Laser induced spark ignition of coaxial methane/oxygen/nitrogen diffusion flames

Transcription

1 Laser induced spark ignition of coaxial methane/oxygen/nitrogen diffusion flames Xiaohui Li, 1,2,3 Yang Yu, 1,2 Xin Yu, 1,2 Chang Liu, 1,2 Rongwei Fan, 1,2 and Deying Chen 1,2,* 1 National Key Laboratory of Science and Technology on Tunable Laser, Harbin Institute of Technology, Harbin, 15000, China 2 Institute of Opto-electronics, Harbin Institute of Technology, Harbin, 15000, China 3 lixiaohuihit@163.com * dychen@hit.edu.cn Abstract: We report the laser induced spark ignition (LSI) of coaxial methane/oxygen/nitrogen diffusion flames using the 64 nm output of a Q-switched Nd:YAG laser. The minimum ignition energy (MIE) and ignition time of the LSI has been determined by measuring the emission signals due to the ignited flames. The effects of the gas mixture properties, including the overall equivalence ratio (Ф), oxygen concentration and flow rate, and the ignition positions on the two parameters have been investigated systematically. The variation of the MIE and ignition time with the experimental conditions has been compared with the existing results and discussed with a special concentration on the effects of the local Ф Optical Society of America OCIS codes: ( ) Laser-induced breakdown; ( ) Combustion diagnostics; ( ) Emission; ( ) Laser induced spark ignition, laser ignition. References and links 1. P. D. Ronney, Laser versus conventional ignition of flames, Opt. Eng. 33(2), (1994). 2. T. A. Spiglanin, A. Mcilroy, E. W. Fournier, R. B. Cohen, and J. A. Syage, Time-resolved imaging of flame kernels: laser spark ignition of H 2 /O 2 /Ar mixtures, Combust. Flame 2(3), 3 32 (1995). 3. T. X. Phuoc, Laser-induced spark ignition fundamental and applications, Opt. Lasers Eng. 44(5), (2006). 4. D. Bradley, C. G. W. Sheppard, I. M. Suardjaja, and R. Woolley, Fundamentals of high-energy spark ignition with lasers, Combust. Flame 13(1-2), (2004). 5. M. Weinrotter, H. Kopecek, E. Wintner, M. Lackner, and F. Winter, Application of laser ignition to hydrogenair mixtures at high pressures, Int. J. Hydrogen Energy 30(3), (2005). 6. L. Zimmer, K. Okai, and Y. Kurosawa, Combined laser induced ignition and plasma spectroscopy: fundamentals and application to a hydrogen-air combustor, Spectrochim. Acta, B At. Spectrosc. 62(12), (2007). 7. J. L. Beduneau, N. Kawahara, T. Nakayama, E. Tomita, and Y. Ikeda, Laser-induced radical generation and evolution to a self-sustaining flame, Combust. Flame 156(3), (2009).. T. X. Phuoc and F. X. White, Laser-induced spark ignition of CH 4 /air mixtures, Combust. Flame 119(3), (1999). 9. M. H. Morsy and S. H. Chung, Laser-induced multi-point ignition with a single-shot laser using two conical cavities for hydrogen/air mixture, Exp. Therm. Fluid Sci. 27(4), (2003).. J. X. Ma, D. R. Alexander, and D. E. Poulain, Laser spark ignition and combustion characteristics of methaneair mixtures, Combust. Flame 112(4), (199). 11. G. Herdin, J. Klausner, E. Wintner, M. Weinrotter, J. Graf, and K. Iskra, Laser ignition - a new concept to use and increase the pontentials of gas engines, in ASME Internal Combustion Engine Division 2005 Fall Technical Conference: AERS-ARICE Symposium on Gas Fired Reciprocating Engines(Ottawa, Canada, 2005). 12. Y.-L. Chen and J. W. L. Lewis, Visualization of laser-induced breakdown and ignition, Opt. Express 9(7), (2001). 13. M. Lackner, S. Charareh, F. Winter, K. Iskra, D. Rüdisser, T. Neger, H. Kopecek, and E. Wintner, Investigation of the early stages in laser-induced ignition by Schlieren photography and laser-induced fluorescence spectroscopy, Opt. Express 12(19), (2004). 14. N. Pavel, M. Tsunekane, and T. Taira, Composite, all-ceramics, high-peak power Nd:YAG/Cr 4+ :YAG monolithic micro-laser with multiple-beam output for engine ignition, Opt. Express 19(), (2011). (C) 2014 OSA February 2014 Vol. 22, No. 3 DOI:.1364/OE OPTICS EXPRESS 3447

2 15. G. Liedl, D. Schuocker, B. Geringer, J. Graf, D. Klawatsch, H. P. Lenz, W. F. Piock, M. Jetzinger, and P. Kapus, Laser induced ignition of gasoline direct injection engines, Proc. SPIE 5777, (2004). 16. S. Brieschenk, S. O Byrne, and H. Kleine, Laser-induced plasma ignition studies in a model scramjet engine, Combust. Flame 160(1), (2013). 17. R. J. Osborne, J. A. Wehrmeyer, H. P. Trinh, and J. W. Early, Evaluation and characterization study of dual pulse laser-induced spark(dplis) for rocket engine ignition system application, in 39th AIAA/ASME/SAE/ASEE Joint Propulsion Conference and Exhibit(AIAA Huntsville, Alabama, 2003), paper T. Razafimandimby, M. De Rosa, V. Schmidt, J. Sender, and M. Oschwald, Laser ignition of a GH 2 /LOX spray under vacuum conditions, in The European Combustion Meeting(2005), pp K. Hasegawa, K. Kusaka, A. Kumakawa, M. Sato, and M. Tadano, Laser ignition characteristics of GOX/GH 2 and GOX/GCH 4 propellants, in 39th AIAA/ASME/SAE/ASEE Joint Propulsion Conference and Exhibit(AIAA, Huntsville, Alabama, 2003), paper L. C. Liou, Laser ignition in liquid rocket engines, in 30th AIAA/SAE/ASME/ASEE Joint Propulsion Conference(AIAA, Indianapolis, IN, 1994), paper F. B. Carleton, N. Klein, K. Krallis, and F. J. Weinberg, Laser ignition of liquid propellants, Symposium (International) on Combustion 23, (1991). 22. M. De Rosa, J. Sender, H. Zimmermann, and M. Oschwald, Cryogenic spary ignition at high altitude conditions, in 42nd AIAA/ASME/SAE/ASEE Joint Propulsion Conference & Exhibit(AIAA, Sacramento, CA, 2006), paper C. Pauly, J. Sender, and M. Oschwald, Ignition of a gaseous methane/oxygen coaxial jet, Prog. Propulsion Phys. 1, (2009). 24. V. Schmidt, U. Wepler, O. Haidn, and M. Oschwald, Characterization of the primary ignition process of a coaxial GH 2 /LOx spray, in 42nd AIAA Aerospace Sciences Meeting and Exhibit(AIAA, Reno, Nevada, 2004). 25. J. L. Beduneau, B. Kim, L. Zimmer, and Y. Ikeda, Measurements of minimum ignition energy in premixed laminar methane/air flow by using laser induced spark, Combust. Flame 132(4), (2003). 26. H. Kopecek, H. Maier, G. Reider, F. Winter, and E. Wintner, Laser ignition of methane-air mixtures at high pressures, Exp. Therm. Fluid Sci. 27(4), (2003). 27. T. X. Phuoc, C. M. White, and D. H. McNeill, Laser spark ignition of a jet diffusion flame, Opt. Lasers Eng. 3(5), (2002). 2. F. B. Carleton, N. Klein, K. Krallis, and F. J. Weinberg, Laser ignition of liquid propellants, Twenty third symposium (International) on combustion 23, (1991). 29. T.-W. Lee, V. Jain, and S. Kozola, Measurements of minimum ignition energy by using laser sparks for hydrocarbon fuels in air: propane, dodecane, and jet-a fuel, Combust. Flame 125(4), (2001). 30. X. Li, B. W. Smith, and N. Omenetto, Laser spark ignition of premixed methane/air mixtures: parameter measurements and determination of key factors for ultimate ignition results, Combust. Flame. Submitted. 31. B. Lewis and G. Von Elbe, Combustion, Flames and Explosions of Gases (Academic Press, 197). 32. A. Lifshitz, K. Scheller, A. Burcat, and G. B. Skinner, Shock-tube investigation of ignition in methane-oxygenargon mixtures, Combust. Flame 16(3), (1971). 33. F. Ferioli, P. V. Puzinauskas, and S. G. Buckley, Laser-induced breakdown spectroscopy for on-line engine equivalence ratio measurements, Appl. Spectrosc. 57(9), (2003). 34. J. Kiefer, J. W. Tröger, Z. S. Li, and M. Aldén, Laser-induced plasma in methane and dimethyl ether for flame ignition and combustion diagnostics, Appl. Phys. B 3(1), (2011). 1. Introduction During the last two decades, laser induced spark ignition (LSI) has been proposed as a promising ignition technique [1 13] with many potential benefits, including easier control of ignition position and ignition timing, no electrodes and thus no heat loss towards the combustion chamber that may lead to extinguishment of the combustion systems [5], wider ignitable equivalence ratio range [11], feasibility of multi-point ignition [9] and ease of synchronization with the diagnostic systems [7]. With the developments of compact and stable solid laser systems [14] for LSI, practical applications of LSI have been demonstrated in several combustion systems, including internal combustion engines [15], natural gas engines [11], model scramjet engine [16], and rocket engines [17 24]. The minimum ignition energy (MIE) and ignition time are two important parameters for the design and evaluation of the combustion systems that apply the LSI technique. Several groups have recently characterized the laser ignition processes, with special concentration on the MIE measurements. Phuoc and White [] successfully ignited methane/air mixtures in a combustion cell using a nanosecond Nd:YAG laser, and found that the lowest ignition energy for equivalence ratios (Ф) in the range of was about 3-4 mj. Beduneau et al. [25] (C) 2014 OSA February 2014 Vol. 22, No. 3 DOI:.1364/OE OPTICS EXPRESS 344

3 investigated the MIE of premixed methane/air mixtures and observed similar trends as those obtained by Phuoc et al. []. Kopecek et al. [26] performed laser ignition of methane/air mixtures in a combustion chamber at elevated initial pressures and reported that the minimum laser pulse energy was in the range of -15 mj and decreased with the increasing pressure and Ф. All the above measurements were performed on premixed fuel/oxidant mixtures, and few researchers have reported the MIE values for the LSI of diffusion flames except for Phuoc et al. [27]. Meanwhile, the ignition time is also an important parameter for the LSI systems, especially for the laser igniters of rocket engines which require rapid and punctual response and control. If the ignition time is too long, the propellants will accumulate in the chamber and may lead to hard starts and even catastrophic explosions. To our knowledge, only Phuoc et al. [27] reported the ignition time of LSI in a jet diffusion methane flame, and detailed results on the effects of the oxygen concentration, flow rate and ignition position on the ignition time have not been reported before. The LSI of coaxial gaseous hydrogen/liquid oxygen spray and gaseous methane/oxygen has been reported by Oschwald and his collaborators [22 24]. However, they mainly concentrated on the investigations of the transient propellant spray and flame behaviors using a high-speed imaging technique, the MIE and ignition time were not reported. In this paper, we performed the laser induced spark ignition of coaxial methane/oxygen/nitrogen diffusion flames to simulate the LSI behaviors on a model rocket engine with coaxial injectors. The MIE and ignition time of the LSI were determined, and the effects of gas mixture properties, including the overall Ф, oxygen concentration and flow rate, and ignition positions on the two parameters were investigated and discussed systematically. 2. Experimental apparatus The experimental apparatus for the LSI of the coaxial methane/oxygen/nitrogen diffusion flames is shown in Fig. 1. It consists of three subsystems: the gas mixing and burner system, laser ignition and diagnostic system, and synchronization system. Fig. 1. Experimental apparatus for the LSI of the methane/oxygen/nitrogen diffusion flames. The burner consists of two coaxial quartz tubes. The inner tube has an inner and outer diameter of 7.2 mm and mm, respectively, while the annulus tube has an outer diameter of 12 mm. The detailed schematic of the coaxial burner is shown in Fig. 2. The oxygen and nitrogen is premixed to a specific oxygen concentration (χ O2 ) which is defined as the mole concentration of the oxygen in the mixture of oxygen and nitrogen, and then flows into the inner tube. Pure methane (>99.5%) is introduced through the outer annulus tube. The flow rate and overall Ф of the gas mixture are controlled by three calibrated mass flow meters (D07-19B, Sevenstar electronics). To enable spatially resolved measurements, the burner is (C) 2014 OSA February 2014 Vol. 22, No. 3 DOI:.1364/OE OPTICS EXPRESS 3449

4 mounted on a three-dimensional translation stage that can be adjusted relative to the laser generated sparks. For the diffusion flame, the ignition position is an important factor for a successful ignition, thus a coordinate system is introduced to address the ignition position. As shown in Fig. 2, the origin of the coordinates is fixed to the axis of the burner on the plane of the burner tip. The positive direction of the X axis is the same with the laser incidence direction. The positive direction of the Y axis is perpendicular to the X axis and to the right when facing the laser incidence direction. The Z axis is then determined using the right-handed coordinate system. The ignition position in the flow field is then given with the coordinates in the form of (X, Y, Z) throughout the whole paper. Fig. 2. Schematic of the coaxial burner and definition of the coordinate system. A laser diode pumped Q-switched Nd:YAG laser (SpitLight DPSSL-250, Innolas) is used for the laser ignition measurements. The 64 nm, ~ nanosecond output of the laser is focused into the methane/oxygen/nitrogen gas flow using a 25mm diameter BK7 lens with a focal length of 150 mm to generate the laser sparks. Assuming that the focused laser beam around the focal spot is cylindrical with a Gaussian beam profile, the diameter and length of the focal region are estimated as.47 μm and μm, respectively, using the formulas given by Beduneau et al. [25]. The laser is operated at a pulse rate of 2 Hz. The input laser pulse energy is measured with energy detector 1 (J-MB-HE, Coherent) through a beam splitter with a 5% energy extraction. The residual energy after the generation of the laser sparks is measured with energy detector 2 (J-50MB-HE, Coherent). The output of the two energy meters is collected using a two-channel energy meter (EPM2000, Molectron) and read out through a RS232-USB interface. The spark energy is then calculated as the difference of the input energy and the residual energy. The success or failure of the laser ignition events is determined by measuring the emission due to the ignited flames during the ignition processes. Typically, in the case of a successful ignition, a stable flame will be generated above the burner after firing of the laser spark, and strong flame emission can be observed. The flame emission is imaged onto the window of a fast photomultiplier tube (PMT, rise time < 2 ns) using a 25 mm diameter, 75 mm focal length BK7 lens. The PMT signal is amplified 25 times using a low noise four-channel preamplifier (SR445A, Stanford Research Systems) and then monitored and collected using a 1 GHz sampling rate digital storage oscilloscope (DPO7014, Tektronix). The PMT signal profiles are then read out and saved through a GPIB-USB port. The synchronization of the whole measurement system is realized using a digital delay generator (DG645, Stanford Research Systems). The external triggers for the flashlamp and Q-Switch of the Nd:YAG laser, the energy meter, and the oscilloscope are all provided by the output of the delay generator. With the synchronization system, the spark energy and the (C) 2014 OSA February 2014 Vol. 22, No. 3 DOI:.1364/OE OPTICS EXPRESS 3450

5 PMT profile of each ignition event can be simultaneously obtained, enabling the correlation of the parameters with each other. 3. Results and discussions 3.1 Definitions and methods The MIE is defined as the smallest spark energy applied for a successful ignition event [2]. For the measurements presented here, since all the flames can burn stably above the burner after been successfully ignited, the MIE measurement is performed in a single-pulse mode [27]. The stable flame is extinguished by switching off the methane flow, and after that the methane flow is resumed and stabilized for 30 seconds for the next ignition test. Considering the stochastic behaviors of the laser induced breakdown processes, the single-pulse measurements are repeated 7 times and the MIE is determined as the average value of the spark energies of all the successful ignition events [25, 29]. The typical flame emission signal profile observed for a successful ignition event is shown in Fig. 3. The strong peak signal observed at the early time is due to the laser plasma emission which usually lasts for about 1-2 μs. It is shown that when the laser spark is generated in the gas mixture, there is a time gap before the flame emission signal emerges. The ignition time is then defined as the time gap between the onset of the laser spark and the time when the flame emission signal emerges. PMT signal (a.u.) Flame emission Ignition time Laser plasma emission Time (μs) Fig. 3. Determination of the success of the LSI and definition of the ignition time. 3.2 MIE and ignition time of different gas mixture properties Overall equivalence ratio effect We measured the MIE and ignition time of the LSI of the methane/oxygen/nitrogen diffusion flames with different overall Ф in the range of The ignition position is fixed to (3, 0, 15), the χ O2 is fixed to 50%, and the total flow rate is fixed to 2.2 liter per minute (LPM). Shown in Fig. 4(a) is the MIE of different overall Ф. The error bars shown in the figure are the standard deviation of the measured spark energies. It is shown that the MIE is around 4 mj and varies little with the overall Ф of and increases to ~7.5 mj with the overall Ф of 0.3. The MIE values obtained here are similar to the results obtained by Phuoc et al. [27] who reported a spark energy of ~4 mj to successfully ignite a methane jet diffusion flame. However, the trend of the variation of the MIE with the overall Ф is different from the results obtained in the LSI of the premixed methane/air mixtures [, 25, 30]. For the premixed methane/air mixtures, the variation of the MIE with the overall Ф is typically of U-shape, i.e., the MIE approaches its minimum value around the stoichiometry, but increases rapidly towards the fuel lean and rich ends [, 25, 30]. The differences between the MIE values of the (C) 2014 OSA February 2014 Vol. 22, No. 3 DOI:.1364/OE OPTICS EXPRESS 3451

6 diffusion flame and the premixed flame is probably due to the deviation of the local Ф at the ignition spot from the nominal overall Ф set by the mass flow meters for the diffusion flame. In the reported relationship of MIE vs. Ф for the premixed methane/air mixtures, the MIE varies little with the Ф in the range of [] or [30], depending on the burner systems applied. Meanwhile, the local Ф region for the invariant MIE might be expanded, since wider flammability limits have been reported in electric spark initiated ignition of the oxygen-enriched gas mixtures [31]. For the overall Ф in the range of , the local Ф may well within the invariant region, thus leading to an almost constant MIE value. When the overall Ф is too lean, the corresponding local Ф may approach the fuel lean end, and as reported in the LSI of premixed methane/air mixtures [, 25, 30], the MIE will then increase accordingly. Detailed distributions of the local Ф in the diffusion flow field will be presented in section 3.3. MIE (mj) Equivalence ratio Ignition time (μs) Equivalence ratio (a) (b) Fig. 4. MIE (a) and ignition time (b) of different overall equivalence ratios. The ignition time of different overall Ф is shown in Fig. 4(b). It is shown that the minimum ignition time is obtained as ~200 μs with overall Ф of When the overall Ф varies towards the fuel lean end, the ignition time firstly increases gradually to ~400 μs with overall Ф of and then increases quickly to ~100 μs with overall Ф of 0.2. The variation of the ignition time with the overall Ф to the fuel rich end has a different trend. The ignition time varies little with the Ф in the range , and keeps at around 400 μs. The ignition time reported here is about one order of magnitude lower than that obtained by Phuoc et al. [27] and Li et al. [30]. This is probably due to the relatively higher oxygen concentration in the gas mixture: the χ O2 is 50% in the measurements presented in this work, while both Phuoc et al. [27] and Li et al. [30] used a χ O2 of 21%. Actually, the ignition time of methane/oxygen mixtures has been investigated in a shock-tube facility [32]. Ignition time of -600 μs was obtained and the ignition time (τ) was reported to be approximately inversely proportional to the oxygen mole concentration in the gas mixture ([O 2 ]) with a relationship of τ [O 2 ] -1.03, which again indicates that higher χ O2 might lead to a shorter ignition time Oxygen concentration effect The MIE and ignition time of the LSI of the methane/oxygen/nitrogen diffusion flames with different oxygen concentrations of the oxygen/nitrogen mixture are shown in Fig. 5. The ignition position is fixed to (3, 0, ). The flow rate is fixed to 2.2 LPM. The overall Ф are 0.7, 1.0 and 1.3, respectively, and the χ O2 is 20%, 40% and 60%, respectively. It is shown that for all the three overall Ф, the MIE does not have a specific trend with the increasing oxygen concentration. The MIE values of different oxygen concentrations are comparable with each other. However, the ignition time has a clear trend with the increasing oxygen concentration. It decreases gradually with the increase of the oxygen concentration. The decrease of the (C) 2014 OSA February 2014 Vol. 22, No. 3 DOI:.1364/OE OPTICS EXPRESS 3452

7 ignition time may be due to the reduced dilution effect of the nitrogen when the oxygen concentration is high. MIE (mj) Oxygen concentration (%) (a) Overall Φ=0.7 Overall Φ=1.0 Overall Φ=1.3 Ignition time (μs) 900 Overall Φ=0.7 Overall Φ= Overall Φ= Oxygen concentration (%) Fig. 5. MIE (a) and ignition time (b) of different oxygen concentrations Flow rate effect The MIE and ignition time of the LSI of the methane/oxygen/nitrogen diffusion flames with different flow rates are shown in Fig. 6. The ignition position is set to (3, 0, ). The χ O2 is fixed to 50%. The overall Ф are 0.7, 1.0, and 1.3, respectively, and the total flow rates are varied from 2.2 LPM to 4.4 LPM. The flow velocities at the ignition position are estimated as cm/s from the gas flow velocities in the inner tube. According to Spiglanin et al. [2] and Beduneau et al. [25], the gas flow can be considered as stagnant during the laser spark generation and flame kernel formation processes, which have a typical expansion speed of the order of 5-6 cm/s [3] and 4 cm/s [25], respectively. Thus the gas flow should have little effect on the laser spark formation. It is shown that the MIE values are comparable with each other for all the flow rates investigated. The MIE values of different overall Ф are also comparable with each other, which is consistent with the results shown in Fig. 4(a). However, the ignition time is sensitive to the variation of the flow rates. The ignition time increases gradually with the increasing flow rates. The higher flow rates may cause more convection losses in the flame kernel [25], and lead to a lower initial flame kernel temperature. The lower initial temperature then lengthens the time needed to reach the critical ignition temperature, thus resulting a longer ignition time. (b) MIE (mj) 1 Overall Φ=0.7 Overall Φ= Overall Φ= Ignition time (μs) Overall Φ=0.7 Overall Φ=1.0 Overall Φ= Flow rate (LPM) Flow rate (LPM) (a) Fig. 6. MIE (a) and ignition time (b) of different flow rates. (b) (C) 2014 OSA February 2014 Vol. 22, No. 3 DOI:.1364/OE OPTICS EXPRESS 3453

8 3.3 MIE and ignition time of different ignition positions For diffusion flames, since the fuel and oxidant diffuses and mixes with each other, the local mixing condition at different spatial positions relative to the burner assembly will vary greatly. Therefore, the ignition position can affect the MIE and ignition time greatly. The MIE and the ignition time of different ignition positions are measured by adjusting the burner assembly relative to the formed laser sparks. The total flow rate is fixed to 2.2 LPM. The overall Ф is set to 1.0, and the χ O2 is set to 50%. During the measurements, the Y coordinate is set to 0, i.e., the laser passes through the center of the inner tube (refer to the illustration of the coordinates system shown in Fig. 2). The horizontal position (X coordinate) is varied in the range of 0-5 mm, i.e., from the burner axis to the outer diameter of the inner tube. The vertical position (Z coordinate) is varied in the range of 1-11 mm. The measurements are performed in the XOZ plane with a spatial grid size of 1 mm 1 mm, and the contour distributions of the MIE and ignition time are obtained. Shown in Fig. 7 is the contour plot of the MIE values at different ignition positions. The white blank areas in the left, lower left and lower right parts of the plot indicate the spatial regions in which the gas mixture is unable to be ignited. It is shown that in the horizontal dimension, the MIE values of the ignition positions near the burner axis (with smaller X values) are smaller than those of the positions far from the burner axis (with larger X values). In the vertical dimension, the MIE varies with the vertical position with different trends in different horizontal regions. In the region with X coordinates of mm, the MIE is generally almost the same throughout the vertical regions investigated. In the region with X coordinates of mm, the MIE firstly increases gradually with the vertical position until Z coordinates of 6- mm, then decreases gradually with the vertical position. In the region with X coordinates of mm, the MIE decreases gradually with the vertical position. The largest MIE value is obtained as 17-1 mj in the region with X coordinates of mm and Z coordinates of mm. 11 Oxygen/Nitrogen Methane Vertical position Z (mm) Horizontal position X (mm) Fig. 7. Contour plot of the MIE values at different ignition positions. The contour plot of the ignition time at different ignition positions is shown in Fig.. It is shown that the shortest ignition time is obtained as less than 140 μs in the region with X coordinates of mm and Z coordinates of mm. Then the ignition time increases gradually towards the smaller and larger X coordinates, especially for the latter case. The longest ignition time is obtained as ~200 μs in the region with X coordinates of mm and Z coordinates of mm. (C) 2014 OSA February 2014 Vol. 22, No. 3 DOI:.1364/OE OPTICS EXPRESS 3454

9 The MIE and ignition time of different ignition positions may be closely related to the local Ф at the corresponding ignition spots. The inner and outer diameters of the inner tube are shown in the contour plots with two dashed lines. Since the oxygen/nitrogen mixture and the pure methane begin to mix after they flow out the separated inner and outer tubes, it is conceivable that the local Ф will vary with the spatial positions. In the inner tube region (X<3.6 mm), since the oxidant is dominant, the gas mixture tends to be fuel lean. Actually, on the axis of the inner tube (X = 0) and in the lower left part of the contour plots, the mixture is too lean to be ignited. Since the pure methane flows through the horizontal region beyond the outer diameter of the inner tube (X>5 mm), the local Ф near the horizontal region with X = 5.0 mm tend to be fuel rich. Actually, in the lower right part of the contour plots, the gas mixture is too rich to be ignited. On the boundary of the oxidant and methane, due to the diffusion of the oxidant and fuel into each other, the local Ф will vary from fuel lean to stoichiometry and then to fuel rich with the increase of the X coordinates. Meanwhile, since the gas mixture will also mix with the ambient air above the burner, the local Ф will decrease with the increasing vertical positions. 11 Oxygen/Nitrogen Methane Vertical position Z (mm) Horizontal position X (mm) Fig.. Contour plot of the ignition time values at different ignition positions. The detailed distributions of the mole fraction of oxygen, mole fraction of methane and local Ф in the diffusion flow field are simulated based on computational fluid dynamics (CFD) calculations. As shown in Fig. 9, the simulation is performed in the spatial regions with X coordinates of 0-20 mm and Z coordinates of 0-96 mm. The flow conditions are the same with that of Fig. 7 and Fig., i.e., the flow rate is 2.2 LPM, the overall Ф is 1.0, and the χ O2 is 50%. Since the burner is axisymmetric, only the distribution on half of the XOZ plane is given. It is shown in Fig. 9(a) that the oxygen mole fraction approaches its highest value in the region near the burner axis and then decreases gradually when the ignition position moves vertically up along the burner axis or away from the burner axis. The lowest oxygen mole fraction is located in the region near the methane outlet. While for the methane flow (see Fig. 9(b)), the highest mole fraction values are obtained near the methane outlet and then it decreases gradually with the diffusion of the methane. The local Ф are calculated using the simulated mole fraction distributions of the oxygen and methane. As shown in Fig. 9(c), there are great variations in the local Ф of the diffusion flow field. The highest local Ф (>) is obtained near the methane outlet with X coordinates of 5-6 mm. Then the local Ф decreases gradually when the ignition position moves outwards from the methane outlet, due to the diffusion and mixing of the methane flow with both the oxidant flow and the ambient air. The lowest local Ф (~zero) is obtained in the region near the burner axis where the oxidant flow outlet locates and in the regions far beyond the methane outlet. It can be seen that the simulated distribution of the local Ф in the flow field are generally consistent with our above qualitative estimations. (C) 2014 OSA February 2014 Vol. 22, No. 3 DOI:.1364/OE OPTICS EXPRESS 3455

10 Vertical position Z (mm) Vertical position Z (mm) Horizontal position X (mm) (a) Horizontal position X (mm) (b) Vertical position Z (mm) Horizontal position X (mm) 0 (c) Fig. 9. Simulated contour distributions of the mole fraction of oxygen (a), mole fraction of methane (b) and local equivalence ratio in the diffusion flow field. The flow rate is 2.2 LPM, the overall Ф is 1.0, and the χ O2 is 50%. The variation of the MIE and ignition time with the Ф in the LSI of the premixed methane/air mixtures has been investigated before [,25,30]. The MIE usually approaches its minimum value near the stoichiometry and increases gradually towards the fuel lean and rich ends. The MIE values of the fuel lean and rich ends are usually one order of magnitude higher than that of the stoichiometry, and the MIE values of the rich end is usually 2-5 times of that of the lean end [30]. The ignition time usually approaches its minimum value at the stoichiometric condition, and then increases towards the fuel rich and lean ends, and the ignition time of the rich end is slightly longer than that of the lean end [30]. It is shown that the measured spatial distributions of the MIE and ignition are generally consistent with the estimations using the relationships of the MIE and ignition time with the Ф obtained for the premixed methane/air mixtures. This again proves that the variation of the MIE and ignition time should be closely related to the local Ф in the diffusion flow field. 3.4 Discussions The local mixing condition within the flow field of the coaxial methane/oxygen/nitrogen diffusion flames can affect the ignition properties to a large extent. The simulation based on the CFD calculations can offer a general estimation on the local mixing conditions in the flow field and explain most of the experimental results. However, it is fair to point out that the CFD calculation presented here has its limitations. It seems that the local Ф are underestimated by the CFD calculations, especially in the regions near the burner axis. In the region with X coordinates of 1-2 mm and Z coordinates of 1- mm, the gas mixture would be unable to be ignited with the very low calculated local Ф values, which contradicts the measured results. Accurate measurement of the local Ф can offset the limitations of the CFD model. Laser induced breakdown spectroscopy (LIBS) technique has been applied for the local Ф measurements [27,30,33,34]. By correlating the local Ф with the intensity ratio of two atomic (C) 2014 OSA February 2014 Vol. 22, No. 3 DOI:.1364/OE OPTICS EXPRESS 3456

11 lines originating from the elements of the fuel and oxidant, respectively, the local Ф can be obtained. We have tried to measure the local Ф by using the line intensity ratio of H α line to the nitrogen atomic triplet lines around 742 nm or to the oxygen atomic triplet lines around 777 nm. However, due to the limitations of our spectrograph system, the above atomic emissions cannot be collected in the same detection window, and thus cannot be measured simultaneously. Therefore, the local Ф was not measured in the present paper. Work to reconstruct another spectrograph system with wider detection window is now in progress, and the local Ф can then be obtained. It should also be noted that the MIE values shown in Fig. 4(a) is lower than those shown in other figures. We think this is mainly due to its relatively higher ignition position. The ignition position in Fig. 4(a) is at (3, 0, 15). At the higher ignition position, the local Ф may be more close to the stoichiometry and thus lower MIE values were obtained. 4. Conclusions Laser spark ignition (LSI) of coaxial methane/oxygen/nitrogen diffusion flames has been achieved using a 64 nm Q-switched Nd:YAG laser. The minimum ignition energy (MIE) and ignition time of the LSI have been obtained by measuring the emission signals due to the ignited flames. The effects of the gas mixture properties, including the overall equivalence ratio (Ф), oxygen concentration, and flow rate, and the ignition positions on the MIE and ignition time have been investigated systematically. Several conclusions draw from our investigations can be summarized as follows: (i) Computational fluid dynamics (CFD) simulations indicate that the local Ф varies greatly within the diffusion flow field. The local Ф approaches its maximum value near the methane outlet, and then decreases gradually when the ignition position moves outwards from the methane outlet, due to the diffusion and mixing of the methane flow with both the oxidant flow and the ambient air. The lowest local Ф is obtained in the region near the burner axis and far beyond the methane outlet. (ii) The spatial contour distributions of the MIE and ignition time have been measured, and they agree generally with the estimations based on the CFD simulations and the existing relationships of the MIE and ignition time with the Ф for the LSI of premixed methane/air mixtures. The MIE values of the ignition positions near the burner axis are smaller than those of the positions far from the burner axis. The shortest ignition time is obtained as less than 140 μs in the horizontal region mm away from the burner axis and vertical region mm above the burner tip. Then the ignition time increases gradually towards the smaller and larger distances away from the burner axis. The variation of the MIE and ignition time is believed to be closely related to the local Ф at the ignition positions. (iii) For the ignitions at (3, 0, 15), the MIE is around 4 mj and varies little with the overall Ф of and increases to ~7.5 mj with overall Ф of 0.3. The ignition time approaches it minimum value ~200 μs with overall Ф of , and increases to ~400 μs with Ф of and , and further to ~100 μs towards the fuel lean end. (iv) For the ignitions at (3, 0, ), the oxygen concentration and flow rate has little effect on the MIE. The ignition time decreases gradually with the increasing oxygen concentration, while increases gradually with the increasing flow rate. The MIE and ignition time values obtained for the LSI of the coaxial diffusion flames can serve as references for design and evaluation of the LSI system for the model rocket engines using coaxial injectors. Acknowledgments The authors thank the financial support from the National Natural Science Foundation of China (Grant No ) and Special Grants for National Key Scientific Instrument and Equipment Development (Project No.2012YQ040164). (C) 2014 OSA February 2014 Vol. 22, No. 3 DOI:.1364/OE OPTICS EXPRESS 3457