- Email: [email protected]
Ignition limits of short-term overloaded electric wires in microgravity
Available online at www.sciencedirect.com
Proceedings of the
Combustion Institute
Proceedings of the Combustion Institute 34 (2013) 2665–2673
www.elsevier.com/locate/proci
Ignition limits of short-term overloaded electric wires in microgravity Yoshitomo Takano, Osamu Fujita ⇑, Naoki Shigeta, Yuji Nakamura, Hiroyuki Ito Division of Mechanical and Space Engineering, Hokkaido University, Kita13 Nishi8, Kita-ku, Sapporo, Hokkaido, Japan Available online 29 June 2012
Abstract Ignition phenomena of electric wires carrying short-term excess electric currents were investigated in microgravity with experiments and calculations. Microgravity experiments were conducted in 100 m and 50 m drop towers and calculations were carried out with a one dimensional cylindrical coordinate system. The experimental results showed that the limiting oxygen concentration (LOC) under a given electric current was much lower in microgravity than that in normal gravity except for extremely large electric current overload cases. According to the calculations, the supplied electric current, the Joule energy supplied to the wire, determined the amount of pyrolysis gas from the insulation and the resulting thickness of the gaseous fuel layer around the sample in gas phase increased. The increased fuel layer thickness resulted in a longer ignition delay, which leads to lower LOC. The changes in the estimated LOC changed as a function of supplied energy and agreed well with the experimental results. Further, the minimum ignition energy causing ignition (ignition limit) is nearly constant under a constant oxygen concentration, which supports experimental findings in previous research. Ó 2012 The Combustion Institute. Published by Elsevier Inc. All rights reserved. Keywords: Ignition limit; Electric wire; Short-circuiting; Fire safety in space; Microgravity
1. Introduction A most likely cause of fires in space is combustion of the wire harness of spacecraft [1,2], and such fires are generally started with short-circuiting or overloading of electric wires. Therefore, it is important to know the ignition characteristics of overloaded wires in microgravity to improve fire safety in space. To know the electric current needed for ignition of electric wires is important in the design of circuit breakers. As carbon diox⇑ Corresponding author. Fax: +81 11 706 6385.
E-mail (O. Fujita).
address:
[email protected]
ide extinguishers are used in the International Space Station [3], limiting oxygen concentration for ignition is essential to design extinguishers based on asphyxiating effects and to establish the ambient oxygen concentration in spacecraft. Knowledge of the ignition characteristics of solid components in microgravity is a basic subject which must be studied to prevent the breaking out of fires, and a number of theoretical and experimental studies have been conducted. One example of solid ignition research in microgravity is piloted ignition of PMMA plates [4,5] and others are non-piloted ignition of thin cellulosic sheets [6–9] or PMMA sheets [10] heated by external radiant sources. However, there are few
1540-7489/$ - see front matter Ó 2012 The Combustion Institute. Published by Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.proci.2012.06.064
2666
Y. Takano et al. / Proceedings of the Combustion Institute 34 (2013) 2665–2673
studies except from our research group [11,12] about spontaneous ignition of electric wires caused by Joule heat generated in the wire core. In the previous research [11], the authors reported dramatic extensions of ignition limits in terms of supplied electric current under microgravity. In that research, excess electric current was continuously supplied until ignition occurred. In other research [12], the authors investigated the ignition characteristics with short-term excess electric currents. There, the length of the current supply was selected as the main test parameter to simulate the status of circuit breaker activation shortly after overloading happens, and delayed ignition of electric wires after short-term overloading was observed in microgravity. In the results with short-term electric current supply, the ignition delay may be longer than with a continuous current supply. The electric current is smaller in microgravity than that on the ground as in the continuous current case. The established findings [11,12] indicate the probability of ignition increases in microgravity. One matter that has not been established in the previous work is the limiting oxygen concentration, below which wire ignition does not occur even with excess electric current. A detailed knowledge of this basic parameter is essential for fire safety in space. The present study, investigates the ignition limit; the minimum value of the electric energy supply causing ignition, and LOC; limiting oxygen concentration, below which ignition does not occur at a given electric current supply for short-term excess electric currents under microgravity, by experiments and calculations. The mechanism to determine the ignition limit and the LOC is discussed based on the numerical results. 2. Experiments
with the sample wire inside, a constant current supply system, and an image recording system. The tested sample in Fig. 2 is a nickel–chrome core wire coated with polyethylene. The outer diameter of the sample is 0.8 mm and the inner core diameter is 0.5 mm. The effective length of the sample is 70 mm; the electric resistance of the core wire is 5.5 X/m. The resistance is the value under room temperature, which is used to calculate the Joule energy, E, generated in the core wire under a given electric current. In the tests, the length of time the current is applied, ta, was varied in the range from 0.2 to 1.5 s to simulate delays in the circuit breaker activation. Oxygen concentration and supplied current values were also varied in the tests. Microgravity experiments were conducted at the MGLAB (Microgravity Laboratory of Japan) in Gifu, Japan, which provides a microgravity environment of 105g0 (g0: gravity level on the ground) for around 4.5 s and at the COSMOTORRE in Hokkaido, Japan, which provides 10 3 g0 environment for around 2.7 s. When light emission in gas phase is observed in the video images, it is defined as “ignition” in the experiment and the “Ignition delay”, tig is defined as the time after the start of current supply till the moment of ignition. 2.1. Method of the numerical calculations The numerical calculations are carried out with a model on a one dimensional cylindrical coordinate system. Fig. 3 is a schematic illustration of the ignition model. At the start of the experiment (time zero) current is supplied to the core wire and insulation starts to degrade to emit pyrolysis gas into the oxygen-containing atmosphere. The evolved gas mixes with the ambient oxygen to form a hot combustible mixture eventually leading to spontaneous ignition. The assumptions adopted for the calculations were as follows:
All experiments were performed in the apparatus similar to that used in previous research [12]. An outline of the experimental set-up is shown in Fig. 1. It is composed of a combustion chamber
1. Both core wire and insulation are thermally thin. 2. The solid fuel thickness is held at the initial value throughout the course of the calculations. The density decreases with time as the reactions degrading the solid phase proceed. 3. Gas generated by the degradation is ethylene.
Fig. 1. Outline of the combustion chamber.o
Fig. 2. Schematic details of the sample wire.o
Y. Takano et al. / Proceedings of the Combustion Institute 34 (2013) 2665–2673 Gas phase
r
rs
Insulated wire
rmax
Δr
Surface 1
j-1
j
r
Inert wall j+1
n
Solid phase
Δ r/2
note the gas phase, the i species, and the initial state, respectively. The solid is composed of core wire as the heat source and the insulation as a solid fuel. The sample in the calculations is the same as that used in the experiments, which is assumed as thermally thin. The governing equations for the solid phase can be expressed as follows: Mass conservation
Δr
m_ ¼ V p Fig. 3. Schematics of ignition model.o
The governing equations for the gas phase can be expressed as follows: Mass conservation ð1Þ
Energy conservation @ðqg C p;g T g Þ 1 @ðrqg uC p;g T g Þ þ @t r @r 1 @ @T g rk ¼ þ qg w_ g r @r @r
dT c Ac ¼ Qin Qcp dt dT p Ap ¼ Qcp Qre Qout qc C p;p dt qc C p;c
ð6Þ ð7Þ
where V and A are the insulation volume and sectional area, respectively. Qin, Qcp, Qre, and Qout are the applied energy by the current supply, the heat conduction between core and insulation, the energy lost by degradation and heat conduction between the insulation and gas phase; m_ is the mass reduction rate; and the subscripts c and p denote the core and insulation, respectively. Qin, Qcp, Qre, and Qout are given as below. Qin ¼ Re I 2 Qcp ¼ 2pkp
Qout ¼ 2pk ð2Þ
ð3Þ
State qT ¼ q1 T 1
ð5Þ
Tc Tp ln ððrp =2Þ þ rc Þ=rc
_ Qre ¼ mDH p
Species conservation @ðqg Y i Þ 1 @ðrqg uY i Þ þ @t @r r 1 @ @Y i rqg D vi w_ g ¼ r @r @r
@qp @t
Energy conservation
4. The gas phase consists of the degradation product, oxygen and nitrogen, and the mixture behaves like an ideal gas. 5. Initially, the ambient atmosphere is quiescent. 6. Atmospheric pressure remains constant at 1 atm throughout the simulation. 7. The thermal properties of wire and insulation are considered constant regardless of temperature. 8. The thermal properties and diffusion coefficient of the gas are constant regardless of temperature and species.
@qg 1 @ðrqg uÞ þ ¼0 r @t @r
2667
ð4Þ
where t and r are the time and radius distance, respectively; q, u, T, and Yi are the gas density, velocity, temperature, and species mass fraction of a species i (i = OX: oxygen, F: fuel gas), respectively; and Cp, k, and D are the specific heat, heat conductivity, and mass diffusivity of the mixture, respectively. The mass diffusivity, D, is assumed to be equal for all species. Then, q and m are the heat release by the gas phase reaction and the stoichiometric coefficient, respectively; and finally x is the reaction rate. The subscripts g, i, and 1 de-
Tp T lnðrp =ððrp =2Þ þ rc ÞÞ
where Re, rp, rc are resistance of inner core per unit length, radius of wire (core radius plus insulation thickness) and radius of inner core, respectively. T is temperature of gas phase of the next grid to the surface. DH p is reaction heat of pyrolysis reaction (in this calculation, 1.2 MJ/kg is used). The chemical reaction in the gas phase is described by a single one-step reaction. The pyrolysis reaction in the solid phase is also described by a single one-step reaction. The reaction rates are given as follows: Gas overall reaction C2 H4 þ 3O2 ! 2CO2 þ 2H2 O Eg w_ g ¼ Bg ðqY F ÞðqY OX Þexp RT g
ð8Þ
Solid fuel pyrolysis reaction ½polyethylenesolid ! ½C2 H4 gas Ep m_ ¼ Bp qp V p exp RT p
ð9Þ
2668
Y. Takano et al. / Proceedings of the Combustion Institute 34 (2013) 2665–2673
where B, R, and E are the frequency factor, the universal gas constant, and the activation energy, respectively. The reaction rate constants, Bg = 125 kJ/mol, Eg = 7.143 109 m3/kg s, Bp = 5.0 1018 s1, Ep = 250 kJ/mol, are as reported in Refs. [13,14]. The initial and boundary conditions are as follows: Initial condition in the solid phase T c ¼ T 1; T s ¼ T 1 in the gas phase T ¼ T 1 ; Y F ¼ Y F ;1 ¼ 0; Y OX ¼ Y OX ;1 ; u ¼ 0 Boundary Conditions
Fig. 4. Example of the appearance of wire combustion directly after ignition in microgravity (E = 7.9 J/cm, O2 – 40%, ta = 1.0 s).
at r ¼ rs T surf ¼ T p
@Y F qusurf Y F ;surf þ qD ¼ qusurf @r r¼rs @Y OX qusurf Y OX;surf þ qD ¼0 @r r¼rs
at r ¼ rmax T ¼ T 1 ; Y F ¼ 0; Y OX ¼ Y OX;1 ; u ¼ 0 where the subscript surf denotes the state of the insulation surface. The calculation domain length is 200 mm and the adopted mesh size is 0.2 mm. The time step adopted in this study is 1.0 105 s and calculations are continued until ignition or an elapsed time of 10 s. In previous studies on the ignition of heated solid materials, many types of the ignition criterion have been proposed for maximum temperature, maximum reaction rate, and others [15]. These different types of criteria gave the similar results regarding the effects of the ambient oxygen, and the derived numerical values themselves were not so different [15]. In this study, ignition is defined to occur when the value of the local heat release rate exceeds 30,000 kW/m3. 3. Limiting oxygen concentration (LOC) by experiments Figure 4 shows a photo of wire combustion immediately after ignition in microgravity. The ignition is initiated at some point and then propagates along the wire. This figure indicates the flame has a cylindrical two-dimensional shape surrounding the sample wire in the absence of buoyancy induced convection. Figure 5 shows Mach-Zehnder interference images for the ignition phenomena in microgravity. The supplied energy is 6.7 J/cm with the current applied for 0.5 s at an oxygen concentration of 40%, which is close to the limiting condition
for ignition. The ignition delay here is 4.134 s and images at every 0.5 s selected from the high speed movie are sequentially shown in Fig. 5. In this figure, fringe shift to the left corresponds to a decrease in the refractive index which is generally caused by decreases in the gas density. Directly after the stop of the current supply, at t = 0.5 s, the fringe shift becomes large near the wire surface (fringes are completely parallel normal to the wire sample when t = 0 s). Then, the area of the large fringe shift spreads outward, and after 3.0 s the point of maximum fringe shift locates away from the sample surface. Finally ignition occurs near the point of maximum fringe shift. Figure 6 shows the LOC as a function of the supplied electrical energy. In the figure, the results at and above 21% oxygen concentration were obtained at MGLAB [12] and at oxygen concentrations below 21% at COSMOTORRE. The plots in Fig. 6 were each obtained after a series of tests at a given electrical energy in a step-bystep decrease in the oxygen concentration to establish the minimum oxygen concentration for ignition, and the tests were repeated three times for one point in the plot on the ground. The repeatability is very high with the test and the LOC can be determined within an accuracy of 1% oxygen. The LOC varies with electric energy, becoming smaller with increases in electric energy. The LOC is different for different gravity conditions, with 10.8 J/cm the LOC is 16% in microgravity and 40% at normal gravity, a significant extension of the flammable limit under microgravity. An explanation could be the elimination of natural convection in microgravity, which results in longer residence time of flammable gas and reduced heat loss from the combustion zone. When the energy supply is 22.0 J/cm, the LOC in microgravity is 12% and that at normal gravity is 13%, which suggests the difference in LOC becomes smaller with increases in supplied energy and the LOC seems to converge to value for both
Y. Takano et al. / Proceedings of the Combustion Institute 34 (2013) 2665–2673
2669
60
900
50
800
40
700
30
Tg [K]
Oxygen Concentration [%]
Fig. 5. Fringe Image (lG, E = 6.7 J/cm, O2 – 40%, ta = 0.5 s, tig = 4.134 s). The wire is designated by the white line and the ignition point is designated by the open circle.
Experiment on the ground
20
t=1.8s
600 t=1.2s
500
10 0
(a)
Experiment in microgravity
6
8
10 12 14 16 18 Electrical Energy [J/cm]
400 20
22
300
Fig. 6. Effect of supplied electric energy on LOC in normal gravity and microgravity.
t=1.0s
t=0.5s 0
1
2
3
4 5 r [mm]
900
6
t=1.4s
7
8
(b)
800 t=1.8s
4. Results of the numerical calculations
700
Tg [K]
gravitational conditions at higher electric energies. Here it must be noted that LOC shows significant differences in the low electric energy condition for different gravity conditions, a situation shown for the first time in this paper. In the following section, the ignition mechanism at the low supplied energy condition is discussed based on numerical calculations.
t=2.0s
600 t=1.2s
500 400 300
0
1
t=1.4s
t=1.0s
t=0.5s 2
3
4 5 r [mm]
6
7
8
Fig. 7. Temperature distribution in (a) ignition condition: E = 7.120 J/cm and (b) no ignition condition: E = 7.119 J/cm (O2 – 21%, ta = 1.0 s).
4.1. Temperature distributions with and without ignition Figure 7 shows temperature distributions for two cases calculated with different amounts of energy supplied. (a) is a case with ignition, 7.120 J/cm and (b) is a case without ignition, 7.119 J/ cm. The current was applied for 1.0 s and the ambient oxygen concentration 21%. In both cases, at the cessation of the current supply, 1.0 s, the maximum temperature is at the wire surface. At 1.4 s, the high temperature region has spread outward, as enough pyrolysis gas is evolved into the gas phase and the position of the maximum temperature is located away from the wire surface. Up to 1.8 s, the temperature in Fig. 7(a) is increasing rapidly in the vicinity of the wire surface and ignition occurs. In Fig. 7(b), the temperature decreases after discontinuing the current and ignition does not occur. If the fringe shift in Fig. 5 is mainly caused by the changing
temperature, the trend in the temperature change given in the calculations resembles the experimental result qualitatively. 4.2. Concept of ignition limit Figure 8 shows the calculated history at the ignition point (at the grid where ignition occurs) of reaction heat and heat loss by heat conduction of (a) the ignition condition (E = 7.120 J/cm) and (b) the no ignition condition (E = 7.199 J/cm). The ambient oxygen concentration is 21% and the current is supplied for 1.0 s. In both cases, the reaction heat keeps increasing after stopping the current supply at 1.0 s and at 1.6 s the reaction heat exceeds heat loss in (a). Here the reaction heat becomes predominant and leads to a thermal runaway. In (b), the value of the reaction heat and
Y. Takano et al. / Proceedings of the Combustion Institute 34 (2013) 2665–2673
Heat release
1.0E+05 1.0E+04
1.0E+03 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0 t [s]
1.0E+07 1.0E+06 1.0E+05
Fuel
0.25 Oxygen 0.20
0.60
0.15 0.40
0.10
0.20
7.120J/cm 21%O2 7.001J/cm 21%O2
1
2
3
4 5 r [mm]
6
900
Heat loss Heat release
1.0E+04
800
Fig. 8. History of reaction heat and heat loss in (a) ignition condition: E = 7.120 J/cm and (b) no ignition condition: E = 7.119 J/cm (O2 – 21%, ta = 1.0 s).
heat loss have not changed from the 1.0 s values at 1.6 s, and the value of the reaction heat does not exceed the heat loss leading to a situation without ignition. These results are consistent with the general ignition concept that ignition is determined by the balance between heat production and heat loss. This establishes that for ignition in an overloaded wire the reaction heat and heat loss by conduction behave like in Fig. 8. 4.3. Effect of electric energy and oxygen concentration on wire ignition Figure 9 shows (a) the fuel and oxygen distribution and (b) the temperature and reaction rate distribution for two different supplied energy cases. The supplied electric energy is 7.210 J/cm or 7.001 J/cm, the ambient oxygen concentration is 21%, and the current is applied for 1.0 s. The fuel layer in (a) becomes thicker in the gas phase with increase in electric energy because more pyrolysis gas is evolved into the gas phase with the higher electric energy supply. After the current supply is discontinued at 1.0 s, the maximum temperatures are very similar but the temperature in the ignition case is higher at 1.8 s. The maximum values of the reaction rates are similar at 1.0 s but there is a big difference right before the ignition, that is, the reaction rate at 1.8 s in the high electric energy condition is larger than that in the low condition.
7
(b)
Tg
8 0.00
1.00E+06
1.00E+02 7.120J/cm 21%O2 7.001J/cm 21%O2 1.00E+00
600 500 RR
300 0
0.05
1.00E+04
t=1.0s
700
1.00E-02
t=1.8s
400
1.0E+03 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0 t [s]
0.30
(a)
t=1.8s
0.80
0.00 0
(b)
t=1.0s
1
2
3
4 5 r [mm]
6
7
Oxygen concentration
Heat loss
1.0E+06
1.00
Ignition
Reaction rate [kg/m3s]
(a)
Fuel concentration
1.0E+07
Tg [K]
Reaction heat or heat loss [W/m3]
Reaction heat or heat loss [W/m3]
2670
8 1.00E-04
Fig. 9. Distributions of (a) fuel and oxygen concentration and (b) temperature and reaction rate around the ignition limit (E = 7.210 J/cm and 7.001 J/cm, O2 – 21%, ta = 1.0 s, in ignition condition tig = 1.801 s).
Figure 10 hows (a) the fuel and oxygen distribution, and (b) the temperature and reaction rate distribution for two different oxygen concentration conditions, 21% or 18%, the supplied electrical energy is 7.210 J/cm and the current applied for 1.0 s. The temperatures at 1.0 s with 21% and 18% oxygen concentration are very similar, while distributions of fuel and oxygen concentrations at 1.0 are different with non-negligible level. This difference causes big differences of temperature and reaction rate at 1.8 s, that is, temperature and reaction rate become larger in the high oxygen concentration. In low electric energy condition, fuel layer becomes thin and it results in increases in air diffusion to the wire surface. Then, gas temperatures in the vicinity of the wire surface decreases, because the surrounding air temperature is lower than the fuel temperature. It causes decreases in the reaction rate. As changing electric energy supply, the thickness of the fuel layer changes. Then, the diffusion time of the surrounding air to the wire surface changes. If the diffusion time becomes shorter, the gas temperature decreases in a short time near the solid surface. So the reaction needs to be faster to achieve ignition, which requires higher oxygen concentrations. When the oxygen concentration is low, the increase rate of the reaction rate becomes smaller. So it needs a longer diffusion time to allow a longer reaction time.
Y. Takano et al. / Proceedings of the Combustion Institute 34 (2013) 2665–2673
(a)
t=1.8s
Oxygen 0.25 0.20
0.60
0.15 0.40
0.10
0.20 0.00
0
1
2
3
4 5 r [mm]
6
900
Tg [K]
7
(b)
Tg
800
8
0.00
1.00E+06 1.00E+04
700
t=1.0s
1.00E+02 7.120J/cm 21%O2 7.120J/cm 18%O2
600 500 RR
1.00E+00
t=1.8s
400 300
0.05
7.120J/cm 21%O2 7.120J/cm 18%O2
0
1
2
3
4 5 r [mm]
1.00E-02 6
7
8
1.00E-04
Fig. 10. Distributions of (a) fuel and oxygen concentration and (b) temperature and reaction rate around the LOC (E = 7.210 J/cm, O2 – 21% and 18%, ta = 1.0 s, in ignition condition tig = 1.801 s).
From these figures, it is noted that no ignition occurs in case where the peak of the reaction rate and the highest temperature point gets very close to the wire surface. As shown in Figs. 5 and 7, the highest temperature point is away from the wire surface when ignition occurs. In such situations, ignition occurs and the heat release keeps exceeding the heat loss until the temperature reaches the one for a thermal runaway. As seen in Fig. 8(a) it takes not so short time to reach ignition. As time elapsed, the highest temperature point shifts to the wire surface, because air diffusion proceeds toward the wire surface. Once the maximum heat release point becomes very close to the wire surface, the chance for ignition is lost because the wire acts as a heat sink and temperature gradient around the maximum heat release point becomes very steep. Then, heat release cannot exceed the heat loss. This is a unique feature of wire ignition by short-term current supply because the ignition delay is longer than the current supply time and wire never works as a heat source at the moment of ignition near the limiting condition. Under these considerations, competition of diffusion time with the delay time for ignition defined as in Fig. 8(a) is very important to understand the onset of ignition. When the supplied electric energy is larger, the thickness of the initial fuel layer increases as shown in Fig. 11. Then the diffusion time increases, which allows a longer ignition delay and lower LOC. When the
oxygen concentration is higher, the ignition delay decreases, which allows a shorter diffusion time corresponding to less supplied energy for ignition. The thickness of the fuel layer right after stopping the current supply is shown in Fig. 11 for some supplied electric energies. The current is applied for 1.0 s and the ambient oxygen concentration is 21%. In this figure, it is obvious that the fuel layer becomes thicker, when the supplied electrical energy increases. Therefore, the diffusion time increases with increases in supplied energy. Then lower LOC is expected with increases in the supplied energy as discussed above. 4.4. Limiting conditions for ignition The calculated LOC as a function of supplied electrical energy is shown in Fig. 12. The results estimated by the numerical calculations showed a similar behavior as the experiments. But a quantitative comparison shows a smaller LOC for the calculations than for the experimental results. One of the causes of the difference may be the radiation heat loss effect, which is not considered in the calculations, and more study is needed on this difference. In this figure, it is obvious that the LOC is strongly dependent on the electric energy at low electric energies, while it depends weakly on the electric energy in the higher energy conditions. It particular, the LOC becomes higher drastically in cases below 7.0 J/cm in the calculated results. Figure 13 shows the ignition limit as a function of the current supply time. This figure shows that in microgravity the ignition limit is almost constant under a constant oxygen concentration. The same has been observed in microgravity experiments [12]. It is indicated that there is specific supplied energy for ignition under a given oxygen concentration. The calculations show the effect of the current supply time, that is, the ignition limit for a 1.0 s of current supply duration
1.00
Fuel concentration
Fuel concentration
0.80
0.30
t=1.0s
Oxygen concentration
Fuel
Reaction rate [kg/m3s]
1.00
2671
At t = 1.0 s (= t a)
0.80 0.60
7.710J/cm 7.400J/cm
0.40 0.20
7.120J/cm
0.00 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 r [mm] Fig. 11. Thickness of fuel layer for different supplied electric energy amounts (E = 7.120 J/cm, 7.400 J/cm, 7.710 J/cm, O2 – 21%, ta = 1.0 s).
Oxygen Concentration [%]
2672
Y. Takano et al. / Proceedings of the Combustion Institute 34 (2013) 2665–2673
60 50 40 30 20
Experiment in microgravity
10 Calculation in microgravity
0
6
8
10
12 14 16 18 Electrical Energy [J/cm]
20
22
Fig. 12. Limiting Oxygen concentrations (LOC) for experiments and calculations.
Ignition limit [J/cm]
8 O2 - 21% 7
6
5 0.0
O2 - 40%
0.2
0.4
O2 - 60%
0.6
0.8 1.0 t a [s]
1.2
1.4
1.6
(1) The LOC in microgravity is much lower than in normal gravity when the supplied energy is low (<14 J/cm), while the value at higher supplied energies in microgravity becomes closer to the value in normal gravity. (2) The thickness of the fuel layer is an important factor for wire ignition phenomena. The thickness of the fuel layer determines the diffusion time in which oxidation reaction proceeds and ignition occurs. When the fuel layer becomes thin, the diffusion time becomes short and the gas temperature decreases in a short time. To achieve ignition reaction needs to be faster, which requires higher oxygen concentrations. (3) The results estimated by the numerical calculations showed similar dependence of LOC on electric energy as the experiments, suggesting the importance of competition between diffusion time and ignition delay time. (4) The ignition limit is almost constant under a constant oxygen concentration even when the current supply duration is different. It is indicated that the supplied energy is the main controlling factor for ignition under a given oxygen concentration.
Fig. 13. Effect of the current supply time on the ignition limit.
Acknowledgments is 1.5% higher, that for 1.5 s is 3.5% higher, than that for a 0.2 s under various oxygen concentrations. According to the results the effect of supply time can be disregarded below 1.0 s. When the current supply duration changes under a constant supplied energy to the wire, the heating rate of wire changes. So temperature distribution around the wire would change a little. In spite of the change, the ignition limit under a certain oxygen concentration is almost the same in these conditions. Therefore, supplied energy is the dominant factor to control the ignition for short-term current supply once oxygen concentration is fixed.
5. Conclusions Ignition phenomena for electric wires with short-term excess electric current supply were investigated by microgravity experiments and numerical calculations to understand the ignition characteristics such as ignition limit and LOC (limiting oxygen concentration). The conclusions may be summarized as follows:
This research is partially supported by JAXA as the candidate experiments for the second phase utilization of JEM/ISS entitled “Quantitative Description of Gravity Impact on Solid Material Flammability as a base of Fire Safety in Space” and is also supported by JAXA Research Working Group to promote space utilization.
References [1] R. Friedman, Fire Mater. 20 (1996) 235–243. [2] NASA Office of Safety and Mission Quality, Flammability, Odor, Off Gassing and Compatibility Requirements and Test Procedures for Materials in Environments that Support Combustion, NHB8060.1C, 1991. [3] G. Ruff, D. Urban, M. King, Proceedings of the 43rd AIAA Aerospace Sciences Meeting and Exhibit, Reno, NV. 2005-341, 2005. [4] S. McAllister, C. Fernandez-Pello, D. Urban, G. Ruff, Proc. Combust. Inst. 32 (2009) 2453–2459. [5] S. McAllister, C. Fernandez-Pello, D. Urban, G. Ruff, Combust. Flame 157 (2010) 1753–1759. [6] O. Fujita, J. Takahashi, K. Ito, Proc. Combust. Inst. 28 (2000) 2761–2768. [7] J. Takahashi, O. Fujita, K. Ito, Proc. Combust. Inst. 30 (2005) 2311–2317.
Y. Takano et al. / Proceedings of the Combustion Institute 34 (2013) 2665–2673 [8] Y. Nakamura, H. Yamashita, T. Takeno, G. Kushida, Combust. Flame 120 (2000) 34–48. [9] S.L. Olson, T. Kashiwagi, O. Fujita, M. Kikuchi, K. Ito, Combust. Flame 125 (2001) 852–864. [10] Y. Nakamura, T. Kashiwagi, S.L. Olson, K. Nishizawa, O. Fujita, K. Ito, Proc. Combust. Inst. 30 (2005) 2319–2325. [11] K. Agata, O. Fujita, Y. Ichimura, T. Fujii, H. Ito, Y. Nakamura, J. Jpn. Soc. Microgravity Appl. 25 (1) (2008) 11–16.
2673
[12] O. Fujita, T. Kyono, Y. Kido, Y. Nakamura, H. Ito, Proc. Combust. Inst. 33 (2011) 2617–2623. [13] K. Nishizawa, Study on Flame Spread over Wire Insulation for Improve Fire Safety in a Spaceship, Ph.D. thesis, Graduate School of Engineering, Hokkaido University, Hokkaido, Japan, 2004.. [14] C.K. Westbrook, F.L. Dryer, Combust. Sci. Technol. 27 (1981) 31–43. [15] T. Kashiwagi, Combust. Sci. Technol. 8 (1974) 225– 236.
Proceedings of the
Combustion Institute
Proceedings of the Combustion Institute 34 (2013) 2665–2673
www.elsevier.com/locate/proci
Ignition limits of short-term overloaded electric wires in microgravity Yoshitomo Takano, Osamu Fujita ⇑, Naoki Shigeta, Yuji Nakamura, Hiroyuki Ito Division of Mechanical and Space Engineering, Hokkaido University, Kita13 Nishi8, Kita-ku, Sapporo, Hokkaido, Japan Available online 29 June 2012
Abstract Ignition phenomena of electric wires carrying short-term excess electric currents were investigated in microgravity with experiments and calculations. Microgravity experiments were conducted in 100 m and 50 m drop towers and calculations were carried out with a one dimensional cylindrical coordinate system. The experimental results showed that the limiting oxygen concentration (LOC) under a given electric current was much lower in microgravity than that in normal gravity except for extremely large electric current overload cases. According to the calculations, the supplied electric current, the Joule energy supplied to the wire, determined the amount of pyrolysis gas from the insulation and the resulting thickness of the gaseous fuel layer around the sample in gas phase increased. The increased fuel layer thickness resulted in a longer ignition delay, which leads to lower LOC. The changes in the estimated LOC changed as a function of supplied energy and agreed well with the experimental results. Further, the minimum ignition energy causing ignition (ignition limit) is nearly constant under a constant oxygen concentration, which supports experimental findings in previous research. Ó 2012 The Combustion Institute. Published by Elsevier Inc. All rights reserved. Keywords: Ignition limit; Electric wire; Short-circuiting; Fire safety in space; Microgravity
1. Introduction A most likely cause of fires in space is combustion of the wire harness of spacecraft [1,2], and such fires are generally started with short-circuiting or overloading of electric wires. Therefore, it is important to know the ignition characteristics of overloaded wires in microgravity to improve fire safety in space. To know the electric current needed for ignition of electric wires is important in the design of circuit breakers. As carbon diox⇑ Corresponding author. Fax: +81 11 706 6385.
E-mail (O. Fujita).
address:
[email protected]
ide extinguishers are used in the International Space Station [3], limiting oxygen concentration for ignition is essential to design extinguishers based on asphyxiating effects and to establish the ambient oxygen concentration in spacecraft. Knowledge of the ignition characteristics of solid components in microgravity is a basic subject which must be studied to prevent the breaking out of fires, and a number of theoretical and experimental studies have been conducted. One example of solid ignition research in microgravity is piloted ignition of PMMA plates [4,5] and others are non-piloted ignition of thin cellulosic sheets [6–9] or PMMA sheets [10] heated by external radiant sources. However, there are few
1540-7489/$ - see front matter Ó 2012 The Combustion Institute. Published by Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.proci.2012.06.064
2666
Y. Takano et al. / Proceedings of the Combustion Institute 34 (2013) 2665–2673
studies except from our research group [11,12] about spontaneous ignition of electric wires caused by Joule heat generated in the wire core. In the previous research [11], the authors reported dramatic extensions of ignition limits in terms of supplied electric current under microgravity. In that research, excess electric current was continuously supplied until ignition occurred. In other research [12], the authors investigated the ignition characteristics with short-term excess electric currents. There, the length of the current supply was selected as the main test parameter to simulate the status of circuit breaker activation shortly after overloading happens, and delayed ignition of electric wires after short-term overloading was observed in microgravity. In the results with short-term electric current supply, the ignition delay may be longer than with a continuous current supply. The electric current is smaller in microgravity than that on the ground as in the continuous current case. The established findings [11,12] indicate the probability of ignition increases in microgravity. One matter that has not been established in the previous work is the limiting oxygen concentration, below which wire ignition does not occur even with excess electric current. A detailed knowledge of this basic parameter is essential for fire safety in space. The present study, investigates the ignition limit; the minimum value of the electric energy supply causing ignition, and LOC; limiting oxygen concentration, below which ignition does not occur at a given electric current supply for short-term excess electric currents under microgravity, by experiments and calculations. The mechanism to determine the ignition limit and the LOC is discussed based on the numerical results. 2. Experiments
with the sample wire inside, a constant current supply system, and an image recording system. The tested sample in Fig. 2 is a nickel–chrome core wire coated with polyethylene. The outer diameter of the sample is 0.8 mm and the inner core diameter is 0.5 mm. The effective length of the sample is 70 mm; the electric resistance of the core wire is 5.5 X/m. The resistance is the value under room temperature, which is used to calculate the Joule energy, E, generated in the core wire under a given electric current. In the tests, the length of time the current is applied, ta, was varied in the range from 0.2 to 1.5 s to simulate delays in the circuit breaker activation. Oxygen concentration and supplied current values were also varied in the tests. Microgravity experiments were conducted at the MGLAB (Microgravity Laboratory of Japan) in Gifu, Japan, which provides a microgravity environment of 105g0 (g0: gravity level on the ground) for around 4.5 s and at the COSMOTORRE in Hokkaido, Japan, which provides 10 3 g0 environment for around 2.7 s. When light emission in gas phase is observed in the video images, it is defined as “ignition” in the experiment and the “Ignition delay”, tig is defined as the time after the start of current supply till the moment of ignition. 2.1. Method of the numerical calculations The numerical calculations are carried out with a model on a one dimensional cylindrical coordinate system. Fig. 3 is a schematic illustration of the ignition model. At the start of the experiment (time zero) current is supplied to the core wire and insulation starts to degrade to emit pyrolysis gas into the oxygen-containing atmosphere. The evolved gas mixes with the ambient oxygen to form a hot combustible mixture eventually leading to spontaneous ignition. The assumptions adopted for the calculations were as follows:
All experiments were performed in the apparatus similar to that used in previous research [12]. An outline of the experimental set-up is shown in Fig. 1. It is composed of a combustion chamber
1. Both core wire and insulation are thermally thin. 2. The solid fuel thickness is held at the initial value throughout the course of the calculations. The density decreases with time as the reactions degrading the solid phase proceed. 3. Gas generated by the degradation is ethylene.
Fig. 1. Outline of the combustion chamber.o
Fig. 2. Schematic details of the sample wire.o
Y. Takano et al. / Proceedings of the Combustion Institute 34 (2013) 2665–2673 Gas phase
r
rs
Insulated wire
rmax
Δr
Surface 1
j-1
j
r
Inert wall j+1
n
Solid phase
Δ r/2
note the gas phase, the i species, and the initial state, respectively. The solid is composed of core wire as the heat source and the insulation as a solid fuel. The sample in the calculations is the same as that used in the experiments, which is assumed as thermally thin. The governing equations for the solid phase can be expressed as follows: Mass conservation
Δr
m_ ¼ V p Fig. 3. Schematics of ignition model.o
The governing equations for the gas phase can be expressed as follows: Mass conservation ð1Þ
Energy conservation @ðqg C p;g T g Þ 1 @ðrqg uC p;g T g Þ þ @t r @r 1 @ @T g rk ¼ þ qg w_ g r @r @r
dT c Ac ¼ Qin Qcp dt dT p Ap ¼ Qcp Qre Qout qc C p;p dt qc C p;c
ð6Þ ð7Þ
where V and A are the insulation volume and sectional area, respectively. Qin, Qcp, Qre, and Qout are the applied energy by the current supply, the heat conduction between core and insulation, the energy lost by degradation and heat conduction between the insulation and gas phase; m_ is the mass reduction rate; and the subscripts c and p denote the core and insulation, respectively. Qin, Qcp, Qre, and Qout are given as below. Qin ¼ Re I 2 Qcp ¼ 2pkp
Qout ¼ 2pk ð2Þ
ð3Þ
State qT ¼ q1 T 1
ð5Þ
Tc Tp ln ððrp =2Þ þ rc Þ=rc
_ Qre ¼ mDH p
Species conservation @ðqg Y i Þ 1 @ðrqg uY i Þ þ @t @r r 1 @ @Y i rqg D vi w_ g ¼ r @r @r
@qp @t
Energy conservation
4. The gas phase consists of the degradation product, oxygen and nitrogen, and the mixture behaves like an ideal gas. 5. Initially, the ambient atmosphere is quiescent. 6. Atmospheric pressure remains constant at 1 atm throughout the simulation. 7. The thermal properties of wire and insulation are considered constant regardless of temperature. 8. The thermal properties and diffusion coefficient of the gas are constant regardless of temperature and species.
@qg 1 @ðrqg uÞ þ ¼0 r @t @r
2667
ð4Þ
where t and r are the time and radius distance, respectively; q, u, T, and Yi are the gas density, velocity, temperature, and species mass fraction of a species i (i = OX: oxygen, F: fuel gas), respectively; and Cp, k, and D are the specific heat, heat conductivity, and mass diffusivity of the mixture, respectively. The mass diffusivity, D, is assumed to be equal for all species. Then, q and m are the heat release by the gas phase reaction and the stoichiometric coefficient, respectively; and finally x is the reaction rate. The subscripts g, i, and 1 de-
Tp T lnðrp =ððrp =2Þ þ rc ÞÞ
where Re, rp, rc are resistance of inner core per unit length, radius of wire (core radius plus insulation thickness) and radius of inner core, respectively. T is temperature of gas phase of the next grid to the surface. DH p is reaction heat of pyrolysis reaction (in this calculation, 1.2 MJ/kg is used). The chemical reaction in the gas phase is described by a single one-step reaction. The pyrolysis reaction in the solid phase is also described by a single one-step reaction. The reaction rates are given as follows: Gas overall reaction C2 H4 þ 3O2 ! 2CO2 þ 2H2 O Eg w_ g ¼ Bg ðqY F ÞðqY OX Þexp RT g
ð8Þ
Solid fuel pyrolysis reaction ½polyethylenesolid ! ½C2 H4 gas Ep m_ ¼ Bp qp V p exp RT p
ð9Þ
2668
Y. Takano et al. / Proceedings of the Combustion Institute 34 (2013) 2665–2673
where B, R, and E are the frequency factor, the universal gas constant, and the activation energy, respectively. The reaction rate constants, Bg = 125 kJ/mol, Eg = 7.143 109 m3/kg s, Bp = 5.0 1018 s1, Ep = 250 kJ/mol, are as reported in Refs. [13,14]. The initial and boundary conditions are as follows: Initial condition in the solid phase T c ¼ T 1; T s ¼ T 1 in the gas phase T ¼ T 1 ; Y F ¼ Y F ;1 ¼ 0; Y OX ¼ Y OX ;1 ; u ¼ 0 Boundary Conditions
Fig. 4. Example of the appearance of wire combustion directly after ignition in microgravity (E = 7.9 J/cm, O2 – 40%, ta = 1.0 s).
at r ¼ rs T surf ¼ T p
@Y F qusurf Y F ;surf þ qD ¼ qusurf @r r¼rs @Y OX qusurf Y OX;surf þ qD ¼0 @r r¼rs
at r ¼ rmax T ¼ T 1 ; Y F ¼ 0; Y OX ¼ Y OX;1 ; u ¼ 0 where the subscript surf denotes the state of the insulation surface. The calculation domain length is 200 mm and the adopted mesh size is 0.2 mm. The time step adopted in this study is 1.0 105 s and calculations are continued until ignition or an elapsed time of 10 s. In previous studies on the ignition of heated solid materials, many types of the ignition criterion have been proposed for maximum temperature, maximum reaction rate, and others [15]. These different types of criteria gave the similar results regarding the effects of the ambient oxygen, and the derived numerical values themselves were not so different [15]. In this study, ignition is defined to occur when the value of the local heat release rate exceeds 30,000 kW/m3. 3. Limiting oxygen concentration (LOC) by experiments Figure 4 shows a photo of wire combustion immediately after ignition in microgravity. The ignition is initiated at some point and then propagates along the wire. This figure indicates the flame has a cylindrical two-dimensional shape surrounding the sample wire in the absence of buoyancy induced convection. Figure 5 shows Mach-Zehnder interference images for the ignition phenomena in microgravity. The supplied energy is 6.7 J/cm with the current applied for 0.5 s at an oxygen concentration of 40%, which is close to the limiting condition
for ignition. The ignition delay here is 4.134 s and images at every 0.5 s selected from the high speed movie are sequentially shown in Fig. 5. In this figure, fringe shift to the left corresponds to a decrease in the refractive index which is generally caused by decreases in the gas density. Directly after the stop of the current supply, at t = 0.5 s, the fringe shift becomes large near the wire surface (fringes are completely parallel normal to the wire sample when t = 0 s). Then, the area of the large fringe shift spreads outward, and after 3.0 s the point of maximum fringe shift locates away from the sample surface. Finally ignition occurs near the point of maximum fringe shift. Figure 6 shows the LOC as a function of the supplied electrical energy. In the figure, the results at and above 21% oxygen concentration were obtained at MGLAB [12] and at oxygen concentrations below 21% at COSMOTORRE. The plots in Fig. 6 were each obtained after a series of tests at a given electrical energy in a step-bystep decrease in the oxygen concentration to establish the minimum oxygen concentration for ignition, and the tests were repeated three times for one point in the plot on the ground. The repeatability is very high with the test and the LOC can be determined within an accuracy of 1% oxygen. The LOC varies with electric energy, becoming smaller with increases in electric energy. The LOC is different for different gravity conditions, with 10.8 J/cm the LOC is 16% in microgravity and 40% at normal gravity, a significant extension of the flammable limit under microgravity. An explanation could be the elimination of natural convection in microgravity, which results in longer residence time of flammable gas and reduced heat loss from the combustion zone. When the energy supply is 22.0 J/cm, the LOC in microgravity is 12% and that at normal gravity is 13%, which suggests the difference in LOC becomes smaller with increases in supplied energy and the LOC seems to converge to value for both
Y. Takano et al. / Proceedings of the Combustion Institute 34 (2013) 2665–2673
2669
60
900
50
800
40
700
30
Tg [K]
Oxygen Concentration [%]
Fig. 5. Fringe Image (lG, E = 6.7 J/cm, O2 – 40%, ta = 0.5 s, tig = 4.134 s). The wire is designated by the white line and the ignition point is designated by the open circle.
Experiment on the ground
20
t=1.8s
600 t=1.2s
500
10 0
(a)
Experiment in microgravity
6
8
10 12 14 16 18 Electrical Energy [J/cm]
400 20
22
300
Fig. 6. Effect of supplied electric energy on LOC in normal gravity and microgravity.
t=1.0s
t=0.5s 0
1
2
3
4 5 r [mm]
900
6
t=1.4s
7
8
(b)
800 t=1.8s
4. Results of the numerical calculations
700
Tg [K]
gravitational conditions at higher electric energies. Here it must be noted that LOC shows significant differences in the low electric energy condition for different gravity conditions, a situation shown for the first time in this paper. In the following section, the ignition mechanism at the low supplied energy condition is discussed based on numerical calculations.
t=2.0s
600 t=1.2s
500 400 300
0
1
t=1.4s
t=1.0s
t=0.5s 2
3
4 5 r [mm]
6
7
8
Fig. 7. Temperature distribution in (a) ignition condition: E = 7.120 J/cm and (b) no ignition condition: E = 7.119 J/cm (O2 – 21%, ta = 1.0 s).
4.1. Temperature distributions with and without ignition Figure 7 shows temperature distributions for two cases calculated with different amounts of energy supplied. (a) is a case with ignition, 7.120 J/cm and (b) is a case without ignition, 7.119 J/ cm. The current was applied for 1.0 s and the ambient oxygen concentration 21%. In both cases, at the cessation of the current supply, 1.0 s, the maximum temperature is at the wire surface. At 1.4 s, the high temperature region has spread outward, as enough pyrolysis gas is evolved into the gas phase and the position of the maximum temperature is located away from the wire surface. Up to 1.8 s, the temperature in Fig. 7(a) is increasing rapidly in the vicinity of the wire surface and ignition occurs. In Fig. 7(b), the temperature decreases after discontinuing the current and ignition does not occur. If the fringe shift in Fig. 5 is mainly caused by the changing
temperature, the trend in the temperature change given in the calculations resembles the experimental result qualitatively. 4.2. Concept of ignition limit Figure 8 shows the calculated history at the ignition point (at the grid where ignition occurs) of reaction heat and heat loss by heat conduction of (a) the ignition condition (E = 7.120 J/cm) and (b) the no ignition condition (E = 7.199 J/cm). The ambient oxygen concentration is 21% and the current is supplied for 1.0 s. In both cases, the reaction heat keeps increasing after stopping the current supply at 1.0 s and at 1.6 s the reaction heat exceeds heat loss in (a). Here the reaction heat becomes predominant and leads to a thermal runaway. In (b), the value of the reaction heat and
Y. Takano et al. / Proceedings of the Combustion Institute 34 (2013) 2665–2673
Heat release
1.0E+05 1.0E+04
1.0E+03 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0 t [s]
1.0E+07 1.0E+06 1.0E+05
Fuel
0.25 Oxygen 0.20
0.60
0.15 0.40
0.10
0.20
7.120J/cm 21%O2 7.001J/cm 21%O2
1
2
3
4 5 r [mm]
6
900
Heat loss Heat release
1.0E+04
800
Fig. 8. History of reaction heat and heat loss in (a) ignition condition: E = 7.120 J/cm and (b) no ignition condition: E = 7.119 J/cm (O2 – 21%, ta = 1.0 s).
heat loss have not changed from the 1.0 s values at 1.6 s, and the value of the reaction heat does not exceed the heat loss leading to a situation without ignition. These results are consistent with the general ignition concept that ignition is determined by the balance between heat production and heat loss. This establishes that for ignition in an overloaded wire the reaction heat and heat loss by conduction behave like in Fig. 8. 4.3. Effect of electric energy and oxygen concentration on wire ignition Figure 9 shows (a) the fuel and oxygen distribution and (b) the temperature and reaction rate distribution for two different supplied energy cases. The supplied electric energy is 7.210 J/cm or 7.001 J/cm, the ambient oxygen concentration is 21%, and the current is applied for 1.0 s. The fuel layer in (a) becomes thicker in the gas phase with increase in electric energy because more pyrolysis gas is evolved into the gas phase with the higher electric energy supply. After the current supply is discontinued at 1.0 s, the maximum temperatures are very similar but the temperature in the ignition case is higher at 1.8 s. The maximum values of the reaction rates are similar at 1.0 s but there is a big difference right before the ignition, that is, the reaction rate at 1.8 s in the high electric energy condition is larger than that in the low condition.
7
(b)
Tg
8 0.00
1.00E+06
1.00E+02 7.120J/cm 21%O2 7.001J/cm 21%O2 1.00E+00
600 500 RR
300 0
0.05
1.00E+04
t=1.0s
700
1.00E-02
t=1.8s
400
1.0E+03 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0 t [s]
0.30
(a)
t=1.8s
0.80
0.00 0
(b)
t=1.0s
1
2
3
4 5 r [mm]
6
7
Oxygen concentration
Heat loss
1.0E+06
1.00
Ignition
Reaction rate [kg/m3s]
(a)
Fuel concentration
1.0E+07
Tg [K]
Reaction heat or heat loss [W/m3]
Reaction heat or heat loss [W/m3]
2670
8 1.00E-04
Fig. 9. Distributions of (a) fuel and oxygen concentration and (b) temperature and reaction rate around the ignition limit (E = 7.210 J/cm and 7.001 J/cm, O2 – 21%, ta = 1.0 s, in ignition condition tig = 1.801 s).
Figure 10 hows (a) the fuel and oxygen distribution, and (b) the temperature and reaction rate distribution for two different oxygen concentration conditions, 21% or 18%, the supplied electrical energy is 7.210 J/cm and the current applied for 1.0 s. The temperatures at 1.0 s with 21% and 18% oxygen concentration are very similar, while distributions of fuel and oxygen concentrations at 1.0 are different with non-negligible level. This difference causes big differences of temperature and reaction rate at 1.8 s, that is, temperature and reaction rate become larger in the high oxygen concentration. In low electric energy condition, fuel layer becomes thin and it results in increases in air diffusion to the wire surface. Then, gas temperatures in the vicinity of the wire surface decreases, because the surrounding air temperature is lower than the fuel temperature. It causes decreases in the reaction rate. As changing electric energy supply, the thickness of the fuel layer changes. Then, the diffusion time of the surrounding air to the wire surface changes. If the diffusion time becomes shorter, the gas temperature decreases in a short time near the solid surface. So the reaction needs to be faster to achieve ignition, which requires higher oxygen concentrations. When the oxygen concentration is low, the increase rate of the reaction rate becomes smaller. So it needs a longer diffusion time to allow a longer reaction time.
Y. Takano et al. / Proceedings of the Combustion Institute 34 (2013) 2665–2673
(a)
t=1.8s
Oxygen 0.25 0.20
0.60
0.15 0.40
0.10
0.20 0.00
0
1
2
3
4 5 r [mm]
6
900
Tg [K]
7
(b)
Tg
800
8
0.00
1.00E+06 1.00E+04
700
t=1.0s
1.00E+02 7.120J/cm 21%O2 7.120J/cm 18%O2
600 500 RR
1.00E+00
t=1.8s
400 300
0.05
7.120J/cm 21%O2 7.120J/cm 18%O2
0
1
2
3
4 5 r [mm]
1.00E-02 6
7
8
1.00E-04
Fig. 10. Distributions of (a) fuel and oxygen concentration and (b) temperature and reaction rate around the LOC (E = 7.210 J/cm, O2 – 21% and 18%, ta = 1.0 s, in ignition condition tig = 1.801 s).
From these figures, it is noted that no ignition occurs in case where the peak of the reaction rate and the highest temperature point gets very close to the wire surface. As shown in Figs. 5 and 7, the highest temperature point is away from the wire surface when ignition occurs. In such situations, ignition occurs and the heat release keeps exceeding the heat loss until the temperature reaches the one for a thermal runaway. As seen in Fig. 8(a) it takes not so short time to reach ignition. As time elapsed, the highest temperature point shifts to the wire surface, because air diffusion proceeds toward the wire surface. Once the maximum heat release point becomes very close to the wire surface, the chance for ignition is lost because the wire acts as a heat sink and temperature gradient around the maximum heat release point becomes very steep. Then, heat release cannot exceed the heat loss. This is a unique feature of wire ignition by short-term current supply because the ignition delay is longer than the current supply time and wire never works as a heat source at the moment of ignition near the limiting condition. Under these considerations, competition of diffusion time with the delay time for ignition defined as in Fig. 8(a) is very important to understand the onset of ignition. When the supplied electric energy is larger, the thickness of the initial fuel layer increases as shown in Fig. 11. Then the diffusion time increases, which allows a longer ignition delay and lower LOC. When the
oxygen concentration is higher, the ignition delay decreases, which allows a shorter diffusion time corresponding to less supplied energy for ignition. The thickness of the fuel layer right after stopping the current supply is shown in Fig. 11 for some supplied electric energies. The current is applied for 1.0 s and the ambient oxygen concentration is 21%. In this figure, it is obvious that the fuel layer becomes thicker, when the supplied electrical energy increases. Therefore, the diffusion time increases with increases in supplied energy. Then lower LOC is expected with increases in the supplied energy as discussed above. 4.4. Limiting conditions for ignition The calculated LOC as a function of supplied electrical energy is shown in Fig. 12. The results estimated by the numerical calculations showed a similar behavior as the experiments. But a quantitative comparison shows a smaller LOC for the calculations than for the experimental results. One of the causes of the difference may be the radiation heat loss effect, which is not considered in the calculations, and more study is needed on this difference. In this figure, it is obvious that the LOC is strongly dependent on the electric energy at low electric energies, while it depends weakly on the electric energy in the higher energy conditions. It particular, the LOC becomes higher drastically in cases below 7.0 J/cm in the calculated results. Figure 13 shows the ignition limit as a function of the current supply time. This figure shows that in microgravity the ignition limit is almost constant under a constant oxygen concentration. The same has been observed in microgravity experiments [12]. It is indicated that there is specific supplied energy for ignition under a given oxygen concentration. The calculations show the effect of the current supply time, that is, the ignition limit for a 1.0 s of current supply duration
1.00
Fuel concentration
Fuel concentration
0.80
0.30
t=1.0s
Oxygen concentration
Fuel
Reaction rate [kg/m3s]
1.00
2671
At t = 1.0 s (= t a)
0.80 0.60
7.710J/cm 7.400J/cm
0.40 0.20
7.120J/cm
0.00 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 r [mm] Fig. 11. Thickness of fuel layer for different supplied electric energy amounts (E = 7.120 J/cm, 7.400 J/cm, 7.710 J/cm, O2 – 21%, ta = 1.0 s).
Oxygen Concentration [%]
2672
Y. Takano et al. / Proceedings of the Combustion Institute 34 (2013) 2665–2673
60 50 40 30 20
Experiment in microgravity
10 Calculation in microgravity
0
6
8
10
12 14 16 18 Electrical Energy [J/cm]
20
22
Fig. 12. Limiting Oxygen concentrations (LOC) for experiments and calculations.
Ignition limit [J/cm]
8 O2 - 21% 7
6
5 0.0
O2 - 40%
0.2
0.4
O2 - 60%
0.6
0.8 1.0 t a [s]
1.2
1.4
1.6
(1) The LOC in microgravity is much lower than in normal gravity when the supplied energy is low (<14 J/cm), while the value at higher supplied energies in microgravity becomes closer to the value in normal gravity. (2) The thickness of the fuel layer is an important factor for wire ignition phenomena. The thickness of the fuel layer determines the diffusion time in which oxidation reaction proceeds and ignition occurs. When the fuel layer becomes thin, the diffusion time becomes short and the gas temperature decreases in a short time. To achieve ignition reaction needs to be faster, which requires higher oxygen concentrations. (3) The results estimated by the numerical calculations showed similar dependence of LOC on electric energy as the experiments, suggesting the importance of competition between diffusion time and ignition delay time. (4) The ignition limit is almost constant under a constant oxygen concentration even when the current supply duration is different. It is indicated that the supplied energy is the main controlling factor for ignition under a given oxygen concentration.
Fig. 13. Effect of the current supply time on the ignition limit.
Acknowledgments is 1.5% higher, that for 1.5 s is 3.5% higher, than that for a 0.2 s under various oxygen concentrations. According to the results the effect of supply time can be disregarded below 1.0 s. When the current supply duration changes under a constant supplied energy to the wire, the heating rate of wire changes. So temperature distribution around the wire would change a little. In spite of the change, the ignition limit under a certain oxygen concentration is almost the same in these conditions. Therefore, supplied energy is the dominant factor to control the ignition for short-term current supply once oxygen concentration is fixed.
5. Conclusions Ignition phenomena for electric wires with short-term excess electric current supply were investigated by microgravity experiments and numerical calculations to understand the ignition characteristics such as ignition limit and LOC (limiting oxygen concentration). The conclusions may be summarized as follows:
This research is partially supported by JAXA as the candidate experiments for the second phase utilization of JEM/ISS entitled “Quantitative Description of Gravity Impact on Solid Material Flammability as a base of Fire Safety in Space” and is also supported by JAXA Research Working Group to promote space utilization.
References [1] R. Friedman, Fire Mater. 20 (1996) 235–243. [2] NASA Office of Safety and Mission Quality, Flammability, Odor, Off Gassing and Compatibility Requirements and Test Procedures for Materials in Environments that Support Combustion, NHB8060.1C, 1991. [3] G. Ruff, D. Urban, M. King, Proceedings of the 43rd AIAA Aerospace Sciences Meeting and Exhibit, Reno, NV. 2005-341, 2005. [4] S. McAllister, C. Fernandez-Pello, D. Urban, G. Ruff, Proc. Combust. Inst. 32 (2009) 2453–2459. [5] S. McAllister, C. Fernandez-Pello, D. Urban, G. Ruff, Combust. Flame 157 (2010) 1753–1759. [6] O. Fujita, J. Takahashi, K. Ito, Proc. Combust. Inst. 28 (2000) 2761–2768. [7] J. Takahashi, O. Fujita, K. Ito, Proc. Combust. Inst. 30 (2005) 2311–2317.
Y. Takano et al. / Proceedings of the Combustion Institute 34 (2013) 2665–2673 [8] Y. Nakamura, H. Yamashita, T. Takeno, G. Kushida, Combust. Flame 120 (2000) 34–48. [9] S.L. Olson, T. Kashiwagi, O. Fujita, M. Kikuchi, K. Ito, Combust. Flame 125 (2001) 852–864. [10] Y. Nakamura, T. Kashiwagi, S.L. Olson, K. Nishizawa, O. Fujita, K. Ito, Proc. Combust. Inst. 30 (2005) 2319–2325. [11] K. Agata, O. Fujita, Y. Ichimura, T. Fujii, H. Ito, Y. Nakamura, J. Jpn. Soc. Microgravity Appl. 25 (1) (2008) 11–16.
2673
[12] O. Fujita, T. Kyono, Y. Kido, Y. Nakamura, H. Ito, Proc. Combust. Inst. 33 (2011) 2617–2623. [13] K. Nishizawa, Study on Flame Spread over Wire Insulation for Improve Fire Safety in a Spaceship, Ph.D. thesis, Graduate School of Engineering, Hokkaido University, Hokkaido, Japan, 2004.. [14] C.K. Westbrook, F.L. Dryer, Combust. Sci. Technol. 27 (1981) 31–43. [15] T. Kashiwagi, Combust. Sci. Technol. 8 (1974) 225– 236.