TNTI is also determined by identifying a threshold value through a visual inspection of the LIF images. This manually determined TNTI is used as a point of comparison for the average conditional vorticity profiles. The average conditional vorticity profiles support the conclusion that an interface internal to the jet is detected when comparing the interface from the algorithm to the interface determined manually. Although not quantified in this thesis, there are moving measurements that show the same section of the TNTI evolving for many frames. This gives the co-moving measurements a clear advantage in measurement time compared to stationary measurements when trying to measure the moving TNTI. Refinements to the experimental setup, the behaviour of these internal interfaces and the detection of the TNTI can be of interest for future research.

The prospect of stricter national and international emission standards for the shipping industry are a driving force in the search for alternative shipping fuels, such as liquefied natural gas (LNG). However, new challenges arise with the widespread use of LNG. For example, there is a desire to use LNG cargo containment systems at lower filling levels. These filling levels are strictly limited to prevent the movement of liquid inside the containment system, which is known as sloshing. Sloshing inside a cargo containment system can result in extreme wave impact events with the potential to cause structural damage. Therefore, a fundamental understanding of these extreme wave impact events is required before studying increasingly complex phenomena.   The study of wave impacts on a wall has been an active area of research for decades. Moreover, the impact of waves upon structures is relevant for many fields such as ocean, coastal, and maritime engineering. The generation of repeatable waves in a laboratory environment is not trivial (Bagnold, 1939). Small changes in the experimental conditions, such as the water depth and the wave generation method, result in significant impact pressure variability. The impact pressure variability is even observed in carefully repeated wave impact experiments with minimal variability of the input parameters. For these measurements, the source of the impact pressure variability is thought to be the instability development on the wave crest. However, the mechanism that is responsible for the formation of these instabilities is still largely unknown.   The aim of this work is to gain insight in the sources of wave impact pressure variability. This is accomplished using direct measurements of the liquid free surface and particle image velocimetry of the surrounding air. The measurements are limited to a single wave (i.e., with a fixed steering signal and water depth) at atmospheric conditions, because of the complexity of the experimental measurements. A plunging breaking wave with a large gas pocket is generated that impacts on a vertical wall. The compression of the large gas pocket induces a significant gas flow between the wave crest and the vertical impact wall, which results in the formation of instabilities on the wave crest.   Quantitative measurements of the liquid free surface are obtained with an extension of the planar laser induced fluorescence (PLIF) method. The newly developed scanning stereo-PLIF measurement technique uses a stereo-camera set-up with a self-calibration procedure adapted for free surface flows. Thereby, the stereo-PLIF technique enables measurements of a free surface over a two-dimensional domain (e.g., y=f(x,z,t)). The system is versatile with a minimal influence on the fluid properties and the measurement domain can be scaled as needed.  A repeatable plunging breaking wave is created in the wave flume of the Hydraulic Engineering Laboratory at the Delft University of Technology. The wave encloses a gas pocket as it approaches the vertical impact wall. Initially, the plunging breaking wave is globally comparable to waves that do not impact on a vertical wall. The aspect ratio of the cross-sectional area of the gas pocket remains relatively constant at Rx/Ry = 1.6 ( ∼√3). Furthermore, the wave velocity (√gh0) and wave tip velocity (1.2√gh0) are initially similar to that of a plunging breaking wave. On the other hand, the trajectory of the wave tip is altered compared to that of a typical plunging breaking wave.   The trajectory of the wave tip is globally similar over repeated wave impact measurements. However, moments before impact the wave tip is deflected by the gas expelled from the gas pocket. The deflection of the wave tip introduces significant variation between the repeated measurements. On close inspection, the wave tip resembles a liquid sheet, that is destabilized by an initial Kelvin-Helmholtz instability (Villermaux et al., 2002). The flapping liquid sheet accelerates the wave tip, which triggers the development of a Rayleigh-Taylor instability. This results in approximately equally spaced liquid filaments (i.e., liquid fingers) over the spanwise direction of the wave. The spanwise wavelength depends on the density ratio (ρa/ρl) and surface tension, which was previously shown to be a source of wave impact pressure variability.  Additionally, particle image velocimetry measurements are performed to determine the interaction between the liquid and gas during a wave impact event. The global gas flow is similar to that of a plunging breaking wave, where a vortex develops on the leeward side of the wave. The vortex consistently separates from the breaking wave and lingers at the back of the breaking wave in the stagnant air. The development of circulation is typical for a vortex that eventually separates at a universal time scale denoted by the formation number (Gharib et al., 1998). However, a typical formation number can not be defined in this particular case, due to the simultaneous change of both the length and velocity scales.  The velocity profile between the wave tip and the vertical impact wall resembles that of a flow past a bluff body. A fit of the measured velocity profile agrees well with the velocity derived from mass conservation. Interestingly, the velocity close to the wave tip is approximately 2 times higher than the bulk velocity estimate. The high velocity close to the wave tip can thus result in an earlier onset of instability development compared to estimates based on the bulk velocity. Furthermore, the flow tends to separate close to the tip just before impact on the vertical wall. The effect of this flow separation on the impact pressure variability depends on the global wave shape prior to impact. For the case with a disturbance on the wave crest, the secondary vortex that forms close to the wave tip tends to break up. On the other hand, if the wave crest is smooth the secondary vortex remains attached. The attached secondary vortex increases the lift on the wave tip, which results in a significant deflection of the wave tip. Consequently, the development of secondary vortices close to the wave tip results in wave impact pressure variability, as the typical deflection is larger than the membrane diameter of a contemporary pressure transducer.   Local phenomena such as flow separation and the development of instabilities define the variability of the peak pressure during wave impacts. On the other hand, the global characteristics of an air-water wave impact on a vertical wall can be retrieved with pressure impulse models (Cooker et al., 1995). The maximum wave impact pressure is relevant for LNG containments systems and wave energy converters. Consequently, numerical models that aim to quantify wave impact pressure variability require accurate models of both the gas phase and the development of free surface instabilities. @en

This thesis describes the investigation of the dynamics of turbulent shear flows over non-smooth surfaces. The research was conducted in two parts, related to the experimental facility used in combination with the applied functional surface. The first part describes the experiments of a turbulent Taylor-Couette flow over a riblet surface. The Taylor-Couette facility proves to be an accurate measurement device to determine the frictional drag of surfaces under turbulent flow conditions. Sawtooth riblets are applied on the inner cylinder surface and have the ability to reduce the total measured drag by 5.3% at Res=4.7x104.  Under these conditions, a small shift is observed in the azimuthal velocity profile that indicates the change in the net system rotation, which on its turn affects the quantity of drag change, the so-called rotation effect. A model based on the angular momentum balance is proposed and quantifies the drag change due to the rotation effect. Using the total measured drag change, the model accurately predicts the velocity shift in the azimuthal direction. In addition to the steady operational conditions, periodically driven Taylor-Couette flows were investigated by modulating the velocity between the two cylinders as a sinusoidal function, while maintaining RΩ = 0. The main scaling parameters are the shear Reynolds number Res, the oscillation Reynolds number Reosc and the Womersley number Wo, such that the required power to overcome the frictional drag becomes equal to <Pd> = f(Res,Reosc, Wo). Large velocity amplitudes A = Reosc/Res > 0.10 induce the growth of frictional drag due to the additional turbulent fluctuations. The required power to overcome the frictional drag is given by <Pd> = <Pd,0>(f(A)+ K*Wo4A2). The first term represents the analytical quasi-steady state solution with the accompanying velocity modulation, while the second term involves the magnitude of the boundary acceleration with the associated velocity fluctuation, where K* is the conditional scaling-factor between the additional drag and the dimensionless acceleration. Riblets are still able to reduce the frictional drag under small accelerations of the periodically driven boundaries, but the effect declines drastically or even enhances the frictional drag when the boundary acceleration becomes more significant.   The second part of this thesis describes the assessment of the applied water tunnel and the interactional behavior between a compliant coating and a turbulent boundary layer flow in the tunnel. In the assessment of the water tunnel, the Clauser chart method showed to be a suitable procedure to quantify the local wall shear stress τw. The interaction between a compliant wall and the near-wall turbulent flow was examined by applying in-house produced visco-elastic coatings with three different stiffnesses.  Two typical flow-surface interaction regimes were identified; the one-way coupled regime and the two-way coupled regime. The one-way coupled regime is valid when the turbulent flow initiates moderate coating surface deformation, while the fluid flow remains undisturbed. All of the three coatings exhibited the one-way coupled interactional behavior, where the surface modulations ζ were smaller than the viscous sublayer thickness δv and scale with the turbulent pressure fluctuations over the coating shear modulus, i.e. ζrms ~ prms/|G*|. In this regime, the surface waves have the propagation velocity in the order of cw = 0.70-0.80 Ub, indicating a strong correlation with the high-intensity pressure fluctuations in the turbulent boundary layer away from the wall. The two-way coupled regime has only been observed for the coating with the lowest shear modulus when Ub > 4.5 m/s, indicating significant surface deformation accompanied by additional fluid motions (u',v') and an increase in the local Reynolds stresses. The velocity profile shifts downwards Δu+ in the log region, which verifies the drag increase due to the significant surface undulations. The visualizations of the surface deformation showed the formation of wave-trains with high amplitudes originating from the initial surface undulations caused by the pressure fluctuations in the turbulent boundary layer (i.e. one-way coupling). When these early surface undulations start to protrude the viscous sublayer, the turbulent flow is capable of transfering more energy towards the coating and initiates the wave-train with high amplitudes. The wave-trains dominate the coating surface incrementally with increasing bulk velocity and propagate with a wave velocity of cw = 0.17-0.18 Ub. The 1-way/2-way regime transition is estimated to occur around ζrms > δv/2. The turbulent flow along the slow-moving wave-trains resembles the classical phenomenon of a turbulent flow over a rigid wavy surface, with a local acceleration and deceleration of the fluid. When the wave-trains start to dominate the coating surface, a linear correlation determines the abovementioned downward shift Δu+, based on the wall-normal velocity component dζ/dt. No frictional drag reduction under turbulent flow conditions was found in this study with this type of visco-elastic compliant coatings.   @en

Fluids above the critical point are widely used in industry. Chemical, pharmaceutical, food industry and energy production are some examples. In the energy production sector they are mainly used as cooling fluids, because they allow to increase the thermal efficiency of the power plants. However, the fundamentals of their heat transfer behavior are still unknown and current heat transfer models fail to predict it. Supercritical (SC) fluids are characterized by strongly varying fluid properties, which are responsible for their particular heat transfer behavior and make them very difficult to model, simulate and experimentally investigate. In past studies, buoyancy was identified as a key cause for the heat transfer deterioration observed in SC fluids. The aim of the research described in this thesis is to investigate the possibility of performing non-intrusive local velocity measurements with the optical technique PIV and to acquire global heat transfer measurements, with strongly changing fluid properties at SC conditions. The experiments were performed in a pure buoyancydriven flow: a Rayleigh-Bénard (RB) flow. The velocity fields of RB convection with strongly varying properties, beyond the so-called Oberbeck-Boussinesq (OB) approximation, were experimentally studied at atmospheric pressure first. An increase of the time-averaged velocity close to the bottom wall of the cell with respect to the top wall of about 13% was found. This finding confirmed experimentally a top-bottom ”broken symmetry” in the velocity field, which was observed in previous numerical and theoretical studies, but it was never experimentally demonstrated before. The heat transfer with strongly variable properties at SC conditions for constant Prandtl and Rayleigh numbers, specifically defined outside the validity range of the OB approximation, was experimentally studied. The measurements were performed at the Max Planck Institute of Dynamics and Self-Organization in Göttingen (Germany), with a European EuHIT project. It was observed that the measured Nusselt number defined for non-OB conditions was different from point to point, showing that merely the Rayleigh and Prandtl numbers are not sufficient to determine the heat transfer through the cell. It was also seen that the measured Nusselt number was 16% larger with respect to the one predicted by the Grossmann-Lohse theory (2000) for the same Rayleigh and Prandtl numbers at OB conditions. A feasibility study of particle image velocimetry (PIV) at SC conditions was done by using the background oriented schlieren technique (BOS). An estimation of the PIV experimental uncertainty at SC conditions was done with the statistical correlation method proposed by Wieneke et al., (2015). PIV was successfully performed at SC conditions. Main difficulties about its applicability were due to blurring and optical distortions in the boundary layer and thermal plumes regions. PIV measurements were performed at three different magnitudes of density difference between top and bottom of the cell. Two of the three experiments were done at similar Rayleigh and Prandtl numbers, defined for non-OB conditions: one towards the liquid phase and the other one towards the gas phase. The former showed a lower large scale circulation (LSC) velocity than the latter. All cases showed the presence of one asymmetric LSC roll, which is different from a typical RB convection flow at OB conditions. Improvements in the accuracy of PIV measurements and the acquisition of more heat transfer data at SC conditions, would help the study of the thermal and viscous boundary layer thicknesses and turbulence modifications that are responsible for different heat transfer regimes in SC fluids.@en

The precise manipulation of particles and droplets is crucial to many microfluidic applications in engineering. The design of microfluidic devices is generally tailored to perform a specific task, with each specific application requiring a unique and fixed design. In this way, using a single device to perform multiple analyses of a wide range of specimens, from biological to chemical specimens, is unfeasible. Here, we address this issue and present a microfluidic approach that dynamically controls the hydrodynamic flow and the streamlines to realize complex multi-particle manipulations within a single device. Our approach combines the design of a flow-through microfluidic flow cell together with an optimization procedure to find a priori optimal particle path-lines, and a Proportion-Integral-Derivative-based (PID) feedback controller to provide real time control over the particle manipulations. In our device, particles are manipulated with hydrodynamic forces, by using a uniform flow through the flow cell and three inlets perpendicular to the flow cell. The streamlines within the device are manipulated by injecting or extracting fluid through the three inlets. We demonstrate the robustness of our approach by performing multiple functions within the device, including particle trapping, particle sorting, particle separation and assembly. We show that the real time control procedure affords accurate particle manipulation, with a maximum error on the order of the diameter of the particle. Our particle manipulation approach is particularly well suited to biological samples and living cells.@en

There is increasing attention for the effects of anthropogenic underwater radiated noise (URN) on marine fauna. This is expected to lead to regulations with respect to the maximum permitted sound emissions of ships. It is known that cavitating tip vortices, generated by ship propellers, are some of the key contributors to URN. Consequently, there is a need to evaluate propeller designs with respect to noise generation in a design stage. Computational fluid dynamics (CFD) has the potential to offer detailed insights into cavitating vortex dynamics and noise sources, at a reasonable cost. URN can be efficiently estimated using CFD in combination with an acoustic analogy. In order to use such predictions in a design process, it is essential to understand and quantify the errors associated with the numerical predictions of noise sources. This thesis investigates the reliability of such evaluations and aims to reduce occurring modelling errors. To compute noise sources, it is necessary to simulate cavitation dynamics using scale-resolving simulations (SRS). Here, part of the turbulence kinetic energy spectrum is resolved in space and time, as opposed to being modelled using Reynolds averaged Navier-Stokes (RANS). The SRS method of choice in this work is the partially averaged Navier-Stokes (PANS) method. Bridging models, such as PANS, exhibit a smooth transition and absence of commutation errors between RANS and large eddy simulation (LES) zones, in contrast to hybrid models such as detached eddy simulation (DES). The formulation allows for a theoretical decoupling of the discretisation and modelling errors, thereby enabling verification and validation processes. PANS allows the user to select the ratios of resolved-to-total turbulence kinetic energy and dissipation (rate). Appropriate settings and methods to estimate these settings a priori are investigated. Furthermore, a new PANS closure is developed, which offers improved convergence behaviour compared to more commonly used models, and is better suited to application for multiphase flows. It has been shown repeatedly in literature that SRS should be accompanied by physical inflow boundary conditions, where time-varying fluctuations, resembling turbulence, should be inserted upstream of the object of interest, to prevent laminar solutions. However, from literature it is clear that for maritime applications this is often neglected. To the knowledge of the author, there is no previous application of such an inflow in combination with cavitation. In this PhD thesis, a synthetic inflow turbulence generator (ITG) is implemented, and tested for several test cases in wetted and cavitating conditions. For these cases, the numerical errors, consisting of discretisation, iterative and statistical errors are evaluated. Firstly, the results when using the ITG are compared against recycling flow results for a turbulent channel flow, using different SRS methods. It was shown that the ITG can deliver a resolved turbulent inflow at lower computational cost. Secondly, the effect of neglecting such an inflow was tested for the Delft Twist 11 hydrofoil, where it was shown that simulating such a flow with a low ratio of resolved-to-total turbulence kinetic energy can lead to flow separation at the wing leading edge. This is in contrast to experimentally observed behaviour. The inclusion of the ITG can reduce this modelling error, although the sheet cavity dynamics remain largely unaffected. Finally, an elliptical wing with a cavitating tip vortex is simulated. The observed vortex dynamics are compared against a semi-analytical model from literature. To obtain vortex dynamics, the ITG was shown to be necessary. The far-field noise generated by the vortex is quantified and related to the cavity dynamics. Some of the main contributions of this research are improved insight in the use of SRS in cavitating conditions, in simulating cavity dynamics and in using an ITG to obtain flow fields representative of experimental conditions. In this way it has enhanced our understanding of the ability and limitations in the prediction of acoustic sources due to cavitation. To improve predictions of cavitation dynamics it is recommended to address the cavitation model and the method which describes the cavity interface, to reduce the discrepancy in average cavity size between simulations and experimental observations.@en

Digital microfluidics emerged and established over the last 20 years. The scope of applications ranges from pharmaceutical to biological, chemical and even thermal processes. The practical use of Taylor droplets as individual reaction chambers, mixing vessels or simply as transportation vehicles on process engineered microfluidic devices allows for precise control and process intensification of precious materials or sparse samples. Despite this widespread use of Taylor droplets, some of the fundamental physics of this flow have not been investigated experimentally.@en

This thesis describes methods and first analyses to study the shape and motion of a large school of approximately 2000 Harengula clupeola (false herring) in three-dimensional space. This school of fish was present at the large-scale and publicly accessible ocean aquarium of the Rotterdam zoo, in the Netherlands known as Diergaarde Blijdorp, from spring 2017 to summer 2020.@en

Synchronization of oscillators is an ubiquitous phenomenon that involves mechanical systems, like pendulum clocks, but also biological systems, like peacemaker cells in the heart or neural activity in the brain. If we consider biological systems at the microscale, namely at the scale of cells, we find that processes like locomotion and fluid transport often exploit synchronization of mechanical oscillators called flagella orcilia. These oscillators at the microscale are whip-like structures extending from the cell body. They are present in a number of micro-organisms like sperm cells, Paramecium or the algae C. reinhardtii. In human, cilia are found in the lungs, the respiratory tract and the middle ear. Cilia are activated in a coordinated way to effectively carry out their function, such as draining mucus. The mechanism behind this cilia coordination is still debated. It is not clear how very simple organisms lacking any feedback system have developed complex oscillatory patterns involving coordination among a multitude of cilia or flagella. There is high interest in understanding the fundamental principles ruling ciliary dynamics, since they would impact medical and engineering applications. The purpose of this thesis is to investigate the mechanisms regulating the synchronization of cilia and flagella.@en