Accurate prediction of discharge coefficients (Cd) in rotating orifices is essential for the design of aero-engine internal air systems, yet existing correlations usually treat axial and radial orifices separately and do not fully represent intermediate wall inclination angles. In this study, steady-state RANS simulations in a rotating reference frame, supported by validation against published data and by rotating orifice experiments, are used to investigate the combined effects of wall inclination angle alpha and length-to-diameter ratio L/d on Cd. The numerical results show that, under typical conditions (N = 3000 rpm, Pi = 1.03, L/d = 1.5), Cd increases from 0.301 to 0.340 as alpha increases from pi/2 to pi, corresponding to a 12.96% increase. Under low rotational speeds and high pressure ratios, the Coriolis force reduces the relative tangential velocity and the incidence angle, thereby increasing Cd with alpha; however, at high rotational speeds and low pressure ratios, the centrifugal resistance to radial inflow becomes dominant, and at N = 7000 rpm, the Cd for the alpha = pi orifice is 38.96% lower than that for the alpha = pi/2 orifice. Increasing L/d promotes flow redevelopment and amplifies the Coriolis-force effect, leading to a larger Cd increase for orifices with larger alpha. Based on these mechanisms, a generalized incidence-angle formulation incorporating Coriolis and centrifugal effects is developed, and a Cd prediction model applicable to pi/2 <= alpha <= pi and different L/d values is proposed. Experimental validation shows that the maximum prediction error is reduced to 2.37%, demonstrating the accuracy of the proposed model for rotating inclined orifices.

To investigate the effects of different feed concentrations on particle dynamics characteristics in the vertical pipe hydraulic conveying system, Particle Image Velocimetry (PIV) and Electrical Resistance Tomography (ERT) techniques were employed to measure the particle velocity and local concentration distributions in this study. Subsequently, the particle fluctuation velocity is analyzed by using Proper Orthogonal Decomposition (POD) and Continuous Wavelet Transform (CWT) to further reveal the particle motion mechanism with different feed concentrations of system. The results show that the particle axial velocity decreases as feed concentration increases, while the local concentration, system pressure, and fluctuation energy of particle velocity increase. Additionally, the peaks of the power spectrum increase with the increasing of feed concentration and are primarily distributed in the lower frequency range, and the quasi-periodic particle flow is formed at the high particle concentration regions. Meanwhile, the energy contribution of each POD mode decreases with the increasing of feed concentration. As feed concentration increases, the small-scale motions of the particles becomes stronger, resulting in a relatively higher fluctuation energy of particle velocity. Furthermore, as feed concentration and POD mode number increase, particle fluctuations deviate more significantly from a Gaussian distribution.

Exploring the impact of alcohol additives on combustion and pyrolysis of ammonia/methane is of great importance in the pursuit of sustainable energy technologies. This work employs Reactive Force Field (ReaxFF) molecular dynamics (MD) simulations to investigate the underlying mechanism of how ethanol and methanol additives affect reaction pathways, NOx emissions and bond energy characteristics in ammonia-methane pyrolysis and combustion processes. It shows that adding alcohols altered NOx formation pathways, reducing the diversity of NOx and shifting the equilibrium toward simpler NOx such as NO and NO2. At 2,000 K, alcohol blends, particularly methanol, demonstrated a notable reduction in NO2 formation. At 3,000 K, both ethanol and methanol suppressed NO production, but the influence of methanol was stronger. Nitric acid production, HNO3, was present at lower temperatures but became negligible at higher temperatures because of the thermal breakdown of the higher-order NOx. These trends confirm that alcohol additives realize a probable role in moderating NOx emissions and stabilizing reaction pathways. The pyrolysis in modified reaction pathways, which facilitated the decomposition of ammonia and methane in these blends, affected the formation of intermediate species, leading to the reduction in peak emissions. In addition, methanol and ethanol showed significant impacts on hydrogen bond energies of the mixture, especially important building block radicals encouraging higher complexity pathways. By leveraging a computationally robust and scalable methodology, this study not only advances a fundamental understanding of alcohol-enhanced ammonia/methane combustion but also informs strategies to optimize these mixtures for practical use in modern propulsion systems.

Recent advances in combustion science have led to the generation of large volumes of data from high-fidelity simulations, detailed chemical-kinetic calculations and engine-relevant measurements and create new opportunities for data-driven modelling across interacting physical and chemical scales. Among these approaches, artificial intelligence has emerged as a promising framework for constructing surrogate models that reduce computational costs, deliver substantial speed-up and support prediction in complex reacting systems. This review provides a state-of-the-art assessment of AI-powered surrogate modelling for multiscale combustion, spanning chemical kinetics, mechanism reduction, turbulent flames, combustors, engines, and emissions prediction. Supervised, unsupervised, and hybrid or physics-guided learning approaches are examined and compared in terms of predictive accuracy, physical consistency, computational efficiency, and generalizability across conditions and scales. The review further discusses key challenges, including limited transferability across fuels and operating regimes, extrapolation errors, inconsistency in datasets and benchmarks, and the difficulty of building robust and trustworthy models for practical combustion workflows. Future opportunities are identified in the development of more reliable, scalable, and physically grounded surrogate frameworks for next-generation combustion research.

Although ammonia-methane (NH3-CH4) combustion is a viable low-carbon energy pathway, high NOx emissions remain a major barrier to practical deployment. In this study, reactive force field (ReaxFF) molecular dynamics (MD) simulations were carried out to investigate how ethanol and methanol additives influence the radical chemistry and NOx formation at 2000 K and 3000 K. Atomic-scale analysis was performed using mean squared displacement (MSD) and radial distribution functions (RDF). The results have revealed that alcohol additions alter species mobility and short-range ordering, weakening N-O correlations while enhancing H-O interactions. These effects promoted the radical redistribution and enabled NOx suppression, with the 10 % ethanol blend achieving the largest reduction (39.6 %) at 3000 K. To extend predictive capability beyond directly simulated cases, machine learning (ML) models were trained on MD-derived descriptors. Among the algorithms tested, Random Forest Regression (RFR) demonstrated the most reliable performance, accurately generalizing to untested alcohol ratios (2 %, 7 %, and 12 %) and reproducing consistent chemical trends. The integrated MD-ML framework shows how atomistic simulations and ensemble learning can together provide predictive insights into cleaner combustion regimes, offering a cost-effective tool for guiding additive selection and optimizing NOx control strategies.

To elucidate the energy-saving mechanism of a swirling flow pneumatic conveying system with variable-pitch blades, this study combines continuous wavelet transform (CWT) and dynamic mode decomposition (DMD) to analyze particle dynamic characteristics from multiple scales. Firstly, the energy-saving performance of variablepitch blades is evaluated in terms of system pressure drop and power consumption coefficient. Compared to conventional axial flow, the variable-pitch blades reduce the optimal conveying velocity and power consumption coefficient by up to 20.12 % and 15.52 %, respectively. Subsequently, the particle concentration and velocity distributions are measured using electrical capacitance tomography (ECT) and particle image velocimetry (PIV), confirming that the variable-pitch blades significantly enhance particle suspension and dispersion in the pipe. Finally, CWT and DMD are employed to conduct a multi-scale analysis of the particle dynamics. The CWT analysis indicates that the variable-pitch blades promote large-scale particle motion while suppressing smallscale motion near the inlet; as the conveying distance increases, particle motion shifts toward smaller scales. The DMD results further support this trend, showing that the blades augment low-frequency modal energy while suppressing high-frequency energy in the upstream region. The analyses of DMD fluctuation intensity, vorticity, and Reynolds shear stress confirm that the variable-pitch blades promote a more uniform particle distribution.

This research introduces an approach for modifying the wake structure of submersibles during descent by integrating flexible appendages onto their surface, with the goal of achieving drag reduction. To quantify the impact of these appendages on hydrodynamic drag and wake patterns, experiments employing a six-component force sensor and high-speed particle image velocimetry (PIV) were conducted on submersible models equipped with appendages of varying lengths. The findings indicate that flexible appendages can reduce drag by up to 8.7% during vertical descent. Analysis of the time-averaged and transient flow fields at a Reynolds number of 109,582 reveals that the appendages disrupt the wake vortices under vertical diving conditions. To elucidate the underlying mechanism, the proper orthogonal decomposition (POD) technique was applied. POD results demonstrate that the flexible appendages modify the vortex structure within the wake, leading to a reduction in vortex energy and a suppression of vortex shedding phenomena.

As carbon-free energy carrier, combustion of ammonia suffers from low burning rates, poor flame stability, and excessive nitrogen oxide (NOx) emissions. Although blending with hydrocarbon fuels such as methane alleviates some drawbacks, NOx formation remains a critical barrier. To address these challenges, we propose a hybrid framework combining reactive force field molecular dynamics (ReaxFF-MD) simulations with machine learning (ML). MD simulations at 2000-3000 K were performed for ammonia-methane blended combustion with 0-10% addition of ethanol or methanol. Adding alcohols suppressed the NOx formation by altering charge redistributions and redirecting nitrogen intermediates into stabilising pathways. Particularly at 3000 K, 10% ethanol and methanol reduced NOx by ∼39.5% and ∼30.1%, respectively. Both chemical and physical descriptors derived from MD were used to train ML models and successfully predicted NOx trends at intermediate compositions (2%, 7%, 12%) with <5% error for ethanol-rich mixtures, though predictions beyond 12% require further validation. This framework reduces reliance on costly simulations while providing mechanistic insights and predictive capability of designing alternative fuels.

This study investigates a novel drag-reduction strategy for submersibles during ascent, achieved by altering the wake structure with surface-mounted flexible attachments in varying numbers. Previous studies have mainly focused on the influence of flexible attachment length on wake modulation, while the effect of attachment number on wake evolution and quantitative drag reduction remains insufficiently understood. To address this knowledge gap, the flow structures around submersibles equipped with different numbers of flexible attachments were analyzed and compared at a Reynolds number of 109 561, with a particular focus on wake characteristics such as Reynolds stress and turbulent kinetic energy (TKE). The wake structure was then analyzed using discrete wavelet transform (DWT) and continuous wavelet transform (CWT) methods. The results demonstrate that flexible attachments stabilize the wake, reduce turbulence intensity, and improve the spatial uniformity of TKE. A maximum drag coefficient reduction of 5.3% is obtained with 20 flexible attachments. Wavelet analyses, which include both DWT and CWT, reveal a dual-scale modulation mechanism: the flexible attachments preferentially break down large-scale vortices while promoting energy cascade toward smaller scales, thereby enhancing turbulent energy dissipation. This multiscale flow modulation ultimately leads to a more stable wake structure, as evidenced by spatial correlation analyses showing significant suppression of velocity fluctuations. However, further increasing the number of flexible attachments beyond this optimum induces wake broadening and adverse secondary flows, resulting in a rebound of drag.

The breakup pattern of a water drop in a high-speed air flow was numerically investigated under conditions which are typical for the stripping-type drop disintegration. Three-dimensional numerical simulations have been performed to investigate the complex interaction of a supersonic shock wave (Ma = 1.47) with a cylindrical water drop using the unsteady Reynolds-averaged Navier-Stokes approach. The Kelvin-Helmholtz Rayleigh-Taylor breakup model is employed to characterize the disintegration of the liquid drop process and a coupled level set/volume of fluids technique is applied to examine the topological changes of the gas/liquid interface. The computational approach has been validated by comparing the predicted displacement/drift with time, acceleration with time, and drag coefficient of the drop against the experimental data. Comparisons, with good agreement, have also been made between the predicted drop length/width/area and the experimentally measured values to elucidate changes in the shape and size of the drop. Comprehensive flow visualization has been used to show the shock-liquid drop interaction process, i.e., the onset of drop compression/flattening, formation of vortices, production/distribution of vorticity, pressure distribution, inception of the downstream separation points and stripping points at the equator followed by their subsequent fusion. These detected qualitative features are consistent with the past experimental observations of stripping. It is demonstrated for the first time that turbulence is generated at the early phase of the shock cylindrical drop interaction process, with the maximum turbulence intensity reaching about 22% around the recirculation zones and the wake region. In addition, the present study shows that the Kelvin-Helmholtz instability plays a more important role than the Rayleigh-Taylor instability in the process succeeding the shock impact on the drop.

As carbon-free energy carrier, combustion of ammonia suffers from low burning rates, poor flame stability, and excessive nitrogen oxide (NOx) emissions. Although blending with methane alleviates some drawbacks, NOx formation remains a critical barrier. To address these challenges, we propose a hybrid framework combining reactive force field molecular dynamics (MD) simulations with machine learning (ML). MD simulations at 2,000–3,000 K were performed for ammonia-methane blended combustion with 0–10% addition of ethanol or methanol. Adding alcohols suppressed the NOx formation by altering charge redistributions and redirecting nitrogen intermediates into stabilising pathways. At 3,000 K, 10% ethanol and methanol reduced NOx by ~ 39.6% and ~ 30.1%, respectively. Both chemical and physical descriptors derived from MD were used to train ML models and extrapolated to unseen conditions (> 10% alcohol) with < 5% error in ethanol-rich mixtures. This framework reduces reliance on costly simulations while providing mechanistic insights and predictive capability of designing alternative fuels.

This paper presents an innovative approach to predicting thermophysical properties of ethanol-octane blends by integrating molecular dynamics (MD) simulations with machine learning (ML) algorithms. The work addresses the growing interest in ethanol-gasoline blends as alternative fuels and the need for efficient computational methods to analyze their properties. Using MD simulations and various ML models such as Decision Tree Regression (DTR), Random Forest Regression (RFR) and Gaussian Process Regression (GPR), the behavior of 660-molecule systems of ethanol-octane mixtures was modeled. The OPLS-AA force field was employed to accurately represent the molecular interactions. Among the ML models, DTR demonstrated the highest accuracy in predicting atomic displacements and velocities. The integration of MD with ML promises rapid and accurate predictions, with error rates consistently below 2.5% across different ethanol concentrations and timesteps. Notably, the ML model showcases remarkable speedup in computational efforts, approximately 1.8, 2.7, and 3.4 times faster for E10, E20 and E85 blends respectively compared with the traditional MD simulations. This approach not only enhances the understanding of ethanol-octane blend properties but also demonstrates the potential for ML to accelerate the complex molecular simulations. The findings of this study have significant implications for the design and optimization of alternative fuels, targeting the sustainable energy demand.

Precise estimation of boiling points in organic fluids is critical for designing efficient and safe thermal systems. This study presents a hybrid molecular dynamic (MD)-machine learning (ML) framework for boiling point estimation in two representative aromatic fluids: biphenyl (C12H10) and diphenyl ether (C12H10O). Two force fields, OPLS-AA and COMPASS, were tested in equilibrium MD simulations. OPLS-AA produced density predictions with a relative error below 2 % compared to experimental values, while COMPASS showed reduced accuracy at elevated temperatures. Boiling point was estimated using a density threshold method (yielding 525.66 K) and a thermodynamically rigorous inflection-point method (508.18 K), revealing similar to 3.3 % deviation between boiling onset and completion. MD data were used to train and evaluate three regression models-Nearest Neighbours Regression (NNR), Neural Network (NN), and Support Vector Regression (SVR). The NNR model achieved the best match with MD data, predicting a boiling point of 524.97 K and density of 0.064 g/cm(3). The NN model accurately estimated boiling temperature (525.3 K) but overestimated density, while SVR underestimated both. This work contributes a novel, interpretable MD-ML framework to integrate the inflection-point detection with data-driven model selection, offering a reproducible and accurate method for boiling point estimation that can be extended to other organic thermal systems.

Diesel trains play a vital role in the UK’s rail passenger transport. Despite efforts to expand electrification, over 10% of the UK’s rail routes will remain non-electrified. To reduce emissions and phase out diesel trains by 2040, the UK rail network is actively exploring alternative fuels. This paper presents a comprehensive technical, economic, and environmental analysis of converting diesel trains to hydrogen-powered trains using a hydrogen combustion engine for the first time. A simulation-based methodology has been developed to assess train performance, fuel consumption, and emissions for both hydrogen and diesel engines. The developed methodology has been validated by comparing the predictions against the available experimental data and a very good agreement has been obtained. A case study involving British Class 195 diesel-powered regional trains on the Manchester Airport to Barrow-in-Furness route is analysed. The simulation results show that hydrogen-powered trains achieve zero carbon emissions and exhibit similar NOx emissions to diesel, with a similar performance. Over the train’s 30-year lifespan, green hydrogen can reduce CO2-equivalent emissions by up to 187.4 kt. The study clearly demonstrates that hydrogen combustion engines offer a practical, mid-term solution for decarbonizing regional rail, with much lower conversion costs compared with fuel cell technology.

This study presents a method to reduce drag and enhance the diving velocity of a submersible through the application of a double-row of silk appendages on its surface. Initially, Computational fluid dynamics (CFD) simulations reveal that boundary layer separation occurs in the midship of a conventional submersible, accompanied by significant vortex formation at the bow and stern. Thus, a double-row of silk appendages is applied to suppress boundary layer separation and modify the vortex structures in the wake. Torque sensors and particle image velocimetry (PIV) systems are employed to measure the drag force and wake flow field. The results show that the silk appendages effectively suppress adverse pressure gradient-induced boundary layer separation. Moreover, the wake structure is optimized, leading to a drag reduction of up to 6.67 %. To investigate the underlying mechanism, continuous wavelet transform (CWT) and dynamic mode decomposition (DMD) methods are applied. The analysis indicates that a double-row of silk appendages with appropriate length reduces low-frequency modal energy, suppresses flow separation and large-scale vortices, promotes small-scale vortex structures, and improves flow field stability. However, longer appendages may intensify local disturbances, resulting in decreased stability.

Three-dimensional (3D) computational fluid dynamic (CFD) simulations have been performed to investigate the complex flow features and stripping of fluid materials from a cylindrical water drop at the late-stage in a Shock Liquid Drop Interaction (SLDI) process when the drop’s downstream end experiences compression after it is impacted by a supersonic shock wave (Ma = 1.47). The drop trajectory/breakup has been simulated using a Lagrangian model and the unsteady Reynolds-averaged Navier–Stokes (URANS) approach has been employed for simulating the ambient airflow. The Kelvin–Helmholtz Rayleigh–Taylor (KHRT) breakup model has been used to capture the liquid drop fragmentation process and a coupled level-set volume of fluid (CLSVOF) method has been applied to investigate the topological transformations at the air/water interface. The predicted changes of the drop length/width/area with time have been compared against experimental measurements, and a very good agreement has been obtained. The complex flow features and the qualitative characteristics of the material stripping process in the compression phase, as well as disintegration and flattening of the drop are analyzed via comprehensive flow visualization. Characteristics of the drop distortion and fragmentation in the stripping breakup mode, and the development of turbulence at the later stage of the shock drop interaction process are also examined. Finally, this study investigated the effect of increasing Ma on the breakup of a water drop by shear stripping. The results show that the shed fluid materials and micro-drops are spread over a narrower distribution as Ma increases. It illustrates that the flattened area bounded by the downstream separation points experienced less compression, and the liquid sheet suffered a slower growth.

This study explores the integration of machine learning (ML) techniques with large eddy simulation (LES) for predicting species mass fraction and flame characteristics in partially premixed turbulent jet flames. The LES simulations, conducted using STAR-CCM+ software, employed the Flamelet Generated Manifold (FGM) approach to effectively capture the interactions between the turbulence and chemical reactions, providing high-fidelity data on flame behaviour and pollutant formation. The simulation was based on the Sandia Flame D specification, utilizing a detailed mesh to accurately represent flow features and flame dynamics. To enhance real-time prediction capabilities, three ML models, Neural Networks (NN), Linear Regression (LR), and Decision Tree Regression (DTR), were trained on the LES data. Comparative analysis using metrics such as Mean Absolute Error (MAE), Mean Squared Error (MSE), Pearson Coefficient (PC), and R-squared (R2) identified the NN model as the most effective one. The NN model demonstrated high accuracy in predicting species mass fractions and flame patterns, significantly outperforming traditional LES solvers in terms of computational efficiency. The study also highlighted the considerable computational speedup achieved by the NN model, making it approximately 17.25 times faster than traditional LES solvers. Despite some limitations, such as handling large dataset fluctuations, the ML models have shown promise for future applications in combustion simulations.

This paper presents an innovative approach to optimising the cold flow dynamics in combustion engines by integrating machine learning (ML) techniques with computational fluid dynamics (CFD). The research focuses on predicting and optimising critical pre-combustion parameters, such as turbulence kinetic energy (TKE) and tumble-y, which are pivotal for enhancing the air-fuel mixing during the intake and compression phases. Three ML models, Random Forest Regression (RFR), Gaussian Process Regression (GPR), and Neural Networks (NN), are evaluated for their predictive capabilities. The GPR model outperforms the others, demonstrating superior accuracy and reduced uncertainty, as highlighted by metrics such as Mean Absolute Error (MAE), Mean Squared Error (MSE), Pearson Coefficient (PC), and R-squared (R2). Additionally, the ML-based approach achieves a remarkable 21.6x speedup compared with traditional CFD solvers, significantly reducing the computational costs while maintaining high fidelity in capturing momentum and thermal characteristics. The optimization results underscore the critical role of TKE and tumble-y in creating favourable conditions for efficient combustion. For instance, as demonstrated in Design #1 (TKE = 396.56 J/kg, Tumble-y = -0.1535, Temp. = 846.42 K, Pres. = 1.52 bar) and Design #2 (TKE = 366.77 J/kg, Tumble-y = -0.1535, Temp. = 549.59 K, Pres. = 2.81 bar), higher TKE and optimized tumble-y values enhance air motion dynamics, promoting better fuel-air mixing and thermal performance. The rigorous assessment of optimization results using the Euclidean distance as a fitness function validates the reliability of the predictions and highlights the potential of ML models for efficient, scalable and cost-effective design exploration. Therefore, the present work provides a beneficial relationship between CFD simulation and experimental findings on cold flow dynamics and how these might play a leading role in precombustion process. Results provide a frame-shifting pathway toward optimization of engine design for the improvement of thermal efficiency, and meeting sustainability targets.

The development of ammonia-methane (NH3-CH4) combustion as a hydrogen-carrier energy source faces major challenges such as significant NOx emissions, hindering its practical implementation. This paper examines how ethanol (C2H6O) and methanol (CH4O) additives influence formation pathways of NOx using ReaxFF molecular dynamics (MD) simulations at temperatures of 2,000 K and 3,000 K. Ten carefully designed fuel mixtures (C1-C10) were evaluated across 0

Three-dimensional (3D) computational fluid dynamics (CFD) simulations have been carried out to investigate the complex interaction of a planar shock wave (Ma = 1.22) with a cylindrical bubble. The unsteady Reynolds-averaged Navier–Stokes (URANS) approach with a level set coupled with volume of fluid (LSVOF) method has been applied in the present study. The predicted velocities of refracted wave, transmitted wave, upstream interface, downstream interface, jet, and vortex filaments are in very good agreement with the experimental data. The predicted non-dimensional bubble and vortex velocities also have great concordance with the experimental data compared with a simple model of shock-induced Rayleigh–Taylor instability (i.e., Richtmyer–Meshkov instability) and other theoretical models. The simulated changes in the bubble shape and size (length and width) against time agree very well with the experimental results. Comprehensive flow analysis has shown the shock–bubble interaction (SBI) process clearly from the onset of bubble compression up to the formation of vortex filaments, especially elucidating the mechanism on the air–jet formation and its development. It is demonstrated for the first time that turbulence is generated at the early phase of the shock cylindrical bubble interaction process, with the maximum turbulence intensity reaching about 20% around the vortex filament regions at the later phase of the interaction process.