In recent years, computational fluid dynamics (CFD) has become an essential design tool across various industries, allowing engineers to tackle complex fluid dynamics problems that would otherwise require costly and time-consuming real-life experiments. For Formula 1 teams, who must experiment within strict time limits in the wind tunnel and on track, the ability to simulate airflow around their race cars under various conditions is crucial for maintaining competitiveness in the fast-paced world of Formula 1 racing. Reynolds-Averaged Navier-Stokes (RANS) simulations remain the industry standard for simulating turbulent flows, as they allow engineers to conduct simulations efficiently. However, this efficiency comes at the expense of accuracy, as RANS cannot resolve all turbulence scales, leading to uncertainties.

Recent advances in data-driven RANS turbulence modeling have enabled partial correction of these uncertainties. However, obtaining a correction that is generalizable under different geometries and flow conditions remains a challenge. Turbulence models are calibrated to fit specific flow regimes, so correcting these models across the entire domain can disturb these calibrations, worsening performance. A solution is to divide the domain into regions based on identifiable physical phenomena and apply local corrections without disturbing calibrated regions. In Formula 1 race car design, the most critical region is the shear layer, where RANS shows the largest discrepancies.

In this thesis, a classifier was developed to distinguish the shear layer from the rest of the domain based on the ratio between turbulent kinetic energy production and destruction, as well as turbulence intensity. Within this classifier region, corrections to the k-ω SST turbulence model are made by extracting model form errors from high-fidelity data using k-corrective-frozen RANS. These corrections include a residual term added to both the k and equations and a term for the anisotropy of the Reynolds stress tensor. The Spars Regression of Turbulent Stress Anisotropy (SpaRTA) framework, based on elastic-net regularization, was used to regress symbolic expressions for the corrections, enabling their application to simulations of unseen test cases.

The models discovered with the SpaRTA framework for the shear layer show promising results, improving the prediction of separation and reattachment positions. These models were tested on various geometries and simulations at different Reynolds numbers, demonstrating a certain level of generalizability. While there is room for further improvement, this thesis shows that integrating targeted model corrections into RANS simulations, informed by isolated shear layer data, can enhance the understanding and prediction of shear layer dynamics in 2D-separated flows.

The Helix approach is a dynamic control method that mitigates wind turbine wake effects by pitching the turbine blades individually. In current literature the method is mainly researched with a high-fidelity actuator line model (ALM). Because of the computational constraints imposed by this model, the effect of the Helix has not been researched for turbine arrays longer than 3. To investigate the effects of the Helix on a farm size scale, a lower fidelity model capturing the Helix is developed in this research. This model is based on a uniform actuator disk model and is called the H-ADM. The H-ADM is validated in laminar conditions at a fine resolution by comparing wake recovery to the ALM. The H-ADM captures both the additional wake mixing and wake displacement caused by the Helix. Furthermore, the H-ADM correctly predicts differences between the counterclockwise (CCW) and clockwise (CW) Helix. Due to modelling assumptions, the H-ADM can run simulations at a coarser resolution and with a smaller time-step than the ALM. Computational gains are achieved with the H-ADM as a result. This research shows that similar simulations are executed over 640 times faster for the H-ADM compared to ALM. The model is used to gain a deeper understanding of the mechanisms causing differences between the CCW and CW Helix. It turns out these differences fade away if no rotational force is exerted by the wind turbine. Finally, the computational gains of the H-ADM are leveraged by applying the Helix on the first row of a large scale wind farm with 48 turbines. For the considered farm the CCW Helix seems to outperform the CW Helix in terms of total power gains.

This thesis contributes to the scientific underpinning of the battle against fruit frost. Fruit frost is the freezing damage to blossoms when in the growing season the night temperature drops below 0±C. This results in damaged or undeveloped fruits, and a yield loss for the fruit farmer. Several techniques against fruit frost have been developed, including sprinkling and wind machines, often in combination with meteorological models, for example, to predict air temperature. However, the contribution of heat exchange with the soil to moderate orchard temperatures is often not included. In this thesis, this heat transfer is investigated, as an increase of heat transfer from the soil to the orchard during the night is a potential remedy against fruit frost. The research is based on measurements for soil temperature, soil heat flux, and soil moisture from two locations (1. Haarweg (Gelderland), The Netherlands 2. Bushland (Texas), The U.S.A.). First, a numerical model is developed to calculate the temperature and soil heat flux profiles for a soil layer. The results are compared to the results of an already developed analytical model. Second, the thermal parameters, that are of influence on the heat transfer, are analyzed by assessing a) their robustness in relation to the model and b) their relation to soil moisture. Because a numerical model is more flexible for shorter periods of data compared to an analytical model (because of underlying assumptions), it can be used to relate the parameters to (daily) varying soil moisture. Third, the numerical model for heat transfer is extended to the vegetation layer, and, again, the results are compared to analytical results. The model is created by assuming homogeneity in both separate layers and by discretizing the governing heat equation over the domain. The results show that the model reproduces temperature and soil heat flux in the soil layer with similar accuracy as the analytical, harmonic model. One thermal parameter, the diffusivity, is robust and does not show a clear dependency on soil moisture. The model is however sensitive to deviations in the other parameter, the heat conductivity. The model shows a clear relation between conductivity and soil moisture, and from this, a site-specific quantitative relation is determined. This relation however is only valid in the investigated region of moisture variation and we recommend future research to cover data in a broader range of soil moisture. Overall, we conclude that the model successfully reproduced the temperature and soil heat flux throughout the full vegetation-soil continuum.

This work aims at finding a second order accurate level-set method which solves the Stefan problem with non-homogeneous Dirichlet boundary conditions in one dimension. The numerical accuracy of the FTCS-scheme, BTCS-scheme and Crank-Nicolson scheme for the discretization of the heat equation was considered, as well as the accuracy of the first order Upwind method, Leapfrog method and the Lax-Wendroff method for the discretization of the advection equation. A level-set method was developed using a finite volume Crank-Nicolson scheme for the discretization of the moving boundary. A second order accurate scheme for solving the advection equation was developed using Lagrange extrapolation polynomials. The moving boundary velocity was estimated using second order Lagrange polynomials. The developed method was found to be second order accurate for a specific range of ratios between time step size and spatial step size.

Background: The Fontan procedure is the last of three stages of congenital heart surgery to treat children born with a single ventricle heart defect. In these patients, a balanced hepatic blood flow distribution (HFD) towards both lungs is important, since a lack of “hepatic factor” has been associated with the formation of pulmonary arteriovenous malformations (PAVMs). With imaging modalities and computational fluid dynamics (CFD), the hepatic blood flow distribution can be studied. Recent CFD studies indirectly quantify HFD to both lungs by tracking ‘hepatic flow’ particles that are uniformly seeded in the Fontan tunnel and analyzing the distribution of these particles towards both lungs (conventional approach). However, this approach is based on the unvalidated assumption that there is a uniform distribution of hepatic blood flow in the Fontan tunnel. Aside from CFD modeling, previous clinical imaging studies showed that respiration has a tremendous impact on the hepatic blood flow in Fontan patients; however, these studies only used Doppler Ultrasound because the resolution of real-time phase-contrast magnetic imaging resonance (PC-MRI) is not yet good enough for direct hepatic flow quantification. Objectives: this study aimed to investigate the distribution of hepatic blood flow in the Fontan tunnel using CFD simulations. Another objective was to quantify the HFD towards both lungs using particle tracking directly from the inlets of the hepatic veins (novel approach) and compare these HFD results to the conventional approach. Furthermore, we performed an in-depth flow analysis and developed a new method to indirectly quantify the hepatic flow in the hepatic veins and the respiratory effects on it using real-time PC-MRI. Methods: Unsteady CFD modeling was used to assess the mixing of hepatic blood flow within the Fontan tunnel and the different HFD quantification methods. Therefore, we created three-dimensional reconstructions of the patient-derived Fontan anatomies based on MRI imaging that included the geometry of the hepatic veins in the computational fluid domain. We created two types of CFD models: 1. CFD models that only included the cardiac effects on blood flow, and 2. CFD models that considered both the cardiac and respiratory effects on blood flow. For the in-depth flow analysis using real-time PC-MRI, we first measured the blood flow in the Fontan tunnel, just above the entrance of the hepatic veins. Second, we derived the IVC blood flow below the hepatic veins. By subtracting these flows, we indirectly quantified the hepatic blood flow. Results and conclusion: The main findings were that hepatic blood flow was non-uniformly distributed within the Fontan tunnel and substantial differences in HFD between the conventional approach and novel approach were found in both types of CFD models that either ignored or considered respiration. Additionally, we showed that it was feasible to indirectly measure hepatic blood flow in Fontan patients while using real-time PC-MRI. These derived flow parameters extracted from real-time PC-MRI acquisitions confirmed what previously has been described that the hepatic blood flow in Fontan patients was heavily influenced by respiration.

Welding is an extensively used technology by different industries such as aerospace, automotive and marine industry. Welding involves the temporal melting of materials in order to join them together. The quality of the weld is known to depend on the fluid flow and heat transfer in the material. A way to influence the heat transfer and fluid flow, and thus the quality of the weld, is by using different power-density distributions. However, the current literature falls short when combining the different power-density distributions with weld pool behaviour. The aim of this thesis is to study the effects of laser power-density distribution on heat and fluid flow in molten metal melting pools. This will be done by running weld pool simulations using a 2D axisymmetric, flat surface model. These simulations use a top-hat, Gaussian or doughnut laser power-density distribution. Besides, the cases have been simulated for a 20 ppm and a 150 ppm sulfur concentration. That the sulfur concentration has an influence on the weld pool shape was already shown in previous studies. However, it was not known how much the influence of sulfur concentrations varied for different laser power-density distributions. The laser power-density distributions are found to have a significant influence on the weld pool shape. The doughnut power-density distribution created a three times deeper weld pool than the Gaussian and top-hat power-density distribution. The maximum temperature of the weld pool was also influenced by the power-density distribution. In the case with a sulfur concentration of 20 ppm a discrepancy of 1000 K has been found between the Gaussian and doughnut distribution. The influence of sulfur concentrations varied for different laser power-density distributions. This influence of the sulfur concentrations has been measured through the difference in aspect ratio for different sulfur concentration cases. For both the top-hat and Gaussian power-density distribution, the aspect ratio differed 0.02 when using a 20 ppm versus a 150 ppm sulfur concentrations. The doughnut power-density distribution reached a higher difference in aspect ratio of 0.1 when using a 20 ppm versus a 150 ppm sulfur concentrations. The use of different power-density distributions in weld pool simulations results in different weld pool behaviour. Besides, the combination of different sulfur concentrations with power-density distributions largely influenced shape, temperature and fluid flow of the weld pool, which results in a great variation of weld pool shapes and sizes. Because there are different requirements for different welding cases, these results are useful in order to construct a quality weld.

In this paper we introduce a setup to investigate aeolian saltation and surface dynamics on a centimetre spatial resolution and a sub second temporal resolution. We develop a Lagrangian saltation model and a high-resolution surface model, which we couple to each other and to a turbulence resolving large eddy simulation model. The simulated transport takes place primarily in the form of aeolian streamers, bursts of elongated transport structures parallel to the wind field, which result in a mass flux signal that is highly heterogeneous both in space and time. The temporal frequency responses up to 1 Hz of the mass flux and wind field share the same characteristics, which indicates a coupling between the two. The system can be in equilibrium, during which the stress profiles induced by the particles, the turbulent fluxes and the imposed large scale pressure gradient balance each other. A bimodal shape is found in the mass flux profile, in which we can distinguish an upper and lower saltation layer. The upper layer is associated with a transitional phase between transport by saltation and suspension that exists in the aeolian streamers. Furthermore, The setup is able to simulate ripples, although future research is needed to investigate the mechanisms that influence the final shape of the ripples.

Porous media are used for a wide variation of applications in energy production and storage. One of those applications is the storage of heat which are of great importance in the renewable energy transition. Most literature reported limit itself only to one single thermal conductivity for the porous media. In reality however, the porous media often consist of multiple materials with different conductivities. This thesis researches numerically natural convective heat transfer in a porous media with both conductive and insulating objects.
The porous medium is simplified to a side heated cavity filled with water as fluid (Pr = 7), 32 aluminum objects with a thermal conductivity ratio of λ* = λ_al/λ_f = 337.33 and 32 wooden objects at thermal conductivity ratio λ* = λ_w/λ_f = 0.29. 4 cases, each with the aluminum and wooden cubes differently configured are simulated at a Rayleigh number(Ra) of 10⁵, 10⁶ and 10⁷. Conjugate heat transfer (the fluid fase and solid fase are modelled as separate regions) simulations are done with direct Navier-Stokes where Ra = 10⁵ and 10⁶ are run steady and Ra = 10⁷ is run transient. There are two different geometries simulated. One when the objects are unattached (heat cannot directly conduct into the objects from the walls) to the wall and when they are attached (heat can directly conduct into the objects) to the wall. From these simulations the temperature profile, the velocity field and the Nu profile is obtained. Also the average Nu at the hot wall and the U_rms are calculated and compared between the different cases. Lastly the thermal disequilibrium %|DT| between the solid temperature and fluid temperature is calculated and compared between place of the object in the cavity and between the cases. The simulations show that conductive objects increased heat transfer at Ra = 10⁵ when they are close to the wall. This is because they decrease the thermal resistance heat travels in the thermal boundary layer. Insulating objects on the other hand show decreased heat transfer as they increase the thermal resistance in the thermal boundary layer. At Ra = 10⁶ and at Ra = 10⁷ the circulation is to strong and the thermal boundary layer is mostly in between the wall and the objects so there is little difference seen in the flow and heat transfer. When the objects are attached to the wall the simulations show very different results for the flow and heat transfer and is very dependent on the conductivity and position of the object. Conductive objects attached at the wall, especially at the bottom of the wall where the temperature-gradient is the biggest, greatly improve the flow and heat transfer. They improve the area heat is transported from the wall to the fluid and thus increase the buoyancy force significantly. Insulting objects attached to the wall on the other hand prevent heat from flowing through the object and also prevents flow from reaching the wall. A big decrease in velocity and heat transfer is seen for all Ra values. For a case where conductive objects are placed against the wall and there are insulating objects in the middle, heat transfer can be improved compared to a fluid-only cavity for Ra ≥10⁶.

Steel is commonly coated to protect it from corrosion. One method of applying this is by using Physical Vapor Deposition, which can be done by using multiple jets. In this process jets next to each other interact. This paper's main aim is to investigate the interaction effect of two rarefied two-dimensional vapor jets in vacuum and how this is influenced by the distance between jets and inlet density. Furthermore, the analytical solution for the collisionless case for a single jet is extended to dual jets. Additionally, the objective is to maximize the processing speed for the use of coating, with a certain uniformity for the parameters researched in this paper and finding an optimal method in doing so. This study is done by using the Direct Simulation Monte Carlo (DSMC) method.
The analytical solution gave the same results as the collisionless DSMC method for both single and dual jets. Simulations with a strong interaction effect resulted in a shock. These behaved similar compared to three-dimensional jets, in the plane of the jets. The shock results in a secondary jet, which has a lower density in the middle. The interaction effect depends primarily on the inlet density. Multiple regimes are observed for different inlet density ranging from small change in properties to a shock wave, with a transitional regime inbetween. The influence of the distance between the jets is found to result in a higher density at the axis of the inlet behind the shock, for bigger distance between jets. However, for very small distances between jets compared to the inlet size the shock is weak. For the optimization, it resulted in the conclusion that the optimal coating in general is applied with the smallest distance between jets. This generally gives a better uniform coating and increased performance. However, this is not always the case when constraining the distance between the jets and sheet, as it only holds if the shock between the jets for this distance. Furthermore, an approximation is found for the optimization, which results in fewer simulations needed.

Offshore wind energy is considered as a powerful form of renewable energy generation. It plays an important role in accelerating the world’s transition towards sustainable energy sources and reducing carbon emissions by fossil fuels. This study focuses on the advancement of service planning, related to this renewable energy source, by researching high resolution metocean modelling. The research aims to assess and advance the modelling performance of typical metocean parameters by using atmospheric large-eddy simulations coupled to a spectral wave model. The GPU-Resident Atmospheric Simulation Platform (GRASP) coupled to Simulating WAves Nearshore (SWAN) was used to simulate the atmospheric- and oceanic conditions in the Gemini wind farm, located in Dutch waters. Large-scale boundary- and initial conditions were provided by the fifth generation of ECMWF’s ReAnalysis (ERA5). Relevant metocean parameters were modelled using two different coupling configurations. The one-way coupled simulation concerns the forcing of SWAN by GRASP friction velocities, for an accurate representation of the one-way momentum exchange to the ocean surface. The two-way coupled simulation concerns the momentum exchange of the friction velocity and roughness length. To accurately represent the sea surface roughness, the parameterization of Taylor and Yelland (2001) was used in this study. Both coupled configurations were used to simulate the first two months of 2017, which were subsequently validated using the available observations.
This study revealed that both coupled simulations caused a reduced value for the roughness length in wind- and wave wake conditions. Furthermore, a spatially averaged reduction in the sea state is observed due to the wake effect, where the magnitude of this wave deficit follows the line of a typical wind turbine thrust curve. The effect is however small compared to a realistic significant wave height. Besides, the two-way coupled simulations provided higher average roughness lengths in comparison to the one-way coupled simulations. This resulted in higher friction velocities and drag coefficients for the two-way coupled simulations, which subsequently reduced the time- and slab averaged wind profiles.
Moreover, the modelling performance of SWAN improves when it is forced by GRASP friction velocities instead of ERA5 wind fields. In addition, the established two-way coupled simulation is proven to be an enhancement for the spectral wave model in comparison to the one-way coupled configuration. The performance of the atmospheric large-eddy simulation could also benefit from the two-way coupled configuration. However, it is sensitive to the implemented roughness length parameterization.

The accurate approximation of the surface tension force is paramount for continuum surface models in the field of computational fluid dynamics for multiphase flow where surface tension is relevant. This involves being able to accurately calculate the curvature at the interface. This study focuses on the use of convolution in smoothing the VOF colour field in order to obtain better approximations of the curvature. Given the sudden jump in values of the VOF colour field, the calculation of its derivative for the curvature is sensitive to errors, given the large values of high order terms that determine the truncation error. To deal with this problem, convolution of this abruptly varying field can be used to create a smoother transition. The curvature approximation of a circular interface improved as the support of the convolution was increased.
It was proven analytically that, for these interfaces, the original curvature is retrieved from the convoluted field. Interfaces along which the curvature varies were also considered, and it was found that there is a critical convolution support that minimizes the error in the curvature, given that the choice of the support length can modify the curvature that is estimated.
An algorithm was implemented in OpenFOAM that calculates the convolution of the VOF colour field. The resulting smoothed field was then used to calculate the curvature, which is needed for the surface tension force of the system. The simulations of a two-dimensional rising bubble resulted in more accurate results for the circularity and the rising velocity, when compared to the original OpenFOAM implementation with no smoothing. With the convolution algorithm, the terminal velocity deviated only 0.01% from a well-accepted benchmark case, a great improvement when compared to the 4.2% difference when no smoothing was used. However, simulations of a static bubble in zero-gravity rapidly resulted in unphysical flow, manifested as a wavy interface, when a convolution support larger than 2 cells was chosen. An improvement of the estimation of the surface tension force direction may be needed for this behaviour to disappear.

With the recent developments in tissue-engineered organ substitutes, there has been an increasing demand for novel culturing techniques to create biological substitutes in vitro, as well as reliable and efficient test methods that can assess their biocompatibility and mechanical performance. In the past two decades, perfusion bioreactors have proved to be important tools in the creation, control, and evaluation of cell cultures and engineered tissues under precisely controlled in vitro conditions that simulate the physiological environment. Due to the demanding requirements for blood vessel substitutes, they have been extensively deployed in the culture and conditioning experiments of novel tissue-engineered vascular grafts (TEVGs) and have been often used to study haemodynamic mechanic stimuli, including the cyclic vessel expansion and wall shear stress. However, a key limitation of the currently available TEVG perfusion bioreactors is their often complex design and operation and the lack of standardisation of the mechanical characterisation.
In this study, the goal was to design, characterise, and construct a TEVG perfusion bioreactor that is foremost simpler to operate and extends the capabilities of particular currently available set-ups. A modular bioreactor design is presented that allows for the toolless mounting of the graft with a diameter up to 6 mm and features completely separated circuits for the intra- and extraluminal side of the scaffold wall. The culture chamber encloses a removable graft frame, mainly machined out of polyether ether ketone (PEEK). In silico simulations have been used to assess local fluid dynamics within the scaffold, to predict the reliability of the simulated culture conditions. The presented design allows the exertion of a broad range of physiologically relevant shear stresses on cultured TEVGs, including arterial shear stresses. A test for the in vitro cytotoxicity showed that the medium extracts of the constructed graft frame did not show any cytotoxic potential. Finally, an assessment method for the estimation of the radial compliance is proposed, which aims to adhere to the relevant international standards on the mechanical characterisation of vascular substitutes.

Metallic droplet deposition is of interest in the industry because of the potential use in additive manufacturing. This work discusses the complex phenomena involved in droplet impingement, especially the effect of temperature dependant surface tension. The volume of fluid (VOF) method is used to solve the axis-symmetric Navier-Stokes equations, which are used to describe the droplet's behaviour. Different temperature dependencies for the surface tension are modelled, to see the effect on the interface and substrate melting. Furthermore, the effect of droplet size, initial temperature on droplet shape and substrate melting is studied. To judge the accuracy of the VOF model, a series of benchmarks are tried. The VOF model used in this work is as accurate or more accurate than previous works. The results show that droplets with a higher percentage of oxygen flatten, this is due to thermo-capillary forces. A higher temperature results in more spreading of the droplet, this is because higher temperatures result in higher surface tensions. These surface tensions keep the droplet together when it first makes contact with the substrate. This causes less air is trapped underneath the droplet, which causes the droplet to spread out more. The air also causes the droplet to cool down less fast, this results in a phenomenon where the hotter droplet is solidified faster than the colder one.