Abstract Sustainable aviation fuels (SAFs) offer a practical route to decarbonize aviation, as they can replace or be blended with conventional kerosene without major engine modifications. However, their impact on operability and thermoacoustic instability in annular combustors is not well documented. The combustion dynamics that drives these instabilities is notably influenced by fuel composition, chemical conversion rate, and unsteady release of heat. This dependence has been recently explored using SAF surrogates, but experimental data on real SAFs are still limited. This study concerns the flame dynamics and thermoacoustic behavior of three nonaromatic SAFs, alcohol-to-jet (AtJ), catalytic hydrothermolysis jet (CHJ), and hydroprocessed esters and fatty acids (HEFA), compared to a dodecane fuel, which serves as a reference. Experiments are carried out in a single-injector combustor using a spray-swirl injector with a swirl number SN=0.74, under equivalent operational conditions. Velocity and particle size profiles are found to be quite similar, but there are notable differences in flame structures. Results highlight important differences in velocity-based flame describing functions (FDFs) and in phase average flame images during a cycle of modulation, indicating that chemical composition and physical properties can influence thermoacoustic behavior and combustor stability.
Abstract The aim of this paper is energy management system (EMS) flexibility for replication in polygeneration grids with a special focus on industrial symbiosis. This innovative EMS, which was developed and tested for application on Eigerøy island (Norway), is based on the integration of optimization software, a scheduler for the energy storage system(s), and a model predictive control (MPC) tool to take into account the dynamic response of the components. Due to the flexibility target of having an easily replicable EMS, this tool was applied to two further grids, as scheduled in the ROBINSON project: the Manna district in Crete (Greece) and the Creed district in the Western Islands (UK). However, following software replication and preliminary simulations, an experimental support was necessary to confirm the EMS robustness and to check its stability during operations in real systems. This was performed using a cyber-physical approach in the Innovative Energy Systems (IES) laboratory of the University of Genoa. It was based on the real-time coupling of hardware components (microturbine, photovoltaic (PV) panels, and a thermal grid) with software tools for the components not physically available in the laboratory. For both replication cases, 24-h tests were performed in cyber-physical mode after a necessary EMS stabilization. The results obtained for these two innovative replication cases were successfully compared with “No EMS” simulations (not including this new optimization tool), showing significant decrease in variable costs (−7.5% for the Manna district and −39.6% for the Creed case).
Abstract Capturing flame blow-off (BO) or extinction limits early in the combustor design process is essential for ensuring reliable performance and reducing development time and cost. This study investigates the applicability of a low-order extinction prediction model based on a stochastic formulation of the imperfectly stirred reactor (sISR) approach to rank different injector designs according to their BO behavior. Using turbulence statistics extracted from Computational Fluid Dynamics (CFD) solutions near blow-off conditions, the framework reproduces air–fuel-ratio-based limits within approximately 20–30% and correctly ranks the injector designs, providing a low-cost pathway for early design screening. Building on this validated baseline, parametric studies across engine-relevant conditions reveal that the critical dissipation rate increases monotonically from idle to climb as pressure and oxidizer temperature rise, reflecting the changing balance between chemical and mixing time scales. Sensitivity analyses show robustness to the CFD-derived inputs to the model and in the number of stochastic realizations. We also assess dimensionless scaling criteria and find that empirically corrected groups provide an approximate collapse across the investigated operating conditions, reducing variability by about 90%. Collectively, the results position sISR as an efficient, physics-based tool for predicting blow-off and for guiding targeted experiments during preliminary design of low-emission aero-engine combustors, at a fraction of the computational cost of high-fidelity large eddy simulation-based methods.
In Abstract In this work, self-excited high frequency thermoacoustic instabilities in a jet-stabilized gas turbine combustor with an axial stage are investigated. High-pressure experiments are performed for fuel mixtures that consist of hydrogen and natural gas in an optically accessible test rig. Various operating conditions are considered in the experiment that vary in fuel composition, air-fuel ratio, and geometry (i.e., the length of the mixing sections in the main burner and the axial stage). Data for pressure oscillations in the combustion chamber as well as high-speed images for the chemiluminescence of the electronically excited hydroxyl radical (OH*) are recorded. Depending on the operating conditions, different frequencies for high-frequency instabilities (HFI) are observed in a range above 2000 Hz. In particular, the length of the mixing section in the main burner, along with the amount of hydrogen in the fuel, appear to be relevant for the occurrence of frequencies that are typically associated with transverse and longitudinal modes. In order to shed light on the type of modes, the experimental data are supplemented by a compressible large-eddy simulation for one operating point.
Abstract Vibration monitoring and health assessment of rotor blades using the blade tip-timing (BTT) method has gained increasing attention in turbine machinery. Identifying synchronous vibration parameters identification of rotor blades can help prevent high-cycle fatigue failure and can also evaluate blade status, such as vibration stress reconstruction and modal identification. However, during acceleration and deceleration, rotor blades passing through synchronous are affected by transient response. This leads to significant errors in the blade vibration parameters identified based on the steady-state response fitting method. To improve the accuracy of blade vibration parameter identification under transient response conditions, the transient vibration displacement response equation of rotor blades during synchronous vibration is derived based on the BTT method monitoring. A whole domain transient response fitting method is proposed by replacing the Faddeeva function with the imaginary error function. To reduce computation time and improve the identification efficiency, an improved whole domain transient response fitting (IWDTRF) method is developed by combining the advantages of Algorithm 916. The IWDTRF method enables engine order (EO) analysis by monitoring blade vibration displacement with only a single sensor based on the BTT method. Simulations and experiments demonstrate that the blade synchronous vibration parameters identified by the proposed method are more accurate and stable, and further confirm its effectiveness for transient vibration parameter identification under variable acceleration conditions, making it suitable for identifying synchronous vibration parameters of rotor blades under actual engine operation conditions.
Abstract Digital twin (DT) technology has become a cornerstone for cyber-physical aero-engine systems, enabling full lifecycle digital replication and intelligent management. To address challenges of model adaptability and generalization under restricted access to fleet data conditions, this article proposes a physics-informed DT modeling framework integrating fleet feature transfer and a fleet-individual cross-attention fusion mechanism. The framework is composed of three main components: (1) a physics-embedded network, where a long short-term memory network captures temporal component dynamics, coupled with a multicomponent attention module to model intercomponent interactions; (2) a fleet-individual cross-attention module designed to extract interactive features between individual engines and the entire fleet; and (3) a pretraining and fine-tuning strategy that transfers generalized knowledge from fleet-level models to individual engines, enabling fast deployment without relying on large-scale historical data. Validation on thrust and exhaust gas temperature (EGT) prediction demonstrates superior accuracy and robustness, achieving the lowest mean absolute percentage errors (MAPEs) of 0.23% for thrust and 2.47% for EGT. Compared with baselines (a general fleet model and individual models without fleet features), the proposed approach reduces average MAPE by 47.80–75.49% (thrust) and 4.43–15.91% (EGT), improving both precision and stability. This framework enhances the scalability and digital management efficiency of aero-engine DT systems.
Abstract In this study, we investigate the coupled dynamics of pressure and heat release during the transition from a single to a double co-rotating detonation wave in a hydrogen-fueled rotating detonation engine (RDE) combustor. We analyze azimuthally decomposed pressure and heat-release-rate fluctuations to characterize the nonlinear multiphysics interactions that govern instability growth and secondary wave formation. We employ high-fidelity simulation datasets validated against experiments for the present analysis. The axial distributions of statistics of fluctuating quantities show a progressive transition from coherent detonation dynamics to shock-controlled combustion. Further analysis is carried out in terms of the modal energies, time–frequency spectra, cross-spectral coherence, thermoacoustic phase synchronization, and higher-order fluctuation statistics. The azimuthal modal energies of pressure and heat release reveal a clear redistribution of energy from the primary mode toward higher-order modes during the onset of wave splitting. Time–frequency analysis and modal coherence maps identify localized frequency locking and strong spatiotemporal coupling between pressure and heat-release-rate fluctuations. The time-lag analysis further shows that pressure fluctuations in the injector and annular wall regions systematically precede heat-release amplification near the detonation front, thus demonstrating a causal feedback mechanism responsible for secondary ignition and wave bifurcation. The combined modal, spectral, phase, and statistical diagnostics distinguish coherent detonation-driven turbulence from incoherent deflagration zones and provide predictive markers for the onset of wave mode transition. These findings advance the understanding of nonlinear shock–turbulence–combustion interaction in RDEs and support the development of reduced-order models for wave control and stabilization in such combustors.
Abstract Injector staging has recently been demonstrated to control thermoacoustic instabilities in an annular combustor, but its effects were investigated at a single operating point. This work extends this concept to a wide operating domain comprising a range of thermal powers and equivalence ratios. Staging consists in mixing two types of injectors, each exhibiting distinct flame responses when operated individually. Experiments carried out on the laboratory-scale annular combustor MICCA map the instability region for several staging patterns and indicate that they affect the instability layout and the limit-cycle amplitude within it. A reduced-order model is applied to the same operating domain and staging configurations. The model, expressed in slow-flow variables, captures the envelope of modal growth and saturation. Time averaging removes fast oscillations, leaving a set of state-dependent delay differential equations governing the nonlinear combustion dynamics. Combined with flame describing functions measured in a single injector combustor, the framework predicts limit-cycle amplitudes, spin ratios, and nodal line angles, which characterize the modal structures of instabilities. Model predictions are compared with measurements across all conditions, reproducing the overall instability trends, distinguishing stable from unstable conditions, and providing reasonable agreement with limit-cycle levels, frequency and mode structure across the tested staging arrangements.
Abstract Conduct forward-design research on engine hail ingestion testing. Starting from air-ground disparities, the study addresses key technical challenges in ground-based testing. Underground static conditions, with the engine in a static state, precise calibration of hail ingestion volume is required, alongside overcoming the technical difficulty of projecting large quantities of hail into the engine. Based on the most stringent spatial environmental criteria from airworthiness regulations, adjustments are made to the hail water content according to the identified disparities. Targeted research on continuous hail projection technology is carried out, employing engineering calculation methods for preliminary design. Cross-sectional frictional pipe flow calculations are used to determine airflow movement, while aerodynamic parameters inform the computation of hail particle forces and motion trajectories. Subsequently, coupled computational fluid dynamics-discrete element method (CFD-DEM) simulations are applied to reveal the acceleration mechanism of hail. Appropriate mesh and time-step sizes were selected based on grid independence and time-step sensitivity studies. Experimental validation of the hail projection apparatus is conducted, analyzing hail projection velocity and radial dispersion. Calculations indicate that the simulated hail velocity exceeds experimental results by 0.3%. Considering both hail projection and air consumption, a hail projection tube length of 4 m and a tube diameter of 45 mm are selected. Applied to a continuous hail ingestion test on a scaled fan/booster rig, the system successfully projects continuous hail into the inlet. The research results provide an approach for the design of hail projection devices used in continuous hail ingestion tests for aero-engines.
Abstract The impact of fuel temperature and composition on flame structure and in situ soot formation is investigated at 1.0 MPa using spatially resolved laser-induced incandescence (LII) measurements. Concurrent hydroxyl chemiluminescence (OH* CL) measurements are used to correlate soot formation and retention with flame heat release. Experiments were conducted over a range of fuel temperatures spanning from 300 K to 575 K, and two liquid hydrocarbon fuels (Jet A and hydroprocessed esters and fatty acids synthetic paraffinic kerosene (HEFA-SPK)). Operating conditions are selected to correspond to mission-relevant low-power pilot-only operation for taxiing and approach, and high-power pilot-and-main operation for cruise. During pilot-and-main operation, fuel aromatic content is vital for the production of soot regardless of fuel temperature. Without aromatics in the fuel, soot production is drastically reduced. For fuels with aromatic content, increasing fuel temperature results in a shift in soot formation structure from small, high-intensity pockets to more frequent, less intense dispersed fields. Pilot-only operation operates locally at an equivalence ratio well above critical sooting limits, resulting in soot formation for Jet A as well as HEFA-SPK. LII signal intensity is ∼3.5 times more intense with aromatics present in the fuel at this condition, indicating their importance in soot formation. The inclusion of aromatics in the fuel results in more rapid soot formation and shifts the location of peak signal intensity upstream. Elevated fuel temperatures for pilot-only operating conditions result in more rapid soot consumption in the downstream region of the combustor only when aromatics are present.
Abstract This work investigates the effects of hydrogen blending on soot formation and flame structure in a laboratory-scale rich–quench–lean (RQL) burner using ethylene–hydrogen mixtures (0–50 vol. % hydrogen) at constant carbon mass flow rate. Laser-induced incandescence (LII) and single-ring and multicyclic polycyclic aromatic hydrocarbons (PAH) planar laser-induced fluorescence (PLIF) were employed to quantify soot and PAH distributions, respectively, while OH* chemiluminescence was used to detect the flame structures and reaction zone location. Hydrogen addition progressively reduced soot by ca. 9%, 36%, and 68% at 10, 30, and 50 vol. % blending, respectively, with multicyclic PAHs decreasing more than single-ring aromatics. The single-ring aromatics were confined to the early parts of the fuel jet, while the multicyclic PAHs spread more downstream. Increasing the percentage of air flowing through the dilution jets results in significant shortening of the flame and reduction in soot, irrespective of the hydrogen content, and results in a difference between single-ring aromatics and multicyclic PAH distributions, possibly due to the reduction of the residence time in rich mixtures. An additional case with helium instead of hydrogen helped to isolate chemical effects from aerodynamic effects. The results suggest that the chemical effects of hydrogen addition dominate over thermal effects. This study shows that hydrogen addition can control particulate emissions. The dataset enables validation of turbulent combustion models for soot.
Abstract Hydrogen is widely viewed as a promising sustainable fuel without direct carbon emissions. However, it introduces new challenges related to pollutant formation, which demand precise and dynamic diagnostics. Accurate measurements of exhaust gas composition are essential, as they provide valuable insights into mechanisms of pollutant formation and serve as a key indicator of combustion efficiency, stability, and overall system performance. This work presents a tunable diode laser absorption spectroscopy (TDLAS)-based sensor adapted for extractive exhaust gas analysis in a full annular combustor rig. The system enables high-temporal-resolution detection of the key combustion species, nitric oxide, nitrogen dioxide, oxygen, and water vapor, which are typical for hydrogen combustion. While these species are traditionally measured using different sensing principles, such as chemiluminescence and paramagnetic detection, TDLAS allows simultaneous measurements of all four molecules. It offers calibration-free operation, independence from mass flow, an unmodified gas matrix, and excellent temporal resolution. For spatially resolved diagnostics, the measurement probe was traversed around the annular exit plane of the combustion chamber in a predefined time. A comparison with conventional measurement equipment under stable combustion conditions demonstrates that the TDLAS system not only reproduces results from established techniques but can even surpass them in terms of temporal resolution and selectivity.
Abstract The present work deals with the numerical and experimental investigation of a novel combustion system for the generation of superheated steam. This combustion system is based on the stoichiometric combustion of pure oxygen and hydrogen in an atmosphere of high-pressure, low-temperature steam (“Oxy-H2 in steam” combustion). For successful steam superheating, it is important that only small quantities of unreacted hydrogen and oxygen remain in the steam. A two-stage design is therefore envisaged, in which the first stage serves mainly the purpose of combustion of hydrogen and oxygen, whereas the second stage is used to provide sufficient time for mixing and for chemistry to reach chemical equilibrium. The total amount of steam is split between these two stages. In the first stage, only a small fraction is added as a diluent to limit the combustion temperature. The bulk of the steam is added in a second stage. The detailed investigation of the first stage, i.e., the burner, is the main focus of the present work. Based on computational fluid dynamics, the combustion process and the conditions in the reaction zone are simulated numerically. Emphasis is given on an analysis of the flame structure and the mixing of the reactants with the dilution steam. For validation, measurement data for the OH* chemiluminescence (OH*-CL) are collected in high-pressure experiments at 10 bar in an optically accessible combustion chamber. The comparison to experimental data shows that the numerical simulation can reproduce key aspects of the combustion accurately.
Abstract Modeling fuel atomization in modern aircraft combustors remains challenging because liquid–gas interface dynamics, ligament formation, and droplet breakup are highly nonlinear. This study presents a systematic numerical framework for primary atomization and reactive flow simulation in airblast injectors. The methodology combines volume-of-fluid (VOF) interface capturing, wall-film modeling, and a discrete phase model (DPM) to resolve liquid breakup and generate nozzle-specific injection files for subsequent reactive simulations using flamelet generated manifolds (FGM) with real-fuel chemistry. A HyChem-based mechanism with 129 species and 880 reactions was employed. Validation was performed using the DLR generic single sector combustor, showing good agreement for both isothermal and reactive cases. Compared with conventional hollow-cone injections requiring empirical tuning, the VOF-Lagrangian wall film (LWF)-DPM approach captures coherent wall-film and ligament dynamics prior to secondary breakup. Adaptive mesh refinement near the liquid–air interface improved spray dispersion and reduced droplet size, demonstrating the balance needed between fidelity and computational cost. The same methodology was applied to Hanwha Aerospace's airblast nozzle, with comparable agreement in spray and single-sector combustion tests. Simulations revealed large coherent structures, including trapped vortices and a precessing vortex core, that enhanced turbulent mixing. The injection model from VOF-LWF-DPM simulation enabled reactive simulations showing differences in flame structure and NOx formation relative to conventional injection models. The overall temperature distributions remained comparable under high-enthalpy conditions. The benefit of realistic spray boundary conditions is expected to increase under low-power and lean blowout regimes.
Abstract In this paper, the impact of cryogenic hydrogen preconditioning on aero engines is investigated using thermodynamic models of three fuel system architectures. Furthermore, highly simplified approaches that are useful for preliminary studies and conceptual design work are developed and evaluated. Fuel systems for cryogenic hydrogen require fuel preheating in addition to the fuel–oil heat exchanger. Three options are investigated: electric, bleed, and combustion preheating. All are modeled in the gasturb performance software using property functions and thermodynamic change-of-state equations. The reduction in thrust-specific energy consumption of the hydrogen engine compared to the kerosene reference engine depends on the required injection temperature. Without preheating, a reduction of 4 % would be achievable in cruise. With preheating to an injection temperature of 273.15 K, a reduction of 1.7 % is achieved with electric preheating, and up to 3 % with bleed or combustion preheating. The results are compared to highly simplified approaches in which the fuel preconditioning is approximated by adjusting mechanical efficiencies, secondary air system inputs, and the fuel heating value. The errors introduced by these simplifications are well below 0.5 % in fuel flow at operating points such as takeoff and cruise, thus, such simplified approaches are often justified. In summary, heat release varies strongly with fuel temperature, especially for cryogenic hydrogen, and the associated power and bleed offtakes for fuel-preheating must be included in performance calculations. This paper presents both simplified approaches and rigorous models to achieve this.
Abstract High-temperature heat pumps (HTHPs) are a key technology for decarbonizing industrial process heat. Within such systems, two-phase water ejectors offer a promising opportunity to function as secondary steam compressors. By entraining hot steam with high-pressure water, they enable simultaneous cooling and pressure increase, thereby reducing power consumption and the number of mechanical compressor stages required. This integration can enhance the energy efficiency and economic viability of HTHPs. However, the internal flow dynamics of two-phase ejectors remain insufficiently understood, particularly under high-pressure and high-temperature operating conditions relevant for heat pump applications. In the mixing region, where the two-phase flow approaches critical conditions, the speed of sound is affected, potentially giving rise to shock structures and variations in thermophysical properties. The lack of experimental insights into these phenomena limits reliable design and optimization of ejectors for industrial applications. This paper investigates the experimental characterization of two-phase water ejectors to support their integration into HTHP systems. A novel closed-loop experimental facility has been developed, enabling continuous operation over representative water and steam conditions. The facility integrates designs for the nozzle and mixing chamber, a water separation and condensation system, and sensors and control strategies. The ejector is based on prior CFD simulations, enabling comparison with simulations. Initial tests provide pressure and entrainment ratios. The results deliver new ejector component designs, an experimental methodology, and initial data that deepen the understanding of two-phase ejector behavior and support their optimized application in advanced HTHP systems.
Abstract A multijet-in-crossflow burner operating with 100% hydrogen was investigated to understand how crossflow swirl intensity, equivalence ratio, and thermal power influence flame stabilization and NOx emissions. Using simultaneous OH- and acetone-PLIF diagnostics, three flame stabilization modes were identified. At high equivalence ratios, anchored jet flames form, which transition into anchored M-shaped flames as swirl increases or equivalence ratio decreases. Both anchored modes are associated with high thermal stress and are therefore undesirable. At sufficiently high airflow velocities, a detached, partially premixed M-shaped flame stabilizes, reducing thermal loads on the injector and offering more favorable operating conditions. NOx emissions and combustor exit temperatures were measured using Fourier transform infrared (FTIR) spectrometry and thermocouples across thermal powers of 3–17 kW, equivalence ratios from 0.1 to 0.9, and swirl numbers up to 1.79. Results show that lower equivalence ratios and higher swirl intensities significantly reduce NOx emissions. A marked decrease in NOx occurs during the transition from jet flames to M-shaped flames, likely due to the formation of an internal recirculation zone (IRZ), while flame lifting itself has little effect. An effective swirl number, introduced in prior work, successfully characterizes the combined influence of swirl and equivalence ratio on NOx formation, enabling identification of an optimal operating window with low emissions and stable lifted flames. Additionally, a scaling law based on the residence time within the flame volume was assessed, showing a good agreement with the experimental dataset.
Abstract In the pursuit of net-zero energy systems, advanced combustion concepts are essential to ensure safe and efficient operation of future gas turbines (GTs) across sustainable fuels and carbon-neutral solutions. Under such unconventional conditions, flames no longer conform to a single regime but span from premixed to non-premixed combustion, depending on fuel injection strategy, turbulence intensity, and local flow field. While scale-resolving reactive Computational Fluid Dynamics (CFD) is central to GT design, modelling multi-regime combustion remains challenging, as most models are tailored to single regimes, limiting accurate prediction of flame stabilisation and emissions. Furthermore, the high Reynolds numbers typical of industrial burners demand accurate treatment of turbulence-chemistry interaction (TCI) to resolve flame dynamics. In Large-Eddy Simulations (LES) based on species transport models, a flame index is employed to identify local regimes, with the Thickened Flame (TF) model applied to premixed combustion. In non-premixed regions, however, thickening is typically deactivated, leaving TCI unresolved and leading to inaccuracies at practical grid resolutions. To address these limitations, this study presents a unified Thickened Flame-Fine-Scales (TF-FS) formalism, coupling the Dynamic Thickened Flame model (DTFLES) with a fine-scales-based TCI closure. Premixed combustion is handled by DTFLES, while non-premixed regions are modelled using an LES-adapted formulation of the Eddy Dissipation Concept (EDC). The hybrid approach, previously assessed on academic configurations, is applied here to a full-scale industrial lean-premixed burner and evaluated against experimental data from the THT Lab in Florence.
Abstract This paper presents the first comprehensive experimental campaign using a rainbow annular cascade of high-pressure turbine (HPT) rotor airfoils using the new Annular Multiframe turbine Blade Rig for Aerothermal analysis (AMBRA) stationary-frame methodology, which reproduces rotor-relative inflow conditions in a nonrotating annulus. A flow-conditioning gauze imposes prescribed radial and circumferential distributions of total pressure and whirl angle, enabling controlled testing of turbine geometries under transonic inflow conditions at exit Mach numbers near 0.85 and 1.00. Operating in a stationary configuration allows use of high-resolution instrumentation—such as dense blade-surface pressure taps, multirake traverses, and detailed flow-angle measurements—typically achievable only in low-speed linear cascades, but applied here to realistic transonic turbine environments where rotating rigs offer limited diagnostic access. Two turbine geometries were investigated over Reynolds numbers from approximately 300,000 to 1,000,000. The experimental suite included simultaneous continuous circumferential traverses of Kiel and five-hole probes to map total pressure, flow angle, and Mach number, together with endwall and blade-surface static pressure measurements; oil-flow visualizations were performed separately. These complementary diagnostics enabled detailed characterization of passage vortices, shock locations, and loading evolution, as well as computation of area-averaged loss coefficients for performance comparison. The results show clear, repeatable differences between geometries, with loss distributions and flow-field features lying well outside the uncertainty bounds. The methodology, therefore, provides high-fidelity aerodynamic discernment under both design and near-sonic off-design conditions. By combining realistic annular inflow with dense spatial measurements, this approach bridges the gap between linear cascades and rotating rigs, offering a powerful framework for systematic evaluation of next-generation transonic turbine designs and secondary flows at Technology Readiness Level (TRL) 3–4.
Abstract Hydrogen is under evaluation as a low-emission fuel candidate for future aviation systems. Despite its environmental advantages, its low volumetric energy density presents substantial engineering challenges, particularly in aircraft design and airport fuel handling infrastructure. As an interim solution, fuel-flexible combustion systems, designed to operate efficiently with both hydrogen and conventional hydrocarbons, offer a viable route toward gradual decarbonization of the aviation sector. However, the inherent complexity of dual-fuel burners, characterized by multi-regime operation and diverse fuel properties, poses several challenges for their numerical modeling. To address these modeling challenges, high-fidelity Large Eddy Simulation (LES) incorporating finite-rate chemistry is one of the most effective tools for capturing the complex dynamics of reactive flows involving hydrogen or hydrocarbon mixtures. Within this framework, the Dynamic Thickened Flame Model (DTFLES) has shown promise for modeling hydrogen flames. However, its applicability to dual-fuel systems, especially those featuring interactions between hydrogen jets and non-premixed hydrocarbon sprays, remains underexplored and requires further investigation. Building on this foundation, the present study evaluates the performance of the DTFLES in simulating reactive flows within laboratory-scale swirled spray burners operating with hydrogen and hydrocarbon spray injection. To mitigate computational expense while preserving chemical fidelity, an analytically reduced mechanism is developed using the ARCANE reduction tool. The model's capability to capture multi-fuel, multi-phase interactions is assessed through direct comparison with experimental diagnostics, including OH-PLIF imaging and flame temperature measurements.
