• Journal of IKEEE
• /
• v.23 no.1
• /
• pp.112-119
• /
• 2019

In this paper, we propose a deep learning structure suitable for embedded system. The flame detection process of the proposed deep learning structure consists of four steps : flame area detection using flame color model, flame image classification using deep learning structure for flame color specialization, $N{\times}N$ cell separation in detected flame area, flame image classification using deep learning structure for flame shape specialization. First, only the color of the flame is extracted from the input image and then labeled to detect the flame area. Second, area of flame detected is the input of a deep learning structure specialized in flame color and is classified as flame image only if the probability of flame class at the output is greater than 75%. Third, divide the detected flame region of the images classified as flame images less than 75% in the preceding section into $N{\times}N$ units. Fourthly, small cells divided into $N{\times}N$ units are inserted into the input of a deep learning structure specialized to the shape of the flame and each cell is judged to be flame proof and classified as flame images if more than 50% of cells are classified as flame images. To verify the effectiveness of the proposed deep learning structure, we experimented with a flame database of ImageNet. Experimental results show that the proposed deep learning structure has an average resource occupancy rate of 29.86% and an 8 second fast flame detection time. The flame detection rate averaged 0.95% lower compared to the existing deep learning structure, but this was the result of light construction of the deep learning structure for application to embedded systems. Therefore, the deep learning structure for flame detection proposed in this paper has been proved suitable for the application of embedded system.

• Journal of Power System Engineering
• /
• v.2 no.1
• /
• pp.17-24
• /
• 1998

The combustion regime was discussed using a laser tomography and flame structure diagram. It was shown first how to represent the turbulent burning velocity and flame structural parameters in the dimensionless plane referred to as the flame structure diagram. And then, turbulent flame structure from the obtained images by laser tomography was compared with combustion regime in the Re-Da plane, one of the diagrams, specified by different researchers. As the result, the $u'/S_{L0}$ ratio at the boundary between the wrinkled laminar flame regime and reactant islands flame regime was found to be about 1.5.

• 한국연소학회:학술대회논문집
• /
• 2015.12a
• /
• pp.245-246
• /
• 2015

In this study, the structure of a edge flame near the flammability limits was explored through several paths approaching the combustion limits of a non-premixed flame: i.e., increase of fuel dilution ratio (FDR), reduction of mean flow velocity and variation of gravity effect. As a result, a unique interesting flame structure was discovered; i.e., a diffusion flame branch was enclosed by two asymmetric premixed flame branches. These structures have been compared for various fuels. Conclusively, each fuel has different flame structure and the meaning of this structure was discussed concerned about our understanding of laminar flame structures.

• Journal of the Korean Society of Combustion
• /
• v.15 no.3
• /
• pp.1-7
• /
• 2010

Edge flames have been interested as a basic structure that is concerned to flame stabilization and re-ignition of non-premixed flames. The edge flame consists of a lean premixed flame, a rich premixed flame, and a diffusion flame. In order to investigate fundamental structures of the edge flames at the conditions near the flammability limits, edge flames were stabilized within a heated narrow channel. Highly diluted partially premixed methane was used, and the flow rates of air and the partially premixed mixture were controlled. Various flame behaviors, including a transition between ordinary edge flames and premixed flames, were observed. Flame stabilization characteristics were examined as well. All flame stabilization conditions in this study showed a similar trend: characteristic time scales were inversely proportional to the equivalence ratio defined at the burner inlet. Finally, an interesting flame structure having a weak diffusion branch enveloped by a closed premixed branch was found near the flammability limits even in a fuel-air mixing layer. This structure was named as a "flame-drop" and the importance of this structure was first suggested.

• Journal of Energy Engineering
• /
• v.11 no.4
• /
• pp.350-358
• /
• 2002

The present work determined the flame structure characteristics of coal syngas combusted inside swirl burners with various nozzle types. Fuel nozzle types are largely classified into two groups of axial and tangential. Experiments were carried out for investigating the effects of fuel nozzle geometry, fuel composition ratio, heating rate, excess air, and degree of swirl on the turbulent diffusion flame structure. To determine the characteristics of the flame structure, axial type fuel nozzle diameter of laboratory-scale combustor is varied to 1.23, 1.96, and 2.95 ㎜ and the direction of tangential type nozzles are varied to radial, clockwise, and counter-clockwise. The comparison of the experimental results was performed to understand functional parameters relating the flame structure. Data analysis showed that the vertical straight flame height generally decreased with increasing swirl number and decreasing axial type nozzle diameter. Flame height established with tangential type nozzle is 3 times shorter than that with vertical type. The flame structures among the 3 different tangential fuel nozzles relatively showed no particular difference. By increasing the heating rate, the width of flame increased generally in both vertical and tangential flame. Within the present experimental parameters of the investigation, flame structure is mainly depends on the nozzle type of the combustor. The visually investigated flame lengths are confirmed through the analysis of temperature profile of each flame.

• Journal of the Korean Society of Combustion
• /
• v.13 no.1
• /
• pp.31-43
• /
• 2008

Oxy-fuel flame has a significantly different structure from that of air-fuel flame because of its high temperature. This study is aimed to find out the difference of the oxy-fuel flame structure in order to understand reaction mechanism closely, which is crucial to design real-scale oxy-fuel combustion system. By examining pictures of counterflow flame and LIF images, we found that oxy-fuel flame had two-zone structure: fuel decomposition region and distributed CO oxidation region. In the oxy-fuel flame, OH radical was distributed intensely through the whole flame due to its higher flame temperature than crossover temperature. For showing those features of the oxy-fuel flame, 1 MW scale IFRF oxy-natural gas burner was simulated by conditional moment closure(CMC) model. Calculation results were compared with experimental data, and showed agreements in trend. In the simulated distributions of fuel decomposition/CO oxidation rates, CO oxidation region was also separated from fuel decomposition zone considerably, which showed the two-zone structure in the oxy-fuel flame.

• Journal of the Korean Society of Safety
• /
• v.25 no.3
• /
• pp.22-27
• /
• 2010

A prediction performance of Fire Dynamics Simulator(FDS) developed by NIST for the diffusion flame structure was validated with experimental results of a laminar slot jet diffusion flame. Two mixture fraction combustion models and two finite chemistry combustion models were used in the FDS simulation for the validation of the jet diffusion flame structure. In order to enhance the prediction performance of flame structure, DNS and radiation model was applied to the simulation. The reaction rates of the finite chemistry combustion models were appropriately adjusted to the diffusion flame. The mixture fraction combustion model predicted the diffusion flame structure reasonably. A 1-step finite chemistry combustion model cannot predict the flame structure well, but the simulation results of a 2-step model were in good agreement with those of experiment except $CO_2$ concentration. It was identified that the 2-step model can be used in the investigation of flame suppression limit with further adjustment of reaction rates

• Journal of Advanced Marine Engineering and Technology
• /
• v.23 no.1
• /
• pp.88-95
• /
• 1999

This study investigates the effects of different kind fuels on the flame structure by using the numerical simulation in triple flame made by a co-flowing fuels-air stream based on the elementary chemical reaction mechanism. Methane and Hydrogen were used as fuel for this study. In order to interpret the result of the study on numerical simulation Skeletal chemistry is employe as the elementary chemical reaction mechanism for methane Gutheil's as an offset ele-mentary chemical reaction mechanism for hydrogen. The result of this study is as follows. In com-parison between the apparent burning velocity change of triple flame and the one-dimensional pre-mixed flame hydrogen fuel flame is higher than methane fuel flame. The flame thrusts out for-ward in the down stream of the boundary between air-fuel mixture and air stream and a part of the flow is bent and forks out in this protruding flame so that a triple flame is originated.

• Journal of ILASS-Korea
• /
• v.3 no.4
• /
• pp.43-49
• /
• 1998

In order to clarify the effect of equivalence ratio and kinds of fule in flame structure, a numerical simulation of triple flame developed in a co-flowing methane-air and air stream was carried out by the elementary chemical reaction mechanism. The following conclusions were obtained. Equivalence ratio at which the apparent burning velocity is maximum is a little larger than that of the one-dimensional premixed flame. Apparent burning velocities are two times higher than that of the one-dimensional premixed flame for the methane-air. The flame thrusts out forward in the downstream of the boundary between mixture and air stream, and a part of the flow is bent and forks out in this protruding flame so that a triple flame is originated; this triple flame is composed of fuel rich and lean premixed flame branches and a diffusion flame branch. Near the equivalence ratio at which the burning velocity of rule-dimensional premixed flame is the largest the effect of one-dimensional premixed flame becomes large and the fuel rich premixed flame advances and becomes vertical to the flow direction.

• Journal of Advanced Marine Engineering and Technology
• /
• v.19 no.4
• /
• pp.34-41
• /
• 1995

The structure of premixed tubulent flames in a constant-volume vessel was investigated using a schlieren method and microprobe method. The schlieren method was used to observe the flame structure qualitatively. The microprobe method, which detects a flamelet by detecting its flame potential signal, was used to investigate the deeper flame structure behind the flame front. The flame potential signal having one to six peaks was obtained in the case of turbulent flames, each of them being regarede as a flamelet existing in the flame zone. Based on this consideration, the flame propagation speed, the thickness of the flame zone, the number of flamelets and the separation distance between adjacent flamelets in the flame zone were measured. Moreover, the thickness of flamelet which could not be attempted in the conventional electrostatic probe method was also investigated. The experimental results of this work suggest the existence of "reactant islands" in the reaction zone, and show that the averaged number of flamelets increases with an increase in the turbulence intensity and/or a decrease in the Damkohler number. The mean thickness of flamelet in the case of turbulent flames was found to be about two times compared to laminar values.ar values.