- Email: [email protected]
Experiment Research on Flame Length Model of Laneway Fire
Available online at www.sciencedirect.com
Procedia Engineering 11 (2011) 61–67
The 5th Conference on Performance-based Fire and Fire Protection Engineering
Experiment Research on Flame Length Model of Laneway Fire CHU Yan-yana,b,* ˈDONG Wen-lia,bˈLIANG Donga,b a
Safety Engineering Research Center, Department of Engineering, Sun Yat-sen University, Guangzhou 510006, China b Guangdong Provincial Key Laboratory of Fire Science and Technology, Guangzhou 510006, China
Abstract Theoretical analysis is presented to laneway flame length model based on dimensional analyses. Analysis results indicate that flame length relates to heat release rate, fire source diameter, combustible matter diffusivity, etc. Based on Similarity principle, the laneway fire experiment plant has been setup. And using a video camera, the data of the experiment is recorded. Experiment results indicate that flame length is directly proportional to heat release rate, fire source diameter and combustible matter diffusivity. And a semi-empirical formula on flame length has been got using the least-square program to fit the experimental data. The results are of importance for flame radiation calculation and fire risk analysis.
© 2011 Published by Elsevier Ltd. Key words: laneway fire; flame length; Experiment research; heat release rate
1. Introduction Laneway fire can burn equipment and cause casualties because of the high temperature of fire. And combustibles in the vicinity of the fire source can be ignited through radiation, which makes fire developing rapidly. Flame radiation is the limited surface radiation relating to the spatial relationship between the area of exposing to radiation and flame. In order to prevent fire propagation and get the value of flame radiation, flame length should be confirmed theoretically. Flame length is a main parameter of combustibility performance assessment index. It is not only the transmission medium of radiation heat to the external space, but also the important index to predict heat release rate of fire source and the development process of fire. Flame length of laneway fire can be used as valuation index of reliability and risk of laneway fire. Nomenclature
ΔT temperature rise of fire area
* Corresponding author. Tel.: +86-20-39332230. E-mail address: [email protected].
1877–7058 © 2011 Published by Elsevier Ltd. doi:10.1016/j.proeng.2011.04.627
%
is coefficient of thermal expansion
CHU Yan-yan et al. / Procedia Engineering 11 (2011) 61–67
62
Q is quantity of heat released by fire area
D is the diameter of fire seat ,
α is dynamic viscous coefficient ν is thermal diffusivity
Y is velocity
N is fuel diffusion rate
2. Theoretical analysis of flame length Most of the flame turbulence buoyancy force diffusion flame in laneway fire. Supposing that upward vertical direction is x axis, the speed of up flow is v, radial direction is the direction of r, air velocity is u. based on Boussinesq approximate criterion, the three fundamental equations about gas in the flame zone, such as mass conservation, momentum conservation and energy conservation are shown as follows:
∂u ∂v + =0 ∂r ∂x
∂v ∂v 1 ∂ § ∂v · +v = ¨νr ¸ + gBΔT ∂r ∂x r ∂r © ∂r ¹
˄2˅
∂ΔT ∂ΔT 1 ∂ § ∂ΔT · +v = ¨ αr ¸+Q ∂r ∂x r ∂r © ∂r ¹
˄3˅
u u
˄1˅
Four independent dimensionless groups as following can be got based on the similarity principle and dimensional analysis:
Π1 =
v D x gBΔTD 3 v Π2 = Π 3 = 0 (Re nold ) Π 4 = (Granshof ) v v0 D v2
The four independent dimensionless criterions should satisfy Equation (4):
f (Π 1 , Π 2 , Π 3 , Π 4 ) = 0 v k Using the dispersal rates of fuel as combustion speed 0 , equation(4) can be expressed : L gBΔTD 3 Q k , , ) ∝ f( D Dk ν v2
˄4˅
˄5˅
Supposing flame length is L. Ignoring the effect of viscous force. Flame length is influenced by fire power, fire v source diameter and fuel diffusion rate etc. Using wind velocity 0 and fire heat release rate Q as fuel diffusion rate k . The flame length is expressed by Equation (6): / ∝ I 4 Y ' ˄6˅ 3. Experiment Design and Results Analysis
3.1. Laneway Fire Experiment Design Based on Similarity principle, the laneway fire experiment plant is setup, shown as Fig.1.The laneway fire simulator is 3 m long and the sectional area is 50×50 cm2. Basing on the theory of zone modelling, the laneway fire simulator is divided into 6 intervals, which are 0.5m long. The traditional camera measuring method is used to
CHU Yan-yan et al. / Procedia Engineering 11 (2011) 61–67
63
measure the flame length. a graduated scale is fixed on the two sides of simulator. And a video camera in front of the simulator is used to record the process of the simulating laneway fire and flame length can be got by candling camera recorder. The definition of flame length is important to the laneway fire simulation experiment. With different definition, experiment results have some deviation. The flame length is the average value of data from fire monitor recorder for more than 3minutes. The front part of fire will fracture into many small flames if the flame length exceeds a certain numerical value. The flame length in the experiments can be defined as the flame length of supervising image at the last moment before fire fracture. The advantage of the flame length definition is to process image simply and the physical conception explicit, but it needs a large amount of calculation.
Fig.1 Diagram of Flame length Experiment System
CHU Yan-yan et al. / Procedia Engineering 11 (2011) 61–67
64
3.2. Experiment Results and Analysis Designs of Verification experiment about research on relation between flame length and velocity are that the rake of laneway fire simulator is zero, 20 ml of diesel is used as fire source , oil pan diameter of fire source is 120 mm and velocity is set as 0.2 mgs-1 and 0.6mgs-1 separately. Designs of Verification experiment about research on relation between flame length and fire power are that velocity is about 0.4mgs-1, oil pan diameter of fire source is 120 mm and fire source are used as 15 ml of diesel and 25 ml of diesel separately. Designs of Verification experiment about research on relation between flame length and fire source diameter are that velocity is about 0.4mgs-1 , 20 ml of diesel is used as fire source and oil pan diameter of fire source is 120 mm and 48mm separately. Change of flame length with velocity, fire source diameter and fire power are shown as Fig. 2̚4. From Fig. 2̚4, it is shown that the process of flame length change is the oscillation process with fire development and flame length can be effect by velocity, fire source heat release and fire source diameter.
0.6m/s 0.2m/s
600 500 400 300 200 100
Flame Length (mm)
25ml 15ml
䭓
0 19 38 57 76 95 114133152 Time(s) Fig.2 Diagram of flame length change
1 20 39 58 77 96 115 134 154 173
Flame Length (mm)
800 700
500 450 400 350 300 250 200 150 100 50 0
☿ ✄ 䭓 ᑺ ˄ mm˅
900
➗⚻ᯊ䯈˄s˅ Time(s)
Fig.3 Diagram of flame length change with different airflow velocity
with different heat release rate
CHU Yan-yan et al. / Procedia Engineering 11 (2011) 61–67
65
600
Fire Source Dia 120mm 500
Flame Length (mm)
Fire Source dia 48mm ) 400 m m ( 嘧 嘖 300 b 嘓 噂 嘰 噇 嘚 200
100 0
1
13 25 37 49 61 73 85 97 109 121 133 146 158 170 182 194 206 218 Time(s)
Fig.4 Diagram of flame length change with different fire source diameter
It is shown that amplitude of flame length with the velocity of 0.6mgs-1 is more than with the velocity of 0.2mgs-1 from Fig.2. It is shown that amplitude of flame length with diesel of 25 ml is more than diesel of 15ml from Fig.3. It is shown that amplitude of flame length with fire source diameter of 120 mm is more than fire source diameter of 48 mm from Fig.4. Experiment results shown the impact of velocity, fire power and fire source diameter to flame length. It is shown that more velocity, fire power and fire source diameter are, energy conversion in combustion process can be more intensely and more unsteady. So more velocity, fire power and fire source diameter are, amplitude of flame length can be lager. Table1 Average flame length, heat release rate, velocity
Average Flame Length
L PP
Fire Source Diameter D/mm
/ '
Heat Release Rate Q/ kW
Velocity v/ mgs-1
CHU Yan-yan et al. / Procedia Engineering 11 (2011) 61–67
66
Minimum value of flame length is defined as can be calculated by / =
/PLQ and time of /PLQ
is defined as
PLQ ,
average flame length
³/ / W ³ W . Average flame lengths are shown as Table 1. PLQ
PLQ
Fig.5 Average Flame length versus heat release rate and airflow velocity
From tab. 1, it is shown that average flame lengths are 75 mm, 119 mm, 240 mm , 280 mm and 310 mm when heat release rates are 1 kw, 2 kw, 3 kw,4 kw and 5 kw , they are 125 mm, 139 mm, 142 mm and 177 mm when velocities are 0.2mgs-1, 0.6 mgs-1, 0.8 mgs-1 and 1.0 mgs-1 and they are 34 mm, 41 mm, 53 mm and 57 mm when fire source diameters are 48 mm, 55 mm, 70 mm and 85 mm. average flame length is always proportionate to heat release rate of fire source, velocity and fire source diameter. But it is independent in velocity and fire source diameter when the fire develops rapidly. Matlab is advanced software with numerical analysis, high performance calculating capability. By Matlab, the least-square program is used to fit the experimental data of flame length with different velocity, fire source diameter and fire power. Fitting curve about flame length with different parameters is shown as Fig. 5. And an average flame length calculating formula can be got by the fitting program, shown as equal (7). It is shown that L is always proportionate to Q , v ,D and it is independent in when
4
is larger
L ( ) = 0.385e 0.4733v Q 0.6209 D
˄7˅
CHU Yan-yan et al. / Procedia Engineering 11 (2011) 61–67
67
4. Conclusion Research on the impact of some factors such as fire source, velocity, etc. to flame length has been developed. Through theoretical analysis and experiment study, it is shown that average flame length is always proportionate to heat release rate of fire source, velocity and fire source diameter but it is independent in velocity and fire source diameter when the fire develops rapidly. And a semi-empirical formula on flame length has been got using the leastsquare program to fit the experimental data. It can help us to calculate radiation intensity of fire in actual fire disaster and risk assess of fire disaster.
Acknowledgements This work was supported by Young Teacher Plan of Sun Yat-sen University (39000-3281402) and Guangdong Provincial Key Laboratory of Fire Science and Technology (2010A06080101.)
References [1]Wang Lizhen ˊStudies on Flame Horizontal Model and Stochastic Model in the Building [D]ˊChong Qing˖Chongqing Universityˈ 2006 [2] Yuan LimingˊThe Height of Linear Turbulent Diffusion Buoyant Flames [J]ˊJournal of University of Science and Technology of Chinaˈ1997ˈ3˖79̚83 [3]Steven J. LuzikˊHow to limit fire and explosion hazards with oil-flooded rotary screw copressors[J]ˊMining Engineeringˈ1987ˈ 9:880̚884 [4]ZangXingkaiˊTheory and Application of Underground Fire [M]ˊBeijing ˖Capital University of Economics and Business press, 1997:132̚139 [5]JvXingangˈ NiJinpingˊImage measurement method for flame length [J]ˊapplied optics,2006,27˄6˅˖525̚528 [6]Wun Junhua etc.ˊrationale of combustion[M]ˊHarbin˖publishing house of Harbin ship School of Engine [7] W.K.ChowˊSimulation of Tunnels Fires Using a Zone ModelˈTunneling and Underground Space Technology[J]ˊ1996(2):221̚236 [8]Huggett CˊEstimation of the rate of heat release by means of oxygen consumption measurementsˊFire and Materials [J]ˈ1980 ( 4 )˖ 61̚65 [9]ZhengXinˊState of the Art of Compartment Fire Modeling [J]ˊ Engineering Scienceˈ2004ˈ3˖68~74 [10]Cui Keqing ˊ Theory and Application of combustion andexplodein Safety Engineering. Beijing:China Metrology Publising ˈ 2005ˊ92̚17 [11] CHU Yanyan, ZHANG Hui, et al. Development a mode to generate a risk map in a building fire. Sci China Ser E-Tech Sci 2010;
Procedia Engineering 11 (2011) 61–67
The 5th Conference on Performance-based Fire and Fire Protection Engineering
Experiment Research on Flame Length Model of Laneway Fire CHU Yan-yana,b,* ˈDONG Wen-lia,bˈLIANG Donga,b a
Safety Engineering Research Center, Department of Engineering, Sun Yat-sen University, Guangzhou 510006, China b Guangdong Provincial Key Laboratory of Fire Science and Technology, Guangzhou 510006, China
Abstract Theoretical analysis is presented to laneway flame length model based on dimensional analyses. Analysis results indicate that flame length relates to heat release rate, fire source diameter, combustible matter diffusivity, etc. Based on Similarity principle, the laneway fire experiment plant has been setup. And using a video camera, the data of the experiment is recorded. Experiment results indicate that flame length is directly proportional to heat release rate, fire source diameter and combustible matter diffusivity. And a semi-empirical formula on flame length has been got using the least-square program to fit the experimental data. The results are of importance for flame radiation calculation and fire risk analysis.
© 2011 Published by Elsevier Ltd. Key words: laneway fire; flame length; Experiment research; heat release rate
1. Introduction Laneway fire can burn equipment and cause casualties because of the high temperature of fire. And combustibles in the vicinity of the fire source can be ignited through radiation, which makes fire developing rapidly. Flame radiation is the limited surface radiation relating to the spatial relationship between the area of exposing to radiation and flame. In order to prevent fire propagation and get the value of flame radiation, flame length should be confirmed theoretically. Flame length is a main parameter of combustibility performance assessment index. It is not only the transmission medium of radiation heat to the external space, but also the important index to predict heat release rate of fire source and the development process of fire. Flame length of laneway fire can be used as valuation index of reliability and risk of laneway fire. Nomenclature
ΔT temperature rise of fire area
* Corresponding author. Tel.: +86-20-39332230. E-mail address: [email protected].
1877–7058 © 2011 Published by Elsevier Ltd. doi:10.1016/j.proeng.2011.04.627
%
is coefficient of thermal expansion
CHU Yan-yan et al. / Procedia Engineering 11 (2011) 61–67
62
Q is quantity of heat released by fire area
D is the diameter of fire seat ,
α is dynamic viscous coefficient ν is thermal diffusivity
Y is velocity
N is fuel diffusion rate
2. Theoretical analysis of flame length Most of the flame turbulence buoyancy force diffusion flame in laneway fire. Supposing that upward vertical direction is x axis, the speed of up flow is v, radial direction is the direction of r, air velocity is u. based on Boussinesq approximate criterion, the three fundamental equations about gas in the flame zone, such as mass conservation, momentum conservation and energy conservation are shown as follows:
∂u ∂v + =0 ∂r ∂x
∂v ∂v 1 ∂ § ∂v · +v = ¨νr ¸ + gBΔT ∂r ∂x r ∂r © ∂r ¹
˄2˅
∂ΔT ∂ΔT 1 ∂ § ∂ΔT · +v = ¨ αr ¸+Q ∂r ∂x r ∂r © ∂r ¹
˄3˅
u u
˄1˅
Four independent dimensionless groups as following can be got based on the similarity principle and dimensional analysis:
Π1 =
v D x gBΔTD 3 v Π2 = Π 3 = 0 (Re nold ) Π 4 = (Granshof ) v v0 D v2
The four independent dimensionless criterions should satisfy Equation (4):
f (Π 1 , Π 2 , Π 3 , Π 4 ) = 0 v k Using the dispersal rates of fuel as combustion speed 0 , equation(4) can be expressed : L gBΔTD 3 Q k , , ) ∝ f( D Dk ν v2
˄4˅
˄5˅
Supposing flame length is L. Ignoring the effect of viscous force. Flame length is influenced by fire power, fire v source diameter and fuel diffusion rate etc. Using wind velocity 0 and fire heat release rate Q as fuel diffusion rate k . The flame length is expressed by Equation (6): / ∝ I 4 Y ' ˄6˅ 3. Experiment Design and Results Analysis
3.1. Laneway Fire Experiment Design Based on Similarity principle, the laneway fire experiment plant is setup, shown as Fig.1.The laneway fire simulator is 3 m long and the sectional area is 50×50 cm2. Basing on the theory of zone modelling, the laneway fire simulator is divided into 6 intervals, which are 0.5m long. The traditional camera measuring method is used to
CHU Yan-yan et al. / Procedia Engineering 11 (2011) 61–67
63
measure the flame length. a graduated scale is fixed on the two sides of simulator. And a video camera in front of the simulator is used to record the process of the simulating laneway fire and flame length can be got by candling camera recorder. The definition of flame length is important to the laneway fire simulation experiment. With different definition, experiment results have some deviation. The flame length is the average value of data from fire monitor recorder for more than 3minutes. The front part of fire will fracture into many small flames if the flame length exceeds a certain numerical value. The flame length in the experiments can be defined as the flame length of supervising image at the last moment before fire fracture. The advantage of the flame length definition is to process image simply and the physical conception explicit, but it needs a large amount of calculation.
Fig.1 Diagram of Flame length Experiment System
CHU Yan-yan et al. / Procedia Engineering 11 (2011) 61–67
64
3.2. Experiment Results and Analysis Designs of Verification experiment about research on relation between flame length and velocity are that the rake of laneway fire simulator is zero, 20 ml of diesel is used as fire source , oil pan diameter of fire source is 120 mm and velocity is set as 0.2 mgs-1 and 0.6mgs-1 separately. Designs of Verification experiment about research on relation between flame length and fire power are that velocity is about 0.4mgs-1, oil pan diameter of fire source is 120 mm and fire source are used as 15 ml of diesel and 25 ml of diesel separately. Designs of Verification experiment about research on relation between flame length and fire source diameter are that velocity is about 0.4mgs-1 , 20 ml of diesel is used as fire source and oil pan diameter of fire source is 120 mm and 48mm separately. Change of flame length with velocity, fire source diameter and fire power are shown as Fig. 2̚4. From Fig. 2̚4, it is shown that the process of flame length change is the oscillation process with fire development and flame length can be effect by velocity, fire source heat release and fire source diameter.
0.6m/s 0.2m/s
600 500 400 300 200 100
Flame Length (mm)
25ml 15ml
䭓
0 19 38 57 76 95 114133152 Time(s) Fig.2 Diagram of flame length change
1 20 39 58 77 96 115 134 154 173
Flame Length (mm)
800 700
500 450 400 350 300 250 200 150 100 50 0
☿ ✄ 䭓 ᑺ ˄ mm˅
900
➗⚻ᯊ䯈˄s˅ Time(s)
Fig.3 Diagram of flame length change with different airflow velocity
with different heat release rate
CHU Yan-yan et al. / Procedia Engineering 11 (2011) 61–67
65
600
Fire Source Dia 120mm 500
Flame Length (mm)
Fire Source dia 48mm ) 400 m m ( 嘧 嘖 300 b 嘓 噂 嘰 噇 嘚 200
100 0
1
13 25 37 49 61 73 85 97 109 121 133 146 158 170 182 194 206 218 Time(s)
Fig.4 Diagram of flame length change with different fire source diameter
It is shown that amplitude of flame length with the velocity of 0.6mgs-1 is more than with the velocity of 0.2mgs-1 from Fig.2. It is shown that amplitude of flame length with diesel of 25 ml is more than diesel of 15ml from Fig.3. It is shown that amplitude of flame length with fire source diameter of 120 mm is more than fire source diameter of 48 mm from Fig.4. Experiment results shown the impact of velocity, fire power and fire source diameter to flame length. It is shown that more velocity, fire power and fire source diameter are, energy conversion in combustion process can be more intensely and more unsteady. So more velocity, fire power and fire source diameter are, amplitude of flame length can be lager. Table1 Average flame length, heat release rate, velocity
Average Flame Length
L PP
Fire Source Diameter D/mm
/ '
Heat Release Rate Q/ kW
Velocity v/ mgs-1
CHU Yan-yan et al. / Procedia Engineering 11 (2011) 61–67
66
Minimum value of flame length is defined as can be calculated by / =
/PLQ and time of /PLQ
is defined as
PLQ ,
average flame length
³/ / W ³ W . Average flame lengths are shown as Table 1. PLQ
PLQ
Fig.5 Average Flame length versus heat release rate and airflow velocity
From tab. 1, it is shown that average flame lengths are 75 mm, 119 mm, 240 mm , 280 mm and 310 mm when heat release rates are 1 kw, 2 kw, 3 kw,4 kw and 5 kw , they are 125 mm, 139 mm, 142 mm and 177 mm when velocities are 0.2mgs-1, 0.6 mgs-1, 0.8 mgs-1 and 1.0 mgs-1 and they are 34 mm, 41 mm, 53 mm and 57 mm when fire source diameters are 48 mm, 55 mm, 70 mm and 85 mm. average flame length is always proportionate to heat release rate of fire source, velocity and fire source diameter. But it is independent in velocity and fire source diameter when the fire develops rapidly. Matlab is advanced software with numerical analysis, high performance calculating capability. By Matlab, the least-square program is used to fit the experimental data of flame length with different velocity, fire source diameter and fire power. Fitting curve about flame length with different parameters is shown as Fig. 5. And an average flame length calculating formula can be got by the fitting program, shown as equal (7). It is shown that L is always proportionate to Q , v ,D and it is independent in when
4
is larger
L ( ) = 0.385e 0.4733v Q 0.6209 D
˄7˅
CHU Yan-yan et al. / Procedia Engineering 11 (2011) 61–67
67
4. Conclusion Research on the impact of some factors such as fire source, velocity, etc. to flame length has been developed. Through theoretical analysis and experiment study, it is shown that average flame length is always proportionate to heat release rate of fire source, velocity and fire source diameter but it is independent in velocity and fire source diameter when the fire develops rapidly. And a semi-empirical formula on flame length has been got using the leastsquare program to fit the experimental data. It can help us to calculate radiation intensity of fire in actual fire disaster and risk assess of fire disaster.
Acknowledgements This work was supported by Young Teacher Plan of Sun Yat-sen University (39000-3281402) and Guangdong Provincial Key Laboratory of Fire Science and Technology (2010A06080101.)
References [1]Wang Lizhen ˊStudies on Flame Horizontal Model and Stochastic Model in the Building [D]ˊChong Qing˖Chongqing Universityˈ 2006 [2] Yuan LimingˊThe Height of Linear Turbulent Diffusion Buoyant Flames [J]ˊJournal of University of Science and Technology of Chinaˈ1997ˈ3˖79̚83 [3]Steven J. LuzikˊHow to limit fire and explosion hazards with oil-flooded rotary screw copressors[J]ˊMining Engineeringˈ1987ˈ 9:880̚884 [4]ZangXingkaiˊTheory and Application of Underground Fire [M]ˊBeijing ˖Capital University of Economics and Business press, 1997:132̚139 [5]JvXingangˈ NiJinpingˊImage measurement method for flame length [J]ˊapplied optics,2006,27˄6˅˖525̚528 [6]Wun Junhua etc.ˊrationale of combustion[M]ˊHarbin˖publishing house of Harbin ship School of Engine [7] W.K.ChowˊSimulation of Tunnels Fires Using a Zone ModelˈTunneling and Underground Space Technology[J]ˊ1996(2):221̚236 [8]Huggett CˊEstimation of the rate of heat release by means of oxygen consumption measurementsˊFire and Materials [J]ˈ1980 ( 4 )˖ 61̚65 [9]ZhengXinˊState of the Art of Compartment Fire Modeling [J]ˊ Engineering Scienceˈ2004ˈ3˖68~74 [10]Cui Keqing ˊ Theory and Application of combustion andexplodein Safety Engineering. Beijing:China Metrology Publising ˈ 2005ˊ92̚17 [11] CHU Yanyan, ZHANG Hui, et al. Development a mode to generate a risk map in a building fire. Sci China Ser E-Tech Sci 2010;