- Email: [email protected]
Effect of high-frequency alternating electric fields on the behavior and nitric oxide emission of laminar non-premixed flames
Fuel 109 (2013) 350–355
Contents lists available at SciVerse ScienceDirect
Fuel journal homepage: www.elsevier.com/locate/fuel
Effect of high-frequency alternating electric fields on the behavior and nitric oxide emission of laminar non-premixed flames Yang Zhang, Yuxin Wu, Hairui Yang, Hai Zhang ⇑, Min Zhu Key Laboratory for Thermal Science and Power Engineering of the Ministry of Education, Department of Thermal Engineering, Tsinghua University, Beijing 100084, PR China
h i g h l i g h t s " We studied the effects of AC electric fields on non-premixed flames. " Flame image, FTIR pattern and chemiluminescent emissions were obtained. " There exist three distinct regimes across the tested voltage range. " The flame behaved non-monotonically as the increasing voltage. " The flame behavior was explained by the competition of three effects.
a r t i c l e
i n f o
Article history: Received 18 December 2012 Received in revised form 26 December 2012 Accepted 27 December 2012 Available online 10 January 2013 Keywords: Alternating electric field Non-premixed flame NO emission Frequency Voltage
a b s t r a c t This paper examined the behavior and NO emission of laminar non-premixed methane/air jet flames when subjected to high frequency alternating electric fields of 10 kHz over the voltage range of 0–4.0 kV. In particular, this paper examined variations of flame shape and luminosity, CO and NO molar fractions in the downstream flue gas, and chemiluminescence from OH⁄ and CH⁄ in the voltageinfluenced flame zone. The results showed that with no application of an alternating electric field, flames were stable at the nozzle exit, bluish at the base and yellowish at the conical tip. However, once applied, different voltage regimes produced different responses from the flame. In the low-voltage regime of 0–1.0 kV, increasing the voltage narrowed the top yellowish zone of the flame and sharpened its conical tip, increased the CO molar fraction in the flue gas, decreased the NO molar fraction in the flue gas, and decreased the chemiluminescence intensity of OH⁄ and CH⁄ in the flame zone by 50%. At 1.0 kV, both CO and NO molar fractions reached extreme values, and the flame was at its weakest. In the mid-voltage regime of 1.0–3.0 kV, increasing the voltage resulted in an inverse response from the flames compared to the low-voltage regime. In the high-voltage regime of 3.0–4.0 kV, increasing the voltage resulted in the gradual disappearance of the top yellowish zone of the flame, increased the CO molar fraction in the flue gas and decreased the NO molar fraction. The transition mechanisms between the regimes are discussed within the context of the high-frequency discharge theory. Three competing effects explain the non-monotonic flame response to the voltage: thermal, ionic wind, and electrical–chemical. The analysis showed that the ionic wind effect majored in the low-voltage regime, the electrical–chemical effect dominated the mid-voltage regime, and all three effects were highly coupled in the high-voltage regime. Ó 2013 Elsevier Ltd. All rights reserved.
1. Introduction The interaction between an electric field and flame behavior has attracted wide attention in the last few decades. Research shows that electric fields affect flames in three major ways: the thermal effect [1–3], which is caused by electrical energy input; the ionic wind effect [4–8]; and the electrical–chemical effect [9–12]. The thermal effect dominates when there is a large current across the
⇑ Corresponding author. Tel.: +86 10 6278 8523; fax: +86 10 6278 1743. E-mail address: [email protected] (H. Zhang). 0016-2361/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.fuel.2012.12.083
electric field, such as a spark plug [3]. The ionic wind effect causes fluid dynamic changes in the flow field, inducing an ‘‘ionic wind’’ [4–8]. The electrical–chemical effect produces high speed electrons, radicals, ions and excited molecules in the pre-flame zone, which directly change the chemistry of the flames [9–12]. Various experimental and numerical studies have been conducted on the influence of direct-current (DC) electric fields on flame behaviors. Experiments showed that the DC electric field strongly affected flame shape [6,13,14], flame propagation speed [5,15], NOx emissions [4,6,14,16] and soot formation [4,13]. Numerical simulations showed that the ionic wind effect dominated [17–20].
Y. Zhang et al. / Fuel 109 (2013) 350–355
Studies on the effects of an alternating-current (AC) electric field on flame behaviors are few. Previous studies mainly focused on the ionic wind effect. Chung et al. published a series of works [21–27] on the effects of the AC electric field on flame stabilization, and produced the following findings. The speed of tribrachial flame propagation increased under AC voltage [24]. The edge structure and the detachment velocity of the nozzle attached non-premixed propane/air flame were both affected by an AC electric field. The detachment velocity changed non-monotonically with the voltage increase at 60 Hz [25]. The lift-off height presented an oscillation characteristic when lower than 30 Hz, due to the directionalternating ionic wind [26]. No oscillation was found in the range of 60–1500 Hz [25]. The height of the diffusion flame decreased with the increasing AC voltage at 400 Hz. Kono et al. [28] noted a delay time in the formation of ionic wind, normally in the order of 10 ms, indicating that the ionic wind did not form when the alternating frequency was higher than 50 Hz. Flame behavior was significantly different to that of the DC and low-frequency AC electric fields when the frequency of the AC electric field was high (e.g. >2000 Hz). This may be because time was insufficient for momentum transfer from the charged particles to the neutral ones. As a result, the electric field did not introduce significant bulk flow motion [25,28]. Alternatively, it may be because the charged particles could not easily reach the electrodes due to high-frequency alternating properties. Therefore, a large amount of charged particles existed within the space between the two electrodes [29]. Studies on the effects of high-frequency alternating electric fields remain limited in number. Consequently, this paper examined the behavior and NO emission of laminar non-premixed methane/air flames when subjected to high frequency AC electric fields of 10 kHz over the voltage range of 0–4.0 kV. In particular, this paper applied different voltages at different fuel and co-flow velocities to study variations of flame shapes and to study CO and NO molar fractions in the downstream flue gas. This research also measured chemiluminescence from two excited radicals, OH⁄ and CH⁄, in order to provide insight into the effects of highfrequency AC electric fields on combustion. 2. Experimental apparatus and methodology 2.1. Experimental apparatus Fig. 1 shows the experimental apparatus, which consisted of three main parts: (1) a coaxial burner and associated gas supply system; (2) a high frequency AC power supply system; (3) a measurement system. 2.1.1. Coaxial burner and associated gas supply system The fuel nozzle was made of a quartz tube with an inner diameter of 8.0 mm and thickness of 1.0 mm (Fig. 1). The tube length was 250 mm to ensure that the velocity profile of the fuel stream was fully developed at the nozzle exit. The fuel used in the experiments was high purity CH4 (>99.99%). Compressed air, as the oxidizer, was supplied around the fuel tube from a concentric quartz tube with an inner diameter of 65 mm. The flow rates of the gas streams were controlled by mass flow meters that were precalibrated by a wet gas meter. 2.1.2. High frequency AC power supply system An AC electric field was applied in the experiments. A metal mesh plate located at 50 mm above the upper rim of the burner was used as the anode and another metal mesh attached to the outside surface of the fuel nozzle was used as the cathode. A sinusoidal AC electric field with 10 kHz alternating frequency
351
was supplied. The magnitude of the field (AC voltage) was adjustable. The AC wave was measured and recorded by an oscilloscope. 2.1.3. Measurement system A flute-shaped probe located 750 mm above the burner sampled the combustion products. A Fourier transform infrared spectrometer (FT-IR) with the model of NETZSCH STA 409C analyzed the flue gas introduced into the probe. The CO and NO molar fractions in the flue gas were measured. A grating spectrometer with the model of ZOLIX SPB300 equipped with a photomultiplier tube (PMT) analyzed the chemiluminescence from the flame. The chemiluminescence intensity of OH⁄ and CH⁄ in the flame zone was measured. A digital camera recorded the flame images (exposure time: 0.5 s; resolution: 72 dpi). 2.2. Experimental conditions Experiments were conducted at four different air flow velocities and four different fuel flow velocities (Table 1). 3. Results and discussion 3.1. AC voltage effect on flame shapes and luminosities Fig. 2 shows images of the non-premixed laminar jet flames at various AC voltages. A typical non-premixed jet flame was stabilized at the exit of the nozzle when no electric field was applied (peak voltage 0.0 kV). The upper portion of the flame was a yellowish color due to the radiation of the soot formations. The base part of the flame was a bluish color due to the radiation of CH radicals. The flame shape and luminosity distinctly changed when an AC electric field was applied. Based on the variation of flame luminosities under different applied voltage, there existed three different regimes depending on the AC voltages: (1) the low-voltage regime with voltage lower than 1.0 kV; (2) the mid-voltage regime with voltage in the range of 1.0–3.0 kV; (3) the high-voltage regime with voltage higher than 3.5 kV. Case 1 is taken as an instance to describe the flame behaviors in each of these three regimes. In the low-voltage regime, as voltage increased, the upper yellowish zone of the flame became narrower and the base bluish zone became wider. Also, the conical tip of flame became sharper. The flame top was sharpest at 1.0 kV. Under these conditions, the yellowish zone was the smallest and the overall flame was the darkest, indicating that soot formation was minimized. The experiments showed that the flame became unstable and that the flow perturbations even extinguished the flame in some instances. In the mid-voltage regime, as voltage increased, there was an inverse response from the flames compared to the low-voltage regime. The upper yellowish zone became larger and the base bluish zone became narrower. The flame top gradually became a plump arc again and the flame was more stable compared to its state at 1.0 kV. In the high-voltage regime, as voltage increased, the AC electric field zone typically started to produce a hissing sound between 3.0 and 3.5 kV. This indicated a slight corona discharge from the sharp tips of the electrodes, which generated a certain amount of plasma. The sound-producing pressure fluctuation affected the flame. The upper yellowish zone of the flame quickly disappeared and the flame tip dispersed (Fig. 2). This result indicated that soot formation was suppressed by the AC corona discharge, similar to the phenomenon found in the dielectric barrier discharge [22]. Furthermore, the flame partially detached from the nozzle and randomly rotated around the nozzle rim. This phenomenon was consistent with the observation by Chung et al. [25] for low AC frequency. However, where Chung et al. observed lift-off as voltage
352
Y. Zhang et al. / Fuel 109 (2013) 350–355
Fig. 1. Schematic of the experimental system.
Table 1 Fuel and air velocities. Case no.
Fuel flow velocity (cm/s)
Air flow velocity (cm/s)
Fuel jet Reynolds number
1 2 3 4
3.2 3.9 4.6 5.3
0.87 0.86 0.84 0.83
115 141 168 194
continued to increase [25], this research found that the flame was extinguished. This research found that the critical voltage between the midvoltage and high-voltage regimes differed for different fuel jet velocities. As fuel jet velocity increased, the critical voltage decreased. The flame surface also became larger with more ions and exited radicals. Therefore, the corona discharge occurred at lower voltage. Since the transition from the mid-voltage to high-voltage regime was caused by the corona discharge, the critical voltage decreased with the increasing fuel jet velocity. However, the difference was within 0.5 kV throughout the experimental range. 3.2. AC voltage effect on flame height Fig. 3 shows the flame height variations and inner fuel zone for the applied AC voltages under different exit velocities of fuel streams (UF,0). The flame boundary was defined using a gray value threshold method [30]. In this paper, the gray value 90 (out of 255 maximum) was determined as the threshold of the flame height location. The flame height was the vertical distance from the nozzle exit to the flame tip (Fig. 3a). The inner fuel zone had visible unburned fuel in a relatively dark color in the flame images (Fig. 3b). The flame height variations for the AC voltage were within 10% for all tested cases, indicating that the effect of the electric field on the flame height was insignificant. However, the inner fuel zone height
Fig. 2. Selected images of the flames at various peak voltages and fuel exit velocities. Three distinct regimes: (1) <1.0 kV; (2) 1.0–3.0 kV; (3) >3.0 kV. (a): UF,0 = 3.2 cm/s; (b): UF,0 = 3.9 cm/s; (c): UF,0 = 4.6 cm/s; (d): UF,0 = 5.3 cm/s).
Y. Zhang et al. / Fuel 109 (2013) 350–355
353
3.3. AC voltage effect on CO and NO formation Fig. 4 shows the infrared patterns of the flue gas analyzed by the FT-IR spectrometer. The peak values of CO and NO significantly differed for different AC voltages. Fig. 5 shows the quantitative results, in which XCO was the CO molar fraction in the flue gas and XNO was the NO molar fraction in the flue gas. XCO increased with UF,0 under a fixed voltage (Fig. 5a). This was expected as more oxidizer causes more complete combustion. However, at a fixed UF,0, voltage variations yielded different XCO in the three regimes. Taking the case of UF,0 = 5.3 cm/s, XCO increased with AC voltage in the low-voltage regime, while it decreased with AC voltage in the mid-voltage regime. When the voltage approached the threshold value of the high-voltage regime, XCO increased and then peaked before the flame extinguished. Flames with other UF,0 values showed similar trends and their extreme values occurred at 1.0 kV. These results were consistent with the previous results (Fig. 2). However, XNO showed an opposite trend against XCO (Fig. 5b). As the AC voltage increased, XNO decreased in the lowvoltage regime, increased in the mid-voltage regime, and decreased in the high-voltage regime. The variation of XNO with UF,0 was non-monotonic (Fig. 5b). This may be caused by variations in the fuel flow rate and air flow rate.
Fig. 3. Variations of flames and inner fuel zone heights for different AC voltages (solid square: UF,0 = 3.2 cm/s; solid triangle: UF,0 = 3.9 cm/s; blank square: UF,0 = 4.6 cm/s; solid circle: UF,0 = 5.3 cm/s;).
3.4. Chemiluminescence emissions from the flame zone under AC fields A grating spectrometer analyzed the chemiluminescence spectrum of the flame zone (white circle) at 0.0, 1.0 and 2.5 kV (Fig. 6). The spectrum pattern represented the integration of the chemilumi-
Fig. 4. FT-IR spectrum patterns of downstream flue gas. (Where UF,0 is the fuel flow velocity at the nozzle exit.) (a) Voltage = 0.5 kV; (b) voltage = 1.0 kV.
was strongly affected by the AC voltage. In the low-voltage regime, it increased linearly with the voltage, whereas in the mid-voltage regime, it gradually decreased. Also, the inner fuel zone height was very sensitive to the flame edge structure. The results were consistent with the previous findings [25]: the flame closest to the nozzle rim was influenced by the electric field over a low frequency range of 60–1500 Hz. The mechanism of the nonmonotonic AC voltage effect is discussed in Section 3.5.
Fig. 5. Variations of CO and NO molar fractions, XCO and XNO, in the flue gas against AC voltages in the flue gas. (a): XCO; (b): XNO (solid square: UF,0 = 3.2 cm/s; blank square: UF,0 = 3.9 cm/s; solid circle: UF,0 = 4.6 cm/s; solid triangle: UF,0 = 5.3 cm/s).
354
Y. Zhang et al. / Fuel 109 (2013) 350–355
Fig. 6. Chemiluminescence Intensity values for the flame at different AC voltages (characteristic peak of OH⁄ locates at approximately 310 nm while the CH⁄ peak locates at approximately 430.5 nm).
nescence intensity of the entire flame zone. The chemiluminescence intensity was strongest when no electric field was applied (0.0 kV). The chemiluminescence intensity dropped 50% when AC voltage increased from 0 to 1.0 kV. The chemiluminescence intensity increased when AC voltage was 2.5 kV, but its value was still lower than the value observed without an electric field. The chemiluminescence intensity values of OH⁄ and CH⁄ were good indicators of the heat release rate. The characteristic peaks of OH⁄ and CH⁄ were located around 310 nm and 430.5 nm, respectively. The chemiluminescence intensity values of OH⁄ and CH⁄ were plotted against AC voltages (Fig. 7). Fig. 7 shows that both intensity values first decreased, then increased, and finally plateaued with the increasing AC voltage. The minimum OH⁄ and CH⁄ chemiluminescence intensity values were both located at 1.0 kV. In this situation, the ionic wind effect strengthened the mixing of the fuel and air at the flame root and decreased the burning intensity. It is generally accepted for methane combustion that the step with the greatest heat release is CO + OH ? CO2 + H. It follows that combustion was quite weak and incomplete at 1.0 kV, resulting in a high, unburned CO emission and a low flame temperature. Thermal NO was suppressed and prompt NO dominated the NO formation because the temperature was low. For the prompt NO formation, the O, CH and CH2 radicals each played a key role. Fig. 7b shows that the CH radical amount was smallest when the voltage was 1.0 kV. Hence, overall NO formation was suppressed by the weak combustion. The non-monotonic nature of the chemiluminescence agreed with those of the CO and NO emissions. 3.5. Transition mechanisms between the regimes Fig. 8 shows the diagram of flame regimes in terms of AC voltage and frequency along with the data obtained from the present work and previous studies [25]. The critical voltages for the regime transitions decreased linearly with an increasing denary logarithm of the frequency (log f). The two linear dashed lines of the critical voltages divide the figure into three regions that correspond to the three regimes (Fig. 2). The transient mechanisms between the regimes are discussed below. 3.5.1. Transition between low-voltage and mid-voltage regimes The ionic wind effect dominated in the low-voltage regime. The AC electric field strongly affected the structure of the flame root and strengthened the mixing of the fuel and air, although no significant oscillating bulk flow was formed with such a high frequency. The unburned combustible gas was diluted and the burning intensity was weakened. The ionic wind effect increased with the
Fig. 7. Variations of chemiluminescence intensity values for OH⁄ and CH⁄ with AC voltages. (a): OH⁄, and (b) CH⁄.
Fig. 8. Diagram of flame regimes in terms of AC voltage and frequency (circles: critical voltages for transition between mid-voltage and high-voltage regimes, solid circles obtained in this work and blank circles from [25]; squares: critical voltages for transition between the low-voltage and mid-voltage regimes, solid squares obtained in this work and blank squares from [25]).
applied voltage. This suppressed soot, increased XCO and decreased XNO in the low-voltage regime. The critical voltage for transition between the low-voltage and mid-voltage regimes decreased as frequency increased. The theory of high-frequency gas discharge explains this. This theory [29] states that there is a critical frequency, fc, below which the ions and electrons induced by the electron avalanche have sufficient time to move to the electrodes. This maintains a low level for the number density of the charged particles in the space between the two electrodes. In this situation, the gas was difficult to discharge unless the voltage was very high. When f > fc, the ions had insufficient time to move to the electrodes before changing direction, whereas the electrons could approach the electrode because of their high mobility. Hence, a large amount of ions accumulated within the space between the electrodes and coupled with the new electron avalanches. The number density of the ions rapidly increased, leading to a silent non-self-sustained discharge. Although this kind of discharge did not generate a high current through the
Y. Zhang et al. / Fuel 109 (2013) 350–355
discharge channel [29], it produced a number of activated radicals and ions to strengthen the electrical–chemical effect. The electrical–chemical effect strengthened the burning intensity and it increased with the increasing voltage. When the voltage increased to a certain value, the electrical–chemical effect prevailed over the ionic wind effect. As a result, a transition occurred from the low-voltage to the mid-voltage regime. When the frequency was higher, the ions accumulated more rapidly and the silent nonself-sustained discharge occurred at a lower voltage. 3.5.2. Transition between mid-voltage and high-voltage regimes During the experiments, a hissing sound as well as streamers near the sharp tips of the electrodes were detected or observed as the AC voltage exceeded the threshold of the corona discharge. Correspondingly, transition occurred from the mid-voltage to the high-voltage regime. For this transition, the effects of the AC electric field were complicated. The ionic wind, electrical–chemical and thermal effects were highly coupled and caused distinctive flame behavior compared with the one in the low-voltage and the mid-voltage regime. More sophisticated experiments are required to assess the mid-voltage to high-voltage transition mechanism.
4. Conclusions The voltage of high frequency alternating electric fields significantly influenced both the appearance and NO emission of laminar non-premixed methane/air jet flames. When no alternating electric field was applied, flames were stable at the nozzle exit, bluish at the base and yellowish in the conical tip. However, once applied, different voltage regimes, ranging from 0–4.0 kV, produced different responses from the flame. In the low-voltage regime (0–1.0 kV), as voltage increased, the upper yellowish zone of the flame became narrower and the base bluish zone became wider. The conical tip of flame became sharper. The CO molar fraction increased in the flue gas but the NO molar fraction decreased in the flue gas. Both OH⁄ and CH⁄ chemiluminescence intensity values measured in the flame zone decreased 50%. In the mid-voltage regime (1.0–3.0 kV), the flame behaved inversely compared to the low-voltage regime. The upper yellowish zone became larger and the base bluish zone became narrower. The flame top gradually became a plump arc again. The CO molar fraction increased in the flue gas and the NO molar fraction decreased in the flue gas. CH⁄ and OH⁄ chemiluminescence intensity values in the flame zone increased but were still lower than the values observed without the application of an alternating electric field. In the high-voltage regime (3.0–4.0 kV), a hissing sound was heard due to the corona discharge occurrence between sharp parts of the electrodes. The upper yellowish zone of the flame disappeared, the CO molar fraction increased in the flue gas and the NO molar fraction decreased in the flue gas. Based on high-frequency discharge theory, the competition between thermal, ionic wind and electrical–chemical effects explained the non-monotonic flame behavior with respect to AC voltage. The ionic wind effect majored in the low-voltage regime while the electrical–chemical effect dominated the mid-voltage regime. In the high-voltage regime, the three effects were highly coupled and the flame response was more complicated. More detailed research into associated chemical kinetics, the effects of fuel type, soot formation and flow mechanics interaction is still required.
355
Acknowledgement This study is supported by NFSC 51076081).
(Nos. 51176095 and
References [1] Starikovskii AY. Plasma supported combustion. Proc Combust Inst 2005;30(2):2405–17. [2] Uddi M, Jiang N, Mintusov E, Adamovich IV, Lempert WR. Atomic oxygen measurements in air and air/fuel nanosecond pulse discharges by two photon laser induced fluorescence. Proc Combust Inst 2009;32(1):929–36. [3] Metghalchi H, Eisazadeh-Far K, Parsinejad F, Keck J. A thermodynamic model for argon plasma kernel formation. Int J Thermodyn 2010;13(4):119–25. [4] Vatazhin AB, Likhter VA, Sepp VA, Shul’Gin VI. Effect of an electric field on the nitrogen oxide emission and structure of a laminar propane diffusion flame. Fluid dyn 1995;30(2):166–74. [5] Marcum SD, Ganguly BN. Electric-field-induced flame speed modification. Combust Flame 2005;143(1–2):27–36. [6] Vega EV, Shin SS, Lee KY. NO emission of oxygen-enriched CH4/O2/N2 premixed flames under electric field. Fuel 2007;86(4):512–9. [7] Altendorfner F, Kuhl J, Zigan L, Leipertz A. Study of the influence of electric fields on flames using planar LIF and PIV techniques. Proc Combust Inst 2011;33(2):3195–201. [8] Vega EV, Lee KY. An experimental study on laminar CH4/O2/N2 premixed flames under an electric field. J Mech Sci Technol 2008;22(2):312–9. [9] Ombrello T, Won SH, Ju Y, Williams S. Flame propagation enhancement by plasma excitation of oxygen. Part I: Effects of O3. Combust Flame 2010;157(10):1906–15. [10] Ombrello T, Won SH, Ju Y, Williams S. Flame propagation enhancement by plasma excitation of oxygen. Part II: Effects of O2 (a1Dg). Combust Flame 2010;157(10):1916–28. [11] Sun WT, Uddi M, Ombrello T, Won SH, Carter C, Ju YG. Effects of nonequilibrium plasma discharge on counterflow diffusion flame extinction. Proc Combust Inst 2011;33(2):3211–8. [12] Sun W, Uddi M, Won SH, Ombrello T, Carter C, Ju Y. Kinetic effects of nonequilibrium plasma-assisted methane oxidation on diffusion flame extinction limits. Combust Flame 2012;159(1):221–9. [13] Saito M, Arai T, Arai M. Control of soot emitted from acetylene diffusion flames by applying an electric field. Combust Flame 1999;119(3):356–66. [14] Sakhrieh A, Lins G, Dinkelacker F, Hammer T, Leipertz A, Branston DW. The influence of pressure on the control of premixed turbulent flames using an electric field. Combust Flame 2005;143(3):313–22. [15] van den Boom J, Konnov AA, Verhasselt A, Kornilov VN, de Goey L, Nijmeijer H. The effect of a DC electric field on the laminar burning velocity of premixed methane/air flames. Proc Combust Inst 2009;32(1):1237–44. [16] Zake M, Turlajs D, Purmals M. Electric field control of NOx formation in the flame channel flows. Global Nest: Int J 2000;2:99–109. [17] Pedersen T, Brown RC. Simulation of electric field effects in premixed methane flames. Combust Flame 1993;94(4):433–48. [18] Hu J, Rivin B, Sher E. The effect of an electric field on the shape of co-flowing and candle-type methane–air flames. Exp Therm Fluid Sci 2000;21(1– 3):124–33. [19] Yamashita K, Karnani S, Dunn-Rankin D. Numerical prediction of ion current from a small methane jet flame. Combust Flame 2009;156(6):1227–33. [20] Belhi M, Domingo P, Vervisch P. Direct numerical simulation of the effect of an electric field on flame stability. Combust Flame 2010;157(12):2286–97. [21] Lee SM, Park CS, Cha MS, Chung SH. Effect of electric fields on the liftoff of nonpremixed turbulent jet flames. Plasma Sci IEEE Trans 2005;33(5):1703–9. [22] Cha MS, Lee SM, Kim KT, Chung SH. Soot suppression by nonthermal plasma in coflow jet diffusion flames using a dielectric barrier discharge. Combust Flame 2005;141(4):438–47. [23] Won SH, Cha MS, Park CS, Chung SH. Effect of electric fields on reattachment and propagation speed of tribrachial flames in laminar coflow jets. Proc Combust Inst 2007;31(1):963–70. [24] Won SH, Ryu SK, Kim MK, Cha MS, Chung SH. Effect of electric fields on the propagation speed of tribrachial flames in coflow jets. Combust Flame 2008;152(4):496–506. [25] Kim MK, Ryu SK, Won SH, Chung SH. Electric fields effect on liftoff and blowoff of nonpremixed laminar jet flames in a coflow. Combust Flame 2010;157(1):17–24. [26] Ryu SK, Kim YK, Kim MK, Won SH, Chung SH. Observation of multi-scale oscillation of laminar lifted flames with low-frequency ac electric fields. Combust Flame 2010;157(1):25–32. [27] Kim MK, Chung SH, Kim HH. Effect of AC electric fields on the stabilization of premixed bunsen flames. Proc Combust Inst 2011;33(1):1137–44. [28] Kono M, Carleton FB, Jones AR, Weinberg FJ. The effect of nonsteady electric fields on sooting flames. Combust Flame 1989;78(3–4):357–64. [29] Raizer YP, Kisin VI, Allen JE. Gas discharge physics. Berlin: Springer-Verlag; 1991. [30] Mikofski MA, Williams TC, Shaddix CR, Blevins LG. Flame height measurement of laminar inverse diffusion flames. Combust Flame 2006;146(1–2):63–72.
Contents lists available at SciVerse ScienceDirect
Fuel journal homepage: www.elsevier.com/locate/fuel
Effect of high-frequency alternating electric fields on the behavior and nitric oxide emission of laminar non-premixed flames Yang Zhang, Yuxin Wu, Hairui Yang, Hai Zhang ⇑, Min Zhu Key Laboratory for Thermal Science and Power Engineering of the Ministry of Education, Department of Thermal Engineering, Tsinghua University, Beijing 100084, PR China
h i g h l i g h t s " We studied the effects of AC electric fields on non-premixed flames. " Flame image, FTIR pattern and chemiluminescent emissions were obtained. " There exist three distinct regimes across the tested voltage range. " The flame behaved non-monotonically as the increasing voltage. " The flame behavior was explained by the competition of three effects.
a r t i c l e
i n f o
Article history: Received 18 December 2012 Received in revised form 26 December 2012 Accepted 27 December 2012 Available online 10 January 2013 Keywords: Alternating electric field Non-premixed flame NO emission Frequency Voltage
a b s t r a c t This paper examined the behavior and NO emission of laminar non-premixed methane/air jet flames when subjected to high frequency alternating electric fields of 10 kHz over the voltage range of 0–4.0 kV. In particular, this paper examined variations of flame shape and luminosity, CO and NO molar fractions in the downstream flue gas, and chemiluminescence from OH⁄ and CH⁄ in the voltageinfluenced flame zone. The results showed that with no application of an alternating electric field, flames were stable at the nozzle exit, bluish at the base and yellowish at the conical tip. However, once applied, different voltage regimes produced different responses from the flame. In the low-voltage regime of 0–1.0 kV, increasing the voltage narrowed the top yellowish zone of the flame and sharpened its conical tip, increased the CO molar fraction in the flue gas, decreased the NO molar fraction in the flue gas, and decreased the chemiluminescence intensity of OH⁄ and CH⁄ in the flame zone by 50%. At 1.0 kV, both CO and NO molar fractions reached extreme values, and the flame was at its weakest. In the mid-voltage regime of 1.0–3.0 kV, increasing the voltage resulted in an inverse response from the flames compared to the low-voltage regime. In the high-voltage regime of 3.0–4.0 kV, increasing the voltage resulted in the gradual disappearance of the top yellowish zone of the flame, increased the CO molar fraction in the flue gas and decreased the NO molar fraction. The transition mechanisms between the regimes are discussed within the context of the high-frequency discharge theory. Three competing effects explain the non-monotonic flame response to the voltage: thermal, ionic wind, and electrical–chemical. The analysis showed that the ionic wind effect majored in the low-voltage regime, the electrical–chemical effect dominated the mid-voltage regime, and all three effects were highly coupled in the high-voltage regime. Ó 2013 Elsevier Ltd. All rights reserved.
1. Introduction The interaction between an electric field and flame behavior has attracted wide attention in the last few decades. Research shows that electric fields affect flames in three major ways: the thermal effect [1–3], which is caused by electrical energy input; the ionic wind effect [4–8]; and the electrical–chemical effect [9–12]. The thermal effect dominates when there is a large current across the
⇑ Corresponding author. Tel.: +86 10 6278 8523; fax: +86 10 6278 1743. E-mail address: [email protected] (H. Zhang). 0016-2361/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.fuel.2012.12.083
electric field, such as a spark plug [3]. The ionic wind effect causes fluid dynamic changes in the flow field, inducing an ‘‘ionic wind’’ [4–8]. The electrical–chemical effect produces high speed electrons, radicals, ions and excited molecules in the pre-flame zone, which directly change the chemistry of the flames [9–12]. Various experimental and numerical studies have been conducted on the influence of direct-current (DC) electric fields on flame behaviors. Experiments showed that the DC electric field strongly affected flame shape [6,13,14], flame propagation speed [5,15], NOx emissions [4,6,14,16] and soot formation [4,13]. Numerical simulations showed that the ionic wind effect dominated [17–20].
Y. Zhang et al. / Fuel 109 (2013) 350–355
Studies on the effects of an alternating-current (AC) electric field on flame behaviors are few. Previous studies mainly focused on the ionic wind effect. Chung et al. published a series of works [21–27] on the effects of the AC electric field on flame stabilization, and produced the following findings. The speed of tribrachial flame propagation increased under AC voltage [24]. The edge structure and the detachment velocity of the nozzle attached non-premixed propane/air flame were both affected by an AC electric field. The detachment velocity changed non-monotonically with the voltage increase at 60 Hz [25]. The lift-off height presented an oscillation characteristic when lower than 30 Hz, due to the directionalternating ionic wind [26]. No oscillation was found in the range of 60–1500 Hz [25]. The height of the diffusion flame decreased with the increasing AC voltage at 400 Hz. Kono et al. [28] noted a delay time in the formation of ionic wind, normally in the order of 10 ms, indicating that the ionic wind did not form when the alternating frequency was higher than 50 Hz. Flame behavior was significantly different to that of the DC and low-frequency AC electric fields when the frequency of the AC electric field was high (e.g. >2000 Hz). This may be because time was insufficient for momentum transfer from the charged particles to the neutral ones. As a result, the electric field did not introduce significant bulk flow motion [25,28]. Alternatively, it may be because the charged particles could not easily reach the electrodes due to high-frequency alternating properties. Therefore, a large amount of charged particles existed within the space between the two electrodes [29]. Studies on the effects of high-frequency alternating electric fields remain limited in number. Consequently, this paper examined the behavior and NO emission of laminar non-premixed methane/air flames when subjected to high frequency AC electric fields of 10 kHz over the voltage range of 0–4.0 kV. In particular, this paper applied different voltages at different fuel and co-flow velocities to study variations of flame shapes and to study CO and NO molar fractions in the downstream flue gas. This research also measured chemiluminescence from two excited radicals, OH⁄ and CH⁄, in order to provide insight into the effects of highfrequency AC electric fields on combustion. 2. Experimental apparatus and methodology 2.1. Experimental apparatus Fig. 1 shows the experimental apparatus, which consisted of three main parts: (1) a coaxial burner and associated gas supply system; (2) a high frequency AC power supply system; (3) a measurement system. 2.1.1. Coaxial burner and associated gas supply system The fuel nozzle was made of a quartz tube with an inner diameter of 8.0 mm and thickness of 1.0 mm (Fig. 1). The tube length was 250 mm to ensure that the velocity profile of the fuel stream was fully developed at the nozzle exit. The fuel used in the experiments was high purity CH4 (>99.99%). Compressed air, as the oxidizer, was supplied around the fuel tube from a concentric quartz tube with an inner diameter of 65 mm. The flow rates of the gas streams were controlled by mass flow meters that were precalibrated by a wet gas meter. 2.1.2. High frequency AC power supply system An AC electric field was applied in the experiments. A metal mesh plate located at 50 mm above the upper rim of the burner was used as the anode and another metal mesh attached to the outside surface of the fuel nozzle was used as the cathode. A sinusoidal AC electric field with 10 kHz alternating frequency
351
was supplied. The magnitude of the field (AC voltage) was adjustable. The AC wave was measured and recorded by an oscilloscope. 2.1.3. Measurement system A flute-shaped probe located 750 mm above the burner sampled the combustion products. A Fourier transform infrared spectrometer (FT-IR) with the model of NETZSCH STA 409C analyzed the flue gas introduced into the probe. The CO and NO molar fractions in the flue gas were measured. A grating spectrometer with the model of ZOLIX SPB300 equipped with a photomultiplier tube (PMT) analyzed the chemiluminescence from the flame. The chemiluminescence intensity of OH⁄ and CH⁄ in the flame zone was measured. A digital camera recorded the flame images (exposure time: 0.5 s; resolution: 72 dpi). 2.2. Experimental conditions Experiments were conducted at four different air flow velocities and four different fuel flow velocities (Table 1). 3. Results and discussion 3.1. AC voltage effect on flame shapes and luminosities Fig. 2 shows images of the non-premixed laminar jet flames at various AC voltages. A typical non-premixed jet flame was stabilized at the exit of the nozzle when no electric field was applied (peak voltage 0.0 kV). The upper portion of the flame was a yellowish color due to the radiation of the soot formations. The base part of the flame was a bluish color due to the radiation of CH radicals. The flame shape and luminosity distinctly changed when an AC electric field was applied. Based on the variation of flame luminosities under different applied voltage, there existed three different regimes depending on the AC voltages: (1) the low-voltage regime with voltage lower than 1.0 kV; (2) the mid-voltage regime with voltage in the range of 1.0–3.0 kV; (3) the high-voltage regime with voltage higher than 3.5 kV. Case 1 is taken as an instance to describe the flame behaviors in each of these three regimes. In the low-voltage regime, as voltage increased, the upper yellowish zone of the flame became narrower and the base bluish zone became wider. Also, the conical tip of flame became sharper. The flame top was sharpest at 1.0 kV. Under these conditions, the yellowish zone was the smallest and the overall flame was the darkest, indicating that soot formation was minimized. The experiments showed that the flame became unstable and that the flow perturbations even extinguished the flame in some instances. In the mid-voltage regime, as voltage increased, there was an inverse response from the flames compared to the low-voltage regime. The upper yellowish zone became larger and the base bluish zone became narrower. The flame top gradually became a plump arc again and the flame was more stable compared to its state at 1.0 kV. In the high-voltage regime, as voltage increased, the AC electric field zone typically started to produce a hissing sound between 3.0 and 3.5 kV. This indicated a slight corona discharge from the sharp tips of the electrodes, which generated a certain amount of plasma. The sound-producing pressure fluctuation affected the flame. The upper yellowish zone of the flame quickly disappeared and the flame tip dispersed (Fig. 2). This result indicated that soot formation was suppressed by the AC corona discharge, similar to the phenomenon found in the dielectric barrier discharge [22]. Furthermore, the flame partially detached from the nozzle and randomly rotated around the nozzle rim. This phenomenon was consistent with the observation by Chung et al. [25] for low AC frequency. However, where Chung et al. observed lift-off as voltage
352
Y. Zhang et al. / Fuel 109 (2013) 350–355
Fig. 1. Schematic of the experimental system.
Table 1 Fuel and air velocities. Case no.
Fuel flow velocity (cm/s)
Air flow velocity (cm/s)
Fuel jet Reynolds number
1 2 3 4
3.2 3.9 4.6 5.3
0.87 0.86 0.84 0.83
115 141 168 194
continued to increase [25], this research found that the flame was extinguished. This research found that the critical voltage between the midvoltage and high-voltage regimes differed for different fuel jet velocities. As fuel jet velocity increased, the critical voltage decreased. The flame surface also became larger with more ions and exited radicals. Therefore, the corona discharge occurred at lower voltage. Since the transition from the mid-voltage to high-voltage regime was caused by the corona discharge, the critical voltage decreased with the increasing fuel jet velocity. However, the difference was within 0.5 kV throughout the experimental range. 3.2. AC voltage effect on flame height Fig. 3 shows the flame height variations and inner fuel zone for the applied AC voltages under different exit velocities of fuel streams (UF,0). The flame boundary was defined using a gray value threshold method [30]. In this paper, the gray value 90 (out of 255 maximum) was determined as the threshold of the flame height location. The flame height was the vertical distance from the nozzle exit to the flame tip (Fig. 3a). The inner fuel zone had visible unburned fuel in a relatively dark color in the flame images (Fig. 3b). The flame height variations for the AC voltage were within 10% for all tested cases, indicating that the effect of the electric field on the flame height was insignificant. However, the inner fuel zone height
Fig. 2. Selected images of the flames at various peak voltages and fuel exit velocities. Three distinct regimes: (1) <1.0 kV; (2) 1.0–3.0 kV; (3) >3.0 kV. (a): UF,0 = 3.2 cm/s; (b): UF,0 = 3.9 cm/s; (c): UF,0 = 4.6 cm/s; (d): UF,0 = 5.3 cm/s).
Y. Zhang et al. / Fuel 109 (2013) 350–355
353
3.3. AC voltage effect on CO and NO formation Fig. 4 shows the infrared patterns of the flue gas analyzed by the FT-IR spectrometer. The peak values of CO and NO significantly differed for different AC voltages. Fig. 5 shows the quantitative results, in which XCO was the CO molar fraction in the flue gas and XNO was the NO molar fraction in the flue gas. XCO increased with UF,0 under a fixed voltage (Fig. 5a). This was expected as more oxidizer causes more complete combustion. However, at a fixed UF,0, voltage variations yielded different XCO in the three regimes. Taking the case of UF,0 = 5.3 cm/s, XCO increased with AC voltage in the low-voltage regime, while it decreased with AC voltage in the mid-voltage regime. When the voltage approached the threshold value of the high-voltage regime, XCO increased and then peaked before the flame extinguished. Flames with other UF,0 values showed similar trends and their extreme values occurred at 1.0 kV. These results were consistent with the previous results (Fig. 2). However, XNO showed an opposite trend against XCO (Fig. 5b). As the AC voltage increased, XNO decreased in the lowvoltage regime, increased in the mid-voltage regime, and decreased in the high-voltage regime. The variation of XNO with UF,0 was non-monotonic (Fig. 5b). This may be caused by variations in the fuel flow rate and air flow rate.
Fig. 3. Variations of flames and inner fuel zone heights for different AC voltages (solid square: UF,0 = 3.2 cm/s; solid triangle: UF,0 = 3.9 cm/s; blank square: UF,0 = 4.6 cm/s; solid circle: UF,0 = 5.3 cm/s;).
3.4. Chemiluminescence emissions from the flame zone under AC fields A grating spectrometer analyzed the chemiluminescence spectrum of the flame zone (white circle) at 0.0, 1.0 and 2.5 kV (Fig. 6). The spectrum pattern represented the integration of the chemilumi-
Fig. 4. FT-IR spectrum patterns of downstream flue gas. (Where UF,0 is the fuel flow velocity at the nozzle exit.) (a) Voltage = 0.5 kV; (b) voltage = 1.0 kV.
was strongly affected by the AC voltage. In the low-voltage regime, it increased linearly with the voltage, whereas in the mid-voltage regime, it gradually decreased. Also, the inner fuel zone height was very sensitive to the flame edge structure. The results were consistent with the previous findings [25]: the flame closest to the nozzle rim was influenced by the electric field over a low frequency range of 60–1500 Hz. The mechanism of the nonmonotonic AC voltage effect is discussed in Section 3.5.
Fig. 5. Variations of CO and NO molar fractions, XCO and XNO, in the flue gas against AC voltages in the flue gas. (a): XCO; (b): XNO (solid square: UF,0 = 3.2 cm/s; blank square: UF,0 = 3.9 cm/s; solid circle: UF,0 = 4.6 cm/s; solid triangle: UF,0 = 5.3 cm/s).
354
Y. Zhang et al. / Fuel 109 (2013) 350–355
Fig. 6. Chemiluminescence Intensity values for the flame at different AC voltages (characteristic peak of OH⁄ locates at approximately 310 nm while the CH⁄ peak locates at approximately 430.5 nm).
nescence intensity of the entire flame zone. The chemiluminescence intensity was strongest when no electric field was applied (0.0 kV). The chemiluminescence intensity dropped 50% when AC voltage increased from 0 to 1.0 kV. The chemiluminescence intensity increased when AC voltage was 2.5 kV, but its value was still lower than the value observed without an electric field. The chemiluminescence intensity values of OH⁄ and CH⁄ were good indicators of the heat release rate. The characteristic peaks of OH⁄ and CH⁄ were located around 310 nm and 430.5 nm, respectively. The chemiluminescence intensity values of OH⁄ and CH⁄ were plotted against AC voltages (Fig. 7). Fig. 7 shows that both intensity values first decreased, then increased, and finally plateaued with the increasing AC voltage. The minimum OH⁄ and CH⁄ chemiluminescence intensity values were both located at 1.0 kV. In this situation, the ionic wind effect strengthened the mixing of the fuel and air at the flame root and decreased the burning intensity. It is generally accepted for methane combustion that the step with the greatest heat release is CO + OH ? CO2 + H. It follows that combustion was quite weak and incomplete at 1.0 kV, resulting in a high, unburned CO emission and a low flame temperature. Thermal NO was suppressed and prompt NO dominated the NO formation because the temperature was low. For the prompt NO formation, the O, CH and CH2 radicals each played a key role. Fig. 7b shows that the CH radical amount was smallest when the voltage was 1.0 kV. Hence, overall NO formation was suppressed by the weak combustion. The non-monotonic nature of the chemiluminescence agreed with those of the CO and NO emissions. 3.5. Transition mechanisms between the regimes Fig. 8 shows the diagram of flame regimes in terms of AC voltage and frequency along with the data obtained from the present work and previous studies [25]. The critical voltages for the regime transitions decreased linearly with an increasing denary logarithm of the frequency (log f). The two linear dashed lines of the critical voltages divide the figure into three regions that correspond to the three regimes (Fig. 2). The transient mechanisms between the regimes are discussed below. 3.5.1. Transition between low-voltage and mid-voltage regimes The ionic wind effect dominated in the low-voltage regime. The AC electric field strongly affected the structure of the flame root and strengthened the mixing of the fuel and air, although no significant oscillating bulk flow was formed with such a high frequency. The unburned combustible gas was diluted and the burning intensity was weakened. The ionic wind effect increased with the
Fig. 7. Variations of chemiluminescence intensity values for OH⁄ and CH⁄ with AC voltages. (a): OH⁄, and (b) CH⁄.
Fig. 8. Diagram of flame regimes in terms of AC voltage and frequency (circles: critical voltages for transition between mid-voltage and high-voltage regimes, solid circles obtained in this work and blank circles from [25]; squares: critical voltages for transition between the low-voltage and mid-voltage regimes, solid squares obtained in this work and blank squares from [25]).
applied voltage. This suppressed soot, increased XCO and decreased XNO in the low-voltage regime. The critical voltage for transition between the low-voltage and mid-voltage regimes decreased as frequency increased. The theory of high-frequency gas discharge explains this. This theory [29] states that there is a critical frequency, fc, below which the ions and electrons induced by the electron avalanche have sufficient time to move to the electrodes. This maintains a low level for the number density of the charged particles in the space between the two electrodes. In this situation, the gas was difficult to discharge unless the voltage was very high. When f > fc, the ions had insufficient time to move to the electrodes before changing direction, whereas the electrons could approach the electrode because of their high mobility. Hence, a large amount of ions accumulated within the space between the electrodes and coupled with the new electron avalanches. The number density of the ions rapidly increased, leading to a silent non-self-sustained discharge. Although this kind of discharge did not generate a high current through the
Y. Zhang et al. / Fuel 109 (2013) 350–355
discharge channel [29], it produced a number of activated radicals and ions to strengthen the electrical–chemical effect. The electrical–chemical effect strengthened the burning intensity and it increased with the increasing voltage. When the voltage increased to a certain value, the electrical–chemical effect prevailed over the ionic wind effect. As a result, a transition occurred from the low-voltage to the mid-voltage regime. When the frequency was higher, the ions accumulated more rapidly and the silent nonself-sustained discharge occurred at a lower voltage. 3.5.2. Transition between mid-voltage and high-voltage regimes During the experiments, a hissing sound as well as streamers near the sharp tips of the electrodes were detected or observed as the AC voltage exceeded the threshold of the corona discharge. Correspondingly, transition occurred from the mid-voltage to the high-voltage regime. For this transition, the effects of the AC electric field were complicated. The ionic wind, electrical–chemical and thermal effects were highly coupled and caused distinctive flame behavior compared with the one in the low-voltage and the mid-voltage regime. More sophisticated experiments are required to assess the mid-voltage to high-voltage transition mechanism.
4. Conclusions The voltage of high frequency alternating electric fields significantly influenced both the appearance and NO emission of laminar non-premixed methane/air jet flames. When no alternating electric field was applied, flames were stable at the nozzle exit, bluish at the base and yellowish in the conical tip. However, once applied, different voltage regimes, ranging from 0–4.0 kV, produced different responses from the flame. In the low-voltage regime (0–1.0 kV), as voltage increased, the upper yellowish zone of the flame became narrower and the base bluish zone became wider. The conical tip of flame became sharper. The CO molar fraction increased in the flue gas but the NO molar fraction decreased in the flue gas. Both OH⁄ and CH⁄ chemiluminescence intensity values measured in the flame zone decreased 50%. In the mid-voltage regime (1.0–3.0 kV), the flame behaved inversely compared to the low-voltage regime. The upper yellowish zone became larger and the base bluish zone became narrower. The flame top gradually became a plump arc again. The CO molar fraction increased in the flue gas and the NO molar fraction decreased in the flue gas. CH⁄ and OH⁄ chemiluminescence intensity values in the flame zone increased but were still lower than the values observed without the application of an alternating electric field. In the high-voltage regime (3.0–4.0 kV), a hissing sound was heard due to the corona discharge occurrence between sharp parts of the electrodes. The upper yellowish zone of the flame disappeared, the CO molar fraction increased in the flue gas and the NO molar fraction decreased in the flue gas. Based on high-frequency discharge theory, the competition between thermal, ionic wind and electrical–chemical effects explained the non-monotonic flame behavior with respect to AC voltage. The ionic wind effect majored in the low-voltage regime while the electrical–chemical effect dominated the mid-voltage regime. In the high-voltage regime, the three effects were highly coupled and the flame response was more complicated. More detailed research into associated chemical kinetics, the effects of fuel type, soot formation and flow mechanics interaction is still required.
355
Acknowledgement This study is supported by NFSC 51076081).
(Nos. 51176095 and
References [1] Starikovskii AY. Plasma supported combustion. Proc Combust Inst 2005;30(2):2405–17. [2] Uddi M, Jiang N, Mintusov E, Adamovich IV, Lempert WR. Atomic oxygen measurements in air and air/fuel nanosecond pulse discharges by two photon laser induced fluorescence. Proc Combust Inst 2009;32(1):929–36. [3] Metghalchi H, Eisazadeh-Far K, Parsinejad F, Keck J. A thermodynamic model for argon plasma kernel formation. Int J Thermodyn 2010;13(4):119–25. [4] Vatazhin AB, Likhter VA, Sepp VA, Shul’Gin VI. Effect of an electric field on the nitrogen oxide emission and structure of a laminar propane diffusion flame. Fluid dyn 1995;30(2):166–74. [5] Marcum SD, Ganguly BN. Electric-field-induced flame speed modification. Combust Flame 2005;143(1–2):27–36. [6] Vega EV, Shin SS, Lee KY. NO emission of oxygen-enriched CH4/O2/N2 premixed flames under electric field. Fuel 2007;86(4):512–9. [7] Altendorfner F, Kuhl J, Zigan L, Leipertz A. Study of the influence of electric fields on flames using planar LIF and PIV techniques. Proc Combust Inst 2011;33(2):3195–201. [8] Vega EV, Lee KY. An experimental study on laminar CH4/O2/N2 premixed flames under an electric field. J Mech Sci Technol 2008;22(2):312–9. [9] Ombrello T, Won SH, Ju Y, Williams S. Flame propagation enhancement by plasma excitation of oxygen. Part I: Effects of O3. Combust Flame 2010;157(10):1906–15. [10] Ombrello T, Won SH, Ju Y, Williams S. Flame propagation enhancement by plasma excitation of oxygen. Part II: Effects of O2 (a1Dg). Combust Flame 2010;157(10):1916–28. [11] Sun WT, Uddi M, Ombrello T, Won SH, Carter C, Ju YG. Effects of nonequilibrium plasma discharge on counterflow diffusion flame extinction. Proc Combust Inst 2011;33(2):3211–8. [12] Sun W, Uddi M, Won SH, Ombrello T, Carter C, Ju Y. Kinetic effects of nonequilibrium plasma-assisted methane oxidation on diffusion flame extinction limits. Combust Flame 2012;159(1):221–9. [13] Saito M, Arai T, Arai M. Control of soot emitted from acetylene diffusion flames by applying an electric field. Combust Flame 1999;119(3):356–66. [14] Sakhrieh A, Lins G, Dinkelacker F, Hammer T, Leipertz A, Branston DW. The influence of pressure on the control of premixed turbulent flames using an electric field. Combust Flame 2005;143(3):313–22. [15] van den Boom J, Konnov AA, Verhasselt A, Kornilov VN, de Goey L, Nijmeijer H. The effect of a DC electric field on the laminar burning velocity of premixed methane/air flames. Proc Combust Inst 2009;32(1):1237–44. [16] Zake M, Turlajs D, Purmals M. Electric field control of NOx formation in the flame channel flows. Global Nest: Int J 2000;2:99–109. [17] Pedersen T, Brown RC. Simulation of electric field effects in premixed methane flames. Combust Flame 1993;94(4):433–48. [18] Hu J, Rivin B, Sher E. The effect of an electric field on the shape of co-flowing and candle-type methane–air flames. Exp Therm Fluid Sci 2000;21(1– 3):124–33. [19] Yamashita K, Karnani S, Dunn-Rankin D. Numerical prediction of ion current from a small methane jet flame. Combust Flame 2009;156(6):1227–33. [20] Belhi M, Domingo P, Vervisch P. Direct numerical simulation of the effect of an electric field on flame stability. Combust Flame 2010;157(12):2286–97. [21] Lee SM, Park CS, Cha MS, Chung SH. Effect of electric fields on the liftoff of nonpremixed turbulent jet flames. Plasma Sci IEEE Trans 2005;33(5):1703–9. [22] Cha MS, Lee SM, Kim KT, Chung SH. Soot suppression by nonthermal plasma in coflow jet diffusion flames using a dielectric barrier discharge. Combust Flame 2005;141(4):438–47. [23] Won SH, Cha MS, Park CS, Chung SH. Effect of electric fields on reattachment and propagation speed of tribrachial flames in laminar coflow jets. Proc Combust Inst 2007;31(1):963–70. [24] Won SH, Ryu SK, Kim MK, Cha MS, Chung SH. Effect of electric fields on the propagation speed of tribrachial flames in coflow jets. Combust Flame 2008;152(4):496–506. [25] Kim MK, Ryu SK, Won SH, Chung SH. Electric fields effect on liftoff and blowoff of nonpremixed laminar jet flames in a coflow. Combust Flame 2010;157(1):17–24. [26] Ryu SK, Kim YK, Kim MK, Won SH, Chung SH. Observation of multi-scale oscillation of laminar lifted flames with low-frequency ac electric fields. Combust Flame 2010;157(1):25–32. [27] Kim MK, Chung SH, Kim HH. Effect of AC electric fields on the stabilization of premixed bunsen flames. Proc Combust Inst 2011;33(1):1137–44. [28] Kono M, Carleton FB, Jones AR, Weinberg FJ. The effect of nonsteady electric fields on sooting flames. Combust Flame 1989;78(3–4):357–64. [29] Raizer YP, Kisin VI, Allen JE. Gas discharge physics. Berlin: Springer-Verlag; 1991. [30] Mikofski MA, Williams TC, Shaddix CR, Blevins LG. Flame height measurement of laminar inverse diffusion flames. Combust Flame 2006;146(1–2):63–72.