Flame Region Research Articles

Structural effects of C3 oxygenated fuels on soot formation in ethylene coflow diffusion flames

This paper investigates how the structure of three C3 oxygenated fuels: dimethyl carbonate (DMC), dimethoxymethane (DMM) and isopropanol (IPA) influences soot formation when the fuels are blended with ethylene in laminar coflow diffusion flames. Up to 20% of the total carbon was substituted with oxygenated fuel. Colour-ratio pyrometry was used to measure the soot volume fraction (fv). IPA caused a strong increase in fv, whereas DMM and DMC both caused an initial increase followed by a progressive decrease in fv as the proportion of oxygenated fuel was increased. Differential mobility spectrometry and thermocouple probes were used to measure the particle size distribution and gas temperature in the flames at 5% blend strength. The hottest region of the 5% flames was consistently about 100 K cooler than the corresponding region of the ethylene flame, indicating a thermal effect of the doping. The 5% flames showed an increase in the maximum centre-line average particle size and fv versus the ethylene flame, with the IPA showing the largest increase. The evolution of the centre-line particle size distributions showed that the 5% flames experienced earlier particle growth compared to the ethylene flame. Consideration of the role of the chemical pathways towards benzene formation suggests that methyl radicals from the decomposition of the oxygenated fuels are responsible for the increase in fv at 5% doping. The difference in fv between the IPA versus DMM and DMC flames is thought to be due to the additional presence of C3 species originating from the carbon-carbon bonded backbone of IPA. Meanwhile, the differences between the DMC versus DMM flames are thought to arise from CO2 produced during the decomposition of DMC, and a corresponding thermal effect where the pyrolysis region of the 5% DMC flame was observed to be about 50 K cooler than the other flames.

Implication of self-throttling on combustion performance in a strut-based scramjet combustor

The present work describes the numerical investigation on the implication of self-throttling on combustion performance of a strut-based scramjet combustor. Firstly, the mixing characteristics of a strut with and without self-throttling are compared and evaluated to portray the effect of self-throttling in a DLR based scramjet combustor. Later, a separate study is performed to investigate the influence of location of self-throttling on the combustion performance of a strut-based scramjet combustor. Our study reveals that the self-throttling increases the enthalpy of formation and the generates the strong shock waves which improves the mixing rate of fuel and air. Further, the mixing commences relatively earlier with reduced pressure losses when the self-throttling is employed. It is observed that the volume of core flame region and the thickness of a flame depends on the distance between the reflected shock and the self-throttling injection slot.Also, the combustion performance deteriorates with the increase in distance between the reflected shock and the injection slot of self-throttling. Further, the combustion and mixing efficiency is found to be optimalwhen self-throttling is employed at the distance of100 mm from the inlet of the combustor.

Prediction of combustion states from flame image in a domestic coal burner

Coal is still a strategic fuel for many developing countries. The environmental impact of emissions resulting from the widespread use of coal worldwide is a matter of serious debate. In this perspective, clean coal burning technologies are in demand. In this study, a measurement system that estimates emission from flame images in a domestic coal burner is proposed. The system consists of a charge-coupled device camera, image processing software (real time image acquisition, noise reduction and extracting features) and artificial intelligence elements (classification of features by neural networks). In feature extraction stage, only five flame region features (Gx, Gy , trace, L 2 and L ∞ norm) is extracted. Gcx and Gcy are the instantaneous change of the horizontal and vertical components of center mass of the flame image. These features are new concepts for emission estimation from the flame image. The proposed system makes a difference with its simpler structure and higher accuracy compared to its counterparts previously presented in the literature.

Numerical Investigation on the Asymmetric Flow Characteristics of Two Propane Fires of Unequal Heat Release Rate in Open Space

Multiple fires of unequal heat release rate (HRR) is a common fire scenario in real fire accidents. Differing from most previous research assumption of identical fire sources, the unequal HRR for those fire sources is regarded as more reasonable in reality. To explore the asymmetric flow characteristics surrounding multiple fires under such circumstances, simulations of two square propane burners with the same side length but different HRRs were carried out. The HRR combination and burner separation distance were varied. The results showed that the asymmetric flow characteristic was found in both the flame and the smoke plume zone. In the flame region, the flame morphology in terms of the tilt angle is a good parameter indicating the asymmetric flow characteristic. In general, the tilt angle of the small fire is larger than that of the big fire. The tilt angle of the small fire decreases while that of the big fire increases until equaling to each other under the HRR ratio reaching unity. In the smoke plume region, the smoke plumes from the small and the big fire will converge at a certain height with some shifting distance, which is another parameter indicating the asymmetric flow characteristic. Besides, the smoke plume merging process driven by asymmetric air entrainment can be divided into three stages, namely the I) Separate stage, II) Converging stage and III) Complete coalescence stage. Correlations of the converging & coalescence height suggesting the starting point of the II and III stages were proposed.

A deep learning framework for autonomous flame detection

This paper proposes a novel framework of flame region-based convolutional neural network for autonomous flame detection. The task of flame detection is especially challenging since flames have greater diversity in colour, texture, and shape than regular rigid objects. To cope with these difficulties due to the various appearances and unclear edges of flames, a proposal generation approach is developed to effectively select candidate flame regions based on two crucial properties of flames, i.e., their dynamics and colours. The candidate flame regions together with a convolutional feature map are further processed by additional layers to output detected flames. The diversity in flame colours is well represented by approximating the distribution using a Dirichlet Process Gaussian mixture model with variational inference. The proposed framework is evaluated on publicly available videos and achieves an average frame-wise accuracy higher than 88%, which outperforms the state-of-the-art methods.

Graphite fluoride as a new oxidizer to construct nano-Al based reactive material and its combustion performance

Reactive materials consisting of aluminum (Al) and fluorine-containing oxidizers have been extensively developed owing to high energy density and potential applications in propellants and explosives. Herein, graphite fluoride (CF) as a new fluorine-containing oxidizer was introduced to construct nano-Al based reactive materials. Systematic investigation and comparison were executed on exothermic reaction behavior and combustion performance of nano-Al/CF and nano-Al/polytetrafluoroethylene (PTFE) based reactive materials. The energy output and exothermic peak of nano-Al/CF were approximately 2367 J/g and 641 °C, which were 300 J/g and 85 °C higher than those of nano-Al/PTFE, respectively. The effects of nano-Al/CF mass ratio and packing density on burning rates were examined to fully understand combustion reaction and flame propagation behavior. Nano-Al/CF exhibited excellent burning rate (1670 m/s), much higher than that of nano-Al/PTFE (930 m/s). More interestingly, flame splitting phenomenon including three distinct flame regions was observed on the critical position with the highest burning rate for nano-Al/CF with the mass ratios of 25/70 and 30/75. The burning rate of nano-Al/CF increased and then decreased with the packing densities ranging from 20.87% to 53.47% TMD. The dominating combustion reaction products (AlF3 and C) were further determined to figure out the reaction mechanism. These results demonstrated that CF as fluorine-containing oxidizer to construct reactive materials has potential applications in propellants, explosives and pyrotechnics.

Effect of 2, 5-dimethylfuran concentration on micro-explosive combustion characteristics of biodiesel droplet

Micro-explosive combustion is considered as an effective way to improve combustion efficiency in combustor. In this work, the micro-explosive combustion characteristics of biodiesel and DMF blended droplet are investigated using high-speed backlit image technique. The droplet is suspended at the intersection of two quartz fiber and ignited by a pair of electrodes. Flame temperature and soot concentration are measured using two-color pyrometry method. Moreover, two new parameters are proposed to analysis droplet micro-explosion, namely droplet instability and breakup index. The results shown that the instability of DMF25BD75, DMF50BD50 and DMF75BD25 droplets is mainly caused by bubbles coalescence, internal circulation of bubbles, micro-explosion and vapor jetting. The low-temperature flame region is mainly located near the droplet surface and flame edge. However, the soot concentration is highest in these two regions. The combustion of blended droplet includes a micro-explosive combustion stage and a stable combustion stage. Four modes of secondary atomization is determined according to droplet breakup index. When the DMF concentration is 50% and 75%, the burning rate is 9% and 37.5% higher than diesel, which is expected to realize rapid and high-efficient combustion.

A Two-Stream CNN Model with Adaptive Adjustment of Receptive Field Dedicated to Flame Region Detection

Convolutional neural networks (CNN) have yielded state-of-the-art performance in image segmentation. Their application in video surveillance systems can provide very useful information for extinguishing fire in time. The current studies mostly focused on CNN-based flame image classification and have achieved good accuracy. However, the research of CNN-based flame region detection is extremely scarce due to the bulky network structures and high hardware configuration requirements of the state-of-the-art CNN models. Therefore, this paper presents a two-stream convolutional neural network for flame region detection (TSCNNFlame). TSCNNFlame is a lightweight CNN architecture including a spatial stream and temporal stream for detecting flame pixels in video sequences captured by fixed cameras. The static features from the spatial stream and dynamic features from the temporal stream are fused by three convolutional layers to reduce the false positives. We replace the convolutional layer of CNN with the selective kernel (SK)-Shuffle block constructed by integrating the SK convolution into the deep convolutional layer of ShuffleNet V2. The SKnet blocks can adaptively adjust the size of one receptive field with the proportion of one region of interest (ROI) in it. The grouped convolution used in Shufflenet solves the problem in which the multi-branch structure of SKnet causes the network parameters to double with the number of branches. Therefore, the CNN network dedicated to flame region detection balances the efficiency and accuracy by the lightweight architecture, the temporal–spatial features fusion, and the advantages of the SK-Shuffle block. The experimental results, which are evaluated by multiple metrics and are analyzed from many angles, show that this method can achieve significant performance while reducing the running time.

Numerical studies on swirling of internal fire whirls with experimental justifications

Numerical studies on internal fire whirls (IFW) generated in a vertical shaft model with a single corner gap were reported in this paper. The generation of IFW, burning rate of fuel and temperature were studied experimentally first. Numerical simulations on medium-scale IFW were carried out using a fully-coupled large eddy simulation incorporating subgrid scale turbulence and a fire source with heat release rates compiled from experimental results. Typical transient flame shape was studied and then simulated numerically by using temperature. The dynamic phenomena of generation and development of IFW were simulated and then compared with experimental results. The predicted results were validated by comparing with experimental data, which demonstrated that an IFW can be simulated by Computational Fluid Dynamics. Numerical results for flame surface, temperature, and flame length agreed well with the experimental results. The IFW flame region and intermittent region were longer than those for an ordinary pool fire. The modified empirical formula for centerline temperature was derived. Variations of vertical and tangential velocity in axial and radial directions were also shown. The vortex core radius was found to be determined by the fuel bed size. Velocity field was not measured extensively due to resources limitation. Comparing measured temperature distribution with predictions is acceptable because temperature is related to the heat release rate, air flow and pressure gradient.

Thermoelectric behavior of Al/PTFE reactive materials induced by temperature gradient

In order to reveal the physical effect of Al/PTFE (mass percentage of 50%/50%) heat conduction and deflagration reaction induced by temperature gradient under the condition of flame heating, the loading and testing systems of cookoff experiment were conducted to measure the thermal response and electrical output characteristics of Al/PTFE before and after deflagration induced by temperature gradient. The thermal shock wave and thermoelectromotive force before deflagration, voltage signal induced by flame plasma expansion after deflagration were measured simultaneously. Meanwhile, the flash signal and the evolutionary process of flame region and temperature after deflagration were obtained. The results show that the thermal shock wave was generated by emergent temperature change in the stage of heat conduction, and the thermoelectromotive force was generated by the superposition of temperature gradients in different regions. The thermoelectromotive force can be equivalent to the series connection of electric source in each striped region. Under the temperature gradient of 6-10 K/mm, the maximum thermoelectromotive force can reach 0.8 V. Before deflagration, the local phase transition of molten or gaseous state induced by thermal decomposition destroyed the stable distribution of the temperature gradient of the sample, which leads to the decrease of the thermoelectric force. Based on the electric output signal of flame plasma, the reaction was divided into two stages including deflagration stage (lasting for 2–3 s) and stable combustion stage (lasting for about 10s). The average temperature of flame first increased and then decreased abruptly. The maximum temperature was 1585 K, and the propagation rate of flame was about 81.6 m/s in the initial stage of deflagration; however, the average temperature of the flame decreased slowly in the stable combustion stage, there was no obvious electrical output and the flash intensity decreased.

Stability limits and NO emissions of premixed swirl ammonia-air flames enriched with hydrogen or methane at elevated pressures

This study reports measurements of stability limits and exhaust NO mole fractions of technically-premixed swirl ammonia-air flames enriched with either methane or hydrogen. Experiments were conducted at different pressures from atmospheric to 5 bar, representative of commercial micro gas turbines. The full range of ammonia fractions in the fuel blend, xNH3, was considered, from 0 (pure methane or hydrogen) to 1 (pure ammonia), covering very lean (φ = 0.25) to rich (φ = 1.60) equivalence ratios. Results show that increasing pressure widens the range of stable equivalence ratios for pure ammonia-air flames. Regardless of pressure, there is a critical ammonia fraction above which the range of stable equivalence ratios suddenly widens. This is because flashback does not occur anymore when the equivalence ratio is progressively increased towards stoichiometric and rich blowout occurs instead. This critical ammonia fraction increases with pressure and is larger for ammonia-hydrogen than for ammonia-methane. Provided that enough hydrogen is blended with ammonia (xNH3 < 0.9), flames with very lean equivalence ratios (φ < 0.7) can be stabilized and these yield competitively low NO emissions (<200 ppm), regardless of pressure. For this reason, very lean swirl ammonia-hydrogen-air flames are promising candidates for micro gas turbines. However, N2O emissions have the potential to be unacceptably large for these operating conditions if heat loss is too large or residence time is too short. As a consequence, the post flame region must be considered carefully. Due to the lower reactivity of methane compared to that of hydrogen, very lean swirl ammonia-methane-air flames could not be stabilized and good NO performance is limited to rich equivalence ratios for ammonia-methane fuel blends. The equivalence ratio above which good NO performance depends on pressure and bulk velocity.

Study of a Premixed Turbulent Counter-Flow Flame with a Large Eddy Simulation Method

The turbulent counter-flow flame (TCF) has proven to be a useful benchmark to study turbulence-chemistry interactions, however, the widely observed bulk flow fluctuations and their influence on the flame stability remain unclear. In the present work, premixed TCFs are studied numerically using a Large Eddy Simulation (LES) method. A transported probability density function (pdf) approach is adopted to simulate the sub-grid scale (sgs) turbulence-chemistry interactions. A solution to the joint sgs-pdf evolution equation for each of the relative scalars is obtained by the stochastic fields method. The chemistry is represented using a simplified chemical reaction mechanism containing 15 reaction steps and 19 species. This work compares results with two meshing strategies, with the domain inside nozzles included and excluded respectively. A conditional statistical approach is applied to filter out the large scale motions of the flame. With the use of digital turbulence, the velocity field in the flame region is well reproduced. The processes of local extinction and re-ignition are successfully captured and analysed together with the strain rate field, and local extinctions are found correlated to the turbulent structures in the reactant stream. The predicted probability of localised extinction is in good agreement with the measurements, and the influence of flame stoichiometry are also successfully reproduced. Overall, the current results serve to demonstrate the capability of the LES-pdf method in the study of the premixed opposed jet turbulent flames.

Carrier-phase DNS of detailed NOx formation in early-stage pulverized coal combustion with fuel-bound nitrogen

Carrier-phase direct numerical simulation of detailed NOx formation in pulverized coal flames (PCC) with fuel-bound nitrogen is conducted in a 3D temporally evolving mixing layer setup where Lagrangian particles (Colombian bituminous coal) in an air stream (upper half of the domain) mix with the products of lean volatile/air combustion in the lower stream. The release of fuel-N is represented by ammonia, hydrogen cyanide, and a lumped nitrogenated tar (pyridine). Devolatilization is modeled by fitting a 2-step pyrolysis approach to the detailed heterogeneous PoliMi kinetics. A comprehensive homogeneous mechanism including all standard pathways of NOx and pyridine oxidation is adopted. Results show a partition of NO in two distinct branches of scatter plots of NO mass fraction vs. volatile mixture fraction after flame establishment, corresponding to NO in the lower stream flame region and hot spots near the upper stream. The contribution of NO2, prompt, and thermal mechanisms to total NOx is limited in the early stages of PCC. The main source of NO is fuel-N, with NH being the most important precursor. Pyridine plays an important role for NO production in the upper stream through CN formed from CHCHCN. CN and ammonia oxidation have the highest contribution to NH production. Regarding NO destruction, NO reactions with HCCO, CHi and C through the reburn process constitute the largest share. NO conversion to N2O by NH followed by conversion of N2O to N2 and NO+N→N2+O are the two most important pathways directly reducing NO to N2.

WUUNet: Advanced Fully Convolutional Neural Network for Multiclass Fire Segmentation

This article describes an AI-based solution to multiclass fire segmentation. The flame contours are divided into red, yellow, and orange areas. This separation is necessary to identify the hottest regions for flame suppression. Flame objects can have a wide variety of shapes (convex and non-convex). In that case, the segmentation task is more applicable than object detection because the center of the fire is much more accurate and reliable information than the center of the bounding box and, therefore, can be used by robotics systems for aiming. The UNet model is used as a baseline for the initial solution because it is the best open-source convolutional neural network. There is no available open dataset for multiclass fire segmentation. Hence, a custom dataset was developed and used in the current study, including 6250 samples from 36 videos. We compared the trained UNet models with several configurations of input data. The first comparison is shown between the calculation schemes of fitting the frame to one window and obtaining non-intersected areas of sliding window over the input image. Secondarily, we chose the best main metric of the loss function (soft Dice and Jaccard). We addressed the problem of detecting flame regions at the boundaries of non-intersected regions, and introduced new combinational methods of obtaining output signal based on weighted summarization and Gaussian mixtures of half-intersected areas as a solution. In the final section, we present UUNet-concatenative and wUUNet models that demonstrate significant improvements in accuracy and are considered to be state-of-the-art. All models use the original UNet-backbone at the encoder layers (i.e., VGG16) to demonstrate the superiority of the proposed architectures. The results can be applied to many robotic firefighting systems.

Hydrogen Micro Jet Diffusion Combustion in a Coaxial Air Jet

In this paper the results of experimental studies of the combustion of a round microjet of hydrogen in a ring subsonic coaxial air jet are presented. It is shown that the combustion scenario for a subsonic hydrogen jet outflow is associated with the presence of a “bottleneck flame region” and nozzle heating, but the initial spherical shape of the flame region is transformed into a cylindrical one. It was found that the combustion of a round microjet of hydrogen in a satellite ring coaxial air jet at a supersonic outflow velocity is accompanied by the presence of supersonic cells.

Spontaneous Raman–LIF–CO–OH measurements of species concentration in turbulent spray flames

This paper presents new measurements of species concentrations, temperature and mixture fraction in selected regions of a turbulent ethanol spray flame. The line-Raman–LIF–COOH setup developed at the Sandia's Combustion Research Facility is utilised to probe regions of a spray flame where laser breakdown of liquid droplets is avoided and the remaining interferences can be corrected. The spray flame is stabilised on the piloted Sydney needle spray burner, where axial translation of the liquid injecting needle in the air-blast stream can transition the spray from dilute to dense. The solution to obtaining successful measurements is found to be multifaceted and includes: the appropriate selection of flame conditions; high sensitivity of the Raman detection system permitting reduced laser energies; development of a pre-processing algorithm to reject strong droplet interferences; and application of the hybrid matrix inversion method combined with wavelet denoising to account for interference corrections and noise at the very low signal levels obtained. Unique and necessary for the successful measurements reported in this paper, a pre-processing algorithm is outlined that removes data points corrupted with strong interferences from droplets. These interferences arise from a range of sources, but the most intense are due to the laser interaction with surrounding mist or liquid fragments, such that measurements near the jet centreline are corrupted and hence discarded. Reliable measurements of mixture fraction, temperature obtained from the sum of the species number densities, and species mole fractions are reported for regions in the flames sufficiently far from the centreline. The paper demonstrates the feasibility of the judicious use of Raman scattering in turbulent spray flames, the results of which will be extremely useful for validating numerical simulations.

Experimental study on the characteristics of buoyancy-driven turbulent flame generated by rectangular fire sources with different length–width ratios in an open space

Fires caused by fuel leakage pose serious hazards to the city’s environment and safety. Different leakage conditions will lead to different shapes of fire sources and produce different hazards to the surrounding environment. To study the characteristics of rectangular fires, this paper conducted experimental studies on rectangular burners with different length–width ratios (n) in an open space. Propane was used as fuel. This study provides for the first time both primary data and new relationships for the temperature and velocity distribution of the flame on the horizontal section at different heights. When the heat release rate is constant, the plume mass flow rate on the horizontal section at the same height increases with the increase of the n. The expressions of temperature and velocity along the rectangular fire centerline are proposed. The empirical equations for calculating the Gaussian half-width of the square and rectangular fire plumes are presented. Finally, combining the expressions above, the correlations of flame mass flow rates in continuous and intermittent flame regions are derived and the calculated values are compared with the experimental data. It is found that the error range is less than 15%. This work can be used to predict the combustion characteristics of buoyancy-driven turbulent flames formed by gaseous fuel. It plays a guiding role in the risk assessment of the gaseous fuel transportation systems. Experimental data and related conclusions also provide important reference data and theoretical support for computational fluid dynamics (CFD) models.

Effects of backwall on inner thermal structure in opposed-flow horizontal flame spread of thick PMMA panel

Flame spread on solid fuel is an important aspect of fire safety engineering. In this work, an unsteady numerical model was developed to investigate the opposed-flow horizontal flame spread mechanism of thick polymethyl methacrylate (PMMA) on backwall with different thermal properties. The temperature fields in both the condensed fuel and backwall material, the fuel gas distribution and the pyrolysis profiles along the fuel bed were analysed in detail. The results showed that the thermal properties of the backwall play an important role in the structure of the melting interface of the PMMA flame front. The molten layer will not penetrate the sample obviously for the case of gypsum backwall. However, the molten liquid penetrates the PMMA sample with more molten liquid generated on the aluminium backwall, thus resulting the higher possibility of molten droplets. It is also indicated that the heat accumulation at the preheated region was enhanced by the heat transfer through the aluminium backwall from the flame and pyrolysis region. On the other hand, the aluminium wall behind the flame tip normally remain at relatively high temperatures, which in turn improved the pyrolysis rate, and lead to higher gas temperature. Therefore, the flame spread of polymer would be accelerated obviously for the case of metal backwall. This work provides important information for the acceleration mechanism of metal backwall on the opposed-flow horizontal flame spread of polymer and the design of polymer components, and also provides an important basis for engineering application to evaluate the fire risk of materials by considering the thermal performance of the backwall.

Flame extension area and temperature profile of horizontal jet fire impinging on a vertical plate

Impinging jet fire of high exit momentum is frequently reported to induce the domino effect in industry fires. Less literature is available for the flame extension area and temperature distribution over a vertical plate impinged by the horizontal jet flame. Thus, this paper conducts a systematic experiment to reveal the effect of nozzle exit velocity, exit diameter and exit-plate spacing on the horizontally impinging jet fire. Experimental observation shows the evolution of flame pattern and color with an increase of exit velocity and a decrease of exit-plate spacing. The flame extension area induced by the horizontal jet flame impingement was measured by a novel method that combines the picture processing technique and the thermal imaging technique. By physical analysis, a new correlation coupling the turbulent Karlovitz stretch factor and the ratio of nozzle exit diameter to exit-plate spacing, was developed for the flame extension area of both horizontally and vertically impinging jet fire. The variation of maximum temperature in the impinging zone with the exit-plate spacing still fits the classic correlation of centerline temperature in the continuous and intermittent flames. However, the horizontally impinging jet fire plume holds a shorter intermittent flame region due to the flame upward bending. It is found that the temperature profile holds a big difference in the upward and downward directions along the vertical plate. A new uniform correlation, with the plume radius as the characteristic length scale, is proposed to well collapse all the temperature data in the impinging zone.

Experimental and kinetic modeling study of ammonia addition on PAH characteristics in premixed n-heptane flames

Due to concerns over pollutant emissions and ever-increasing energy crisis, ammonia is being considered as a promising alternative fuel to control soot particles and their precursors (polycyclic aromatic hydrocarbons, PAHs) emissions in combustion for diesel engines. This paper aims to investigate the chemical effects of NH3 addition on PAH formation in n-heptane laminar premixed flames. The laser-induced fluorescence (LIF) technique was used to measure the relative PAH fluorescence intensities qualitatively. A kinetic modeling study was performed to predict PAH concentrations by comparing with the experimental results, and then clarify the underlying effects of NH3 on PAH formation. Both experimental and numerical results revealed that PAH concentrations decreased monotonically as ammonia fraction increased, suggesting a PAHs-inhibiting role for NH3 in n-heptane baseline flames. Reaction pathway analysis revealed that the suppression of NH3 on PAH formation could be attributed to the competition for H, OH and O radicals in fuels' decomposition, and chemical interactions of amine radicals with the C1-C2 species. For this reason, the NH3 addition produced a relatively decreased C2H2, thereby inhibiting the formation of propargyl. As a result, benzene formation was inhibited through C3H3 self-recombination reaction, resulting sequentially in the suppression of PAH formation and growth. Besides, the decreased C2H2 yield in the post-flame region of NH3 doping flames also inhibited the formation of PAHs through the hydrogen-abstraction-acetylene-addition (HACA) mechanism.