Search

The transition to low-carbon energy systems has heightened interest in hydrogen and ammonia as sustainable alternatives to traditional hydrocarbon fuels. However, the development and operation of combustors utilizing these fuels, like other combustion systems, are challenged by thermoacoustic instabilities arising from the interaction between unsteady heat release and acoustic wave oscillations. Among many different methods for studying thermoacoustic instabilities, thermoacoustic network models have played an important role in analyzing the essential dynamics of these instabilities in combustors operating with low-carbon fuels. This paper provides a comprehensive review of thermoacoustic network modeling techniques, focusing specifically on their application to hydrogen- and ammonia-based combustion systems. We outline the key mathematical frameworks derived from fundamental equations of motion, along with experimental validations and practical applications documented in existing studies. Furthermore, current research gaps are identified, and future directions are proposed to improve the reliability and effectiveness of thermoacoustic network models, contributing to the advancement of efficient and stable low-carbon combustors. [ABSTRACT FROM AUTHOR]

In this work, we explore the combustion mechanism of single micron-sized aluminum particles using a numerical model. The Burcat database and Catoire mechanism is considered as the thermodynamic data and the kinetic mechanism for our numerical model of aluminum combustion. Two independent experiments, including particle temperature profiles, ignition delay and burning time, are selected to evaluate the performance of the numerical model. The model shows great agreement for all considered properties. A parametric study is further conducted to identify the effect of involved physical parameters on the combustion process. The diffusion coefficient (D) of oxidizers and the activation energy of surface kinetics (Esurf) and evaporation coefficient (α) of aluminum impact the particle temperature the most. Burning time is most sensitive to the activation energy of surface kinetics (Esurf). The optical measurement in a solid propellant combustion indicates that the contact angle of the oxide cap on Al particle is between 10° and 20°. It is found that the selection of contact angle of the oxide cap significantly impacts the prediction of combustion time and residual of active aluminum. The current work highlights the importance of physical properties on the prediction of Al combustion, suggesting that more detailed evaluation from experiments and theory is encouraged.

Oxygen-enriched combustion is increasingly recognized as a viable approach for clean energy production and carbon capture, offering substantial benefits for boosting combustion efficiency and mitigating pollutant emissions, which makes it widely adopted in various industrial applications. Liquefied petroleum gas (LPG), predominantly consisting of propane (C3H8), is commonly utilized in numerous combustion systems, yet its emissions of soot particulates have raised considerable environmental concerns. This study delves into the combustion dynamics and soot formation behavior of propane, the principal component of LPG, under oxy-fuel combustion conditions, with the inclusion of H2O and CO2, utilizing both experimental techniques and numerical simulations. The results reveal that CO2 and H2O suppress soot formation through distinct mechanisms. CO2 decreases soot nucleation and surface growth by lowering flame temperature and H atom concentration, but it minimally enhances soot oxidation. H2O significantly reduces soot formation by chemically increasing OH radical concentration, thereby enhancing soot oxidation. A detailed decoupling analysis further shows that CO2's influence is predominantly thermal and chemical, resulting in lower OH levels and an elongated flame shape. In contrast, H2O's substantial thermal and chemical effects decrease flame height and promote soot reduction. These insights advance the understanding of soot formation control in oxy-fuel combustion, offering strategies to optimize combustion efficiency and minimize environmental impact. [ABSTRACT FROM AUTHOR]

The term "hybrid draft biomass cookstove" (HDBC) refers to a stove that employs both natural and forced drafts for its primary and secondary air supply. The HDBC optimization analysis is carried out in this work using computational fluid dynamics (CFD), considering the effect of forced draft secondary air mass flow rates to achieve the optimum (lower emissions and higher efficiency) HDBC performance. The analysis includes combustion and soot modeling to evaluate the thermal and emission performance of the stove. The CFD results are validated against the experimental data obtained from the water boiling test (WBT 4.2.3). The extremely low or high secondary air mass flow rate degrades the stove's performance; hence, its optimization is essential. The optimal mass flow rate identified in this investigation was 0.017 kg/s, at which the thermal and emission performances of HDBC were evaluated. An eddy dissipation model is employed to simulate biomass combustion, considering the reaction kinetics of wood volatiles reacting with oxygen, and a one-step soot model is used to predict soot formation. The combustion model agreed well with the experimental data for efficiency and temperature, with an average deviance of 10% and 20%, respectively. However, it significantly overestimated CO emissions by 81% and achieved Tier 4 performance in simulation as opposed to Tier 0 in the experiment. The PM2.5 emissions were under predicted by the soot model, but both the CFD and the experiment results achieved Tier 4 performance as per the World Health Organization (WHO) standards. Hence, a one-step soot model was found to be effective in predicting soot formation. [ABSTRACT FROM AUTHOR]

The filtered tabulated chemistry (FTACLES) approach utilizes data from pre-tabulated explicitly filtered 1D flame profiles for closure of the LES-filtered transport terms. Different methodologies are discussed to obtain a suitable progress variable c from detailed chemistry calculations of a methane/air flame. In this context, special focus is placed on the analytical modeling of the reaction source term using series of parameterized Gaussians. For increasing effective filter sizes in LES (i.e. including the flame thickening) the precise shape of the reaction rate profile becomes less and less relevant. In particular, it is shown that for one-step chemistry, a single Gaussian is sufficient to derive an explicitly expressible 1D flame profile with a prescribed laminar flame speed and thermal flame thickness. The resulting artificial flame profile is shown to have similarities with profiles based on carbon chemistry and detailed reaction mechanisms. Next, the behavior of the filtered c-transport equation is analyzed and several possible closure methods are compared for a wide range of filter widths. It is shown that the unclosed contribution of the filtered diffusion term can be combined with the subgrid convection term, thus simplifying the FTACLES formulation. The model is implemented in OpenFOAM and validated in 1D for a variety of LES filter sizes in combination with artificial flame thickening. A power-law-based wrinkling model is modified for use with artificial flame thickening and combined with the FTACLES model to enable 3D simulations of a premixed turbulent Bunsen burner. The comparison of 3D Large Eddy Bunsen flame simulations at increasing levels of turbulence intensity shows a good match to experimental results for most investigated cases. In addition, the results are mostly insensitive to the variation of the mesh size. [ABSTRACT FROM AUTHOR]

The need to refuel future missions to Mars and the Moon via in situ resource utilization (ISRU) requires the development of LOX/LCH4 engines, which are complex and expensive to develop and improve. This paper discusses how the use of digital engineering—specifically physics-based modeling (PBM)—can aid in developing, testing, and validating a LOX/LCH4 engine. The model, which focuses on propulsion performance and heat transfer through the engine walls, was created using Siemens' STAR-CCM+ CFD tool. Key features of the model include Eulerian multiphase physics (EMP), complex chemistry (CC) using the eddy dissipation concept (EDC), and segregated solid energy (SSE) for heat transfer. A comparison between the complete GRI 3.0 and Lu's reduced combustion mechanisms was performed, with Lu's mechanism being chosen for its cost-effectiveness and similar output to the GRI mechanism. The model's geometry represents 1/8th of the engine's volume, with a symmetric rotational boundary. The performance of this engine was investigated using NASA's chemical equilibrium analysis (CEA) and STAR-CCM+ simulations, focusing on thrust levels of 125 lbf and 500 lbf. Discrepancies between theoretical predictions and simulations ranged from 1.4% to 28.5%, largely due to differences in modeling assumptions. While NASA CEA has a zero-dimensional, steady-state approach based on idealized conditions, STAR-CCM+ accounts for real-world factors such as multiphase flow, turbulence, and heat loss. For the 125 lbf case, a 9.2% deviation in combustion chamber temperature and a 15.0% difference in thrust were noted, with simulations yielding 113.48 lbf compared to the CEA's 133.52 lbf. In the 500 lbf case, thrust reached 488 lbf, showing a 2.4% deviation from the design target and an 8.6% increase over CEA predictions. Temperature and pressure deviations were also observed, with the highest engine wall temperature at the nozzle throat. Monte Carlo simulations revealed that substituting LNG for LCH4 affects combustion dynamics. The findings emphasize the need for advanced modeling approaches to enhance the prediction accuracy of rocket engine performance, aiding in the development of digital twins for the CROME. [ABSTRACT FROM AUTHOR]

High-speed microjet hydrogen–air diffusion flames are investigated computationally. The focus is on the prediction of the so-called bottleneck phenomenon. The latter has been previously observed as a specific feature of the present flame class and has not yet been investigated computationally. In the configuration under consideration, the nozzle diameter is 0.5 mm and six cases with mean nozzle injection velocities (U) between 306 m/s and 561 m/s are considered. The flow in the nozzle lance is analyzed separately to obtain detailed inlet boundary conditions for the flame calculations. It is confirmed by calculation that the phenomenon is mainly determined by the transition to turbulence in the initial parts of the free jet. The transitional turbulence proves to be the biggest challenge in predicting this class of flames, as the generally available turbulence and turbulent combustion models reach the limits of their validity in transitional flows. In a Reynolds-Averaged Numerical Simulation framework, the Shear Stress Transport model is found to perform better than alternative two-equation models and is used as the turbulence model. By neglecting the interactions between the turbulence and chemistry (no-model approach), it is possible to predict the morphology of the bottleneck flame and its dependence on U qualitatively. However, the position of the bottleneck is overpredicted for U < 561 m/s. The experimental flames in the considered U range are all attached to the nozzle. This is also predicted by the no-model approach. The Eddy Dissipation Concept (EDC) used as the turbulence combustion model predicts, however, lifted flames (with increasing lift-off height as U decreases). With the EDC, no bottleneck morphology is observed for U = 561 m/s. For lower U, the EDC results for the bottleneck position are generally closer to the measurements. It is demonstrated that accuracy in predicting the bottleneck position can be improved by ad hoc modifications of the turbulent viscosity. [ABSTRACT FROM AUTHOR]

The solution of reactive computational fluid dynamics (CFD) simulations is accelerated by the implementation of a hybrid central processing unit/graphics processing units (CPU/GPU) Finite Volume solver based on the operator‐splitting strategy, where the chemistry integration is treated independently of the flow solution. The integration of ordinary differential equations (ODEs) describing the finite‐rate chemical kinetics is solved by an adaptive multi‐block explicit solver on GPUs, while the load of the fluid solution is distributed on a multicore CPU algorithm. The resulting speed‐up for reactive CFD simulations is up to 10×$$ \times $$; the performance gain increases with the size of the mechanism. The proposed implementation is general and can be applied to any CFD problem where the governing equations for the fluid transport are coupled with an ODE system. Code validation is performed against reference solutions on a selection of test cases involving reacting flows. [ABSTRACT FROM AUTHOR]

Drop-in fuels for aviation gas-turbine engines have been introduced recently to mitigate global warming. Despite their similarity to the fossil fuel Jet A-1, their combustion in traditional combustors should be thoroughly analyzed to maintain engine health and low emissions. The paper introduces criteria for assessing the impact of the chemical composition of fuels on combustion in the DEGN 380 turbofan. Based on previous emission-test results, the power functions of carbon monoxide and its emission index were adopted as the model of combustion. Based on the general notation of chemical reactions leading to the production of CO in combustion, the regression coefficients were given a physical meaning by linking them with the parameters of the kinetic equations, i.e., the reaction rate constant of CO and CO2 formation expressed as exponential functions of combustor outlet temperature and the concentration of O2 in the exhaust gas, as well as stoichiometric combustion reactions. The obtained empirical functions show that, in the entire range of engine operating parameters, synthetic components affect the values of the rate constants of CO and CO2 formation. It can be explained by the change in activation energy determined for all chain-of-combustion reactions. The activation energy for the CO formation chain changes in the range between 8.5 kJ/mol for A0 and 24.7 kJ/mol for A30, while for the CO2 formation chain between 29.8 kJ/mol for A0 and 30.8 kJ/mol for A30. The reactivity coefficient lnαiCOACODCO changes between 2.29 for A0 and 6.44 for A30, while lnαiCO2ACO2DCO2 changes between 7.90 for A0 and 8.08 for A30. [ABSTRACT FROM AUTHOR]

The need to refuel future missions to Mars and the Moon via in situ resource utilization (ISRU) requires the development of LOX/LCH4 engines, which are complex and expensive to develop and improve. This paper discusses how the use of digital engineering—specifically physics-based modeling (PBM)—can aid in developing, testing, and validating a LOX/LCH4 engine. The model, which focuses on propulsion performance and heat transfer through the engine walls, was created using Siemens’ STAR-CCM+ CFD tool. Key features of the model include Eulerian multiphase physics (EMP), complex chemistry (CC) using the eddy dissipation concept (EDC), and segregated solid energy (SSE) for heat transfer. A comparison between the complete GRI 3.0 and Lu’s reduced combustion mechanisms was performed, with Lu’s mechanism being chosen for its cost-effectiveness and similar output to the GRI mechanism. The model’s geometry represents 1/8th of the engine’s volume, with a symmetric rotational boundary. The performance of this engine was investigated using NASA’s chemical equilibrium analysis (CEA) and STAR-CCM+ simulations, focusing on thrust levels of 125 lbf and 500 lbf. Discrepancies between theoretical predictions and simulations ranged from 1.4% to 28.5%, largely due to differences in modeling assumptions. While NASA CEA has a zero-dimensional, steady-state approach based on idealized conditions, STAR-CCM+ accounts for real-world factors such as multiphase flow, turbulence, and heat loss. For the 125 lbf case, a 9.2% deviation in combustion chamber temperature and a 15.0% difference in thrust were noted, with simulations yielding 113.48 lbf compared to the CEA’s 133.52 lbf. In the 500 lbf case, thrust reached 488 lbf, showing a 2.4% deviation from the design target and an 8.6% increase over CEA predictions. Temperature and pressure deviations were also observed, with the highest engine wall temperature at the nozzle throat. Monte Carlo simulations revealed that substituting LNG for LCH4 affects combustion dynamics. The findings emphasize the need for advanced modeling approaches to enhance the prediction accuracy of rocket engine performance, aiding in the development of digital twins for the CROME.

Manifold-based models are an efficient modeling framework for turbulent combustion but, in their basic formulation, do not account for the compressibility effects of high-speed flows. To include the effects of compressibility, ad hoc corrections have been proposed but result in an inconsistent thermodynamic state between the manifold and flow simulation. In this work, an iterative algorithm to consistently incorporate compressibility effects into manifold-based models is developed. The manifold inputs (fuel and oxidizer temperatures and pressure) are determined iteratively to reflect the nonnegligible variations in thermodynamic state (expressed in terms of density and internal energy in flow simulations) that are characteristic of supersonic combustion. The algorithm is demonstrated on data from simulations of high-speed reacting mixing layers and is significantly more accurate than established approaches that only partially couple the manifold in compressible flow simulations. The proposed approach eliminates partial coupling approximation errors in excess of 10 and 20% for temperature and water source term. [ABSTRACT FROM AUTHOR]

In this study, the effects of intake port injection of methane and direct of injection diesel on emissions and the combustion of a light-duty RCCI engine were numerically researched. In this way, AVL Boost software was used for 1-dimensional simulation of the combustion process and emission estimation. Higher octane number methane gas was mixed with air from the intake port, while lower octane number diesel fuel was injected directly into the combustion chamber during the compression stroke. Methane gas was injected at a rate of 65% and diesel fuel at a rate of 35%. The diesel injected directly into the combustion chamber was sprayed at a rate of 5% between 0 and 35 °CA with 8 different injection timings after the main injection. In the model engine, 50% high EGR rate was applied in all combustion modes. The results showed that these parameters have significant effects on performance and emissions. The results in summary; NOx emission reduced at all engine speeds with delayed diesel post-injection timing. The maximum drop of NOx emission was 57.34% with Post15. Although the addition of methane slightly increased the soot emission, it was significantly reduced by the simultaneous addition of methane and application of post-injection strategies of diesel fuel. The soot emission was reduced by 58.85% with the Post35 injection strategy.

In the last few years, car manufacturers are moving toward electrified powertrains, like Battery Electric Vehicles (BEV) and Hybrid Electric Vehicles (HEV), to reduce the levels of CO2 in the atmosphere. A promising alternative to BEVs to reduce CO2 emissions, while maintaining an existing and well-developed technology, could be the use of hydrogen for internal combustion engines (ICEs). Particularly, the zero-carbon content of hydrogen allows for guaranteeing very clean combustion and near-zero carbon footprint. Within this context, the present study aims to deeply investigate the effects of pure hydrogen fueling of Spark Ignition (SI) engines, especially on specific fuel consumption and NOx emissions. Particularly, the benefits of pure hydrogen fueling were assessed in a production small Direct Injection (DI) SI engine. The engine is intended to operate in steady state conditions as an Auxiliary Power Unit (APU) for a series hybrid powertrain. The effects of pure hydrogen fueling were investigated at two engine speeds, namely 2000 and 3000 rpm. All tests were carried out at unthrottled conditions and lean mixture. A 1-D engine model was developed within a commercial software. Combustion, heat transfer, and NOx emissions embedded sub-models were tuned based on the experimental data, to accurately reproduce the engine behavior in a wide range of operating conditions. Both experimental and simulation results demonstrated that hydrogen fueling, together with lean mixture operation, allow for significantly improving engine thermal efficiency. The results of 1-D model simulations supported the energy management strategies of the hybrid powertrain in selecting the most suitable ICE operating condition. Particularly, the best engine operating conditions in terms of fuel consumption are achieved at 2500 rpm, with λ = 2.0, while the lowest NOx emissions are reached when λ is increased up to 2.4. [ABSTRACT FROM AUTHOR]

During the last years, the sales of vehicles equipped with Spark-Ignition (SI) engines have increased, to the detriment of those powered by Compression-Ignition (CI) engines. However, SI engines provide lower efficiency than CI engines, due to the low compression ratio and stoichiometric mixture operation. A way to increase the efficiency of SI engines is lean combustion, as it allows to increase the specific heats ratio (γ) and reduces pumping losses at part loads. Nevertheless, the operation with extremely lean mixtures deals with ignition issues, promoting Cycle-to-Cycle Variability (CCV) and increasing the probability of misfire. The prechamber ignition concept, also known as Turbulent Jet Ignition (TJI), is a promising solution for enabling the implementation of lean combustion, without its drawback in SI engines. Such a concept can be implemented according to two approaches: In active TJI, there is an additional fuel supply system inside the prechamber, while in passive TJI there isn't. Therefore, the main advantage of passive TJI is its simplicity, as the prechamber can be installed into a conventional spark plug body, with obvious benefits in terms of packaging and costs. In this work, the benefits of passive TJI on combustion and performance are assessed by simulation analyses. Particularly, a 1-D engine model was developed to simulate the TJI combustion and was validated versus experimental data. Afterward, a 0-D method was applied to assess the impact of the relative air-fuel ratio on CCV. The analysis was carried out in a high speed and load operating condition, namely 4500 rpm and 13 bar of IMEP, under both stoichiometric and lean mixture. Experimental and numerical results demonstrate the effectiveness of the passive TJI concept in promoting faster and more stable combustion also in lean-burn conditions. [ABSTRACT FROM AUTHOR]

Simulations are presented for a generic, round supersonic combustor. Turbulence is modeled in the combustor using a wall-modeled large eddy simulation approach. Combustion is modeled using a small quasi-global mechanism and a more detailed skeletal mechanism. Both mechanisms are used in conjunction with two variations of the partially-stirred reactor model for sub-grid turbulence chemistry interactions. Sensitivity of the solutions to grid resolution is investigated. It is found that in order to achieve reasonable grid convergence in the mean wall pressure, the model constant that appears in the partially-stirred reactor model must be a function of both the chemistry mechanism and the grid resolution. Most of the combinations of mechanism and turbulent combustion model tested can be tuned in order to predict the location of the pre-combustion shock train and the peak mean pressure in the combustor. It is found that while the different models are able to reproduce the mean wall pressure, there are significant differences in the mean temperature and heat release rate fields. The sensitivity of the different combinations of mechanisms and partially-stirred reactor formulation is quantified and some combinations are found to be more prone to blowout. Two of the tuned models were tested across several fuel equivalence ratios with a single value of the partially-stirred reactor model constant. One model produced reasonable predictions of shock location and peak mean pressure for each equivalence ratio. The second model captured the global trends in the mean wall pressure, but was unable to quantitatively predict the shock location and peak mean pressure for all equivalence ratios tested. [ABSTRACT FROM AUTHOR]

Modern diagnostic tools in turbulent combustion allow for highly-resolved measurements of reacting flows; however, they tend to generate massive data-sets, rendering conventional analysis intractable and inefficient. To alleviate this problem, machine learning tools may be used to, for example, discover features from the data for downstream modeling and prediction tasks. To this end, this work applies generative adversarial networks (GANs) to generate realistic flame images based on a time-resolved data set of hydroxide concentration snapshots obtained from planar laser induced fluorescence measurements of a model combustor. The generative model is able to generate flames in attached, lifted, and intermediate configurations dictated by the user. Using k-means clustering and proper orthogonal decomposition, the synthetic image set produced by the GAN is shown to be visually similar to the real image set, with recirculation zones and burned/unburned regions clearly present, indicating good GAN performance in capturing the experimental data statistical structure. Combined with techniques for controlling the configuration of generated flames, this work opens new avenues towards tractable statistical analysis and modeling of flame behavior, as well as rapid and inexpensive flame data generation.

The automotive sector is greatly contributing to pollutant emissions and recent regulations introduced the need for a major control of, and reduction of, internal combustion engine emissions. Artificial intelligence (AI) algorithms have proven to hold the potential to be the thrust in the state-of-the-art for engine-out emission prediction, thus enabling tailored calibration modes and control solutions. More specifically, the scientific literature has recently witnessed strong efforts in AI applications for the development of nitrogen oxides (NOx) virtual sensors. These latter replace physical sensors and exploit AI algorithms to estimate NOx concentrations in real-time. Still, the calibration of the algorithms, together with the appropriate choice of the specific metric, strongly affects the prediction capability. In the present paper, a machine learning-based virtual sensor for NOx monitoring in diesel engines was developed, based on the Extreme Gradient Boosting (XGBoost) machine learning algorithm. The latter is commonly used in the literature to deploy virtual sensors due to its high performance, flexibility and robustness. An experimental campaign was carried out to collect data from the engine test bench, as well as from the engine electronic control unit (ECU), for the development and calibration of the virtual sensor at steady-state conditions. The virtual sensor has, since then, been tested throughout on an on-road driving mission to assess its prediction performance in dynamic conditions. In stationary conditions, its prediction accuracy was around 98%, whereas it was 85% in transient conditions. The present study shows that AI-based virtual sensors have the potential to significantly improve the accuracy and reliability of NOx monitoring in diesel engines, and can, therefore, play a key role in reducing NOx emissions and improving air quality. [ABSTRACT FROM AUTHOR]

In this study, the effects of intake port injection of methane and direct of injection diesel on emissions and the combustion of a light-duty RCCI engine were numerically researched. In this way, AVL Boost software was used for 1-dimensional simulation of the combustion process and emission estimation. Higher octane number methane gas was mixed with air from the intake port, while lower octane number diesel fuel was injected directly into the combustion chamber during the compression stroke. Methane gas was injected at a rate of 65% and diesel fuel at a rate of 35%. The diesel injected directly into the combustion chamber was sprayed at a rate of 5% between 0 and 35 °CA with 8 different injection timings after the main injection. In the model engine, 50% high EGR rate was applied in all combustion modes. The results showed that these parameters have significant effects on performance and emissions. The results in summary; NOx emission reduced at all engine speeds with delayed diesel post-injection timing. The maximum drop of NOx emission was 57.34% with Post15. Although the addition of methane slightly increased the soot emission, it was significantly reduced by the simultaneous addition of methane and application of post-injection strategies of diesel fuel. The soot emission was reduced by 58.85% with the Post35 injection strategy. [ABSTRACT FROM AUTHOR]

Techniques such as exhaust gas recirculation (EGR) and water-in-fuel emulsions (WFEs) can significantly decrease NOx emissions in diesel engines. As a disadvantage of adopting EGR, the afterburning period lengthens owing to a shortage of oxygen, lowering thermal efficiency. Meanwhile, WFEs can slightly reduce NOx emissions and reduce the afterburning phase without severely compromising thermal efficiency. Therefore, the EGR–WFE combination was modeled utilizing the KIVA-3V code along with GT power and experimental results. The findings indicated that combining EGR with WFEs is an efficient technique to reduce afterburning and enhance thermal efficiency. Under the EGR state, the NO product was evenly lowered. In the WFE, a considerable NO amount was created near the front edge of the combustion flame. Additionally, squish flow from the piston's up–down movement improved fuel–air mixing, and NO production was increased as a result, particularly at high injection pressure. Using WFEs with EGR at a low oxygen concentration significantly reduced NO emissions while increasing thermal efficiency. For instance, using 16% of the oxygen concentration and a 40% water emulsion, a 94% drop in NO and a 4% improvement in the Indicated Mean Effective Pressure were obtained concurrently. This research proposes using the EGR–WFE combination to minimize NO emissions while maintaining thermal efficiency. [ABSTRACT FROM AUTHOR]

Computational Fluid Dynamics (CFD) modeling of gasoline spark-ignited engine combustion has been extensively discussed using both detailed chemistry mechanisms (e.g., SAGE) and flamelet models (e.g., the G -equation). The models have been extensively validated under normal operating conditions; however, few studies have discussed the capability of these models in capturing DISI combustion under cold-start conditions. A cold-start differs from normal operating conditions in various respects, such as (1) having highly retarded spark timing to help generate a high heat flux in the exhaust for a rapid catalyst light-off; (2) having split-injection strategies to ensure a favorable stratification at the vicinity of the spark plug and reduced film formation; and (3) having optimized valve timings for reduced NOx emissions via increased internal residuals and reduced hydrocarbon (HC) emissions via prolonged oxidation of the combustion products. The retarded spark timing introduces the adverse effect of a decaying turbulence field, which results in a reduced turbulent flame speed. The analysis of all these factors happening inside the cylinder appears complicated at first glance; however, it could be made possible by efficient use of the existing CFD models. The current study explored the capability of the SAGE detailed chemistry model in capturing cold-start flame travel in a DISI engine. The results were then compared against the G -equation-based GLR model, which has been validated for excellent predictions of the DISI cold-start combustion as shown by Ravindran et al. The flame travel was captured on a Borghi-Peters diagram to find that the flame travels through corrugated, wrinkled, and laminar regimes. In order to fully evaluate the capability of the detailed chemistry model in predicting such changing turbulence-chemistry interactions, it will need to be studied individually in each regime; however, the scope of the current paper is limited to the study of the model behavior in the laminar regime, which will be shown to be important for DISI engine cold-start. The SAGE detailed chemistry model, with a toluene reference fuel (TRF) mechanism validated for gasoline laminar flame speeds, was found to significantly under-predict the flame propagation speeds because of the effects of numerical viscosity and discrepancies in capturing molecular diffusion. The causes and effects of this under-prediction and the ways in which this can be improved are presented in the paper. [ABSTRACT FROM AUTHOR]

Estimation of combustion phasing and power production is essential to ensuring proper combustion and load control. However, archetypal control-oriented physics-based combustion models can become computationally expensive if highly accurate predictive capabilities are achieved. Artificial neural network (ANN) models, on the other hand, may provide superior predictive and computational capabilities. However, using classical ANNs for model-based prediction and control can be challenging, since their heuristic and deterministic black-box nature may make them intractable or create instabilities. In this paper, a hybridized modeling framework that leverages the advantages of both physics-based and stochastic neural network modeling approaches is utilized to capture CA50 (the timing when 50% of the fuel energy has been released) along with indicated mean effective pressure (IMEP). The performance of the hybridized framework is compared to a classical ANN and a physics-based-only framework in a stochastic environment. To ensure high robustness and low computational burden in the hybrid framework, the CA50 input parameters along with IMEP are captured with a Bayesian regularized ANN (BRANN) and then integrated into an overall physics-based 0D Wiebe model. The outputs of the hybridized CA50 and IMEP models are then successively fine-tuned with BRANN transfer learning models (TLMs). The study shows that in the presence of a Gaussian-distributed model uncertainty, the proposed hybridized model framework can achieve an RMSE of 1.3 × 10−5 CAD and 4.37 kPa with a 45.4 and 3.6 s total model runtime for CA50 and IMEP, respectively, for over 200 steady-state engine operating conditions. As such, this model framework may be a useful tool for real-time combustion control where in-cylinder feedback is limited.

Summary: In this study, a diesel engine combustion was modeled to estimate engine performance and emissions for the first time in the field of engine combustion. Four different algorithms including Grasshopper Optimization Algorithm, Ant Lion Optimizer, and Gray Wolf Optimization as well as Artificial Neural Network were employed to predict thermal efficiency, fuel consumption, CO, HC, and NOx emissions of a diesel engine fueled with diesel‐biodiesel blend emulsions containing water and cerium oxide nano additives. The models proposed were developed using two inputs (fuel type and engine load). The results showed that the Gray Wolf Optimization led to maximum coefficient correlation (0.9940 and 0.9966) and minimum Mean Square Error compared with the other employed algorithms for brake thermal efficiency and brake‐specific fuel consumption. The best results were obtained for Gray Wolf Optimization, Ant Lion Optimizer, and Grasshopper Optimization Algorithm, respectively. The same sequence was also found for estimating engine emissions. However, Gray Wolf Optimization showed slightly better result for estimation of engine emission than engine performance. In overall, the results of Gray Wolf Optimization were in perfect agreement with the experimental values compared to the other nature‐inspired algorithms as well as Artificial Neural Network in predicting fuel combustion. The model proposed can find application in fuel and engine manufacturers. [ABSTRACT FROM AUTHOR]

Thrust regulation is applied to maintain the performance of the liquid propellant rocket engine. The thrust level of a rocket engine can be readily controlled by adjusting the number of propellants introduced into the combustion chamber. In this study, a gas generator design is proposed in which thrust regulation is maintained by performing secondary combustion in the divergent section of the nozzle of a gas generator. Tangential and normal injection techniques have also been studied for better combustion analyses. A normal injection technique is used for the experiment and CFD results are validated with the experimental data. Chemical equilibrium analyses are also performed by minimizing Gibbs free energy with the steepest descent method augmented by the Nelder–Mead algorithm. These equilibrium calculations give the combustion species as obtained through the CFD results. Performance evaluation of the rocket engine, with and without secondary combustion in the gas generator, led to an increase of 42% thrust and 46.15% of specific impulse with secondary combustion in the gas generator. [ABSTRACT FROM AUTHOR]

Superior fuel economy, higher torque and durability have led to the diesel engine being widely used in a variety of fields of application, such as road transport, agricultural vehicles, earth moving machines and marine propulsion, as well as fixed installations for electrical power generation. However, diesel engines are plagued by high emissions of nitrogen oxides (NOx), particulate matter (PM) and carbon dioxide when conventional fuel is used. One possible solution is to use low-carbon gaseous fuel alongside diesel fuel by operating in a dual-fuel (DF) configuration, as this system provides a low implementation cost alternative for the improvement of combustion efficiency in the conventional diesel engine. An initial step in this direction involved the replacement of diesel fuel with natural gas. However, the consequent high levels of unburned hydrocarbons produced due to non-optimized engines led to a shift to carbon-free fuels, such as hydrogen. Hydrogen can be injected into the intake manifold, where it premixes with air, then the addition of a small amount of diesel fuel, auto-igniting easily, provides multiple ignition sources for the gas. To evaluate the efficiency and pollutant emissions in dual-fuel diesel-hydrogen combustion, a numerical CFD analysis was conducted and validated with the aid of experimental measurements on a research engine acquired at the test bench. The process of ignition of diesel fuel and flame propagation through a premixed air-hydrogen charge was represented the Autoignition-Induced Flame Propagation model included ANSYS-Forte software. Because of the inefficient operating conditions associated with the combustion, the methodology was significantly improved by evaluating the laminar flame speed as a function of pressure, temperature and equivalence ratio using Chemkin-Pro software. A numerical comparison was carried out among full hydrogen, full methane and different hydrogen-methane mixtures with the same energy input in each case. The use of full hydrogen was characterized by enhanced combustion, higher thermal efficiency and lower carbon emissions. However, the higher temperatures that occurred for hydrogen combustion led to higher NOx emissions. [ABSTRACT FROM AUTHOR]

To accurately model the Direct Injection Spark Ignition (DISI) combustion process, it is important to account for the effects of the spark energy discharge process. The proximity of the injected fuel spray and spark electrodes leads to steep gradients in local velocities and equivalence ratios, particularly under cold-start conditions when multiple injection strategies are employed. The variations in the local properties at the spark plug location play a significant role in the growth of the initial flame kernel established by the spark and its subsequent evolution into a turbulent flame. In the present work, an ignition model is presented that is compatible with the G-Equation combustion model, which responds to the effects of spark energy discharge and the associated plasma expansion effects. The model is referred to as the Plasma Velocity on G-surface (PVG) model, and it uses the G-surface to capture the early kernel growth. The model derives its theory from the Discrete Particle Ignition (DPIK) model, which accounts for the effects of electrode heat transfer, spark energy, and chemical heat release from the fuel on the early flame kernel growth. The local turbulent flame speed has been calculated based on the instantaneous location of the flame kernel on the Borghi-Peters regime diagram. The model has been validated against the experimental measurements given by Maly and Vogel,1 and the constant volume flame growth measurements provided by Nwagwe et al.2 Multi-cycle simulations were performed in CONVERGE3 using the PVG ignition model in combination with the G-Equation-based GLR4 model in a RANS framework to capture the combustion characteristics of a DISI engine. Good agreements with the experimental pressure trace and apparent heat-release rates were obtained. Additionally, the PVG ignition model was observed to substantially reduce the sensitivity of the default G-sourcing ignition method employed by CONVERGE. [ABSTRACT FROM AUTHOR]

This paper provides a numerical study on n-dodecane flames using Large-Eddy Simulations (LES) along with the Flamelet Generated Manifold (FGM) method for combustion modeling. The computational setup follows the Engine Combustion Network Spray A operating condition, which consists of a single-hole spray injection into a constant volume vessel. Herein we propose a novel approach for the coupling of the energy equation with the FGM database for spray combustion simulations. Namely, the energy equation is solved in terms of the sensible enthalpy, while the heat of combustion is calculated from the FGM database. This approach decreases the computational cost of the simulation because it does not require a precise computation of the entire composition of the mixture. The flamelet database is generated by simulating a series of counterflow diffusion flames with two popular chemical kinetics mechanisms for n-dodecane. Further, the secondary breakup of the droplet is taken into account by a recently developed modified version of the Taylor Analogy Breakup model. The numerical results show that the proposed methodology captures accurately the main characteristics of the reacting spray, such as mixture formation, ignition delay time, and flame lift-off. Additionally, it captures the "cool flame" between the flame lift-off and the injection nozzle. Overall, the simulations show differences between the two kinetics mechanisms regarding the ignition characteristics, while similar flame structures are observed once the flame is stabilised at the lift-off distance. [ABSTRACT FROM AUTHOR]

Research in the field of Internal Combustion (IC) engines focuses on the drastic reduction of both pollutant and greenhouse gas emissions. A promising alternative to gasoline and diesel fuel is represented by the use of gaseous fuels, above all green hydrogen but also Natural Gas (NG). In previous works, the authors investigated the performance, efficiency, and emissions of a supercharged Spark Ignition (SI) engine fueled with mixtures of gasoline and natural gas; a detailed research involving the combustion process of this kind of fuel mixture has been previously performed and a lot of experimental data have been collected. Combustion modeling is a fundamental tool in the design and optimization process of an IC engine. A simple way to simulate the combustion evolution is to implement a mathematical function that reproduces the mass fraction burned (MFB) profile; the most used for this purpose is the Wiebe function. In a previous work, the authors proposed an innovative mathematical model, the Hill function, that allowed a better interpolation of experimental MFB profiles when compared to the Wiebe function. In the research work presented here, both the traditional Wiebe and the innovative Hill function have been calibrated using experimental MFB profiles obtained from a supercharged SI engine fueled with mixtures of gasoline and natural gas in different proportions; the two calibrated functions have been implemented in a zero-dimensional (0-D) SI engine model and compared in terms of both Indicated Mean Effective Pressure (IMEP) and cyclic pressure variation prediction reliability. It was found that the Hill function allows a better IMEP prediction for all the operating conditions tested (several engine speeds, supercharging pressures, and fuel mixtures), with a maximum prediction error of 2.7% compared to 4.3% of the Wiebe function. A further analysis was also performed regarding the cyclic pressure variation that affects all the IC engines during combustion and may lead to irregular engine operation; in this case, the Hill function proved to better predict the cyclic pressure variation with respect to the Wiebe function. [ABSTRACT FROM AUTHOR]

Numerical simulations of combustion processes in rocket engines requires a long run time of supercomputer systems even for a very short physical time. Therefore, creating digital twins of rocket engines needs enormous processor time, and is not very effective. This computational time surpasses the actual physical time of the process in many orders of magnitude. To speed up numerical simulations the paper presents a solution of the chemical kinetics problem using artificial neural network approach. Using the architecture of a multilayer neural network with bypass connections, namely residual network, it is possible to obtain a fast and reliable solution to the problem. The neural network is trained to predict the state of the system only one time step ahead. Using it in a recursive mode, it is possible to forecast for thousands of steps without loss of accuracy. • Simulation of a chemical kinetic problem by the neural network method was carried out. • The network was trained on a wide range of data. • A shortened neural network with preservation of accuracy was obtained. • The considered network used a small amount of RAM (Random Access Memory). [ABSTRACT FROM AUTHOR]

This work analyses, experimentally and numerically, the combustion behavior of three aviation fuels: a standard Jet A-1, a high aromatic content fuel, and an isoparaffinic Alcohol to Jet (ATJ) fuel. The goal is to demonstrate the ability of a chemical kinetic model to capture the chemistry underlying the combustion behavior of a wide range of jet fuels, starting from compositional information. Real fuels containing up to hundreds of components are modeled as surrogates containing less than 10 components, which represent the chemical functionalities of the real fuel. By using an in-house numerical optimizer, the fuel components and their relative quantities are selected, and a semi-detailed kinetic model (containing about 450 species) is used to simulate the formation of the main oxidation products and reaction intermediates. Calculations are compared with species profiles measured in a laminar flow reactor to validate the model and provide insights into the reactivity of the fuels. Finally, starting from the results, general observations on the strengths and limits of the approach are provided, highlighting areas where further investigations are required.

Combustion science is an interdisciplinary study that involves nonlinear physical and chemical phenomena in time and length scales, including complex chemical reactions and fluid flows. Combustion widely supplies energy for powering vehicles, heating houses, generating electricity, cooking food, etc. The key to study combustion is to improve the combustion efficiency with minimum emission of pollutants. Machine learning facilitates data-driven techniques for handling large amounts of combustion data, either obtained through experiments or simulations under multiple spatiotemporal scales, thereby finding the hidden patterns underlying these data and promoting combustion research. This work presents an overview of studies on the applications of machine learning in combustion science fields over the past several decades. We introduce the fundamentals of machine learning and its usage in aiding chemical reactions, combustion modeling, combustion measurement, engine performance prediction and optimization, and fuel design. The opportunities and limitations of using machine learning in combustion studies are also discussed. This paper aims to provide readers with a portrait of what and how machine learning can be used in combustion research and to inspire researchers in their ongoing studies. Machine learning techniques are rapidly advancing in this era of big data, and there is high potential for exploring the combination between machine learning and combustion research and achieving remarkable results.

Estimation of combustion phasing and power production is essential to ensuring proper combustion and load control. However, archetypal control-oriented physics-based combustion models can become computationally expensive if highly accurate predictive capabilities are achieved. Artificial neural network (ANN) models, on the other hand, may provide superior predictive and computational capabilities. However, using classical ANNs for model-based prediction and control can be challenging, since their heuristic and deterministic black-box nature may make them intractable or create instabilities. In this paper, a hybridized modeling framework that leverages the advantages of both physics-based and stochastic neural network modeling approaches is utilized to capture CA50 (the timing when 50% of the fuel energy has been released) along with indicated mean effective pressure (IMEP). The performance of the hybridized framework is compared to a classical ANN and a physics-based-only framework in a stochastic environment. To ensure high robustness and low computational burden in the hybrid framework, the CA50 input parameters along with IMEP are captured with a Bayesian regularized ANN (BRANN) and then integrated into an overall physics-based 0D Wiebe model. The outputs of the hybridized CA50 and IMEP models are then successively fine-tuned with BRANN transfer learning models (TLMs). The study shows that in the presence of a Gaussian-distributed model uncertainty, the proposed hybridized model framework can achieve an RMSE of 1.3 × 10−5 CAD and 4.37 kPa with a 45.4 and 3.6 s total model runtime for CA50 and IMEP, respectively, for over 200 steady-state engine operating conditions. As such, this model framework may be a useful tool for real-time combustion control where in-cylinder feedback is limited. [ABSTRACT FROM AUTHOR]

In this work, the effect of shear on turbulent lean premixed flames under stratified burning conditions including the flame TSF-A-r and TSF-D-r of the Darmstadt stratified burner, respectively, is numerically investigated. For this purpose, the large eddy simulation (LES) technique, coupled with tabulated chemistry and an artificially thickening approach, is applied. First, numerical predictions of time averaged data are compared with available experimental data in order to establish the prediction capability of the numerical approach. Then, the simulation results are exploited to analyze and characterize the stratified mixing layers and their interaction with the flame. In particular, coherent patterns in the mixing layers and their influence on the flame are pointed out. Finally, the onset and intensity of stratification of the two test cases are evaluated and compared in terms of equivalence ratio gradient, strength of the mixing and reaction layers, and scalar dissipation rates. It turns out that (1) a higher shear tends to attenuate the level of stratification, (2) the mixing and reaction layers correlate earlier and stronger with increasing turbulence, (3) a dominant premixed mode can rather be promoted with a stronger turbulence as can be observed from the ratio of the dissipation rate of the mixture fraction and that of the reaction progress variable. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Central South University is the property of Springer Nature and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Thrust regulation is applied to maintain the performance of the liquid propellant rocket engine. The thrust level of a rocket engine can be readily controlled by adjusting the number of propellants introduced into the combustion chamber. In this study, a gas generator design is proposed in which thrust regulation is maintained by performing secondary combustion in the divergent section of the nozzle of a gas generator. Tangential and normal injection techniques have also been studied for better combustion analyses. A normal injection technique is used for the experiment and CFD results are validated with the experimental data. Chemical equilibrium analyses are also performed by minimizing Gibbs free energy with the steepest descent method augmented by the Nelder–Mead algorithm. These equilibrium calculations give the combustion species as obtained through the CFD results. Performance evaluation of the rocket engine, with and without secondary combustion in the gas generator, led to an increase of 42% thrust and 46.15% of specific impulse with secondary combustion in the gas generator.

Developing a profound understanding of the combustion characteristics of the cold-start phase of a Direct Injection Spark Ignition (DISI) engine is critical to meeting increasingly stringent emissions regulations. Computational Fluid Dynamics (CFD) modeling of gasoline DISI combustion under normal operating conditions has been discussed in detail using both the detailed chemistry approach and flamelet models (e.g. the G-Equation). However, there has been little discussion regarding the capability of the existing models to capture DISI combustion under cold-start conditions. Accurate predictions of cold-start behavior involves the efficient use of multiple models - spray modeling to capture the split injection strategies, models to capture the wall-film interactions, ignition modeling to capture the effects of retarded spark timings, combustion modeling to accurately capture the flame front propagation, and turbulence modeling to capture the effects of decaying turbulent kinetic energy. The retarded spark timing helps to generate high heat flux in the exhaust for the faster catalyst light-off during cold-start. However, the adverse effect is a reduced turbulent flame speed due to decaying turbulent kinetic energy. Accordingly, developing an understanding of the turbulence-chemistry interactions is imperative for accurate modeling of combustion under cold-start conditions. In the present work, combustion characteristics during the cold-start, fast-idle phase is modeled using the G-Equation flamelet model and the RANS turbulence model. The challenges associated with capturing the turbulent-chemistry interactions are explained by tracking the flame front travel along the Borghi-Peters regime diagram. In this study, a modified version of the G-Equation combustion model for capturing cold-start flame travel is presented. [ABSTRACT FROM AUTHOR]

This paper assesses the accuracy of partially-premixed turbulent combustion models based on the tabulation of chemical kinetics, under multi-injection diesel engine-relevant conditions. For this purpose, 2-D direct numerical simulation (DNS) is carried out. Pockets of gaseous n-heptane are randomly distributed in a turbulent field of a partially burnt n-heptane/air mixture. The burnt gases composition and enthalpy correspond to the partial oxidation of a pilot injection that precedes the main injection, represented here by the fresh fuel pockets. The DNS domain is enclosed in a larger volume, permitting quasi-constant pressure autoignition. Chemical kinetics is modeled by a 29-species skeletal reaction mechanism for n-heptane/air mixture autoignition and flame propagation. A homogeneous isotropic turbulence spectrum is used to initialize the velocity field in the domain. A DNS database is generated varying the progress of the pilot injection combustion c 0 and the velocity fluctuation level u ′ of the turbulence spectrum. Three different modeling approaches are tested a priori against the DNS data: (1) the tabulated homogeneous reactor, which is a direct exploitation of the chemistry tabulation ignoring any local mixture heterogeneity; (2) the presumed conditional moment model, which includes a separate statistical description for the mixture and the combustion progress; (3) the approximated diffusion flame model, which considers the heterogeneous turbulent reactor as a diffusion flame. Since the same chemical kinetics mechanism is used for the generation of the chemistry tabulation, the study is entirely focused on the evaluation of the different modeling assumptions. Results show that accounting for initial progress variable of the mixture ( c 0 ) is mandatory for such models. They also indicate the omission of mixture fraction Z and progress variable c heterogeneities and the assumption of the statistical independence of Z and c as the main responsible for model discrepancies under the studied conditions. [ABSTRACT FROM AUTHOR]

Combustion phenomena were discovered still in far ancient time of China. From the 50's of the last century, owing to the fast development of energy and power, aeronautical and astronautical, chemical and metallurgical engineering, combustion theory started to be studied in China. The Chinese scientists studied the theory of ignition, laminar flame propagation, droplet combustion, and spray combustion. Later, from the 80's of the last century, numerical modeling of combustion started to be studied in China, including turbulence modeling, turbulent combustion modeling, two-phase turbulence modeling and two-phase combustion modeling, in the approaches of Reynolds Navier–Stokes (RANS) modeling, large-eddy simulation (LES), and direct numerical simulation (DNS) of combustion. Due to the limitation of a paper size, this paper gives only a review of studies on theory and modeling of droplet and spray combustion in China. [ABSTRACT FROM AUTHOR]

This investigation outlines a data-assisted approach that employs random forest classifiers for local and dynamic submodel assignment in turbulent-combustion simulations. This method is demonstrated in simulations of a single-element GOX/GCH4 rocket combustor; a priori as well as a posteriori assessments are conducted to (i) evaluate the accuracy and adjustability of the classifier for targeting different quantities of interest (QoIs), and (ii) assess improvements, resulting from the data-assisted combustion model assignment, in predicting target QoIs during simulation runtime. Results from the a priori study show that random forests, trained with local flow properties as input variables and combustion model errors as training labels, assign three different combustion models – finite-rate chemistry (FRC), flamelet progress variable (FPV) model, and inert mixing (IM) – with reasonable classification performance even when targeting multiple QoIs. Applications in a posteriori studies demonstrate improved predictions from data-assisted simulations, in temperature and CO mass fraction, when compared with monolithic FPV calculations. An additional a posteriori data-assisted simulation of a modified configuration demonstrates that the present approach can be successfully applied to different configurations, as long as thermophysical behavior can be represented by the training data. These results demonstrate that this data-driven framework holds promise for dynamic combustion submodel assignments in reacting flow simulations. [ABSTRACT FROM AUTHOR]

The present paper provides a contribution to the CFD modelling of reacting flows in IC engines fueled with natural gas. Despite the fact that natural gas has been widely investigated into in the last decades, the literature still lacks reliable models and correlations to be exploited so as to efficiently support the design of internal combustion engines. The paper deals with the development of an accurate CFD model, capable of capturing the effects of the engine working conditions and mixture compositions on the combustion process. The CFD model is based on the Extended Coherent Flame Model combustion model coupled to a laminar flame speed one through a user subroutine, which replaces the commonly adopted empirical correlations. The flame speed values have been derived from the application of a reaction mechanism for natural gas-air-residual gases mixtures.In the second part of the paper, the model is validated and applied to the investigation of the dependence of the combustion quality on the fuel doping with hydrogen as well as on the mixture dilution with EGR. As a matter of fact, the attractiveness of the mixture dilution with EGR relies on the potential in containing engine-out NOx emissions as well as in reducing the pumping losses, thus further abating fuel consumption at part loads. Finally, the effects of fuel blending with H2 on the EGR tolerance is discussed in the paper.

Combustion simulations are the key aspect for full-scale 3D simulations of modern and perspective engines for aerospace propulsion. The present work studies an opportunity to solve chemical kinetics problems using artificial neural networks. Using classical numerical methods, sets of training data were produced. Choosing among different architectures of multi-layer neural networks and tuning their parameters, a rather simple model was worked out capable to solve this problem. The neural network obtained works in a recursive mode, and it can predict the behavior of chemical multi-species dynamical system by many steps. • The kinetic mechanism is resolved using the classical numerical method. • Training datasets received. • A neural network was selected and trained. • Recursive neural network can predict the behavior of chemical kinetics. [ABSTRACT FROM AUTHOR]

In this study, a novel furnace concept for in-flight particle spheroidization is presented, characterized and evaluated. A natural gas fired burner with a continued staged air principle and internal recirculation (COSTAIR), was used to provide the required temperature for the spheroidization process. Therefore, a numerically inexpensive CFD model for the calculation of combustion and multiphase flow is proposed. Particularly for the calculation of particle trajectories and particle peak temperatures of non-spherical (chiseled and flaky) slag particles, the presented CFD model differs in two modifications from the current state-of-the-art CFD models: first, a numerically efficient combustion model with a detailed chemical reaction mechanism was used in order to calculate the temperature profile of the furnace. While in most current state-of-the-art CFD models the numerically expensive and time consuming eddy dissipation concept (EDC) model or the insufficient eddy dissipation model (EDM) are used. Second, the discrete phase model, which is based on a numerically efficient Euler-Lagrangian approach, is used for multiphase modeling. Non-spherical particles are considered by application of a suitable particle drag model from literature. Although, the assumption of spherical particles is more common in current state-of-the-art CFD models for multiphase modeling. It was concluded that the presented furnace concept is applicable for the semi-industrial scale production of spherical boiler slag particles. The numerical results show that the assumption of non-spherical particles, compared to spherical particles, is more suitable for the calculation of particle trajectories and particle peak temperatures in the presented furnace. • A novel furnace concept for particle spheroidization is presented. • Time saving combustion modelling using a flamelet approach. • Characterization of temperature and flow conditions in the furnace. • Application of a numerically inexpensive Euler-Lagrangian approach. • Comparison between CFD and measurements. [ABSTRACT FROM AUTHOR]

The probabilty density function (PDF) of the mixture fraction is of integral importance to a large number of combustion models. Here, a novel modelling approach for the PDF of the mixture fraction is proposed which employs dissipation elements. While being restricted to the commonly used mean and variance of the mixture fraction, this model approach individually considers contributions of the laminar regions as well as the turbulent core and the turbulent/non-turbulent interface region. The later region poses a highly intermittent part of the flow which is of high relevance to the non-premixed combustion of pure hydrocarbon fuels. The model assumptions are justified by means of the gradient trajectory based analysis of high fidelity direct numerical simulation (DNS) datasets of two turbulent inert configurations and a turbulent non-premixed jet flame. The new dissipation element based model is validated against the DNS datasets and a comparison with the beta PDF is presented. [ABSTRACT FROM AUTHOR]