- Email: [email protected]
methane diffusion flames
JID: PROCI
ARTICLE IN PRESS
[m;October 6, 2016;16:38]
Available online at www.sciencedirect.com
Proceedings of the Combustion Institute 000 (2016) 1–8 www.elsevier.com/locate/proci
Flame base structures of micro-jet hydrogen/methane diffusion flames Jian Gao, Akter Hossain, Yuji Nakamura∗ Department of Mechanical Engineering, Toyohashi University of Technology, 1-1 Hibarigaoka, Tempaku, Toyohashi 441-8580, Japan Received 4 December 2015; accepted 12 August 2016 Available online xxx
Abstract In this study, a comparison of flame base structures of hydrogen/methane–air diffusion flames formed over a tiny-jet is made numerically for both isothermal and thermal conductive burner conditions, in order to clarify the fuel dependent flame stabilization mechanisms. It is found that, unlike a methane flame, the flame base of a hydrogen flame always attaches to the burner. The analyses indicate that the dominant intermediateconsumption steps have significantly lower activation energies for the hydrogen flame as compared to a methane flame. More importantly, one of the HO2 production reactions (R43f: H + O2 + M → HO2 + M), which has a dominant role in sustaining reactivity at the flame base, shows a negative temperature dependence, causing the heat release rate in the flame base kernel to increase as the burner wall temperature decreases. With a thermal conductive burner (thermal conductivity of 16 W/m-K) over a wide range of fuel jet velocities (0.5– 4.0 m/s), it is found that the burner tip is heated to a significantly higher temperature by a hydrogen flame due to its unique stabilization mechanism. The mixing effects of hydrogen and methane are then considered. It is found that the burner tip temperature can be reduced by adding methane into the fuel flow. This is because, according to the investigation of the structures of the hydrogen/methane jet diffusion flames, the reaction rate of R43f is suppressed due to the included intermediates (e.g., CH3 , CH2 O) consumption steps of methane. It is expected that the flame attachment feature associated with the flame base structure can be easily controlled by mixing hydrogen and methane, making it possible to control the burner tip temperature in advance. © 2016 by The Combustion Institute. Published by Elsevier Inc. Keywords: Micro flame; Extinction; Jet diffusion flame; Hydrogen flame; Flame-burner interaction
1. Introduction Given that hydrocarbon and oxygenated hydrocarbon fuels have much higher specific energy densities than batteries (e.g., lithium ion batteries) ∗
Corresponding author. Fax: +81 532 44 6647. E-mail address: [email protected] (Y. Nakamura).
[1], development of combustion based micro scale power generation systems (such as micro scale engines and turbines) are especially favored to meet the increasing demands for portable power generators, micro-satellite thrusters, micro unmanned aircrafts, micro reactors, micro sensors, and others. In order to design a compact, stable, and long life combustor or burner for those devices, a deep understanding of the fundamentals of micro scale
http://dx.doi.org/10.1016/j.proci.2016.08.034 1540-7489 © 2016 by The Combustion Institute. Published by Elsevier Inc.
Please cite this article as: J. Gao et al., Flame base structures of micro-jet hydrogen/methane diffusion flames, Proceedings of the Combustion Institute (2016), http://dx.doi.org/10.1016/j.proci.2016.08.034
JID: PROCI
2
ARTICLE IN PRESS
[m;October 6, 2016;16:38]
J. Gao et al. / Proceedings of the Combustion Institute 000 (2016) 1–8
combustion is essential. Both premixed and diffusion combustion strategies should be potential options to utilize thermal energy, while according to the recent reviews [2,3] studies regarding micro diffusion flames are relatively less compared with those for premixed flames. The micro jet diffusion flame was first studied by Ban et al. [4] about twenty years ago. Fuel was ejected into a quiescent air environment through a submillimeter-diameter burner. Such a tiny flame is stabilized by the burner, and controlled by diffusion and convection whilst the buoyancy effect is minor. Additionally, due to the “inherent” large surface to volume ratio (S/V) feature in miniaturization, flame-burner interaction plays a significant role in the stability of the microflames. For this reason, the stabilization mechanism of micro-jet diffusion flames has been studied by numerous researchers [5–17]. In general, local extinction occurs in the vicinity of the burner wall, and hence there exists a gap (flame quenching zone) between the flame base and the burner wall, as observed in the experiments by Cheng et al. [8,10,11], Kuwana et al. [13] and Fujiwara et al. [14] and Fujiwara and Nakamura [15] for methane flames. Such a gap has been also clearly observed in micro-jet diffusion flames for general carbon-contained fuels, such as ethane, ethylene, and acetylene [4], propane [5], and ethanol [17]. On the contrary, as notified in the previous studies by Cheng et al. [6,9] and our recent numerical work [16] using hydrogen as the fuel, it seems that a hydrogen diffusion flame readily attaches to the burner wall and local extinction may not occur in the vicinity of the burner wall, invoking that the stabilization mechanism for a hydrogen flame may differ from that of a methane flame. Once the flame is fully attached to the burner, substantial heat can be transported (recirculated) to the fuel to improve the flame stability. On the other hand, the burner tip temperature can increase dramatically, causing structural damage (thermal damage to burner). In this regard, a strategy to control the flame base structure (thus the burner tip temperature) is important, especially for microflame technology. The objective of this study is to clarify the stabilization mechanism of micro-jet hydrogen/methane diffusion flames by numerical approaches. Computations are performed by adopting both isothermal and thermal conductive burner wall boundary conditions to understand the effect of burner wall temperature on the flame attachment mechanism. Fuel mixing effects on the flame base structure are also examined and a strategy to control the burner tip temperature and the degree of heat recirculation is suggested. 2. Numerical approach The numerical method is similar to our recent work [16]. A set of governing equations of mass, momentum, energy and species are solved
Fig. 1. Schematic of computation domain and boundary conditions.
in a two-dimensional (r–z) cylindrical coordinate by the commercial software FLUENT 14.5 CFD [18] based on the finite volume method (FVM) with the SIMPLE algorithm. Figure 1 illustrates the schematic of the computation domain and boundary conditions. Fuel is ejected into the ambient air through a tube (burner) with an inner diameter of 0.8 mm and an outer diameter of 1.2 mm. The pressure outlet boundary conditions are imposed at the far field boundaries. The burner wall is assumed to be non-slip, and chemically inert (because the recent study by Kizaki et al. [19] has proved that the effects of surface quenching reactions are minor at normal pressure; catalytic surface reaction assisted diffusion flames could be our future work however it is not considered in this study). Two kinds of thermal boundary conditions are considered at the burner wall. For the isothermal burner conditions, the inner surface and the top surface of the burner wall are assumed to be adiabatic, while the outer surface is set to a temperature of Tb , which varies from 300 K to 1800 K, because in reality a flame shall attach (if it can) to the outer surface of the burner wall, and therefore the outer surface temperature is critical for the flame-burner attachment. For the thermal conductive burner conditions, grid is put in the burner (solid) zone, namely, the thermal conduction equation in the burner zone (solid) is solved as well. Conjugate thermal boundaries are set at the inner surface, the outer surface and the top surface of the wall, while 300 K is imposed at the bottom of the burner wall to allow the heat loss from the domain. Although radiation may more or less affect flame temperatures at near extinction conditions, it has been shown that [8,9] the microjet diffusion flame structures could be predicted satisfactorily by the numerical models excluding radiation heat loss. Thus, radiation effects are also
Please cite this article as: J. Gao et al., Flame base structures of micro-jet hydrogen/methane diffusion flames, Proceedings of the Combustion Institute (2016), http://dx.doi.org/10.1016/j.proci.2016.08.034
JID: PROCI
ARTICLE IN PRESS
[m;October 6, 2016;16:38]
J. Gao et al. / Proceedings of the Combustion Institute 000 (2016) 1–8
3
Fig. 2. Computed temperature contours and heat release rate (HRR) isopleths for CH4 flames at fuel jet velocity of Vf = 0.4 m/s for (a) Tb = 1800 K (b) Tb = 300 K, and for H2 flames at Vf = 1.0 m/s for (c) Tb = 1800 K (d) Tb = 300 K.
neglected in this study considering that our concern here is the significant differences in flame structure between methane and hydrogen flames. A semi-detailed mechanism including 17 species and 58 steps [20] is employed for the chemical kinetics for the flames of methane, hydrogen, and their mixtures. The mechanism, thermodynamic and transport input files are available in the CHEMKIN database [21]. The species transport properties are evaluated using mixture averaged formulas. At first, we performed several computations to simulate the previous experiments in [9,10,15]. It was confirmed that both the methane flame and hydrogen flame structures can be well predicted by this mechanism. The total number of grid points are 103 and 168 in radial and axial directions, respectively, over a physical domain of 15 mm× 25 mm. The minimum grid spacing of 0.025 mm is placed near the exit of the burner, and gradually stretched toward the far boundaries. No apparent difference in the computation results (flame structure, maximum temperature, etc.) is observed using a nearly double finer grid system, indicating the present grid system is fine enough. Time-dependent calculations are performed to obtain steady state solutions. At first, a high temperature spot (∼2000 K) is placed near the burner exit to achieve ignition, and then it is removed. A steady state is declared until the fluctuation in the maximum temperature is less than 2 K. 3. Isothermal burner 3.1. Comparison of methane and hydrogen flames Figure 2 shows the computed temperature contours and heat release rate (HRR) isopleths for CH4
flames at a fuel jet velocity of Vf = 0.4 m/s for (a) Tb = 1800 K (b) Tb = 300 K, and for H2 flames at Vf = 1.0 m/s for (c) Tb = 1800 K (d) Tb = 300 K. At Tb = 1800 K (Fig. 2a), the CH4 flame attaches to the outer surface of the burner. As Tb decreases from 1800 K to 300 K, the flame temperature (defined as the maximum gas temperature in the computation domain in this study) decreases from 2206 K to 1966 K, and the maximum HRR decreases from 1.3E9 W/m3 to 7.4E8 W/m3 . As expected, at Tb = 300 K, local extinction occurs in the vicinity of the outer surface of the burner wall, resulting in a gap (flame quenching zone) between the burner and the stabilized flame (see Fig. 2b). On the contrary, the H2 flame shows a different behavior compared to the methane flame. As Tb decreases from 1800 K to 300 K, although the flame temperature decreases from 2091 K to 1770 K, local extinction does not occur and the flame base kernel keeps attaching to the outer surface of the burner wall (Fig. 2c and d). Moreover, the HRR in the flame base kernel significantly increases from 7.5E8 W/m3 to 2.8E9 W/m3 as Tb decreases from 1800 K to 300 K, namely, the flame base kernel reactivity shows a negative temperature dependence. 3.2.. The predominant reactions Figure 3 shows the heat release rates from elementary reactions in the flame base kernels for both the CH4 flame and the H2 flame at Tb = 300 K. For the CH4 flame, the reactions related to CH3 radical and CH2 O are the main contributors to the total HRR. The reaction R6f (CH3 + O → CH2 O + H) predominately contributes to the total HRR, agreeing with the previous study by Takahashi et al. [22].
Please cite this article as: J. Gao et al., Flame base structures of micro-jet hydrogen/methane diffusion flames, Proceedings of the Combustion Institute (2016), http://dx.doi.org/10.1016/j.proci.2016.08.034
JID: PROCI
4
ARTICLE IN PRESS
[m;October 6, 2016;16:38]
J. Gao et al. / Proceedings of the Combustion Institute 000 (2016) 1–8
Fig. 3. Comparison of HRRs from elementary reactions in the flame base kernels for the CH4 flame (in Fig. 2b) and the H2 flame (in Fig. 2d) at Tb = 300 K.
It is noticed that, most of the listed reactions for the CH4 flame are chain-terminating (e.g., R1f, R6f, R7f, R26f and R27f), and the activation energy for the reaction R27f (CH2 O + H → HCO + H2 ) is 12,600 J/mole, corresponding to an activation temperature of 1500 K. Therefore, a high temperature is essential in order to sustain the reactivity in the flame base kernel, and that the flame quenching zone exists in the vicinity of a lower temperature burner wall (e.g., in Fig. 2b for Tb = 300 K). For the H2 flame, HRRs from the reactions R52f (H + OH + M → H2 O + M), R43f (H + O2 + M → HO2 + M), and R45f (H + HO2 → 2OH) are significantly higher than other reactions. Note that, the activation energies of R52f and R43f are zero, and the activation energy of R45f is 4506 J/mole (corresponding to an activation temperature of 540 K), which is much lower than that of R27f in the CH4 flame. Further, R43f and R45f act as a chain-propagation reaction pathway, which could sustain the reactivity in the flame base kernel. The initiation reaction R43f is a recombination reaction requiring a third body, and its reaction rate constant is expressed as k = 3.6 × 1017 T−0.72 exp(0/T) cm3 mol−1 s−1 , showing a negative temperature dependence. According to the computation results, as Tb decreases from 1800 K to 300 K, the net reaction rate of R43 in the flame base kernel increases from 0.8 kmol/m3 -s to 4.0 kmol/m3 -s, showing the same trend as for the total HRR (see Fig. 2c and d). It is implied that R43 might be the dominant reaction for the reactivity in the flame base kernel.
3.3. Verification by a modified kinetic model In order to verify that R43f is the dominant reaction for the reactivity in the flame base kernel for a H2 flame, additional computation is performed using a modified kinetic model, in which the reaction rate constant of the reaction R43f is revised from k = 3.6 × 1017 T−0.72 exp(0/T) cm3 mol−1 s−1 to k = 3.6 × 1017 T0 exp(−5035/T) cm3 mol−1 s−1 . In this way, the negative temperature dependence of the reactivity of R43f is lost, and a high temperature is required to sustain its reactivity due to the elevated activation energy. The computed temperature contours and HRR isopleths are shown in Fig. 4 for (a) Tb = 1800 K and (b) Tb = 300 K. It can be seen that, such a “H2 flame” attaches to the burner at Tb = 1800 K, whereas extinction occurs in the vicinity of the burner for Tb = 300 K, and hence a quenching gap exists, namely, such a “H2 flame” shows a similar behavior to the CH4 flame in Fig. 2a and b. Now, it is verified that the reaction R43 plays a dominant role in sustaining the flame base kernel reactivity for a H2 flame. It also can be concluded that the unique stabilization mechanism of H2 flames is caused by its kinetics, but not other factors such as its higher mass diffusivity. In this paper the critical role of the reaction R43f (H + O2 + M → HO2 + M) in determining the flame base structure, making a hydrogen flame always attach to the burner is revealed. In addition, the effects of recombination reactions (such as R52f and R43f) on hydrogen flame-wall interaction have also been reported by several other
Please cite this article as: J. Gao et al., Flame base structures of micro-jet hydrogen/methane diffusion flames, Proceedings of the Combustion Institute (2016), http://dx.doi.org/10.1016/j.proci.2016.08.034
ARTICLE IN PRESS
JID: PROCI
[m;October 6, 2016;16:38]
J. Gao et al. / Proceedings of the Combustion Institute 000 (2016) 1–8
Fig. 4. Computed temperature contours and HRR isopleths for “H2 flames” at Vf = 1.0 m/s for (a) Tb = 1800 K and (b) Tb = 300 K, using a modified kinetic model, in which the rate constant of the reaction R43f is revised from k = 3.6 × 1017 T−0.72 exp(0/T) cm3 mol−1 s−1 to k = 3.6 × 1017 T0 exp(−5035/T) cm3 mol−1 s−1 .
researchers. Through one-dimensional numerical computations for premixed hydrogen/air flames impinging on isothermal walls, Aghalayam and Vlachos [23] indicated that the reaction R43f would weaken the thermal coupling between flame and wall, and decrease the H concentration to increase the NO2 concentration near an inert wall. Also by employing a simplified one-dimensional configuration, Dabireau et al. [24] studied the transient quenching processes for hydrogen/oxygen premixed and diffusion flames near a low-temperature wall (700 K), showing that the low-activation-energy recombination reactions could lead to high fluxes to the wall just before flame quenching. Further, Gruber et al. [25] found that, even for a turbulent hydrogen/air premixed flame, these recombination reactions also play an important role on flame-wall convective heat transfer and could contribute up to 70% to the total heat release rate at a cold wall. Beyond agreements with these previous studies, the two-dimensional computation results in our study imply that, such effect is more pronounced and could dominate the structure of a microscale hydrogen diffusion flame. Probably, this is because of the enlarged S/V ratio and that sufficient H and O2 could diffuse to the flame base kernel to sustain the reactivity there. 4. Thermal conductive burner 4.1. Structure of hydrogen/methane flames For the thermal conductive burner conditions, the burner material is assumed to be stainless
5
steel and the thermal conductivity is set to λb = 16 W/m K. Figure 5 shows the computed temperature contours and HRR isopleths for (a) H2 flame at Vf = 2.0 m/s, (b) H2 /CH4 80%/20% flame at Vf = 1.2 m/s, (c) H2 /CH4 20%/80% flame at Vf = 0.6 m/s, and (d) CH4 flame at Vf = 0.5 m/s. Note that, for comparison purpose, we have adjusted the fuel jet velocities to have close flame temperatures (2028– 2034 K). For the CH4 flame (Fig. 5d), due to the heat loss through the burner wall, a quenching zone is observed in the vicinity of the burner. For the H2 flame (Fig. 5a), as described above in Section 3, the flame attaches to the burner. From Fig. 5b and c, one can see that, as CH4 is added into H2 , the flame detaches from the burner wall. Figure 6 compares the flame structures of (a) H2 flame at Vf = 2.0 m/s, and (b) H2 /CH4 80%/20% flame at Vf = 1.2 m/s. It is shown that, compared to the H2 flame, the mole fraction of the H radical is dramatically reduced in the H2 /CH4 80%/20% flame. Comparison of the contours of the net production rate of H radical (the right sides in Fig. 6a and b) indicates that, a large amount of H radicals are consumed in the fuel side as CH4 is added. Figure 7 shows the computed radial profiles of the net reaction rates of R6, R27, R43, and R45 at z = 10 mm for the H2 flame and the H2 /CH4 80%/20% flame. For the H2 /CH4 80%/20% flame, R6 and R27 are reactive in the fuel side (from r = 0.6 mm to 1.5 mm), where the net reaction rates of R43 and R45 are decreased compared to those for the H2 flame. This might be because the intermediates (CH3 and CH2 O) consumption reactions (such as R6 and R27) for CH4 consume a certain amount of H radicals in the fuel side, and thus the reaction rates of R43 and R45 are both suppressed accordingly. As shown in Fig. 6 (isopleths in the left sides), the net reaction rate of R43 in the flame base kernel is significantly reduced from 1.6 kmol/m3 -s to 1.0 kmol/m3 -s as 20% CH4 is added. Consequently, the negative temperature dependence of the reactivity in the flame base kernel is weakened. Further increasing the volumetric fraction of CH4 in the fuel flow can make the flame detach from the burner (as shown in Fig. 5c for the H2 /CH4 20%/80% flame). 4.2. Controlling the burner tip temperature Figure 8 shows the temperature profiles on the axis for various H2 /CH4 mixtures. The dashed line at z = 10 mm indicates the location of the burner exit. It is shown that, as the volumetric fraction of H2 increases, the temperature inside the burner for the fuel flow increases. This is because the flameburner attachment is enhanced as H2 is added, producing a higher burner tip temperature (as shown in Fig. 5), and consequently more heat is recirculated into the fuel flow. Therefore, adding H2 into CH4 can elevate the heat recirculation degree, and flame stability is expected to be improved.
Please cite this article as: J. Gao et al., Flame base structures of micro-jet hydrogen/methane diffusion flames, Proceedings of the Combustion Institute (2016), http://dx.doi.org/10.1016/j.proci.2016.08.034
JID: PROCI
6
ARTICLE IN PRESS
[m;October 6, 2016;16:38]
J. Gao et al. / Proceedings of the Combustion Institute 000 (2016) 1–8
Fig. 5. Computed temperature contours and HRR isopleths at the thermal conductive burner condition (the thermal conductivity of burner wall is λb = 16 W/m-K) for (a) H2 flame at Vf = 2.0 m/s, (b) H2 /CH4 80%/20% flame at Vf = 1.2 m/s, (c) H2 /CH4 20%/80% flame at Vf = 0.6 m/s, (d) CH4 flame at Vf = 0.5 m/s. Here the flame temperatures range between 2028 K and 2034 K.
Fig. 6. Computed contours of H radical mole fraction, isopleths of the net reaction rate of R43, and isopleths of the net production rate of H radical at the thermal conductive burner condition (λb = 16 W/m-K), for (a) H2 flame at Vf = 2.0 m/s, (b) H2 /CH4 80%/20% flame at Vf = 1.2 m/s.
Finally, the relationship between flame temperatures and burner tip temperatures is summarized for the H2 flame, the H2 /CH4 80%/20% flame, and the H2 /CH4 60%/40% flame over the fuel jet velocity range from Vf = 0.5 m/s to 4.0 m/s (the Reynolds number from 3.8 to 91.9) in Fig. 9. As shown in the figure, for each fuel mixture, as Vf increases, the flame temperature increases, while the burner tip temperature first increases and then decreases, namely, there exists a maximum burner tip temperature for each kind of fuel composition. It clearly shows that the maximum burner tip temperature
is decreased as the CH4 concentration increases. Therefore, adding CH4 into H2 can reduce the burner temperature and effectively prevent thermal damage to the burner tip, which could be a potential problem in a hydrogen microflame.
5. Conclusions In this study, we perform a systematic numerical study to investigate the stabilization mechanisms of micro-jet hydrogen/methane diffusion
Please cite this article as: J. Gao et al., Flame base structures of micro-jet hydrogen/methane diffusion flames, Proceedings of the Combustion Institute (2016), http://dx.doi.org/10.1016/j.proci.2016.08.034
JID: PROCI
ARTICLE IN PRESS J. Gao et al. / Proceedings of the Combustion Institute 000 (2016) 1–8
Fig. 7. Comparison of radial profiles of the net reaction rates of R6, R27, R43, and R45 at z = 10 mm (burner exit) for the H2 flame at Vf = 2.0 m/s, and the H2 /CH4 80%/20% flame at Vf =1.2 m/s.
[m;October 6, 2016;16:38]
7
flames. For the hydrogen flame, it is understood that the dominant intermediate-consumption steps have significantly lower activation energies compared to the methane flame. The reaction R43f (H + O2 + M → HO2 + M), which plays a dominant role in sustaining the reactivity at the flame base, shows a negative temperature dependence, causing the heat release rate in the flame base kernel to increase as the burner wall temperature decreases. This allows the hydrogen flame base to remain attached to the burner as the burner wall temperature decreases from 1800 K to 300 K. Further computations are then performed in order to clarify the effect of fuel mixing (hydrogen and methane). It is shown that the burner tip temperature can be controllable by replacing some amount of hydrogen with methane, due to the inhibition effect on R43f by the included intermediates (e.g., CH3 , CH2 O) consumption steps of methane. It is suggested that mixing hydrogen and methane could be an effective strategy to enhance heat recirculation or prevent potential thermal damage to the burner, thus improving the combustion performance in micro diffusion flames. Acknowledgments
Fig. 8. Temperature profiles on the axis for H2 /CH4 flames. (1): H2 flame at Vf = 2.0 m/s, (2): H2 /CH4 80%/20% flame at Vf = 1.2 m/s, (3) H2 /CH4 60%/40% flame at Vf = 0.85 m/s, (4): H2 /CH4 40%/60% flame at Vf = 0.7 m/s, (5): H2 /CH4 20%/80% flame at Vf = 0.6 m/s, (6) CH4 flame at Vf = 0.5 m/s. (Here flame temperatures are within the range from 2028 K to 2034 K).
This work is supported by Research Foundation for the Electro-technology of Chubu (RFEC), Iwatani-Naoji Memorial Foundation, Asahi Glass Foundation, Grant-in-Aid for Scientific Research(B) (no. 26289041). Fruitful discussion with Dr. Umeda (Toho Gas Co. Ltd.) through the joint research project on “Hybrid hydrogen energy supply system” organized by Chubu Region Institute for Social and Research is greatly appreciated. The authors also acknowledge Mr. Lindsay Prescott (Technical Advisor at TUT-RAC) for his language editing. References
Fig. 9. Burner tip temperatures as a function of flame temperature for the H2 flame, the H2 /CH4 80%/20% flame, and the H2 /CH4 60%/40% flame over the fuel jet velocity range from Vf =0.5 m/s to 4.0 m/s.
[1] A.C. Fernandez-Pello, Proc. Combust. Inst. 29 (2002) 883–899. [2] K. Maruta, Proc. Combust. Inst. 33 (2011) 125–150. [3] Y. Ju, K. Maruta, Prog. Energy Combust. Sci. 37 (2011) 669–715. [4] H. Ban, S. Venkatesh, K. Saito, J. Heat Transf. 116 (4) (1994) 954–959. [5] L.M. Matta, Y. Neumeier, B. Lemon, B.T. Zinn, Proc. Combust. Inst. 29 (2002) 933–939. [6] T.S. Cheng, Y.C. Chao, C.Y. Wu, et al., Proc. Combust. Inst. 30 (2005) 2489–2497. [7] Y. Nakamura, H. Yamashita, K. Saito, Combust. Theory Model. 10 (6) (2006) 927–938. [8] T. Cheng, C.-P. Chen, C.-S. Chen, Y.-H. Li, C.-Y. Wu, Y.-C. Chao, Combust. Theory Model. 10 (5) (2006) 861–881. [9] T.S. Cheng, C.Y. Wu, C.P. Chen, et al., Combust. Flame 146 (1–2) (2006) 268–282. [10] C.P. Chen, Y.C. Chao, T.S. Cheng, G.B. Chen, C.Y. Wu, Proc. Combust. Inst. 31 (2007) 3301–3308.
Please cite this article as: J. Gao et al., Flame base structures of micro-jet hydrogen/methane diffusion flames, Proceedings of the Combustion Institute (2016), http://dx.doi.org/10.1016/j.proci.2016.08.034
JID: PROCI
8
ARTICLE IN PRESS
[m;October 6, 2016;16:38]
J. Gao et al. / Proceedings of the Combustion Institute 000 (2016) 1–8
[11] T.S. Cheng, Y.C. Chao, C.P. Chen, C.Y. Wu, Combust. Flame 152 (3) (2008) 461–467. [12] Y. Nakamura, H. Ban, K. Saito, T. Takeno, Progress in Scale Modeling, Springer, 2008, pp. 293–306. [13] K. Kuwana, N. Tagami, S. Mizuno, T. Ida, Proc. Combust. Inst. 32 (2009) 3115–3121. [14] K. Fujiwara, Y. Nakamura, T. Hirasawa, J. Therm. Sci. Technol. 6 (2011) 235–246. [15] K. Fujiwara, Y. Nakamura, Combust. Flame 160 (8) (2013) 1373–1380. [16] A. Hossain, Y. Nakamura, Proc. Combust. Inst. 35 (2015) 3413–3420. [17] Y. Gan, Y. Luo, M. Wang, Y. Shi, Y. Yan, Appl. Therm. Eng. 89 (2015) 306–315. [18] ANSYS, FLUENT 14.0, ANSYS Inc., Canonsburg, 2011.
[19] Y. Kizaki, H. Nakamura, T. Tezuka, S. Hasegawa, K. Maruta, Proc. Combust. Inst. 35 (2015) 3389–3396. [20] R.W. Bilger, S.H. Stårner, R.J. Kee, Combust. Flame 80 (2) (1990) 135–149. [21] CHEMKIN-PRO release 15101, Reaction Design Inc., San Diego, 2010. [22] F. Takahashi, W. John Schmoll, V.R. Katta, Proc. Combust. Inst. 27 (1998) 675–684. [23] P. Aghalayam, D. Vlachos, AIChE J. 44 (9) (1998) 2025–2034. [24] F. Dabireau, B. Cuenot, O. Vermorel, T. Poinsot, Combust. Flame 135 (1) (2003) 123–133. [25] A. Gruber, R. Sankaran, E. Hawkes, J. Chen, J. Fluid Mech. 658 (2010) 5–32.
Please cite this article as: J. Gao et al., Flame base structures of micro-jet hydrogen/methane diffusion flames, Proceedings of the Combustion Institute (2016), http://dx.doi.org/10.1016/j.proci.2016.08.034
ARTICLE IN PRESS
[m;October 6, 2016;16:38]
Available online at www.sciencedirect.com
Proceedings of the Combustion Institute 000 (2016) 1–8 www.elsevier.com/locate/proci
Flame base structures of micro-jet hydrogen/methane diffusion flames Jian Gao, Akter Hossain, Yuji Nakamura∗ Department of Mechanical Engineering, Toyohashi University of Technology, 1-1 Hibarigaoka, Tempaku, Toyohashi 441-8580, Japan Received 4 December 2015; accepted 12 August 2016 Available online xxx
Abstract In this study, a comparison of flame base structures of hydrogen/methane–air diffusion flames formed over a tiny-jet is made numerically for both isothermal and thermal conductive burner conditions, in order to clarify the fuel dependent flame stabilization mechanisms. It is found that, unlike a methane flame, the flame base of a hydrogen flame always attaches to the burner. The analyses indicate that the dominant intermediateconsumption steps have significantly lower activation energies for the hydrogen flame as compared to a methane flame. More importantly, one of the HO2 production reactions (R43f: H + O2 + M → HO2 + M), which has a dominant role in sustaining reactivity at the flame base, shows a negative temperature dependence, causing the heat release rate in the flame base kernel to increase as the burner wall temperature decreases. With a thermal conductive burner (thermal conductivity of 16 W/m-K) over a wide range of fuel jet velocities (0.5– 4.0 m/s), it is found that the burner tip is heated to a significantly higher temperature by a hydrogen flame due to its unique stabilization mechanism. The mixing effects of hydrogen and methane are then considered. It is found that the burner tip temperature can be reduced by adding methane into the fuel flow. This is because, according to the investigation of the structures of the hydrogen/methane jet diffusion flames, the reaction rate of R43f is suppressed due to the included intermediates (e.g., CH3 , CH2 O) consumption steps of methane. It is expected that the flame attachment feature associated with the flame base structure can be easily controlled by mixing hydrogen and methane, making it possible to control the burner tip temperature in advance. © 2016 by The Combustion Institute. Published by Elsevier Inc. Keywords: Micro flame; Extinction; Jet diffusion flame; Hydrogen flame; Flame-burner interaction
1. Introduction Given that hydrocarbon and oxygenated hydrocarbon fuels have much higher specific energy densities than batteries (e.g., lithium ion batteries) ∗
Corresponding author. Fax: +81 532 44 6647. E-mail address: [email protected] (Y. Nakamura).
[1], development of combustion based micro scale power generation systems (such as micro scale engines and turbines) are especially favored to meet the increasing demands for portable power generators, micro-satellite thrusters, micro unmanned aircrafts, micro reactors, micro sensors, and others. In order to design a compact, stable, and long life combustor or burner for those devices, a deep understanding of the fundamentals of micro scale
http://dx.doi.org/10.1016/j.proci.2016.08.034 1540-7489 © 2016 by The Combustion Institute. Published by Elsevier Inc.
Please cite this article as: J. Gao et al., Flame base structures of micro-jet hydrogen/methane diffusion flames, Proceedings of the Combustion Institute (2016), http://dx.doi.org/10.1016/j.proci.2016.08.034
JID: PROCI
2
ARTICLE IN PRESS
[m;October 6, 2016;16:38]
J. Gao et al. / Proceedings of the Combustion Institute 000 (2016) 1–8
combustion is essential. Both premixed and diffusion combustion strategies should be potential options to utilize thermal energy, while according to the recent reviews [2,3] studies regarding micro diffusion flames are relatively less compared with those for premixed flames. The micro jet diffusion flame was first studied by Ban et al. [4] about twenty years ago. Fuel was ejected into a quiescent air environment through a submillimeter-diameter burner. Such a tiny flame is stabilized by the burner, and controlled by diffusion and convection whilst the buoyancy effect is minor. Additionally, due to the “inherent” large surface to volume ratio (S/V) feature in miniaturization, flame-burner interaction plays a significant role in the stability of the microflames. For this reason, the stabilization mechanism of micro-jet diffusion flames has been studied by numerous researchers [5–17]. In general, local extinction occurs in the vicinity of the burner wall, and hence there exists a gap (flame quenching zone) between the flame base and the burner wall, as observed in the experiments by Cheng et al. [8,10,11], Kuwana et al. [13] and Fujiwara et al. [14] and Fujiwara and Nakamura [15] for methane flames. Such a gap has been also clearly observed in micro-jet diffusion flames for general carbon-contained fuels, such as ethane, ethylene, and acetylene [4], propane [5], and ethanol [17]. On the contrary, as notified in the previous studies by Cheng et al. [6,9] and our recent numerical work [16] using hydrogen as the fuel, it seems that a hydrogen diffusion flame readily attaches to the burner wall and local extinction may not occur in the vicinity of the burner wall, invoking that the stabilization mechanism for a hydrogen flame may differ from that of a methane flame. Once the flame is fully attached to the burner, substantial heat can be transported (recirculated) to the fuel to improve the flame stability. On the other hand, the burner tip temperature can increase dramatically, causing structural damage (thermal damage to burner). In this regard, a strategy to control the flame base structure (thus the burner tip temperature) is important, especially for microflame technology. The objective of this study is to clarify the stabilization mechanism of micro-jet hydrogen/methane diffusion flames by numerical approaches. Computations are performed by adopting both isothermal and thermal conductive burner wall boundary conditions to understand the effect of burner wall temperature on the flame attachment mechanism. Fuel mixing effects on the flame base structure are also examined and a strategy to control the burner tip temperature and the degree of heat recirculation is suggested. 2. Numerical approach The numerical method is similar to our recent work [16]. A set of governing equations of mass, momentum, energy and species are solved
Fig. 1. Schematic of computation domain and boundary conditions.
in a two-dimensional (r–z) cylindrical coordinate by the commercial software FLUENT 14.5 CFD [18] based on the finite volume method (FVM) with the SIMPLE algorithm. Figure 1 illustrates the schematic of the computation domain and boundary conditions. Fuel is ejected into the ambient air through a tube (burner) with an inner diameter of 0.8 mm and an outer diameter of 1.2 mm. The pressure outlet boundary conditions are imposed at the far field boundaries. The burner wall is assumed to be non-slip, and chemically inert (because the recent study by Kizaki et al. [19] has proved that the effects of surface quenching reactions are minor at normal pressure; catalytic surface reaction assisted diffusion flames could be our future work however it is not considered in this study). Two kinds of thermal boundary conditions are considered at the burner wall. For the isothermal burner conditions, the inner surface and the top surface of the burner wall are assumed to be adiabatic, while the outer surface is set to a temperature of Tb , which varies from 300 K to 1800 K, because in reality a flame shall attach (if it can) to the outer surface of the burner wall, and therefore the outer surface temperature is critical for the flame-burner attachment. For the thermal conductive burner conditions, grid is put in the burner (solid) zone, namely, the thermal conduction equation in the burner zone (solid) is solved as well. Conjugate thermal boundaries are set at the inner surface, the outer surface and the top surface of the wall, while 300 K is imposed at the bottom of the burner wall to allow the heat loss from the domain. Although radiation may more or less affect flame temperatures at near extinction conditions, it has been shown that [8,9] the microjet diffusion flame structures could be predicted satisfactorily by the numerical models excluding radiation heat loss. Thus, radiation effects are also
Please cite this article as: J. Gao et al., Flame base structures of micro-jet hydrogen/methane diffusion flames, Proceedings of the Combustion Institute (2016), http://dx.doi.org/10.1016/j.proci.2016.08.034
JID: PROCI
ARTICLE IN PRESS
[m;October 6, 2016;16:38]
J. Gao et al. / Proceedings of the Combustion Institute 000 (2016) 1–8
3
Fig. 2. Computed temperature contours and heat release rate (HRR) isopleths for CH4 flames at fuel jet velocity of Vf = 0.4 m/s for (a) Tb = 1800 K (b) Tb = 300 K, and for H2 flames at Vf = 1.0 m/s for (c) Tb = 1800 K (d) Tb = 300 K.
neglected in this study considering that our concern here is the significant differences in flame structure between methane and hydrogen flames. A semi-detailed mechanism including 17 species and 58 steps [20] is employed for the chemical kinetics for the flames of methane, hydrogen, and their mixtures. The mechanism, thermodynamic and transport input files are available in the CHEMKIN database [21]. The species transport properties are evaluated using mixture averaged formulas. At first, we performed several computations to simulate the previous experiments in [9,10,15]. It was confirmed that both the methane flame and hydrogen flame structures can be well predicted by this mechanism. The total number of grid points are 103 and 168 in radial and axial directions, respectively, over a physical domain of 15 mm× 25 mm. The minimum grid spacing of 0.025 mm is placed near the exit of the burner, and gradually stretched toward the far boundaries. No apparent difference in the computation results (flame structure, maximum temperature, etc.) is observed using a nearly double finer grid system, indicating the present grid system is fine enough. Time-dependent calculations are performed to obtain steady state solutions. At first, a high temperature spot (∼2000 K) is placed near the burner exit to achieve ignition, and then it is removed. A steady state is declared until the fluctuation in the maximum temperature is less than 2 K. 3. Isothermal burner 3.1. Comparison of methane and hydrogen flames Figure 2 shows the computed temperature contours and heat release rate (HRR) isopleths for CH4
flames at a fuel jet velocity of Vf = 0.4 m/s for (a) Tb = 1800 K (b) Tb = 300 K, and for H2 flames at Vf = 1.0 m/s for (c) Tb = 1800 K (d) Tb = 300 K. At Tb = 1800 K (Fig. 2a), the CH4 flame attaches to the outer surface of the burner. As Tb decreases from 1800 K to 300 K, the flame temperature (defined as the maximum gas temperature in the computation domain in this study) decreases from 2206 K to 1966 K, and the maximum HRR decreases from 1.3E9 W/m3 to 7.4E8 W/m3 . As expected, at Tb = 300 K, local extinction occurs in the vicinity of the outer surface of the burner wall, resulting in a gap (flame quenching zone) between the burner and the stabilized flame (see Fig. 2b). On the contrary, the H2 flame shows a different behavior compared to the methane flame. As Tb decreases from 1800 K to 300 K, although the flame temperature decreases from 2091 K to 1770 K, local extinction does not occur and the flame base kernel keeps attaching to the outer surface of the burner wall (Fig. 2c and d). Moreover, the HRR in the flame base kernel significantly increases from 7.5E8 W/m3 to 2.8E9 W/m3 as Tb decreases from 1800 K to 300 K, namely, the flame base kernel reactivity shows a negative temperature dependence. 3.2.. The predominant reactions Figure 3 shows the heat release rates from elementary reactions in the flame base kernels for both the CH4 flame and the H2 flame at Tb = 300 K. For the CH4 flame, the reactions related to CH3 radical and CH2 O are the main contributors to the total HRR. The reaction R6f (CH3 + O → CH2 O + H) predominately contributes to the total HRR, agreeing with the previous study by Takahashi et al. [22].
Please cite this article as: J. Gao et al., Flame base structures of micro-jet hydrogen/methane diffusion flames, Proceedings of the Combustion Institute (2016), http://dx.doi.org/10.1016/j.proci.2016.08.034
JID: PROCI
4
ARTICLE IN PRESS
[m;October 6, 2016;16:38]
J. Gao et al. / Proceedings of the Combustion Institute 000 (2016) 1–8
Fig. 3. Comparison of HRRs from elementary reactions in the flame base kernels for the CH4 flame (in Fig. 2b) and the H2 flame (in Fig. 2d) at Tb = 300 K.
It is noticed that, most of the listed reactions for the CH4 flame are chain-terminating (e.g., R1f, R6f, R7f, R26f and R27f), and the activation energy for the reaction R27f (CH2 O + H → HCO + H2 ) is 12,600 J/mole, corresponding to an activation temperature of 1500 K. Therefore, a high temperature is essential in order to sustain the reactivity in the flame base kernel, and that the flame quenching zone exists in the vicinity of a lower temperature burner wall (e.g., in Fig. 2b for Tb = 300 K). For the H2 flame, HRRs from the reactions R52f (H + OH + M → H2 O + M), R43f (H + O2 + M → HO2 + M), and R45f (H + HO2 → 2OH) are significantly higher than other reactions. Note that, the activation energies of R52f and R43f are zero, and the activation energy of R45f is 4506 J/mole (corresponding to an activation temperature of 540 K), which is much lower than that of R27f in the CH4 flame. Further, R43f and R45f act as a chain-propagation reaction pathway, which could sustain the reactivity in the flame base kernel. The initiation reaction R43f is a recombination reaction requiring a third body, and its reaction rate constant is expressed as k = 3.6 × 1017 T−0.72 exp(0/T) cm3 mol−1 s−1 , showing a negative temperature dependence. According to the computation results, as Tb decreases from 1800 K to 300 K, the net reaction rate of R43 in the flame base kernel increases from 0.8 kmol/m3 -s to 4.0 kmol/m3 -s, showing the same trend as for the total HRR (see Fig. 2c and d). It is implied that R43 might be the dominant reaction for the reactivity in the flame base kernel.
3.3. Verification by a modified kinetic model In order to verify that R43f is the dominant reaction for the reactivity in the flame base kernel for a H2 flame, additional computation is performed using a modified kinetic model, in which the reaction rate constant of the reaction R43f is revised from k = 3.6 × 1017 T−0.72 exp(0/T) cm3 mol−1 s−1 to k = 3.6 × 1017 T0 exp(−5035/T) cm3 mol−1 s−1 . In this way, the negative temperature dependence of the reactivity of R43f is lost, and a high temperature is required to sustain its reactivity due to the elevated activation energy. The computed temperature contours and HRR isopleths are shown in Fig. 4 for (a) Tb = 1800 K and (b) Tb = 300 K. It can be seen that, such a “H2 flame” attaches to the burner at Tb = 1800 K, whereas extinction occurs in the vicinity of the burner for Tb = 300 K, and hence a quenching gap exists, namely, such a “H2 flame” shows a similar behavior to the CH4 flame in Fig. 2a and b. Now, it is verified that the reaction R43 plays a dominant role in sustaining the flame base kernel reactivity for a H2 flame. It also can be concluded that the unique stabilization mechanism of H2 flames is caused by its kinetics, but not other factors such as its higher mass diffusivity. In this paper the critical role of the reaction R43f (H + O2 + M → HO2 + M) in determining the flame base structure, making a hydrogen flame always attach to the burner is revealed. In addition, the effects of recombination reactions (such as R52f and R43f) on hydrogen flame-wall interaction have also been reported by several other
Please cite this article as: J. Gao et al., Flame base structures of micro-jet hydrogen/methane diffusion flames, Proceedings of the Combustion Institute (2016), http://dx.doi.org/10.1016/j.proci.2016.08.034
ARTICLE IN PRESS
JID: PROCI
[m;October 6, 2016;16:38]
J. Gao et al. / Proceedings of the Combustion Institute 000 (2016) 1–8
Fig. 4. Computed temperature contours and HRR isopleths for “H2 flames” at Vf = 1.0 m/s for (a) Tb = 1800 K and (b) Tb = 300 K, using a modified kinetic model, in which the rate constant of the reaction R43f is revised from k = 3.6 × 1017 T−0.72 exp(0/T) cm3 mol−1 s−1 to k = 3.6 × 1017 T0 exp(−5035/T) cm3 mol−1 s−1 .
researchers. Through one-dimensional numerical computations for premixed hydrogen/air flames impinging on isothermal walls, Aghalayam and Vlachos [23] indicated that the reaction R43f would weaken the thermal coupling between flame and wall, and decrease the H concentration to increase the NO2 concentration near an inert wall. Also by employing a simplified one-dimensional configuration, Dabireau et al. [24] studied the transient quenching processes for hydrogen/oxygen premixed and diffusion flames near a low-temperature wall (700 K), showing that the low-activation-energy recombination reactions could lead to high fluxes to the wall just before flame quenching. Further, Gruber et al. [25] found that, even for a turbulent hydrogen/air premixed flame, these recombination reactions also play an important role on flame-wall convective heat transfer and could contribute up to 70% to the total heat release rate at a cold wall. Beyond agreements with these previous studies, the two-dimensional computation results in our study imply that, such effect is more pronounced and could dominate the structure of a microscale hydrogen diffusion flame. Probably, this is because of the enlarged S/V ratio and that sufficient H and O2 could diffuse to the flame base kernel to sustain the reactivity there. 4. Thermal conductive burner 4.1. Structure of hydrogen/methane flames For the thermal conductive burner conditions, the burner material is assumed to be stainless
5
steel and the thermal conductivity is set to λb = 16 W/m K. Figure 5 shows the computed temperature contours and HRR isopleths for (a) H2 flame at Vf = 2.0 m/s, (b) H2 /CH4 80%/20% flame at Vf = 1.2 m/s, (c) H2 /CH4 20%/80% flame at Vf = 0.6 m/s, and (d) CH4 flame at Vf = 0.5 m/s. Note that, for comparison purpose, we have adjusted the fuel jet velocities to have close flame temperatures (2028– 2034 K). For the CH4 flame (Fig. 5d), due to the heat loss through the burner wall, a quenching zone is observed in the vicinity of the burner. For the H2 flame (Fig. 5a), as described above in Section 3, the flame attaches to the burner. From Fig. 5b and c, one can see that, as CH4 is added into H2 , the flame detaches from the burner wall. Figure 6 compares the flame structures of (a) H2 flame at Vf = 2.0 m/s, and (b) H2 /CH4 80%/20% flame at Vf = 1.2 m/s. It is shown that, compared to the H2 flame, the mole fraction of the H radical is dramatically reduced in the H2 /CH4 80%/20% flame. Comparison of the contours of the net production rate of H radical (the right sides in Fig. 6a and b) indicates that, a large amount of H radicals are consumed in the fuel side as CH4 is added. Figure 7 shows the computed radial profiles of the net reaction rates of R6, R27, R43, and R45 at z = 10 mm for the H2 flame and the H2 /CH4 80%/20% flame. For the H2 /CH4 80%/20% flame, R6 and R27 are reactive in the fuel side (from r = 0.6 mm to 1.5 mm), where the net reaction rates of R43 and R45 are decreased compared to those for the H2 flame. This might be because the intermediates (CH3 and CH2 O) consumption reactions (such as R6 and R27) for CH4 consume a certain amount of H radicals in the fuel side, and thus the reaction rates of R43 and R45 are both suppressed accordingly. As shown in Fig. 6 (isopleths in the left sides), the net reaction rate of R43 in the flame base kernel is significantly reduced from 1.6 kmol/m3 -s to 1.0 kmol/m3 -s as 20% CH4 is added. Consequently, the negative temperature dependence of the reactivity in the flame base kernel is weakened. Further increasing the volumetric fraction of CH4 in the fuel flow can make the flame detach from the burner (as shown in Fig. 5c for the H2 /CH4 20%/80% flame). 4.2. Controlling the burner tip temperature Figure 8 shows the temperature profiles on the axis for various H2 /CH4 mixtures. The dashed line at z = 10 mm indicates the location of the burner exit. It is shown that, as the volumetric fraction of H2 increases, the temperature inside the burner for the fuel flow increases. This is because the flameburner attachment is enhanced as H2 is added, producing a higher burner tip temperature (as shown in Fig. 5), and consequently more heat is recirculated into the fuel flow. Therefore, adding H2 into CH4 can elevate the heat recirculation degree, and flame stability is expected to be improved.
Please cite this article as: J. Gao et al., Flame base structures of micro-jet hydrogen/methane diffusion flames, Proceedings of the Combustion Institute (2016), http://dx.doi.org/10.1016/j.proci.2016.08.034
JID: PROCI
6
ARTICLE IN PRESS
[m;October 6, 2016;16:38]
J. Gao et al. / Proceedings of the Combustion Institute 000 (2016) 1–8
Fig. 5. Computed temperature contours and HRR isopleths at the thermal conductive burner condition (the thermal conductivity of burner wall is λb = 16 W/m-K) for (a) H2 flame at Vf = 2.0 m/s, (b) H2 /CH4 80%/20% flame at Vf = 1.2 m/s, (c) H2 /CH4 20%/80% flame at Vf = 0.6 m/s, (d) CH4 flame at Vf = 0.5 m/s. Here the flame temperatures range between 2028 K and 2034 K.
Fig. 6. Computed contours of H radical mole fraction, isopleths of the net reaction rate of R43, and isopleths of the net production rate of H radical at the thermal conductive burner condition (λb = 16 W/m-K), for (a) H2 flame at Vf = 2.0 m/s, (b) H2 /CH4 80%/20% flame at Vf = 1.2 m/s.
Finally, the relationship between flame temperatures and burner tip temperatures is summarized for the H2 flame, the H2 /CH4 80%/20% flame, and the H2 /CH4 60%/40% flame over the fuel jet velocity range from Vf = 0.5 m/s to 4.0 m/s (the Reynolds number from 3.8 to 91.9) in Fig. 9. As shown in the figure, for each fuel mixture, as Vf increases, the flame temperature increases, while the burner tip temperature first increases and then decreases, namely, there exists a maximum burner tip temperature for each kind of fuel composition. It clearly shows that the maximum burner tip temperature
is decreased as the CH4 concentration increases. Therefore, adding CH4 into H2 can reduce the burner temperature and effectively prevent thermal damage to the burner tip, which could be a potential problem in a hydrogen microflame.
5. Conclusions In this study, we perform a systematic numerical study to investigate the stabilization mechanisms of micro-jet hydrogen/methane diffusion
Please cite this article as: J. Gao et al., Flame base structures of micro-jet hydrogen/methane diffusion flames, Proceedings of the Combustion Institute (2016), http://dx.doi.org/10.1016/j.proci.2016.08.034
JID: PROCI
ARTICLE IN PRESS J. Gao et al. / Proceedings of the Combustion Institute 000 (2016) 1–8
Fig. 7. Comparison of radial profiles of the net reaction rates of R6, R27, R43, and R45 at z = 10 mm (burner exit) for the H2 flame at Vf = 2.0 m/s, and the H2 /CH4 80%/20% flame at Vf =1.2 m/s.
[m;October 6, 2016;16:38]
7
flames. For the hydrogen flame, it is understood that the dominant intermediate-consumption steps have significantly lower activation energies compared to the methane flame. The reaction R43f (H + O2 + M → HO2 + M), which plays a dominant role in sustaining the reactivity at the flame base, shows a negative temperature dependence, causing the heat release rate in the flame base kernel to increase as the burner wall temperature decreases. This allows the hydrogen flame base to remain attached to the burner as the burner wall temperature decreases from 1800 K to 300 K. Further computations are then performed in order to clarify the effect of fuel mixing (hydrogen and methane). It is shown that the burner tip temperature can be controllable by replacing some amount of hydrogen with methane, due to the inhibition effect on R43f by the included intermediates (e.g., CH3 , CH2 O) consumption steps of methane. It is suggested that mixing hydrogen and methane could be an effective strategy to enhance heat recirculation or prevent potential thermal damage to the burner, thus improving the combustion performance in micro diffusion flames. Acknowledgments
Fig. 8. Temperature profiles on the axis for H2 /CH4 flames. (1): H2 flame at Vf = 2.0 m/s, (2): H2 /CH4 80%/20% flame at Vf = 1.2 m/s, (3) H2 /CH4 60%/40% flame at Vf = 0.85 m/s, (4): H2 /CH4 40%/60% flame at Vf = 0.7 m/s, (5): H2 /CH4 20%/80% flame at Vf = 0.6 m/s, (6) CH4 flame at Vf = 0.5 m/s. (Here flame temperatures are within the range from 2028 K to 2034 K).
This work is supported by Research Foundation for the Electro-technology of Chubu (RFEC), Iwatani-Naoji Memorial Foundation, Asahi Glass Foundation, Grant-in-Aid for Scientific Research(B) (no. 26289041). Fruitful discussion with Dr. Umeda (Toho Gas Co. Ltd.) through the joint research project on “Hybrid hydrogen energy supply system” organized by Chubu Region Institute for Social and Research is greatly appreciated. The authors also acknowledge Mr. Lindsay Prescott (Technical Advisor at TUT-RAC) for his language editing. References
Fig. 9. Burner tip temperatures as a function of flame temperature for the H2 flame, the H2 /CH4 80%/20% flame, and the H2 /CH4 60%/40% flame over the fuel jet velocity range from Vf =0.5 m/s to 4.0 m/s.
[1] A.C. Fernandez-Pello, Proc. Combust. Inst. 29 (2002) 883–899. [2] K. Maruta, Proc. Combust. Inst. 33 (2011) 125–150. [3] Y. Ju, K. Maruta, Prog. Energy Combust. Sci. 37 (2011) 669–715. [4] H. Ban, S. Venkatesh, K. Saito, J. Heat Transf. 116 (4) (1994) 954–959. [5] L.M. Matta, Y. Neumeier, B. Lemon, B.T. Zinn, Proc. Combust. Inst. 29 (2002) 933–939. [6] T.S. Cheng, Y.C. Chao, C.Y. Wu, et al., Proc. Combust. Inst. 30 (2005) 2489–2497. [7] Y. Nakamura, H. Yamashita, K. Saito, Combust. Theory Model. 10 (6) (2006) 927–938. [8] T. Cheng, C.-P. Chen, C.-S. Chen, Y.-H. Li, C.-Y. Wu, Y.-C. Chao, Combust. Theory Model. 10 (5) (2006) 861–881. [9] T.S. Cheng, C.Y. Wu, C.P. Chen, et al., Combust. Flame 146 (1–2) (2006) 268–282. [10] C.P. Chen, Y.C. Chao, T.S. Cheng, G.B. Chen, C.Y. Wu, Proc. Combust. Inst. 31 (2007) 3301–3308.
Please cite this article as: J. Gao et al., Flame base structures of micro-jet hydrogen/methane diffusion flames, Proceedings of the Combustion Institute (2016), http://dx.doi.org/10.1016/j.proci.2016.08.034
JID: PROCI
8
ARTICLE IN PRESS
[m;October 6, 2016;16:38]
J. Gao et al. / Proceedings of the Combustion Institute 000 (2016) 1–8
[11] T.S. Cheng, Y.C. Chao, C.P. Chen, C.Y. Wu, Combust. Flame 152 (3) (2008) 461–467. [12] Y. Nakamura, H. Ban, K. Saito, T. Takeno, Progress in Scale Modeling, Springer, 2008, pp. 293–306. [13] K. Kuwana, N. Tagami, S. Mizuno, T. Ida, Proc. Combust. Inst. 32 (2009) 3115–3121. [14] K. Fujiwara, Y. Nakamura, T. Hirasawa, J. Therm. Sci. Technol. 6 (2011) 235–246. [15] K. Fujiwara, Y. Nakamura, Combust. Flame 160 (8) (2013) 1373–1380. [16] A. Hossain, Y. Nakamura, Proc. Combust. Inst. 35 (2015) 3413–3420. [17] Y. Gan, Y. Luo, M. Wang, Y. Shi, Y. Yan, Appl. Therm. Eng. 89 (2015) 306–315. [18] ANSYS, FLUENT 14.0, ANSYS Inc., Canonsburg, 2011.
[19] Y. Kizaki, H. Nakamura, T. Tezuka, S. Hasegawa, K. Maruta, Proc. Combust. Inst. 35 (2015) 3389–3396. [20] R.W. Bilger, S.H. Stårner, R.J. Kee, Combust. Flame 80 (2) (1990) 135–149. [21] CHEMKIN-PRO release 15101, Reaction Design Inc., San Diego, 2010. [22] F. Takahashi, W. John Schmoll, V.R. Katta, Proc. Combust. Inst. 27 (1998) 675–684. [23] P. Aghalayam, D. Vlachos, AIChE J. 44 (9) (1998) 2025–2034. [24] F. Dabireau, B. Cuenot, O. Vermorel, T. Poinsot, Combust. Flame 135 (1) (2003) 123–133. [25] A. Gruber, R. Sankaran, E. Hawkes, J. Chen, J. Fluid Mech. 658 (2010) 5–32.
Please cite this article as: J. Gao et al., Flame base structures of micro-jet hydrogen/methane diffusion flames, Proceedings of the Combustion Institute (2016), http://dx.doi.org/10.1016/j.proci.2016.08.034