Pilot Flame Research Articles

Experimental investigation of flow field features and spark ignition process in a multi-swirl airblast injector

Lean combustion has been applied in aero gas turbine engines in response to rising concerns regarding the effects of pollutant emissions on the environment. As one of low emissions combustion technologies, stratified partially premixed combustion is capable of reducing NOx emissions while still providing robust pilot flames. In this work, the processes of liquid fuel atomization, fuel–air mixing and spark ignition in a multi-swirl staged combustor with airblast atomizer are experimentally investigated to further illustrate the effects of flow field features on flame propagation and the underlying reasons leading to successful and failed ignitions (misfires). A trajectory processing algorithm is developed to identify the detailed flame propagation behaviors in successful and misfire scenarios. PIV, PMie and High-speed imaging technologies are applied to provide the velocity distribution, droplets distribution and time-resolved flame images, respectively. Evolution of projected flames shows that the mechanism of flame motions in the early phase following spark can be explained by non-reacting flow structures and spray distributions. It is demonstrated that the early phase propagating trajectories play an important role in determining overall successful ignition. The variation regarding the flow field, droplets distribution, and flame evolution under pressure drops varying from 1% to 6% are analyzed. The ignition and combustion reach a relatively stable state when the pressure drop of injector exceeds 4%, which reflects in the minimum LLO FAR varies around 0.021 ± 0.001 and the time of flame development stabilizes at 40 ± 5 ms.

The effectiveness of a novel coal-igniting-fuel technology and application in a direct current burner

As alternative energy sources increase the proportion of intermittent energy supplies in power grids, coal-fired power plants have come to play an important role in flexible peak power regulation. The burner is the core of the pulverized-coal boiler, and so its performance greatly affects the ability to achieve stable combustion at low boiler loads and also determines the load adjustment range. A novel coal-igniting-fuel technology using intermediate frequency induction heating has been proposed as a means of obtaining better combustion stability at low and variable loads. This method involves the combustion of a small quantity of coal particles in the ignitor to form a central pilot flame that subsequently ignites the surrounding pulverized coal in the main burner to achieve stable combustion. The present work utilized numerical simulations and full-scale experiments to systematically investigate the coal ignition and combustion characteristics associated with this technology. The numerical results show that coal particles inside the ignitor are rapidly heated and ignited, after which the resulting flame is gathered and accelerated by the flame extension tube, thereby forming a pilot flame near the ignitor outlet. When applied to a direct current (DC) burner, this ignitor is able to ignite the primary coal/air flow in a timely manner to form a second high-temperature zone near the burner nozzle. Compared with the absence of an ignitor, the flame center area appears roughly 1.6 m ahead. In an experiment using a single ignitor, it is found that the induction heating coil can quickly heat coal particles, thereby generating a stable high-temperature heat source within 80 s. The application of this ignitor to a DC burner allows the rapid start of the burner at a low initial ambient temperature, and the load can be adjusted over wide ranges of 17.6% – 100% and 13.6% – 100% when burning Shenhua coal and Blended coal, respectively. The degree of coal burnout is found to gradually improve as the thermal load of the burner is increased, reaching a maximum of approximately 93% at full load.

Combustion performance in a cavity-based combustor under subatmospheric pressure

Combustion performance and flame stabilization mechanism of the flameholder under the subatmospheric pressure are the foundation for the design of high-altitude operating aircraft. An experimental cavity-based combustor was developed to explore the combustion performance in low pressure. Four low pressures of 0.03, 0.04, 0.05 and 0.06 MPa were validated to investigate the effect of inlet pressure on the flame stability limits, lean ignition and blowout process, combustion efficiency, and outlet temperature profiles. A unique lean spark ignition process was discovered at 0.03 MPa that the flame downstream of the V-type flameholder promotes the formation of the piloted flame in the cavity. Results indicate that the pressure reduction decreases the spray flame propagation, subsequently weakening flame stability and combustion efficiency. With the inlet pressure decreases, the pilot fuel injected into the cavity still maintains effective combustion, while the declining combustion is achieved in the mainstream fuel. This phenomenon suggests that increasing the pilot fuel proportion in the combustion chamber will enhance the combustion in ultra-low pressure.

Machine learning tabulation of thermochemistry in turbulent combustion: An approach based on hybrid flamelet/random data and multiple multilayer perceptrons

A new machine learning methodology is proposed for speeding up thermochemistry computations in simulations of turbulent combustion. The approach is suited to a range of methods including Direct Numerical Simulation (DNS), Probability Density Function (PDF) methods, unsteady flamelet, Conditional Moment Closure (CMC), Multiple Mapping Closure (MMC), Linear Eddy Model (LEM), Thickened Flame Model, the Partially Stirred Reactor (PaSR) method (as in OpenFOAM) and the computation of laminar flames. In these methods, the chemical source term must be evaluated at every time step, and is often the most expensive element of a simulation. The proposed methodology has two main objectives: to offer enhanced capacity for generalisation and to improve the accuracy of the ANN prediction. To accomplish the first objective, we propose a hybrid flamelet/random data (HFRD) method for generating the training set. The random element endows the resulting ANNs with increased capacity for generalisation. Regarding the second objective, a multiple multilayer perceptron (MMP) approach is developed where different multilayer perceptrons (MLPs) are trained to predict states that result in smaller or larger composition changes, as these states feature different dynamics. It is shown that the multiple MLP method can greatly reduce the prediction error, especially for states yielding small composition changes. The approach is used to simulate flamelets of varying strain rates, one-dimensional premixed flames with differential diffusion and varying equivalence ratio, and finally the Large Eddy Simulation (LES) of CH4/air piloted flames Sandia D, E and F, which feature different levels of local extinction. The simulation results show very good agreement with those obtained from direct integration, while the range of problems simulated indicates that the approach has great capacity for generalisation. Finally, a speed-up ratio of 12 is attained for the reaction step.

OH, PAH, and sooting imaging in piloted liquid-spray flames of diesel and diesel surrogate

Comparison of the flame structure, polycyclic aromatic hydrocarbons (PAH), and soot formation characteristics of commercial diesel and a diesel surrogate is investigated using planar laser-induced fluorescence (PLIF) of OH and PAH at 283 nm, and laser-induced incandescence (LII) of soot at 1064 nm. The diesel and diesel surrogate flames present a similar overall appearance, with a “dark core” region of fine droplets exiting the injector and a light blue outer flame near the base. Hot soot regions appear downstream, indicated by bright yellowish flame luminosity. Comparison of simultaneously acquired single-laser-shot PLIF images of OH and PAH indicates that smaller PAH reside mostly in the central regime in the ~20–60% height of the flame, sandwiched between the flame fronts. In contrast, larger PAH and soot particles are formed further downstream. Soot-LII signals are present only in the upper ~40% height of the flame for both the fuels. The distribution of mean PAH-PLIF and soot-LII signals along the flame height is also found to be similar for both fuels, except diesel is shown to produce approximately 1.5× more soot than the diesel surrogate. Based on the analysis of probability distributions, the maximum OH-PLIF signals in both fuels have similar radial and axial spatial distributions, while maximum PAH-PLIF signals occur at a higher axial location for the diesel surrogate. However, for the diesel surrogate, maximum soot-LII signals are frequently appeared at lower heights. Therefore, exploration of more appropriate diesel surrogates may be necessary to closely mimic the PAH and soot formation behaviors of diesel fuel. The present study shows that the combined optical diagnostic techniques and the statistical analysis can be used for studying commercial diesel and diesel surrogates, and thus establish a robust framework for detailed studies of chemical kinetics, combustion dynamics, and soot formation in liquid hydrocarbon spray combustion.

CFD modeling of pyrolysis oil combustion using finite rate chemistry

This paper reports the first Computational Fluid Dynamics (CFD) model developed for biomass pyrolysis oil spray combustion using Finite-Rate Chemistry (FRC) approach. To make the CFD calculations feasible, a reduced mechanism for modeling the combustion of biomass Fast Pyrolysis Oil (FPO) based on the POLIMI 1412 mechanism and a model for eugenol oxidation was developed. The reduced mechanism consisted of 200 reactions and 71 species. This level of complexity was found to be a good tradeoff between predictive power and computational cost such that the reduced model could be used in CFD modeling. The predictive power of the reduced mechanism was demonstrated via 0D (adiabatic, premixed, constant pressure reactor), 1D (laminar counterflow flame) and 3D (CFD of a methane-air flat-flame piloted FPO spray flame) calculations. Results from CFD were compared against experimental data from non-intrusive optical diagnostics. The reduced model was successfully used in CFD calculations—the computational cost was approximately 2 orders of magnitude higher than that of a simplified model. Using the reduced mechanism, the concentration of pollutants, minor combustion products, and flame radicals could be predicted—this is added capability compared to already existing models. The CFD model using the reduced mechanism showed quantitative predictive power for major combustion products, flame temperature, some pollutants and temperature, and qualitative predictive power for flame radicals and soot.

Effect of pilot hydrogen on the formation of dynamic flame in an ethylene-fueled scramjet with a cavity

Pilot hydrogen is used for the ignition of gaseous ethylene to improve combustion efficiency at low flight Mach numbers. The process of dynamic flame stabilization has been investigated experimentally in a scramjet equipped with a cavity by using a variety of fuel injection strategies involving the injection of pure hydrogen, pure ethylene, and combinations of fuels to understand the effects of pilot hydrogen on supersonic combustion. This study reports experiments conducted in a direct-connected facility that simulated flight conditions at a Mach number of 4, a total temperature of 953 K, and a total pressure of 0.82 MPa. The data were collected by using hydroxyl planar laser-induced fluorescence (OH-PLIF), methylene (CH), luminosity images, and 10-kHz static pressure transducers. The results show that the internal flow oscillated at a dominant frequency of approximately 450 Hz without fuel injection. Oscillation in internal flow was suppressed when pilot hydrogen was injected into the cavity, whereas the influence of other injection strategies was minor. The OH-PLIF images indicate that the pilot flame was unstable, with different locations and occupying different areas. The normalized pressure was apparently higher when pilot hydrogen was injected ahead of the cavity step. The pure ethylene flame was difficult to stabilize without the application of pilot hydrogen. Once the pilot hydrogen had been turned off, the ethylene flame moved toward the cavity ramp and its area expanded. Thus, pilot hydrogen is a precondition for the stabilization of the ethylene flame.

Influence of angled dispersion gas on coaxial atomization, spray and flame formation in the context of spray-flame synthesis of nanoparticles

Liquid atomization determines the initial conditions for flame formation and particle synthesis. Without a stable flame, high droplet velocities and thus short droplet residence time in the flame may lead to droplets being injected into an extinguished flame, which influences synthesis and final particle output. An experimental investigation of spray formation and flame stability is performed through high-speed visualization. Targeted variation of nozzle geometry is applied to improve spray-flame interaction and compared to a standardized burner. Timescales of spray density and flame fluctuations are quantified and compared, where the latter were significantly larger and hence not correlated. Instead, dispersion gas forms a barrier between spray phase and pilot flame; hence, ignition depends on large liquid lumps with high radial momentum to break through the dispersion gas for spray ignition. Angling of dispersion gas flow increases radial shear and turbulence and leads to refined atomization and improved flame stability. To investigate the nozzle influence on particle formation, particle characteristics are examined by online and offline analytics with focus on particle structures and product purity. The modified nozzle produced smaller primary particle sizes, thus indicating a sensitivity of sintering dominance on the nozzle geometry. Impurities impact the examination of particle structures and general particle functionality. Carbon contamination was apparent in synthesized particles and also indicated sensitivity to nozzle geometry. Discrepancies to literature data are discussed regarding differences in flame activity and droplet characteristics. The report highlights, how product characteristics can differ crucially due to changes in nozzle geometry despite comparable operating conditions.Graphic abstract

PIV/PLIF investigation of unsteady turbulent flow and mixing behind a model gas turbine combustor

The present paper reports on the investigation of turbulent transport and fuel mixing under nonreacting conditions for a model gas-turbine swirl burner based on a design by Turbomeca. Two regimes of fuel (methane with acetone vapor) injection are compared. Specifically, the injection between the vanes of the radial swirler organizes lean, well-premixed combustion in a primary zone. The fuel supply as a central jet from the swirler’s centre-body provides a pilot flame. A combination of stereoscopic PIV and acetone PLIF systems is used to measure the velocity and concentration fields in the flows with a Reynolds number of 3 × 104. The effect of the central jet density is tested by replacing methane with air and neon. The data are processed using the POD method to extract coherent flow structures and quantify large-scale variations in concentration produced by them. In both cases, the flow dynamics were associated with the movement of large-scale vortex structures in the inner and outer mixing layers. The coherent flow fluctuations provide a substantial contribution to the turbulent transport, which locally exceeds 60% for the Reynolds shear stresses and Reynolds fluxes. The mixing is analysed based on the local probability of the fuel concentration pulsations. The probability distributions for the coherent and stochastic components of the pulsations are considered separately. For the premixed operation regime, the variation in the local equivalence ratio at the nozzle exit of the swirl burner reached 35% with the contribution of coherent pulsations up to 7%. In contrast, the large-scale vortex structures for the diffusion pilot jet flame primarily contributed to the advection of the fuel but not mixing. This paper provides relevant data for validating numerical methods for simulating unsteady turbulent mixing in swirl combustors.

Flame stability analysis of flame spray pyrolysis by artificial intelligence

Flame spray pyrolysis (FSP) is a process used to synthesize nanoparticles through the combustion of an atomized precursor solution; this process has applications in catalysts, battery materials, and pigments. Current limitations revolve around understanding how to consistently achieve a stable flame and the reliable production of nanoparticles. Machine learning and artificial intelligence algorithms that detect unstable flame conditions in real time may be a means of streamlining the synthesis process and improving FSP efficiency. In this study, the FSP flame stability is first quantified by analyzing the brightness of the flame’s anchor point. This analysis is then used to label data for both unsupervised and supervised machine learning approaches. The unsupervised learning approach allows for autonomous labeling and classification of new data by representing data in a reduced dimensional space and identifying combinations of features that most effectively cluster it. The supervised learning approach, on the other hand, requires human labeling of training and test data but is able to classify multiple objects of interest (such as the burner and pilot flames) within the video feed. The accuracy of each of these techniques is compared against the evaluations of human experts. Both the unsupervised and supervised approaches can track and classify FSP flame conditions in real time to alert users of unstable flame conditions. This research has the potential to autonomously track and manage flame spray pyrolysis as well as other flame technologies by monitoring and classifying the flame stability.

Multi-physics modeling of the ignition of polymer matrix composites exposed to fire

Advanced polymer matrix composites are widely used in modern aircraft and as such have to withstand fire to comply with safety regulations. The thermal degradation of these lightweight materials is a complex process involving chemical reactions in the solid and gas phases, potentially leading to flaming combustion. Here, we numerically investigate the heat feedback from the ignition of outgassing at the surface of a structural polymer matrix composite exposed to a pilot flame. Our model couples the thermal, physical and chemical processes in the solid phase to compute the composite degradation, as well as the production and movements of pyrolysates. A counterflow diffusion flame is used to model the ignition of the combustible effluent close to the sample surface, where the hot jet of the pilot flame impinges. The predictive capabilities of the method are demonstrated considering a carbon/epoxy composite subjected to a small-scale fire test. As input data for our model, material properties have been measured for virgin and partially burned samples via Differential Scanning Calorimetry, Xenon Flash Analysis and Thermogravimetric Analysis. The model is able to predict the backface temperature and time-to-ignition of the composite sample exposed to a flame with a good degree of accuracy.

Experimental Investigation on the Effects of the Geometry of the Pilot Burner on Main Flame

Various kinds of pilot burners were experimentally investigated to examine the effects of their geometry and their location relative to the main burner of a real size combustor. In addition, a wide range of fuel equivalence ratios were investigated to analyze the feasibility of the novel pilot burner for the conventional burner application. From the results, it is shown that the novel pilot burner with multi air holes had a thin, straight, long and stable pilot flame, while the conventional pilot burner had a thick, lifted, short and unstable flame. It is also shown that the novel pilot burner with an upper air flow hole had a straight pilot flame which led to less thermal damage to the burner combustor. This study suggests that not only pilot burner flame shape but also the vertical location of the pilot burner from the main burner combustor has a significant effect on combustor durability.

A continuous flow liquid propellant strand burner for high pressure monopropellant and bipropellant combustion studies

A combustion assembly capable of continuously burning monopropellant and bipropellant liquid fuels at pressures up to 80 bars (1145 psig) was designed and constructed. The assembly is based on a liquid propellant strand burner where a manifold maintains small positive differential pressures on the fuel to maintain a steady supply into the reaction vessel. Optical ports enable direct visualization of the flame and will allow for future spectroscopic and imaging studies of the flame. The strand burner design was tested using nitromethane with both air and inert environments in the reaction vessel. Continuous combustion was sustained for almost 8min in air (34 bars/500 psig) and more than 6min in N2 (70 bars/1000 psig). A unique outcome from the initial testing of this device is the ability to ignite liquid nitromethane in an inert environment without the use of a pilot flame started in air.

Effects of pressure on soot production in piloted turbulent non-premixed jet flames

Laser-induced incandescence (LII) was used to quantify the soot volume fraction in five piloted turbulent non-premixed C2H4-N2 jet flames at elevated pressures. In one series of flames, the bulk jet velocity was held constant as the pressure was increased from 1 bar (Re = 10,000) to 3 bar (Re = 30,000), and then 5 bar (Re = 50,000). In the other series, a Reynolds number of 10,000 was held constant at 1 bar, 3 bar, and 5 bar. LII measurements were calibrated with laminar diffusion flames at each pressure using 2D diffuse line-of-sight attenuation. Mean and RMS soot volume fractions along the flame centerline exhibit Gaussian profiles for all of the flames. The magnitude of soot volume fraction and the spatial soot distribution are enhanced by increases in pressure. In both series of flames, the peak mean soot volume fraction scales with the pressure as p2.2. In the constant velocity series, the scaling drops to p1.5 if the volume-integrated mean soot volume fraction on a per fuel mass basis is considered instead. At 3 bar, a threefold increase in Re leads to a decrease in the mean soot volume fraction despite the decrease in soot intermittency. At 5 bar, a fivefold increase in Re results in soot intermittency near zero at the flame’s center and a slight increase in the mean soot volume fraction.

Effects of high-pressure methane jet on premixed ignited flame in constant-volume bomb

Direct injection (DI) technology imparts natural gas (NG) engines with high thermal efficiency and reduces emissions. Furthermore, NG directly jets into the pilot flame. The interaction between the NG jet flow and pilot flame has a significant influence on the initial ignition and flame stability. This study investigates the effects of a high-pressure methane jet on the pilot ignited flame, using a constant-volume bomb (CVB) test rig. The interval between the methane injection and the ignition determines the premixed flame and jet combustion. The high-pressure methane jet flow collides with acts on the pilot premixed flame, which results in two phenomena: 1) Flame quenching: the premixed flame is blown off by the jet; and 2) Jet flame: the methane jet is ignited by the premixed flame. In the jet flame, laminar and turbulent flames co-exist. The development of the jet flame occurs in three stages: Stage I—the no-jet stage, wherein a laminar flame appears; Stage II—the interaction stage, wherein the jet interacts with the premixed flame, which results in either flame quenching or stable jet flame; and Stage III—the stable stage, wherein the flame becomes stable.

Validation of an Eulerian Stochastic Fields Solver Coupled with Reaction\u2013Diffusion Manifolds on LES of Methane/Air Non-premixed Flames

The Eulerian stochastic fields (ESF) combustion model can be used in LES in order to evaluate the filtered density function to describe the process of turbulence–chemistry interaction. The method is typically computationally expensive, especially if detailed chemistry mechanisms involving hydrocarbons are used. In this work, expensive computations are avoided by coupling the ESF solver with a reduced chemistry model. The reaction–diffusion manifold (REDIM) is chosen for this purpose, consisting of a passive scalar and a suitable reaction progress variable. The latter allows the use of a constant parametrization matrix when projecting the ESF equations onto the manifold. The piloted flames Sandia D–E were selected for validation using a 2D-REDIM. The results show that the combined solver is able to correctly capture the flame behavior in the investigated sections, although local extinction is underestimated by the ESF close to the injection plate. Hydrogen concentrations are strongly influenced by the transport model selected within the REDIM tabulation. A total solver performance increase by a factor of 81% is observed, compared to a full chemistry ESF simulation with 19 species. An accurate prediction of flame F instead required the extension of the REDIM table to a third variable, the scalar dissipation rate.

Effect of air-assistant on ignition and flame-holding characteristics in a cavity-strut based combustor

For promoting the expansion of pilot flame in high-subsonic flows, radial flame-holders are commonly introduced to combine with the pilot flame-holder, which may inversely bring about issues to the pilot ignition and flame stability. To simultaneously obtain satisfactory ignition and flame-propagation performance, a novel air-assistant design is proposed for the cavity-strut based combustor. Furthermore, comparison experiments are conducted under ambient pressure to investigate the influence of air-assistant on combustor characteristics. The discrepancies are experimentally explored in terms of lean ignition, blowout, and combustion characteristics in three cavity-based combustors accompanied with numerical flow fields. The results indicate that the air-assistant design can widen the ignitable space for ignition, extend the flammable Mach number, and decrease the lean ignition and blowout fuel/air ratio by 18.6% and 17.9% respectively. The air-assistant method can also accelerate the kernel generation and flamelet growth and suppress the flame quenching in the cavity-strut structure, which enhances the flame stability in either strut-stabilizing or cavity-stabilizing mechanism. The introduction of air-assistant regulates the flame structure and eventually increases the temperature rise of combustor by 250–350 K within the full range of operating fuel/air ratio. Moreover, the temperature distribution altered by air-assisting is more suitable for engineering combustors with higher requirements on circumferential flame propagation.

Smoke Emission Properties of Floor Covering Materials of Furnished Apartments in a Building.

The paper presents results of tests related to smoke optical density conducted on four various textile floor coverings for the needs of building interior design. Smoke emission is one of basic elements that characterize the fire environment. Consequently, the objective of the paper was to carry out a comparative analysis of smoke generation of chosen floor coverings for selected thermal exposures and in the presence or absence of a stimulus igniting the volatile gaseous phase (pilot flame). For the needs of our experimental research use was made of polypropylene, polyester, composite of wool, cotton, viscose and polyamide floor coverings. The highest value of the maximum specific optical density of smoke (494.7) was recorded for the floor covering consisting of 100% polypropylene (with higher fiber) under flameless combustion conditions (without the pilot flame). The polypropylene floor covering without underlay proved to be the best material from among all the tested ones with respect to smoke generating properties, and its samples offered the lowest value of optical density after 4 min for testing variants without the application of a pilot burner, with the flammable phase of decomposition products of this sample during the testing in which the burner was used to ignite at the latest. Experimental research has been carried out based on the standard ISO 5659–2:2017–08. The tests results were compared with international optical smoke density requirements for the interior design of ships and trains.

Anchoring of turbulent premixed hydrogen/air flames at externally heated walls

A joint experimental and numerical investigation of turbulent flame anchoring at externally heated walls is presented. The phenomenon has primarily been studied for laminar flames and micro-combustion while this study focuses on large-scale applications and elevated Reynolds number flows. Therefore, a novel burner design is developed and examined for a diverse set of operating conditions. Hydroxyl radical chemiluminescence measurements are employed to validate the numerical method. The numerical investigation evaluates the performance of various hydrogen/air kinetics, Reynolds-averaged turbulence models and the eddy dissipation concept (EDC) as a turbulence-chemistry interaction model. Simulation results show minor differences between detailed chemical mechanisms but pronounced deviations for a reduced kinetic. The baseline k-ω turbulence model is assessed to most accurately predict flame front position and shape. Universal applicability of EDC modeling constants is contradicted. Conclusively, the flame anchoring concept is considered a promising approach for pilot flames in continuous combustion devices.

Deep Residual Networks for Flamelet/progress Variable Tabulation with Application to a Piloted Flame with Inhomogeneous Inlet

ABSTRACT In this work, a deep neural network is presented which is trained on flamelet/progress variable (FPV) tables and validated in a combustion large eddy simulation (LES) of the Sydney/Sandia flame with inhomogeneous inlets. Using data scaling and transformation techniques, as well as novel network architectures as the residual skip connection we are able to store all combustion relevant quantities in one single network, thus reducing effectively the memory footprint compared to the FPV tables, while keeping data retrieval times similar to table interpolation. The accuracy of the deep neural networks (DNN) is compared to its FPV counterpart and achieves excellent agreement. The DNN is also able to accurately predict the stiff progress variable source term in the thin reaction zone. Finally, the DNN thermochemistry representation is validated in combustion simulations of a 2D laminar premixed flame and a 3D LES of the Sydney/Sandia piloted jet flame with inhomogeneous inlet. The DNN results are in very good agreement with the conventional tabulated FPV simulations and show a promising way to efficiently reduce the storage size of high dimensional pretabulated thermochemical state space in reactive flow simulations. Abbreviations: ANN: Artificial neural network; API: Application programming interface; CFD: Computational fluid dynamics; CFL: Courant-Friedrichs-Lewy; CPU: Central processing unit; DL: Deep learning; DNN: Deep neural network; FGM: Flamelet generated manifold; FLOPS: Floating point operation per second; FPV: Flamelet/progress variable; GPU: Graphics processing unit; HPC: High-performance computing; ILDM: Intrinsic low-dimensional manifold; LES: Large eddy simulation; MAE: Mean absolute error; MPI: Message passing interface; MSE: Mean squared error; PDF: Probability density function; SGS: Subgrid scale; SOM: Self-organizing map; TCI: Turbulence–chemistry interaction.