To reduce humanity’s dependency on fossil fuels, hydrogen is increasingly investigated as a promising energy carrier. When converting the chemical energy of hydrogen to the desired form, hydrogen reacts with oxygen to form water. This makes hydrogen very promising for achieving net-zero carbon emissions. For the aviation sector, hydrogen combustion has been suggested as a possible solution.
As with any technology, challenges remain. Whilst combustion modelling using Computational Fluid Dynamics (CFD) is not a solved problem to begin with, hydrogen combustion is made more complex by differential diffusion. This occurs when different species do not diffuse at the same rate. Hydrogen tends to diffuse faster than other species, causing local changes in mixture fraction, in addition to super-adiabatic temperatures and super-equilibrium product concentrations. This effect is stronger in strained or curved flames. Flame strain and curvature also influences the local reaction rate. These phenomena occur on a very small length and time scale. To simulate this directly, very fine meshes and very precise transport equations for each species are required. This can be prohibitively expensive for large scale designs or for design iteration.
In cases where simulating this behaviour directly is not possible, an alternative model is required. In this paper, five variants of such a model are assessed. These models are based on the assumption that the flame can be approximated as an ensemble of 1-dimensional flamelets. These flamelets can be analysed before the CFD simulations are performed. The results are then parametrized by a number of control variables, using a presumed filtered probability density function which aims to include sub-grid scale effects. By changing the flamelet conditions and control variable definition, different manifolds can be generated. In this paper, a number of these Flamelet-Generated Manifolds (FGMs) are compared to a higher fidelity model (using an Eulerian Stochastic Fields (ESF) approach). These FGMs use the mixture fraction, the progress variable and their sub-grid variances as control variables. Five different FGMs are tested, characterized by the progress variable definition (water vs hydrogen mass fraction) and flamelet strain rate (0 (unstrained) vs 3000 vs 6000 vs 13000 s^-1.
The performance of these models is assessed by means of a Large Eddy Simulation (LES) of a combustor with a bluff-body, using the open source CFD software OpenFOAM. The bluff-body causes a recirculation zone in its wake. Additionally, it causes a strain on the flame front, where differential diffusion is expected to cause super-adiabatic temperatures and super-equilibrium mixture fractions.
The results show that the FGMs give a good prediction of the conditionally averaged reaction rates. They are also capable of qualitatively predicting the increase in mixture fraction, super-adiabatic temperature and super-equilibrium product mass fractions. They can therefore be used in a strained combustor setting with lean, premixed hydrogen. However, there are challenges regarding the prediction of temperature, especially for FGMs using a hydrogen based progress variable. This should be investigated in the future. Furthermore, the FGMs predict the reaction rate well despite over-predicting the local effects of strain and differential diffusion.

A growing concern over the irreversible effects of global warming has led to the creation of international carbon neutrality goals by 2050. The mounting demand placed on the aviation sector creates growing pressure on new and innovative solutions to offset the growth in predicted aircraft emissions. Among many, hydrogen combustion offers a uniquely auspicious yet challenging approach to carbon neutrality in aviation. This project aims to study the effect of low temperature hydrogen and air injection on fuel mixing quality, the emission of secondary combustion pollutants, and the reaction zone interactions with the flow field. Large Eddy Simulations using detailed combustion models and a skeletal chemical kinetic mechanism are implemented to simulate the combustion of a hydrogen enriched jet A-1 spray flame. Using a scale model of the Auxiliary Propulsion and Power Unit (APPU) combustor, two similar simulations are set up, the first with an air/hydrogen inlet static temperature of 250\,K and the second with an air/hydrogen inlet static temperature of 400\,K. The final results are post processed and analysed. The results show a clear increase in mixing quality between hydrogen and jet A-1 when the temperature is decreased for this combustor geometry, strongly corroborating the initial hypothesis. Additional results show that the combined effect of lower injection velocity and lower temperature injection of air/hydrogen leads to an increase in the stretch rate applied to the reaction zones closest to the injector and a decrease in flame structure length. Analysis of species distributions reveals significant spatial variations across the combustor. Still, only a significant and conclusive reduction in CO emissions with a decrease in inlet temperature could be established. The limited computational resources associated with this research consequently limit the quantification of the observed phenomena, however, the analysed trends provide sufficient evidence to support the use of lower temperature hydrogen as a potential solution to the issues associated with the mixing of both fuels and towards lower emissions combustion.

This thesis addresses thermoacoustic instabilities and flashback in hydrogen combustion. A simplified model of Ansaldo Energia's GT36 reheat combustor was simulated at 20 bar using Large Eddy Simulation (LES), revealing unsteady flame dynamics driven by strong pressure oscillations. To detect flashback precursors, LES-derived time series were sampled at the combustor wall after an analysis identified suitable monitoring locations, representing a step toward practical sensor placement.

Fourteen thermodynamic, velocity, and species mass fraction signals were reduced via autoencoders with 2–4 latent variables; the three-latent representation emerged as optimal, isolating transition sharpness and mid-frequency modes. Clustering of this space with a modularity-based algorithm consistently identified precursors with maximum lead times of ~42 μs and virtually no false positives. In one case, a flashback was successfully predicted and suppressed. Robustness analyses confirmed generalization across locations and noise levels, demonstrating that wall-based latent clustering advances predictive flashback control toward real-world deployment.

This thesis explores the magnetic stabilisation of strained hydrogen flames, focusing on mitigating thermodiffusive instabilities in lean premixed laminar hydrogen flames while preserving their inherent NOx reduction benefits. Utilising comprehensive numerical simulations of counterflow flame configurations, the study reveals that radially decreasing magnetic field gradients modify flame structure, predominantly through indirect mech-
anisms. These mechanisms involve magnetically induced alterations to the bulk flow field, which subsequently couple with differential diffusion effects to influence the distribution of reactants, affecting mixture fraction and the temperature field. Key findings demonstrate that the representative thermal thickness of the flame remains effectively unchanged, with a kernel-density (KDE) mode shift of 0.16 %. Crucially, total kinematic stretch is reduced by approximately 5-13 % near the axis, driven primarily by the curvature-induced contribution. Tangential strain exhibits a distinct two-regime response, decreasing by about 5-8 % in the preheat zone but increasing by 6-8 % in the reaction zone. Furthermore, a two-regime velocity-field redistribution is observed, where the axial component broadly weakens, while the radial velocity decreases for low progress variable (C) near the axis and strengthens for C ≳ 0.5. This research clarifies that direct magnetic effects are intrinsically weak and negligible
for practical control, with effective manipulation relying on these indirect pathways. The study also highlights that the efficacy of magnetic control diminishes with increasing strain rates and richer equivalence ratios. This work provides novel insights into the fundamental mechanisms governing magnetic control in premixed hydrogen flames, offering a framework for advancing sustainable combustion technologies.

Hydrogen technologies show promise in reducing the effects of aviation on anthropogenic climate forcing. This report aims to develop the design of a robust, low-maintenance, low NOx emission hybrid hydrogen-electric powertrain for retrofitting the Beechcraft 1900D by 2035. An optimal sizing of the two powerplants: high-temperature proton-exchange membrane fuel cells and a gas turbine with a rich-burn, quick-mix, lean-burn combustor, was performed, minimizing both the cost and climate performance of the design. In this optimization, the storage system, electrical system and thermal management system were sized utilizing technology projections for 2035. Additionally, exhaust water from the fuel cell stack is injected into the combustion chamber to further reduce NOx emissions. The resulting design has a passenger capacity of 15 with a range of 707 km, at a ticket price of €221 and a climate impact of 29.29 kgCO2,eq per passenger per flight, which is four times more sustainable than the current Beechcraft 1900D at a similar price point. The aircraft produces 1.1 grams of NOx on a typical flight, and never emits more than 8 ppm at any stage. The fuel cell provides all of the aircraft power during cruise and other low power flight phases, whereas the combustion chamber provides additional power during takeoff and climb. It is recommended to monitor the development of hydrogen technologies so they may be implemented in this retrofit by 2035.

Within the context of hydrogen combustion, the trapped vortex combustor (TVC) combined with a Rich-Quench-Lean (RQL) combustion strategy holds great promise for achieving ultra-low emissions and advancing sustainable combustion technologies. However, ensuring thorough and sufficient fuel-air mixing before the completion of the chemical reactions currently remains a critical challenge. This study focused on addressing this problem by exploring hydrogen’s injection temperature as the potential solution. Therefore, a combination of 1D simulations and Large-Eddy simulations were conducted. Subsequently, the role of this temperature modification on the chemical reactivity was evaluated, and the overall impact on the RQL's effectiveness was assessed. The results indicated that lowering hydrogen injection temperature from 300 K to 150 K reduced chemical reactivity by 15% to 25%. However, this reduction was insufficient to suppress or alter the combustion mode within the cavity significantly. Nevertheless, the overall temperature reduction within the TVC led to a significant decrease in NOx emissions of about 25%

Rotating Detonation Combustors present a promising advancement in propulsion and power generation with potential for improved thermal efficiency and power density. However, RDC operating characteristics depend significantly on the injector design, with injector pressure losses and mixing quality having a strong influence on wave-mode selection, wave speed, and pressure gain performance. Understanding these dependencies, and implementing more performant injector designs, are essential steps in maturing the technology.

This thesis, conducted at the TU Berlin Hermann-Föttinger Institute for Fluid Dynamics, examines the influence of the injector design on RDC operation. First, a comparative study evaluates a novel low-pressure-loss Twin-Co-Axial injector against a well-documented Radial-Inward-Slot baseline, assessing reduced injector pressure losses and improved mixing quality. Second, the effect of injector pressure-balance is investigated, a parameter governing post-wave recovery and impacting the balance between mixture stratification and parasitic burning. Together, these studies advance the understanding of injector-driven wave behaviour and performance.

Hydrogen represents a promising pathway for decarbonizing heavy-duty transport; however, accurately modeling its injection and combustion behavior in dual-fuel engines remains challenging. This work addresses two critical aspects of hydrogen high-pressure direct injection (HPDI) systems: the characterization of the injection process and the physical mechanisms governing combustion initiation.

A CFD-based methodology was developed to reconstruct injection profiles from apparent heat release rate (aHRR) data and to apply nozzle flow theory for approximating injector behavior, validated through experimental comparison. The findings indicate that classical convergent-nozzle theory fails to capture the observed injection trends. While convergent nozzle theory predicts variable, pressure-dependent mass flow rates, the reconstructed mass flow profiles exhibit nearly constant injection rates when the needle is open, consistent with convergent–divergent nozzle theory predictions. However, the observed variation in maximum injection rates across different cases suggests these deviations may arise not only from geometric constraints but also from aerodynamic phenomena such as boundary layer separation or recirculation within the injector. Such flow features can induce pressure-dependent effects that limit injector performance beyond what nozzle geometry and convergent-nozzle theory alone would predict.

Regarding combustion, the study reveals that hydrogen ignition, triggered by a small diesel pilot, is dominated by localized high-temperature regions produced by the diesel flame. This accelerated autoignition contrasts with alternative hypotheses involving radical transport or direct flame interaction.

Overall, these results advance the understanding of injection and ignition phenomena in hydrogen HPDI engines, providing valuable insights for refining CFD models and supporting the development of efficient hydrogen-powered heavy-duty engines.

This thesis explores the dynamics of hydrogen-enriched methane flames in a swirl-stabilized combustor with axial air injection, a configuration crucial for advancing low-emission combustion technologies. The study focuses on enhancing the understanding of flame behaviour under varying hydrogen enrichment levels, with particular attention to the effects of heat loss and strain on flame shape, mixing, and emissions. Utilizing computational methods, the research investigates key operating scenarios from the APPU project, refining thermal boundary predictions through an improved Heat Resistance Tuning approach. Additionally, it assesses the performance of a necessary flashback-prevention method, Axial Air Injection, quantifying its impact on mixing and flame shape, including their effects on temperature and pollutant formation, with varying results based on the level of hydrogen enrichment in the fuel. The results offer valuable insights that could inform control strategies for the stable and efficient operation of hydrogen-enriched combustion systems, supporting the development of fuel-flexible and environmentally sustainable propulsion technologies.

The motion of liquid metals is described by the equations of magnetohydrodynamics (MHD), that com bine the Maxwell equations and the Navier-Stokes equations. In these type of flows, the magnetic field interacting with the conductive metal induces large pressure losses and unconventional turbulence states such as quasi 2D turbulence, turbulence suppression and flow anisotropy. Currently this turbu lence behaviour can be captured in higher fidelity Computational Fluid Dynamics (CFD) simulations such as Large Eddy Simulation (LES) and Direct Numerical Simulation (DNS), but the high computa tional cost of these simulations make them impractical for industrial applications compared to Reynolds AveragedNavier-Stokes(RANS).However, theeddyviscositymodelswhicharetypicallyusedinRANS are not able to capture anisotropic turbulence states occurring in MHD flows, which results in sig nificant discrepancies in the mean velocity field. Hence, this work presents a data-driven approach to model MHD turbulence. To achieve this time-averaged LES data of annular pipe flow cases at differ ent Hartmann numbers are used to derive corrections for the k − ω SST model. Two correction fields are obtained through a frozen RANS simulation in which the mean LES fields are inserted into the RANS equations. The Reynolds stress anisotropy term is approximated with a modified Tensor Basis Neural Network (TBNN). Moreover, for modelling the turbulence production correction a Scalar Basis Neural Network (SBNN) is proposed and compared to a Sparse Algebraic Regression using the SpaRTA approach. The resulting data driven models are able to reduce the error of the Reynolds stress anisotropy values and the mean flow velocity fields, and can generalise to annular flow cases with different Hartmann numbers from those of the training cases.

With climate change posing increasing risks, the Advisory Council for Aviation Research and Innovation in Europe (ACARE) aims to reduce CO2 emissions by 75% and NOx emissions by 90% per passenger kilometer by 2050, compared to a baseline aircraft from 2000. The ARENA project addresses this by developing a waste heat recovery system using aircraft engine exhaust gases, improving fuel efficiency. This thesis focuses on integrating the condenser of such a system into the propulsion unit, aiming to minimize drag from the ram air cooling duct by evaluating different heat exchanger topologies and duct designs.

The study uses the IMOTHEP Distributed fans Research Aircraft with electric Generators by ONERA (DRAGON) concept, a hybrid electric aircraft with two tail-mounted turbogenerators. The research proceeds in several stages. Initially, a multipass-condenser's potential to reduce pressure drop was examined by adjusting the heat exchanger blockage factor per pass, but results showed no reduction in pressure drop.

Next, a lumped parameter model was developed to analyze drag, pressure drop, temperature increase, and ram air duct length. This model evaluated the sensitivity of duct geometrical parameters on the drag recovery factor—a dimensionless number indicating net thrust. Findings revealed that inclining the heat exchanger efficiently increases the drag recovery factor, while the diffuser area ratio has a similar effect but is less space-efficient. The mass flow rate ratio showed less sensitivity, and fin height or pitch had the smallest effect on drag recovery.

Using the lumped parameter model, an optimal preliminary ram air duct design was identified for different spatial constraints. The study found that inline plain tube bundle and flat tube offset strip fin heat exchangers provided more compact solutions with higher drag recovery factors. For large diffuser-blocked area fractions, optimal duct geometry remained independent of heat exchanger type, characterized by a 70-degree maximum inclination angle, a 0.7 mass flow rate ratio, and maximized diffuser area ratio within spatial constraints.

A verification study of the lumped parameter model was conducted using a two-dimensional Reynolds Averaging Navier Stokes (RANS) computational fluid dynamics (CFD) analysis with k-ω SST turbulence and a porous zone to mimic the heat exchanger. The CFD analysis confirmed the lumped parameter model's predictions, with a 0.8% difference in drag recovery factor for optimal duct geometry. Across various geometries, the mean absolute difference in drag recovery factor between models was 1.082%, with a standard deviation of 0.584%.

In conclusion, integrating the condenser in the propulsion unit within spatial constraints results in positive thrust, thus supporting Meredith's 1935[2] claim. This integration reduces net drag and improves overall efficiency, contributing to significant emission reductions in line with ACARE's targets.

This thesis introduces a newly-developed turbulent combustion model, as a next step towards modelling hydrogen combustion in aircraft engines. The proposed model (FGM-ESF) merges the Flamelet Generated Manifold approach's tabulated chemistry with the Eulerian Stochastic Field method's statistical treatment of flame-turbulence interactions at the subgrid scales, which are not resolved in LES. This hybrid model excels in managing complex combustor dynamics, high turbulence, and both premixed and non-premixed combustion modes, all while maintaining computational efficiency. Validated with a lifted turbulent H2/N2 jet flame in vitiated coflow, reflecting typical combustor conditions, the FGM-ESF model produces accurate predictions of mean velocity, temperature, and mixture in close agreement with the experiments. Comparatively, its performance matches the more costly, fully transported chemistry ESF model, showing limited sensitivity to the number of stochastic fields. The balance between computational efficiency and precision in the FGM-ESF model highlights its importance in the advancement of hydrogen-powered aircraft engines.

This study investigated the effect of tangential strain on the stability of perturbed laminar lean premixed hydrogen flames in a counterflow reactants-to-products configuration. Laminar premixed flames are highly susceptible to intrinsic instabilities, including hydrodynamic and thermodiffusive instabilities, which can cause perturbations in the flame front to grow. However, the impact of tangential strain, introduced by velocity gradients along the flame front in a counterflow setup, had not been fully explored in the context of flame stability. The present work bridged this gap by combining the counterflow setup used in previous studies of NOx reduction with the perturbation analysis commonly employed in flame stability research.

Direct Numerical Simulations were performed to study the behaviour of sinusoidal perturbations imposed on the flame front, focusing on the growth rates of these perturbations under two different strain rates (2000 1/s and 4000 1/s). The results indicated that the tangential strain improved the stability of the flame front, as the strain rate led to a reduction in both the maximum observed growth rate and the growth rate after the perturbation reached its peak. The growth rates observed were in the non-linear regime, characterised by continual variation over time. At higher strain rates, the flame front stabilised more quickly, suggesting a strong correlation between strain rate and perturbation growth dynamics. The strain-induced velocity gradients displaced the flame front, reducing its effective curvature and wavelength, further contributing to the stabilisation process.

In addition to the strain rate, the study investigated the effects of varying the amplitude and wavelength of initial perturbations. It was found that the amplitude of the initial perturbation had little impact on the maximum growth rate, but variations in the initial wavelength significantly influenced both the growth dynamics and the maximum growth rate of the perturbations.

Overall, the results provided a comprehensive understanding of how tangential strain affected flame stability and highlighted the importance of wavelength variations in determining perturbation growth rates. These findings provide insights into potential strategies to improve flame stability in practical combustion systems aimed at reducing emissions, such as in lean hydrogen combustion.

TOP contoured nozzles, with large area-ratios, are commonly employed in rocket propulsion systems as they feature an excellent thrust-to-weight ratio. A significant shortcoming to this design is that, during the startup and shutdown transients of a LRE, the internal nozzle flow progresses through a series of overexpanded flow states - Free Shock Separation (FSS) and Restricted Shock Separation (RSS), which produce critical loads associated with SWBLI and asymmetric flow separation. Exacerbated by FSI, this operational phase is known to generate the highest vibroacoustic loads at which payload and vehicle structures are subjected to when the engine is operated at off-design conditions.

Motivated by the importance of understanding how the interaction between the developing flow and the vibrating nozzle walls has an effect on supersonic noise generation and propagation, this work has studied the effect that wall compliancy has on the vibroacoustic loading of TOP contoured nozzles. This is demonstrated by means of cold flow tests carried out in the High Speed Laboratories of the Delft University of Technology on a stiff-walled aluminum nozzle, which serves as a baseline test case, and on a urethane-based compliant walled nozzle. Tests are conducted under comparable flow conditions and test parameters are measured by means of acoustic and optical techniques. Simultaneous recordings are performed and include the nozzle-wall deformation, by means of stereoscopic tracking of tracers on the nozzle lip, the imprint of the near-field acoustic signature, by means of arrays of pressure-microphones, and Schlieren imaging of the jet plume.

Measurement data allows for a Fourier decomposition of the nozzle lip displacement and of the acoustic pressure field in azimuth. Reconstruction of the instantaneous plume development enables the identification of the main flow structures responsible for noise generation.

Comparison of results between the two test articles highlights a different spectral content and directivity pattern. Correlation between the structural displacements and the acoustic signal, together with the use of DMD, quantitatively aids the investigation of how FSI has an impact on the generation of an aeroelastic tone at 180 Hz. Findings suggest that its production is the result of the periodic thickening and thinning of the shear layer owing to the heightened flapping motion of the nozzle lip preceding RSS transition, driven by an intensified shock foot instability.

Recent years have seen an increase in studies focusing on data-driven techniques to enhance modelling approaches like the two-equation turbulence models of Reynolds-averaged Navier-Stokes (RANS). Different techniques have been implemented to improve the results from these simulations. In particular, the main focus has been on overcoming the limitations implied by the Boussinesq assumption. This has been approached by using machine learning techniques as a way of discovering new formulations that could overperform when compared to traditional models.
Despite promising results for steady RANS simulations, little has yet been investigated in URANS applications. In this dissertation, this lack of research will be addressed. The main ideas are then, first, to see how the available information in URANS simulations can be used to improve the anisotropic Reynolds stress tensor prediction, and second if and how this can be done by using a sparse regression technique, whose framework is known as SpaRTA. A procedure involving the triple decomposition of High-Fidelity velocity fields is applied, aiming at finding a model exclusively for the stochastic component of the anisotropy. The test case which is considered is the flow around a cylinder at Re=3900. The High-Fidelity data was collected by running Large Eddy Simulations in OpenFOAM, after which the velocity was split into its different components through Proper Orthogonal Decomposition.
A priori results have shown good performance of the trained models, outperforming the Boussinesq assumption both in the prediction of turbulence componentiality and also on the values of the single anisotropy components.

Large eddy simulation paradigm is employed to analyse the internal flow field of a lean premixed swirlstabilized combustor with axial air injection at both non-reacting and reacting conditions for a methane and a methane-hydrogen fuel mixture. The Thickened Flame combustion model with the kinetic mechanism GRI 3.0 and the detailed chemical kinetics solver SAGE available in ConvergeCFD is employed to simulate the flow. An adaptive mesh strategy is used to maximise the mesh resolution in the flame and boundary layer regions. The numerical results are first validated against in-house experimental velocity measurements obtained via particle image velocimetry, and then leveraged to provide further insights on the flow behaviour. Significant reductions in CO2, CO and NOx emissions are observed when changing the fuel to the CH4/H2 mixture. From a POD analysis is observed that a Precessing Vortex Core is present in both the reacting and non-reacting conditions. Flashback and severe local extinctions are not observed during the simulated time.