Liquid hydrogen propulsion presents itself as one of the most promising technologies for a zero emission commercial aviation future. However, two of the main challenges are the low volumetric energy density and the cryogenic storage temperature of the fuel. In collaboration with Airbus and the ZEROe project, this work develops a model to predict the response of the liquid hydrogen tank when the thermal insulation is lost and very high amount of heat are transferred to the tank and its pressure is increased. The model includes the most important physics driving the response, and enables the study of the system and its improvement, maximizing the amount of fuel carried while being able to safely control the pressure rise.

This research investigates the adjoint-based optimization of bare tube heat exchangers, aiming to minimize air-side pressure drop. The study focuses on optimizing tube shapes in an in-line bare tube heat exchanger configuration, selected for its inherent advantages in reducing pressure drop. The optimization targets flow conditions characterized by a Reynolds number of approximately 1000, based on the tube diameter. The accuracy of simulations was found to depend significantly on the turbulence modelling approach. Different models were benchmarked against empirical correlations to ensure accuracy, with the most accurate model used to recompute the flow fields of the optimized designs as a post-processing step. The optimized round tubes achieved a pressure drop reduction of up to 20%. For elliptical tubes, the optimization resulted in a pressure drop reduction of up to 10%. This 10% reduction adds to the notable improvement gained by replacing round tubes with elliptical ones.

Analysing and developing turbomachinery at off-design conditions requires the usage of robust and efficient Computational Fluid Dynamics (CFD) solvers. With the introduction of the computer, many advancements have been made in the field of numerical methods involving these flow problems. These numerical methods are used in order to solve these complex flow structures that are formed due to the interaction between the rotating machinery and the fluid. With the emergence of new design paradigms using computer-aided optimisation, novel solver methods are required which are able to deal with the increase in computational cost and the convergence difficulties following off-design conditions. One of the CFD solvers that is used for turbomachinery design is the open-source software SU2. SU2 is currently developed partially at the Delft University of Technology, where an increase in solver performance with respect to turbomachinery aerodynamics is greatly desired. The current work provides a numerical study of SU2's current performance with respect to turbomachinery analysis, as this is currently unknown.

The current performance of SU2 is to be analysed using steady RANS turbomachinery simulations. The research conducted by Xu et al. \cite{Xu2020} will be used as reference data in order to validate SU2's performance together with data obtained from the CFD solvers CFX and Numeca. The research conducted by Xu et al. resulted in the development of their NUTSCFD solver, which showed strong performance with respect to turbomachinery analysis. Four test cases are set up and used in order to analyse SU2. The test cases that are considered for the numerical study include: the NACA 0012 airfoil, the LS89 turbine cascade, the MTU centrifugal compressor and the 1.5 stage ETH turbine. The first three test cases are validated using the results obtained by the NUTSCFD solver, where the ETH turbine is validated using CFX and Numeca. Following these test cases, SU2's performance is analysed using the residual behaviour. Xu et al. dedicate their performance increase to be the result of the Newton-Krylov method that is implemented in their NUTSCFD solver. SU2 also includes a Newton-Krylov solver, but its performance with respect to turbomachinery is also unknown. A performance assessment with respect to SU2's Newton-Krylov method involving turbomachinery analysis is therefore conducted as well.

The results obtained using the NACA 0012 and LS89 test cases show a discrepancy in solver performance involving SU2. With respect to SU2's Newton-Krylov solver, the NACA 0012 test case shows a reduction in non-linear iterations where this is not found for the LS89 test case. Both results showed however a large difference in performance when compared to the NUTSCFD solver, where SU2's standard solver was also unable to match NUTSCFD. The results obtained following these test cases have led to the development of the MTU test case, where the MTU test case was to be used in order to provide a more accurate comparison between SU2 and NUTSCFD. Instead, SU2 was unable to run the MTU test case, where the solver showed stalling behaviour...

The Flying-V is a novel aircraft concept that has improved fuel efficiency through its distinctive V-shaped structure containing the pressurised cabin, fuel tanks and cargo hold. This configuration presents several key design challenges, one of them being the design of the landing gear. Previous studies indicate that the Flying-V landing gear significantly contributes to the total mass of the aircraft and its placement influences overall aircraft design, both of which impact the viability of the concept. Landing gear designs have been proposed, but are based on outdated requirements, and have notable shortcomings when it comes to feasibility, reliability of mass estimates, aircraft integration and family design commonality. This research aims to explore landing gear designs to identify the best solution for the Flying-V family.

A novel conceptual landing gear design framework is developed, incorporating a gear positioning and sizing optimisation algorithm and physics-based structural sizing. The gear positioning and sizing algorithm allows for efficient exploration of various design features, and can quickly adapt to changing aircraft requirements, making it particularly suitable for the conceptual design phase and the development of family design derivatives. The algorithm is validated using five different reference aircraft with distinct specifications and landing gear configurations, and demonstrated its ability to accurately predict gear positions and lengths. The use of physics-based structural sizing reduces the reliance on statistics-based mass estimation methods, which are known to yield inaccurate results for large and unconventional aircraft. Reference aircraft gear mass data is used to derive a finite-element mass correction equation, improving the accuracy of the mass estimation.

Several Flying-V landing gear concepts are evaluated, each designed with different features, or optimised for a different set of aircraft requirements representing any of the Flying-V family members. For the FV-1000, this resulted in a final design that has a double folding strut, a four-wheel articulated bogie and a strut shortening mechanism, offering a feasible solution with optimal mass and stowage properties. The gear weighs 13.6 tonnes, which is 5.1% of the aircraft maximum take-off mass (MTOM). A derivative design for the FV-900 features the same structural components, but without an articulation mechanism and includes dedicated rolling stock sizing, resulting in a gear weight of 13.3 tonnes, or 5.7% of MTOM. The FV-800 landing gear design follows the same concept as for the FV-1000, but with dedicated structural and rolling stock sizing, significantly reducing mass compared to using the FV-1000-sized common gear. The FV-800 is more constrained by fuel tank capacity and range, which justifies deviating from the commonality principle. The dedicated FV-800 gear weighs 9.2 tonnes which is 5.0% of the MTOM.

This document contains the final Master Thesis Report to obtain the Master of Science on Aerospace Engineering (Flight Performance & Propulsion - Propulsion & Power) at the Delft University of Technology.

The research work focuses on characterizing the interaction of a radial inlet turbine with a downstream diffuser through the size of the rotor tip gaps. In the implementation of power turbines it is common to add a diffuser downstream of it to lower the rotor exhaust pressure and thus increase power extraction for the same inlet conditions. However, diffusersare bulky and take a lot of space in the assembly. This lowers the power density of the machine, increases installation weight in transport applications, and installation costs in ground based operations. In the quest to obtain compact diffusers, researchers noticed that the non-dimensional static pressure recovery (Cp) of this device was higher when operating downstream of a turbine than with uniform inlet conditions. The publications in this field are relatively scarce, and thus there is a knowledge gap with immediate practical application.

Some researchers have focused on the interaction of turbine vortical structures with the boundary layer of the diffuser. They have found out that under this conditions the boundary layer can support steeper pressure gradients without detaching. This is only applicable in very steep diffusers that would stall in isolation. Notably, all the publications in this field deal with axial machines, and as the text will show the picture changes considerably when applying the theory to radial turbines.

This work studies another side of the problem. The text will focus on stable diffusers, and thus there is no boundary layer that needs reinforcement. Turbine rotor tip gaps generate an increase of entropy and reduce turbine power generation. However, these gaps also cast powerful vortexes that affect the diffuser flowfield. This project will study the effect of the tip gap configuration on the pressure recovery of the diffuser, in order to better understand the trade-offs and support the design process. Tip gap sizes are
usually determined from mechanical constraints. This work provides insight into the real cost of a tip gap by analyzing the assembly turbine + diffuser, and thus it guides the designer into where to focus his efforts when working with these mechanical constraints.

The general conclusion of the project is that there is a range of tip gap sizes where the performance of the diffuser in enhanced. In this range, the performance cost in the rotor for increasing the tip gaps is partially compensated by a better pressure recovery in the diffuser. Furthermore, it has been discovered that in a radial turbine it is more influential to close the leading edge gap (axial gap), even when this implies opening the trailing edge (radial) one. This information will guide engineers when choosing the bearings for the machine and dealing with design trade-offs.

As a byproduct of this research, a novel study in radial turbine flow structures is carried out. Not a lot is known about the vortical structures in these machines, and this work provides a first qualitative description about origin and behaviour of this vortexes. Furthermore, different flow modes are identified and correlated to a simple geometrical parameter that is readily available in 0D designs. The combination of this information with previous tip gap flow models, and diffuser-vortex interaction models, will allow the
obtention of improved losses models including the effect of the diffuser from the conceptual design stage. This data could also feed into noise models for radial turbines, a novel field of interest. This work not only provides useful design guidelines, but also sets down the basis for future research leading to a better
understanding and modeling of radial inflow turbines.

The need for more efficient aircraft has led to the conception of innovative engine configurations, such as the combined-cycle engine proposed by Delft University of Technology or the Water-enhanced turbofan developed by MTU and Pratt & Whitney. The evaluation of the potential benefits of these novel engine concepts requires detailed performance studies considering weight and drag penalties as figures of merit. At present, no weight estimation tools with sufficient level of accuracy and flexibility are publicly available, apart from WATE++, a tool developed by NASA. However, this software is subject to export control restrictions and cannot be used outside the USA. For this reason, the development of a new component-based preliminary engine design tool, "Weight EStimation of gas Turbine engines" (WEST), was started. The goal of WEST is to predict the weight of novel engine architectures with a reasonable level of accuracy, accounting for design parameters like turbine inlet temperature, overall pressure ratio, mass flow rate, and turbomachinery configuration, with a minimal set of geometry inputs specified by the user. A previous work demonstrated that WEST can effectively predict the weight of turbofan gas generators. As only the main components are modeled, the WEST estimates account for approximately 70-90% of the actual engine weight.

The capabilities of WEST were expanded in this study to allow for the design of small-scale turboshaft engines. To this purpose, a methodology to design radial compressor disks was implemented, and successfully verified against FEM results. Regarding the modeling of the complete turboshaft, it was found that the predictions of the tool account for only 60-70% of the actual engine weight. Such result was, however, expected, as WEST does not take into account the particle separator and the integrated gearbox, which may account for up to $30\%$ of the engine's total weight.

The most effective way to decrease the cost of energy of wind turbines is to increase the rotor diameter. Therefore, this calls for the need of aerodynamic research on large wind turbines. This thesis project will tackle the topics addressed by the IEA Wind Task 47 which is focused on the aerodynamic analysis and comparison of the newly proposed IEA 15-MW reference wind turbine using varying fidelity tools. Therefore, the main goal of this work is to analyze whether a low-fidelity approach is still feasible for aerodynamic analysis in comparison with a high-fidelity approach. It was found that there are quantitative differences between the different fidelity methods but also qualitative similarities. It can be concluded that the low-fidelity methods are relatively less suitable for a large turbine design, especially for high TSR values, which should be confirmed by further research including different fidelity tools and turbines.

Organic Rankine Cycle systems are gaining increasing attention as a modular, cost-effective and decentralised thermal energy recovery solution. The design of an efficient expander in the ORC system is crucial to its rapid uptake by the industry. Compared to volumetric expanders, a radial turbine provides advantages in terms of flexibility and better part-load performances. The design of ORC radial turbines is faced with two challenges. Firstly, the expansion in the stator occurs close to the vapour-liquid dome and the critical point, where the ideal gas assumption is no longer valid. Such expansions fall under the Non-ideal thermodynamic regime. Secondly, due to the lower speed of sound in the working fluid, the expansions are supersonic and are accompanied by compressible flow features such as shock waves, expansion fans which generate entropy and reduce the performance of the expander. There exists significant validated design methodologies and empirical relations for conventional turbomachinery. However this is not the case for unconventional turbomachinery where established loss correlations are not applicable. Hence, there is a knowledge gap in validated design tools for unconventional turbomachinery like ORC radial turbines. The thesis forms a part of the validation campaign of the open-source flow solver, SU2 and approaches the exercise by studying expansions in two paradigmatic test cases, namely (i) a linear stator cascade and (ii) a converging - diverging nozzle.

The study of expansions through a linear cascade is a preceding step to the study of a rotating radial turbine. Consequently, a RANS simulation on the single channel blade passage is studied under the validated assumption of flow periodicity. The expansion takes place from inlet total conditions of 18.4 bara and 525 K (Z =0.558) to a static pressure of 1.95 bara (Z = 0.951) and exit flow of Mach 2. Two types of response quantities are studied namely direct and system response quantities. Direct response quantities are those that can be measured directly such as the pressure, Mach, density. System response quantities are derived from direct measurements and provide information to characterise the performance of the stator. The selected system response quantities are: (i) Pressure loss coefficient (C_p= 0.074), (ii) Kinetic energy loss coefficient (ζ_KE= 0.115), (iii) Entropy loss coefficient (ζs=0.118), (iv) Base pressure loss coefficient (C_pb= -0.065), (v) Standard deviation of exit flow angle (σ_β2 = 1.244) and (vi) Standard deviation of exit Mach (σ_M2= 0.033). Experimentally, it is possible to measure the pressure loss coefficient, base pressure loss coefficient and the flow uniformity at the outlet using a combination of pressure and direct velocity measurements. Through a Design of Experiments approach, the sensitivity of the flow to input uncertainties was studied through a stochastic collocation based forward propagation of the uncertainty. The input uncertainties considered are the inlet total pressure, fluid viscosity and critical point properties. The Sobol indices from the uncertainty quantification indicate the more dominant influence of the critical point properties over other inputs considered. The results also validate the use of a constant viscosity assumption for the RANS simulation. The subsequent planned unit test case was to characterise the expansion through an optimised stator blade row. To this end, a deterministic adjoint based optimisation was performed with the objective function of minimising entropy generation. The resulting geometry was studied and resulted in a pressure loss coefficient improvement of around 4%, although stator flow uniformity at the exit was compromised. The uncertainty quantification performed on the optimised blade geometry yielded robustness improvements on the pressure loss coefficient and reduction in uncertainties associated with direct response quantities, although no strong correlations can be drawn between the improvements in uncertainty and the deterministic optimisation. Given current machining tolerances, the realisation of such a negligible geometry change is not viable from an experimental
perspective.
The second section of the thesis deals with experimental investigation of expansions in a converging-diverging nozzle through a matrix of isobars with increasing degree on non-ideality. Two isentropes with inlet pressure of 2.73 bara (Z = 0.9526) and 6 bara (Z = 0.901) with pressure ratio of 8.76 were performed with recording of pressure, temperature, flowrate and density measurements along the ORCHID. The flow field was visualised using the schlieren method using a z-type layout. The thesis reports the first measurements of total pressure before the nozzle inlet and vapour density and flowrate measurements. The experimental data was post-processed to identify steady-state and quantify the Type A and Type B uncertainties. The schlieren images were used to extract information on the Mach distribution along the nozzle mid-plane using an inhouse line extraction tool. Lastly, a 2.5^o wedge at the exit of the nozzle is used to generate oblique shock waves that are then manually measured. The flow field at the exit of the nozzle is in the ideal region where the shock angle only depends on upstream Mach number. Hence, the experimentally observed oblique shock angles for the off-design case are close to on-design (Inlet pressure = 18.4 bara) predictions.

Adverse pressure gradients, separation and other forms of non-equilibrium flows are often encountered in flows of interest. In these type of flows, the Boussinesq hypothesis does not hold and often leads to erroneous predictions by eddy viscosity models. In an attempt to capture these non-equilibrium effects, lag parameter models introduce a lag parameter, which is derived from an elliptic blending Reynolds stress model. A novel deterministic machine learning algorithm, referred to as Sparse Regression of Turbulent Stress Anisotropy (SpaRTA), has been used with the objective of developing a data-driven turbulence model based on the elliptic blending k − ω lag parameter model and evaluating its performance in terms of generalizability, interpretability and its ability to infer the quantities of interest. Corrective terms are introduced and computed directly from high-fidelity data to account for the model-form error by using the k-corrective-frozen-RANS approach. SpaRTA is then used to infer algebraic stress models for these corrective terms using a Galilean invariant integrity basis. It was shown that the k-corrective-frozen-RANS framework has the ability of representing the mean flow features by propagating the corrective terms through a CFD model in OpenFOAM and comparing its result to high-fidelity data. Cross-validation is used to test the performance of the models on unseen data using three flow cases that involve separation, namely periodic hills (Re=10595), converging-diverging channel (Re=12600) and curved backward-facing step (Re=13700). In order to assess the impact of the additional transport equation of the lag parameter, the same data-driven approach was applied to the conventional two-equation k − ω model. Utilizing an additional transport equation for the lag parameter in this data-driven approach did not result in any significant improvements in terms of predictive capability or generalizability, as both data-driven approaches showed a similar performance, although the data-driven k − ω models were more numerically stable. It was found that corrective terms formulated using a reduced integrity basis yields data-driven models that have a similar predictive capability compared to models that used the full integrity basis to construct the corrective terms. A significant portion of the resulting data-driven models showed an improvement in predictive capability over the standard (non-data-driven) k − ω model. Furthermore, most of the models were able to generalize their predictions to two-dimensional flow cases that had different complexity and showed a significant improvement over the baseline k − ω model.

Increased public awareness and political concern about the environmental footprint of aviation have created very ambitious goals to reduce emissions in the future. This objective of this research project is to develop a multidisciplinary turbofan engine design and sizing tool for conceptual engine design. The tool will be used in a parametric analysis to determine how the main design parameters impact engine performance and sizing. Methods to improve the propulsive and thermal efficiency have been analyzed to assess the effect on engine characteristics. The bypass ratio, fan pressure ratio, overall pressure ratio and turbine inlet temperature have been analyzed in a parametric analysis. It was found that higher values of these design parameters improved engine performance at the cost of engine size and weight.

Flow unsteadiness caused by impeller rotation, vortex-shedding, secondary flows etc. can lead to the generation of acoustic waves within the turbomachinery cascade. This causes pressure loading on the impeller. When acoustic resonance occurs, i.e. the frequency of acoustic wave excitation matches with the structural natural frequencies of the impeller, high fatigue and vibrations are encountered, which can lead to structural failure. Centrifugal compressor applications like turbocharging and process engineering require an advanced understanding of the aeroacoustic excitation mechanisms as these have been suspected of playing a significant role in structural failures. Given that the current state-of-the-art analysis methods are incapable of explaining various instances of structural failures, a novel Lattice Boltzmann Method (LBM) based approach is explored. For the first time, aerodynamic and performance predictions, along with aeroacoustic amplitudes from the LBM based approach will be compared with a conventional Unsteady Reynolds Averaged Navier Stokes (URANS) based approach as well as test rig data. This will assess the feasibility of the LBM model for analysing forced response behaviour of a centrifugal compressor operating at conditions of acoustic resonance. The research compressor has been chosen based on an aeromechanic test campaign where high impeller blade trailing edge vibrations were measured. The computational domain consists of full compressor wheel including the hub and the shroud cavities. An attempt will be made to quantify the resonant amplification factor by simulating off-resonant conditions. The findings from this research would result in the development of a numerical framework for assessing the physics and severity of resonant excitations in centrifugal compressors. It will also highlight the importance of accounting for aeroacoustic mechanisms in the aeromechanical design of centrifugal compressor stages.

Laminar-turbulent transition (LTT) is the process through which smooth laminar flow transits into chaotic turbulent flow. Investigation of the paths taken to transit into turbulence is a front-runner among other methods followed to characterise turbulent flows. This is of particular importance in aerospace and energy industries for the design of wings and gas turbines. Early research used Linear Stability Theory (LST) to analyse the stability of the flow; with the increase in computational power, Direct Numerical Simulation (DNS) has been developed to solve the flow field entirely. Most of the research on LTT has been centered on ideal fluids with limited focus on the effects of high temperature. The impact of other strong non-ideal effects on LTT such as dense gas effects have not been investigated. This work aims to study the effects of dense gas on LTT for boundary layer flows of toluene over a flat plate.Flows over a flat plate boundary layer are investigated in 3 stages. First, the base flow is solved for ideal air and non-ideal toluene for 6 different Eckert numbers (Ec). Second, the base flow is provided as an input to solve the eigenvalues of the stability equations for both fluids and each Ec using an in-house MATLAB code. The unstable eigenmode is identified and tracked. The growth rates and phase velocities are calculated and compared between ideal air and toluene. Third, DNS simulations are performed using a FORTRAN code, to solve the governing equations of compressible flows for different Ec. The simulations are performed on a pre-processed base flow solution, subjected to 2D sinusoidal perturbations forced into the computation domain at the wall. A no-slip and adiabatic wall boundary conditions are applied to the flat plate with a sponge region at the outlet and the top of the computation domain. The growth rates and phase velocities of these perturbations are calculated and validated with the predictions made by LST. Finally, a perturbation energy budget analysis is conducted to study the nature of the unique results obtained.The LST results show that as Ec number increases, all flows over a flat plate become more stable. For toluene flows, the stabilising effect of increasing Ec is more pronounced and all flows with Ec > 0.15 are stable and have no modal instabilities. The results from the DNS simulations validate these predictions from LST and match perfectly in growing conditions, but deviate from one another in stable conditions. The deviation in results are hypothesised to be the contribution of multiple decaying modes to stable behaviour of the flow. Furthermore, perturbation energy budget analysis showed that for Ec = 0.05 and 0.10, the spatial growth of perturbations are positive due to a positive production term and negative for Ec = 0.15, with a negative production term. The negative production term is attributed to the negative integral of the perturbation profile.

Large-scale liquid rocket engines require high-speed turbopumps to inject cryogenic propellants into the combustion chamber. Modeling the impact of secondary path and leakage flows on the performance and axial thrust of the pump is critical during the early design phase. Previous studies derived simplified models and empirical correlations to model leakage loss and determine the pressure distribution in the side gaps. The prediction capabilities of these models are subject to limitations and uncertainties mainly due to the interaction between impeller exit flow and sidewall gaps especially at partload operation. The effects of the leakage injection on the impeller flow are not fully understood and not captured by analytical correlations from literature. These can have destabilizing and detrimental effects on the hydraulic performance and the head curve, leading to a divergence betweenmodel predictions and experimental results. In this project a reduced order model is developed to capture the effects of the sidewall gaps on turbopump performance and determine the axial thrust on the rotor. High fidelity calculations of the pump and volute are performed to identify the key physical mechanisms associated with loss generation in leakage paths and improve upon existing analytical models found in literature. Axial thrust and leakage losses are modeled numerically with a single passage geometry to capture the interaction of impeller flow and leakage flow through the side gaps at multiple operating points. It is found that the recirculation of pressurized fluid, leakage mass flow, is the most impactful effect on the performance. The volumetric efficiency is captured within 16% with the analytical model and 6% with the single passage numerical calculations at nominal operation. At partload operation however, the accuracy of both models reduces. Disk friction is under predicted by the investigated analytical model and axial thrust is over predicted. For both effects the reduced numerical model shows promising results that improve the prediction capability around nominal operation and offer a higher flexibility required for the early design phase as compared with the high fidelity, full annulus calculations. For the investigated cases the leakage injection to the impeller flow does not show a destabilizing impact on the characteristic behavior.

The project concerns uncertainty reduction of parameters of a thermodynamic equation of state for a dense gas, using Bayesian inference. The dense gas considered is D6 siloxane and the equation of state used is the polytropic van der Waals equation. The shock tube data comes from the flexible asymmetric shock tube (FAST) experiment. This is modeled using the quasi-one-dimensional Euler equations with a source term that depends on time. A surrogate model based on sparse grids and a sensitivity analysis using Sobol' indices are both applied. The Markov chain Monte Carlo technique is applied to sample from the posterior probability distribution on the chosen parameters of the computer model. The results indicated that some of the thermodynamic parameters were identified, but that their mean values showed a disagreement with the true values in the literature.

Turbopumps are used in large-scale liquid rockets to inject the cryogenic propellants into the combustor. The turbopumps operate at high rotational speed, which in turn leads to cavitation at the inlet of the impeller. Cavitation can cause performance issues, instabilities and mechanical damage of the blades.
This thesis presents and validates a shape optimization framework for the design of splitter blades that extends the operative range under cavitation while maintaining the wetted performance. For a target turbopump application, the optimization framework allows for independent changes to the blade angle distributions across the span and to the pitchwise position of the splitter blades while preserving the thickness distributions.
The optimization is conducted with a surrogate-based gradient method. The geometry is optimized at a fixed cavitation number corresponding to a 5% head coefficient drop-off, while constraints are imposed on the wet pump performance. It is found that this approach, coupled with the optimal design points distribution provided by the Design of Experiment method, reduces the computational cost of the optimization process by minimizing the number of multiphase calculations.
The numerical results show that the optimized splitter blades successfully increase the pump operative range by 2.2% and increase the head coefficient by 5.3% compared to the baseline case with non-optimized splitters. These results are corroborated by experiments conducted in a closed-loop water test facility. Several pump geometries are tested through rapid prototyping using additive manufacturing. The experimental data validate the optimization framework, demonstrating a 4.7% increase of pump operative range and a 7.6% increase in head coefficient. The calculations are used to gain insight in the physical mechanisms for the performance improvement. The analysis of the results indicates that the improved performance is due to the optimized position and shape of the splitter blades which increase the pump slip factor.

Curved exhaust ducts are used in aero engine applications for different purposes, including thrust vectoring, shielding of parts from the exhaust or improving stealth properties. Their integration, however, regularly represents a design problem due to flow separation and high aerodynamic losses occurring in the bend. Curved duct flows for both compressible and incompressible conditions have been studied extensively in the past. However, no experimental results of a high Reynolds number flow through a turbine exit guide vane (EGV) followed by a curved duct have yet been published. An in-depth CFD analysis of the aerodynamic effects is therefore carried out, using the RANS solvers TRACE (at MTU Aero Engines) and SU2, to analyze the flow field and geometric sensitivities of a high Reynolds number flow through an EGV followed by a 90 degree bent duct. The geometry of interest is investigated at a Reynolds number of Re=10^6 and a ratio of bend radius to duct diameter of one. The inflow conditions are prescribed to closely resemble typical flow conditions at the low pressure turbine exit plane of a turbojet engine. After validating the solver with experimental data using the test case of a 90 degree bent duct, an initial CFD analysis of the combined exit guide vane geometry with curved duct is carried out to identify the dominant flow phenomena and mutual effects of the EGV and the bend. Three zones of flow separations are found, each at the convex and concave sides of the bend and one at the lower side of the plug. Secondary flows caused by the bend are found to have an effect on the flow upstream of the EGV, leading to a non-uniform flow inlet angle and aerodynamic loading of the individual blades. Separation downstream of the EGV is influenced by the presence of low velocity wakes from the EGV. Subsequently, a sensitivity study is carried out to find the effect of different geometrical parameters on the flow field. Main investigated parameters are the ratio of bend radius to diameter (R_c/D) and the distance between EGV and bend (l/D). Additionally, the aerodynamic effects of the circumferential EGV positioning, swirl and the plug shape are investigated. It is found that an increase in both bend radius and distance between bend and EGV improves aerodynamic efficiency, while swirl can decrease pressure losses in the duct for small bend radii of R_c/D<0.8. Improving the plug shape and rotating the EGV allows to further increase the aerodynamic efficiency without weight increase. For further optimization of the geometry, it is recommended to include duct geometries with a non-constant bend radius and outer duct diameter to increase the design space.

Two-phase thermal control systems are used for payloads that require temperature stability under high and fluctuating heat output. However, the complex interaction between the two phases in a flow makes it hard to predict the performance of such a system. The goal was to develop a numerical model able to predict the transient behaviour of such a system within a 30% error margin. An existing explicit one-dimensional finite-difference simulation was expanded using new correlations from literature. The performance of this model was validated using experimental measurements gathered with the NLR two-phase demonstrator setup. The resulting numerical simulation can predict the transient behaviour of the pressure drop and liquid level in the accumulator with a mean error of 7.7% and 1.3% respectively. The goal therefore has been reached.While room remains for further improvement and validation, this model can significantly shorten the development time of new two-phase systems.