Combustion Oscillation Characteristics and Flame Structures in a

Subscriber access provided by BOSTON UNIV

Article

Combustion Oscillation Characteristics and Flame Structures in a Lean Premixed Prevaporized Combustor Peifeng Sun, Yiren Yuan, Bing Ge, Yinshen Tian, Zilai Zhang, and Shusheng Zang Energy Fuels, Just Accepted Manuscript • DOI: 10.1021/acs.energyfuels.7b01302 • Publication Date (Web): 02 Aug 2017 Downloaded from http://pubs.acs.org on August 5, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Energy & Fuels is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

Combustion Oscillation Characteristics and Flame Structures in a Lean Premixed Prevaporized Combustor

Peifeng Sun, Yiren Yuan, Bing Ge, Yinshen Tian, Zilai Zhang, and Shusheng Zang* Institute of Turbomachinery, School of Mechanical Engineering, Shanghai Jiao Tong University, Shanghai 200240, People’s Republic of China

ABSTRACT: The combustion oscillation characteristics in the LPP combustor at different equivalence ratios and inlet velocities of the combustor were experimentally investigated. The acoustic modal analysis was carried out to identify the acoustic eigenmodes in the LPP combustor. The flame structures under different operating conditions were investigated. The evolution laws of flame structures, coordinates and velocities of the flame centers, and heat release rates throughout a combustion oscillation cycle were investigated by phase-locked OH-PLIF. The result shows that there is a periodical process of flame roll-up for the flame close to the LPP combustor axis and a periodical process of separation and consolidation for the flame far from the LPP combustor axis. The spatial distribution of the normalized Rayleigh index in the LPP combustor was calculated to determine the driving zones of the combustion oscillation. The mechanism of combustion oscillation in this LPP combustor was identified by the analysis of the flame structures, heat release rates throughout a combustion oscillation cycle, and the spatial distribution of the normalized Rayleigh index.

1 ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 23

1. INTRODUCTION With environmental regulations becoming more stringent in recent years, the importance of low NOx combustion has increased. The LPP combustor has the great advantage of low NOx emission. However, the LPP combustor suffers from detrimental combustion oscillation,1,

2

which limits the

applicability of this technology. Some researches have recently been conducted on combustion oscillation in the LPP combustor. Driscoll et al.3, 4 studied combustion oscillation of the TAPS injector and concluded that the equivalence ratio fluctuation, the shear layer vortex shedding, and the flame-flame interaction may be the main causes of combustion oscillation in the LPP combustor. Nguyen5 measured the correlation between the equivalence ratio fluctuation and combustion oscillation in a jet-A fueled LPP combustor. He founded that there were two mild trends between the equivalence ratio fluctuation and combustion oscillation. When the equivalence ratio fluctuation became bigger, the combustion oscillation intensity became larger or sometimes smaller. Allouis et al.6 investigated the relationship between the dominant frequency of combustion oscillation and the equivalence ratio. They found that the dominant frequencies would shift upward with a decreasing equivalence ratio. Armitage et al.1 identified the frequencies of the combustion oscillation in the LPP combustor using the turbulent combustion code TARTAN. Tang et al.7 reported the application of the large eddy simulation methodology in the study of combustion oscillation in the LPP combustor. The LES predictions could provide a wealth of information on the unsteady turbulent nature of the flow. Tran et al.8 and Dowling et al.9 analyzed the influence of the acoustic boundary conditions on combustion oscillation. The influences of the flowfield4, the coherent vortex structures10 and the nozzle configuration11 on combustion oscillation in the LPP combustor were also 2 ACS Paragon Plus Environment

Page 3 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

studied. Although there are some researches on combustion oscillation in an LPP combustor, studies on combustion oscillation characteristics and flame structures throughout a combustion oscillation cycle in an LPP combustor have been relatively few. In particular, the flame structures throughout a combustion oscillation cycle in the LPP combustor have not been well studied. It is of great importance to study the flame structures throughout a combustion oscillation cycle because the variation of flame structures will alter the heat release rate distribution in the combustor and influence the coupling between the heat release rate and pressure. This study mainly focuses on combustion oscillation characteristics at different equivalence ratios and inlet velocities, together with the evolution law of flame structures throughout a combustion oscillation cycle in order to better understand the mechanism of combustion oscillation in the LPP combustor.

2. EXPERIMENTAL ARRANGEMENT AND MEASUREMENTS 2.1. Injector and Combustor. The LPP combustion oscillation experimental setup is composed of the air heating and supply system, fuel feeding system, LPP combustion test rig shown in Figure 1, and the phase-locked OH-PLIF diagnostic system shown in Figure 2. The vane angles of the mixer swirler and the combustor swirler are 55° and 35°, respectively. A quartz combustor was used to visualize the flame structure by OH-PLIF. A choked orifice was located upstream of the preheated air chamber to create a well-defined acoustic boundary condition. An oil pump was used to feed the kerosene. The spray cone angle of the main nozzle is 60°. The maximum flow rate of the kerosene in this experiment was 2.90 g/s, corresponding to a thermal power of 120 kW. The combustor was operated at 3 ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 23

atmospheric pressure. The working process of the LPP combustor can be described as follow: the air was heated to 500 K by a 36 kW electric heater and then supplied to the premixer, in which the kerosene droplets injected from the main nozzle were vaporized by the heated air. At the same time as the evaporation of kerosene droplets, mixing of fuel and air was completed. Finally, the premixed and prevaporized combustible mixture of fuel and air will enter the LPP combustor and start to burn.

Figure 1. Experimental setup of LPP combustor.

2.2. Evaporation Time and Residence Time of Kerosene Droplets in the Premixer. Kerosene (C10H20) was used as the fuel in the experiment and its distillation temperature is from 458 K to 488 K. The evaporation time of the kerosene droplets in the premixer can be computed by the D2-law, which can be expressed as follow:12

D 2 = D02 − Kt

(1)

where D0 is the initial diameter of droplets (mm) and K is the evaporation rate constant (m2/s) and can be 4 ACS Paragon Plus Environment

Page 5 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

expressed as:

K=

8k g

ρl c pg

ln( Bq + 1)

(2)

where kg is thermal conductivity (W/(m.K)), Bq is the Spalding transfer number, ρl and c pg are the density of liquid fuel (kg/m3) and the heat capacity (kJ/(kmol.K)), respectively. The subscripts l and g refer to the liquid phase and the gaseous phase. The mean diameter of kerosene droplets injected from the main nozzle is about 30µm, the thermal conductivity is 0.0438 W/(m.K), the Spalding transfer number is 1.4532, the density of the liquid kerosene and the heat capacity are 800 kg/m3 and 1017.23 kJ/(kmol.K), respectively. The calculated evaporation time of kerosene droplets is about 5.4 ms according to the formula and factors listed above. The range of axial velocity of the preheated air in the premixer during the experiment was between 12.5 m/s and 25.6 m/s. The distance between the main injector and the outlet of the premixer is about 225 mm. So the residence time of kerosene in the premixer was approximately between 9 ms and 18 ms. Considering the evaporation time and residence time of the kerosene droplets in the premixer, it can be concluded that the kerosene droplets were completely vaporized before entering the combustor.

2.3. Phase-locked OH-PLIF Diagnostics. The LPP flame was imaged by OH-PLIF in this experiment, as OH-PLIF has been successfully used for measurements of the heat release rate in turbulent premixed flame.13 An Andor iStar 334 ICCD camera was used and the resolution of the chip was 1024 × 1024 pixels. The field of view was a 100 mm × 100 mm rectangular test section. The excitation laser was derived from a pulsed Nd: YAG laser pumping a tunable dye laser with Rhodamine 6G as the dye solution before going through a frequency double crystal. The output ultraviolet laser beam


Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

had a wavelength of 281.46 nm with a pulse duration of 20 ns, and was used to excite OH radicals. The laser beam was expanded by a set of spherical and cylindrical lenses, and formed a laser sheet with a thickness of less than 0.5 mm. The laser sheet was vertically guided through the center of the test section in the combustor. The fluorescence was then collected around a wavelength of 310 nm by the ICCD camera. Three hundred single-shot OH-PLIF raw images were recorded for the post-processing of each flame. The phase-locked OH-PLIF was adopted to characterize the variations of the flame structures throughout an oscillation cycle. The dynamic pressure of the combustor, which was obtained by a Kulite pressure transducer, was used as the reference signal to trigger the phase-locked measurements. The reference pressure and the corresponding phase-locked triggering signal sequences are shown in Figure 3. p0 is set as the critical value for triggering. When the pressure reaches p0, a positive signal is sent to the delay generator, which generates a triggering signal for the laser and the ICCD camera. The phase corresponding to p0 is defined as phase 0. The system continuously triggers the delay generator with a delay time every 80 pressure oscillation cycles. For phase 1, the time delay, t, is added to the system when p’ reaches p0, where t = 1 ( f ⋅ N ) and N is the number of phase angles specified. For phase 2, the delay time is set as 2t and in the same manner for the following phases. So the evolution of phase-averaged flame structures at different phases could be obtained. Quantities of interest such as the heat release rate and Rayleigh index can then be calculated.


Page 6 of 23

Page 7 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

Figure 2. Schematic of phase-locked OH-PLIF.

Figure 3. Pressure oscillations and phase-locked triggering signal sequecnes.

3. RESULTS AND DISCUSSION 3.1. Combustion Oscillation Characteristics of the LPP Combustor. A wide range of stable and unstable regions of the LPP combustor are systematically mapped as shown in Figure 4 by varying the equivalence ratio (0.42 - 0.98) and combustible mixture inlet velocities of the combustor (25 m/s - 49 m/s). The size of each point is representative of the RMS pressure fluctuation amplitude. Flame stabilization is achieved down to an equivalence ratio of 0.42 and for the combustible mixture inlet velocity of the combustor from 31 m/s to 49 m/s. There is no flashback when the inlet velocity is higher than 19 m/s.


Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

It can be seen in Figure 4 that at a constant inlet velocity there is a range of equivalence ratios in which the combustion oscillation can be driven. The equivalence ratio region with combustion oscillation decreases with an increase of inlet velocities. The combustion oscillation disappears when the inlet velocity is higher than 37 m/s. Under a certain equivalence ratio, a critical inlet velocity exists, below which the combustion oscillation can be driven.

Figure 4. Operation diagram of the LPP combustor. Figure 5 shows the combustion oscillation intensity represented by the RMS pressure fluctuation at a constant equivalence ratio of 0.66 and different inlet velocities. The amplitude of the RMS pressure fluctuation decreases from 1698 Pa to 1493 Pa when the inlet velocity increases from 25 m/s to 31 m/s. This means that with an increase of the inlet velocity, the combustion oscillation intensity will decrease while the dominant frequency of the combustion oscillations will increase from 134 Hz to 136 Hz. When the inlet velocity increases to 37 m/s, the combustion becomes stable with an RMS pressure fluctuation of 326 Pa. When the inlet velocity further increases to 43 m/s and 49 m/s, the combustion remain stable and the RMS pressure fluctuations are 336 Pa and 111 Pa respectively. The result shows that whether the combustion oscillation will be excited or not and the combustion oscillation intensity in the LPP 8 ACS Paragon Plus Environment

Page 8 of 23

Page 9 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

combustor is dependent on the inlet velocity of the combustor when the equivalence ratio is 0.66.

Figure 5. Effect of inlet velocity on the RMS pressure amplitude.

Figure 6. Effect of equivelence ratio on the RMS pressure amplitude.

Figure 6 shows the combustion oscillation intensity represented by the RMS pressure fluctuation at a constant inlet velocity of 37 m/s and different equivalence ratios. The combustion oscillations are driven when the equivalence ratios are 0.54 and 0.58 as shown in Figure 6. The intensities and the dominant frequencies of the two oscillation cases are 1317 Pa, 626 Pa and 136 Hz, 138 Hz respectively. It is clear that with an increase of the equivalence ratio, the combustion oscillation intensity will decrease while the dominant frequency of the combustion oscillations will increase when the combustion oscillation is excited in the LPP combustor. The combustion remain stable under other equivalent ratios, and the amplitude of the RMS pressure fluctuation will be less than 200 Pa. This means that there is a range of equivalence ratios, in which the combustion oscillation can be driven at a constant inlet velocity.

3.2. Acoustic Modal Analysis. The identification of the acoustic eigenmodes of the combustion test rig is necessary for combustion oscillation study because the most problematic type of oscillation in a gas turbine combustor involves the coupling between acoustic motion and flame dynamics. So the natural acoustic eigenmodes of the combustion test rig were analyzed by ANSYS


Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

APDL. The computational domain of the combustion test rig for acoustic modal analysis is divided into 3 zones, which are the preheated air chamber, the combustor, and the exhaust pipe. The dimensions and flow properties of the 3 zones for acoustic modal analysis are listed in Table 1. The inlet of the preheated air chamber was treated as acoustically closed due to the choked inlet isolating the acoustics from the upstream flow. The outlet of the system is treated as an open boundary. The first 4 longitudinal acoustic frequencies of the combustion test rig are listed in Table 2, because the oscillations driven in the LPP combustor are low-frequency combustion oscillations. It can be seen that the predicted frequency of the second longitudinal mode well agrees with the ones measured in the experiment. So the driven mode in the LPP combustor corresponds to the second longitudinal acoustic mode of the choked-open combustion system.

3.3 Flame Structures in the LPP Combustor. The flame structures under different operating conditions were investigated. It was found in the experiment that the LPP flame structures take a V-shape or an M-shape (flat) as shown in Figure 7 and exhibit different oscillation characteristics under different conditions. The operating conditions in the experiment are listed in Table 3.

Figure 7. V-shaped (a) and flat (b) LPP flame. Figure 8 shows the time-averaged flame structures under different inlet velocities at constant


Page 10 of 23

Page 11 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

equivalence ratios of 0.54 and 0.66. At a constant equivalence ratio of 0.54, the flame structures are flat when the inlet velocities are lower than 43 m/s, while the flame structure changes to be a V-shape when the inlet velocity is 43 m/s. At a constant equivalence ratio of 0.66, the flame structures are flat when the inlet velocities are 25 m/s and 31 m/s, while the flame structures change to be a V-shape when the inlet velocities are 37 m/s and 43 m/s. This is because the velocity gradient in the mixing layer of the combustor will decrease at lower inlet velocities, making the flame more susceptible to perturbations and becoming unstable. It can be seen from Figure 4 that the combustion oscillations are driven when the inlet velocities are lower than 43 m/s at a constant equivalence ratio of 0.54 and when the inlet velocities are 25 m/s and 31 m/s at a constant equivalence ratio of 0.66. This means that the combustion oscillation driven in the LPP combustor changes the flame structure from a V-shape to flat.

Figure 8. Time-averaged flame structures at different inlet velocities. The time-averaged flame structures under different equivalence ratios at constant inlet velocities of 25 m/s and 37 m/s are shown in Figure 9. At a constant inlet velocity of 25 m/s, the flame structures are flat when the equivalence ratios are higher than 0.50, while the flame structures assume a V-shape under


Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

the equivalence ratio of 0.50. At a constant inlet velocity of 37 m/s, the flame structures are flat under equivalence ratios of 0.54 and 0.58, while the flame structures assume a V-shape under other equivalence ratios. It can be seen from Figure 4 that there is a range of equivalence ratios (0.54 - 0.90) at the inlet velocity of 25 m/s and a range of equivalence ratios (0.54 - 0.58) at the inlet velocity of 37 m/s, in which the combustion oscillations are driven. This is because the energy from combustion will not be enough to excite the combustion oscillation if the equivalence ratio is below the low limit of the equivalence ratio range. It will move towards rich combustion if the equivalence ratio is higher than the upper level of the equivalence ratio range and the combustion will again be stable. This validates that the flame structure is closely correlated with whether the combustion oscillation is driven or not. The flame structure assumes a V-shape when the combustion is stable, whereas the flame structure is flat when the combustion is unstable.

Figure 9. Time-averaged flame structures at different equivalence ratios. According to the flame structures shown in Figures 8 and 9, it is concluded that the combustion oscillation driven in the LPP combustor will compact the flames to be a flat flame and make the flame


Page 12 of 23

Page 13 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

close to the injector exit which is very close to the acoustic anti-node located in the upstream end of the combustor. Thus the unsteadiness of the flat LPP flame may be mainly due to their proximity of the heat release to the acoustic anti-node just the same as the methane fueled lean premixed flame.14,15

3.4 Mechanism of Combustion Oscillation. 3.4.1 Evolution of Flame Structures throughout a Combustion Oscillation Cycle. From Figures 4 - 6, it is concluded that combustion oscillation intensity of case 5 is strongest. So the flame structures throughout a combustion oscillation cycle of case 5 were investigated by phase-locked OH-PLIF. The time resolved pressure fluctuation and dominant frequency of pressure fluctuation are shown in Figure 10. It can be seen that the pressure fluctuation presents a regular waveform and the amplitude of the time resolved pressure fluctuation is about 2.6 % (average to peak).

Figure 10. Time resolved and amplitude spectra of pressure fluctuations of case 5. 13 ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The phase-averaged flame structures throughout an oscillation cycle of case 5 are shown in Figure 11. The OH fluorescence intensity along the green line, which can best explain the variation of the flame structures, at phase angle 1350, 1800 and 2700 has been plotted in Figure 12. The flame in the LPP combustor can be approximately divided into two parts along the Y axis for the convenience of analysis. The first is that of the flame close to the combustor axis (at the bottom of the phase-averaged flame images), while the second is that of the flame far from the combustor axis (at the top of the phase-averaged flame images). For the flame close to the LPP combustor axis, there is a periodical process of convective roll-up of flame structures. The flame roll-up seems to occur at phase angle 1350 and is evident at phase angle 1800. The flame roll-up causes a very rapid reduction in flame area and decreases the heat release rate to the lowest value throughout the combustion oscillation cycle, as shown in Figure 15. The rolled-up flame moves radially outward from a phase angle 1800 to 3150. At the same time, the rolled-up flame gradually grows stronger and merges with the other flame. For the flame far from the LPP combustor axis, there is a periodical process of separation and consolidation. At phase angle 00, the flame in this region is an integral flame, while at phase angle 1350, the flame is almost completely separated in this region and there is a valley value of OH fluorescence intensity shown in Figure 12. The trend of the flame motions here begins to change from separation to consolidation from a phase angle 1350 to 2700. Finally, the flame again becomes an integral flame at phase angle 2700. As can be seen in Figure 12, there is no obvious valley value of OH fluorescence intensity in this area at phase angle 2700.


Page 14 of 23

Page 15 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

Figure 11. Phase-averaged flame structures throughout an oscillation cycle of case 5.

Figure 12. Profiles of the normalized OH fluorescence intensity along the defined lines. The variations of coordinates of the flame centers throughout an oscillation cycle are shown in Figure 13. The flame center is defined as the point with the highest value of OH fluorescence intensity. The Y coordinate of the flame center increases from 47.4 mm to 73.9 mm when the phase angle changes from 00 to 2250. This is because the flame moves towards the positive Y axis of the combustor. There is an obvious jump of the Y coordinate at phase angle 2700. The Y coordinate of the flame center dramatically changes from 72.9 mm to 39.5 mm. This is because the flame close to the LPP combustor axis grows stronger than the other flame. The Y coordinate of the flame center moves to 45.5 mm at 15 ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

phase angle 3150 which is a little lower than the value at phase angle 00. The variation of the X coordinate of the flame centers throughout the oscillation cycle is not as obvious as the Y coordinate, but the fluctuation trend is the same.

Figure 13. Variations of coordinates of the flame centers throughout the oscillation cycle of case 5. The velocities of the flame centers at different phase angles are shown in Figure 14. From a phase angle 450 to 2250, the Y component of the velocity of the flame centers fluctuates between 2 m/s and 10 m/s. But it dramatically decreases to -36 m/s at phase angle 2700 because of the transfer of the flame center from the flame far from the LPP combustor axis to the flame close to the LPP combustor axis which is shown in Figure 11 and Figure 13. At the next phase angle, the Y component of the velocity of the flame center rapidly increases to 6 m/s. The X component of the velocity of the flame centers fluctuates between -3 m/s and 5 m/s, and there is no sudden change of the X component of the velocity of the flame center. This means that the flame center moves back and forth along the combustor axis.


Page 16 of 23

Page 17 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

Figure 14. Variations of velocities of the flame centers throughout an oscillation cycle of case 5. The periodical process of roll-up of flame structures together with the periodical process of separation and consolidation of the flame in the LPP combustor changes the heat release rate distribution in the combustor, resulting in a periodical heat release rate fluctuation. The combustion oscillation will be driven when the periodical heat release rate fluctuation couples with the periodical pressure fluctuation. That all the flame structures throughout an oscillation cycle are flat validates the conclusion that the combustion oscillation will cause the flame structure to be flat.

3.4.2 Coupling of Heat Release Rate and Pressure Fluctuation. Figure 15 shows that the heat release rate and pressure of case 5 reach their minimum value at phase angle 1350 and 1800 respectively. That means the heat release rate lags the pressure by 450 and the Rayleigh criterion is satisfied. According to the Rayleigh theory16, the combustion oscillation can be driven when the phase lag between the heat release rate fluctuation and the pressure fluctuation is less than 900.


Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 23

Figure 15. Variations of heat release rate and pressure throughout an oscillation cycle of case 5. The Rayleigh index is a very useful parameter because it can characterize the mutual coupling between the pressure field and heat release rate distribution. The Rayleigh index can be defined as:16

R=

1 T p′q′dt T ∫0

(3)

where T , p ′ and q′ are the period of the combustion oscillation, pressure fluctuation and heat release rate fluctuation. The local Rayleigh index is calculated through multiplying the oscillating heat release rate by the oscillating pressure at each pixel. The combustion oscillation will be amplified if

R > 0 or damped if R < 0 .

Figure 16. Spatial distribution of the normalized Rayleigh index. The spatial distribution of the normalized local Rayleigh index of the LPP combustor is plotted in Figure 16. The areas with a large positive Rayleigh index are mainly the two zones centered at the coordinate point (45, 40) and (63, 90), respectively. That means there are two main drivers of the 18 ACS Paragon Plus Environment

Page 19 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

combustion oscillation in the LPP combustor. There is a zone with a large negative Rayleigh index in the middle of the two zones with a large positive Rayleigh index. To some extent, the spatial distribution of the normalized Rayleigh index in this LPP combustor is similar with the conclusion of Lee et al.17 They found the driving zones of combustion oscillation reside in both the corner and inner recirculation zones, whereas the damping zone remains along the shear layer. Figure 15 validates that the combustion oscillation in the LPP combustor is driven by the coupling of the heat release rate fluctuation and pressure fluctuation.

4. CONCLUSION The combustion oscillation characteristics under different equivalence ratios and combustible mixture inlet velocities of the combustor were investigated. The evolution of flame structures throughout a combustion oscillation cycle in the LPP combustor was investigated. The following conclusions can be extracted from this study. 1. There is a critical inlet velocity, below which the combustion oscillation can be driven at a constant equivalence ratio. There is a range of equivalence ratios, in which the combustion oscillation can be driven at a constant inlet velocity. In this experiment, the critical inlet velocity is 31 m/s when the equivalence ratio is 0.65, and the range of equivalence ratios is between 0.54 and 0.58 when the inlet velocity is 37 m/s. 2. The flame structure assumes a V-shape when the combustion is stable. The combustion oscillation in the LPP combustor will compact the flame structure to be flat. The phase-averaged flame structures throughout an oscillation cycle shows that there is a process of flame roll-up for the flame close to the LPP combustor axis and there is a process of separation and consolidation for the flame far from the LPP 19 ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

combustor axis. The periodical variation of the flame structures in the LPP combustor throughout an oscillation cycle will change the heat release rate distribution; resulting in the periodical fluctuation of the heat release rate. The combustion oscillation will be driven when the periodical heat release rate fluctuation couples with the periodical pressure fluctuation. 3. The spatial distribution of the normalized local Rayleigh index of the LPP combustor validates that the combustion oscillation is driven by the coupling of the heat release rate fluctuation and pressure fluctuation.

Notes The authors declare no competing financial interest.

REFERENCES (1) Armitage, C. A.; Cant, R. S.; Dowling, A. P.; Hynes, T. P. Linearised theory for LPP combustion dynamics. Linearised theory for LPP combustion dynamics. Proceeding of ASME Turbo Expo 2003; Atlanta, Georgia, June 16-19, 2003. (2) Lieuwen, T.; McManus, K. Introduction: combustion dynamics in lean-premixed prevaporized (LPP) gas turbines. Journal of Propulsion and Power 2003, 19, 721-721. (3) Dhanuka, S. K.; Temme, J. E.; Driscoll, J. F. Unsteady aspects of lean premixed prevaporized gas turbine combustors: flame-flame interactions. Journal of Propulsion and Power 2011, 27, 631-641. (4) Temme, J. E.; Allison, P. M.; Driscoll, J. F. Combustion instability of a lean premixed prevaporized gas turbine combustor studied using phase-averaged PIV. Combustion and Flame 2014, 161, 958-970. 20 ACS Paragon Plus Environment

Page 20 of 23

Page 21 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

(5) Nguyen, Q. V. Measurements of equivalence ratio fluctuations in a lean premixed prevaporized (LPP) combustor and its correlation to combustion instability. Proceeding of ASME Turbo Expo 2002, Amsterdam, June 3-6, 2002. (6) Allouis, C.; Beretta, F.; Amoresano, A. Experimental study of lean premixed prevaporized combustion fluctuations in a gas turbine burner. Combustion Science and Technology 2008, 180, 900-909. (7) Tang, G.; Yang, Z.; McGuirk, J. J. LES predictions of aerodynamic phenomena in LPP combustors. Proceeding of ASME Turbo Expo 2001; New Orleans, Louisana, June 4-7, 2001. (8) Tran, N.; Ducruix, S.; Schuller, T. Analysis and control of combustion instabilities by adaptive reflection coefficients. 13th AIAA/CEAS Aeroacoustics Conference, Rome, May 21-23, 2007. (9) Dowling, A. P.; Stow, S. R. Acoustic analysis of gas turbine combustors. Journal of Propulsion and Power 2003, 19, 751-764. (10) Lohrmann, M.; Büchner, H. Prediction of stability limits for LP and LPP gas turbine combustors. Combustion science and technology 2005, 177, 2243-2273. (11) Bernier, D.; Lacas, F.; Candel, S. Instability mechanisms in a premixed prevaporized combustor. Journal of Propulsion and Power 2004, 20, 648-656. (12) Turns, S. R. An introduction to combustion: Concepts and Application. McGraw-hill: New York, 1996. (13) Kaminski, C. F.; Bai, X. S.; Hult J.; Dreizler, A.; Lindenmaier, S.; Fuchs, L. Flame growth and wrinkling in a turbulent flow. Applied Physics B 2000, 71, 711-716. (14) Therkelsen, P. L.; Portillo, J. E.; Littlejohn, D.; Martin, S. M.; Cheng, R. K. Self-induced unstable 21 ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 23

behaviors of CH4 and H2/CH4 flames in a model combustor with a low-swirl injector. Combustion and Flame 2013, 160, 307-321. (15) Huang, Y.; Sung, H. G.; Hsieh, S. Y.; Yang, V. Large-eddy simulation of combustion dynamics of lean-premixed swirl-stabilized combustor. Journal of Propulsion and Power 2003, 19, 782-794. (16) Rayleigh, J. W. S. The theory of sound. Dover, New York, Vol.2, 1945. (17) Lee, S. Y.; Seo, S.; Broda, J. C.; Pal, S.; Santoro, R. J. An experimental estimation of mean reaction rate and flame structure during combustion instability in a lean premixed gas turbine combustor. Proceedings of the Combustion Institute 2000, 28, 775-782.

TABLE Table 1. Dimensions and Flow Properties of the Combustion Rig for Acoustic Analysis dimensions and properties

preheated air chamber

combustor

exhaust pipe

length (mm). diameter (mm). temperature (K). pressure (bar). density (kg/m3). sonic velocity (m/s)

800 365 500 1 0.706 449

550 260 1200 1 0.294 695

4840 124 700 1 0.504 531

Table 2. Measured and Predicted Acoustic Frequencies

mode

measured frequencies (Hz)

predicted frequencies (Hz)

1L 2L 3L 4L

none 134-138 none none

76 139 226 288


Page 23 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

Table 3. Operating Conditions in the Experiment. test case

equivalence ratio

inlet velocity (m/s)

thermal power (kW)

case 1 case 2 case 3 case 4

0.54 0.54 0.54 0.54

25 31 37 43

44 55 66 77


0.66 0.66 0.66 0.66

25 31 37 43

53 66 79 93


0.50 0.54 0.58 0.62

25 25 25 25

41 44 47 50


0.50 0.54 0.58 0.62

37 37 37 37

61 66 71 76


Combustion Oscillation Characteristics and Flame Structures in a

Recommend Documents