Fluidic Flame Holder
Flame Stabilization in a high speed premixed environment requires the presence of a bluff body to stabilize the flame. Bluff bodies or geometrical flame holders introduce a sudden expansion in the flow forming an energetic mechanism for turbulence production and a low speed recirculation zone where the flame is anchored. However, geometrical flame holders form a low pressure region which causes thrust reduction. A fluidic-based flame holder using a transverse slot jet issuing into a cross flow offers potential thrust and efficiency benefits for propulsion. The transverse slot jet flame holder has been shown to develop a low-speed recirculation zone capable of stabilizing a stationary flame, analogous to a rearward-facing step typically found in a ramjet engines. The current study explores the nonreacting and reacting flow field of the fluidic flame holder. A half wall-mounted V-gutter was also studied to provide a comparison to a more traditional flame holder. Digital particle image velocimetry was used to study the flow field of the fluidic flame holder. The fluidic flame holder demonstrated the capability to optimize for flame holding performance and combustion efficiency through the control of the momentum ratio of the jet to the channel flow, scaling the geometric dimensions of the mean recirculation. The reacting flow field of the fluidic flame holder shows a development of a co-flowing shear region due to the accelerated hot combustion products, indicating a higher heat release rate than the V-gutter. The effects of combustion on the mean and turbulent flow fields for both flame holders indicate larger integral length scales and higher turbulent burning velocities produced by the fluidic flame holder having potential to utilize shorter combustors, hence reducing weight and improving efficiency. Dilatation rate as physical measurements of the volumetric expansion due to heat release rate were studied as an assessment of heat release rate showing higher heat rates produced by the fluidic flame holder suggesting efficient combustion.
The schematic of the ramjet model is shown in Fig. 1. The figure shows three sections of the experiment: the nozzle, diffuser, and combustion section. Note that the diffuser is included to replicate realistic initial conditions for a real propulsion system as the projected surface area in the streamwise direction is critical for thrust production; the diffuser will result in enhanced thrust relative to a step configuration that is commonly used for flame anchoring. The analogous sudden expansion based burner would have the same area ratio as the diffuser, but would also incur higher total pressure losses. The operating conditions were such that the mean velocity at the diffuser inlet was nominally 33 m/s. The Reynolds number based on the nozzle mean exit velocity and nozzle height was held constant at approximately 30000.
Fig. 1 Schematic of the experiment
The geometric aspects of the recirculation zone were measured through interpretation of the streamline distribution to document the increase in size of the recirculation zone with increasing momentum ratio. Figure 2 shows the change of the normalized recirculation length l/H and height h/H with momentum ratio MR. A previous study of a transverse slot jet established that the momentum ratio is the appropriate ndependent parameter for scaling different sizes slot jets. The accelerated core of the quasi-parabolic mean velocity profile exiting the diffuser for the present study altered the distribution of the mean flow momentum, leading to a slightly reduced penetration of the jet compared to a uniform mean flow obtained in the previous experiment. However; at low MR cases, higher jet penetration was experienced compared to the previous experiment due to the lower crossflow momentum near the wall.
Fig. 2 The (a) length and (b) height of the recirculation zone in the midspan plane as a function of MR
The stability curve was constructed for the fluidic flame holder to explore the lean and rich blowout limits, and is presented in Fig. 3. A mixture of fuel and air was used for the fluidic stream of the transverse jet. The stability performance of the fluidic flame holder was explored at three different relative fluidic stream mixtures. Initially a fluidic stream mixture of equivalence ratio matching that of the main combustor flow, fj/fm = 1, was considered. Comparatively, the wall mounted V-gutter shows a slightly broader stability performance than the fluidic flame holder, as shown in Fig. 3. Leaner stability limits compared to the V-gutter where achieved by injecting a slightly richer fluidic mixture than the main flow fj/fm = 1.25, which will promote the possibility of operating the fluidic dump combustor at lean conditions hence enhancing fuel efficiency by controlling temperature for NOx control and extending the operational envelope. Richer limits where established by injecting a slightly leaner fluidic mixture than the main flow fj/fm = 0.833, extending the stability curve to richer limits if needed. The simplicity of controlling the fluidic mixture as well as the jet blowing rate shows the possibility of expanding the stability performance and the operational domain allowing higher fuel efficiency than traditional flame holders. An additional benefit for the fluidic flame holder is the possibility of inducing a highly flammable fuel (a fuel characterized with high dissociation and heat release rates) in the fluidic stream while maintaining a more traditional fuel in the main stream which will improve the stability performance of the fluidic dump combustor. It is clear that the stability curve of the fluidic flame holder is highly dependent on the chemical nature and the composition mixture of the fluidic stream.
Fig. 3 Flame stabilization boundaries of a V-gutter and a fluidic flame holder at fj/fm = 1, fj/fm = 1.25, and fj/fm = 0.833
Figure 4 shows the mean unfiltered chemiluminescence intensity and schematic of a stabilized mean flame at MR = 0.9 and f = 0.919. The stabilized flame is stretched from above the recirculation zone and spreads downstream and reaches the top wall of the combustor.
Fig. 4 Image of a stabilized flame (a) at MR = 0.9, f = 0.92, fj/fm = 1, and Uo = 11 m/s, (b) schematic
Movie 1 shows a Schlieren movie of the ignition process. The fluidic stream is injected from the lower left corner crossing the oncoming main flow from the left. The fluidic stream mixture is at an equivalence ratio matching that of the main flow at f = 0.86. The images show the development of a localized flame at the spark plug which grows downstream and upstream igniting the main flow; the entrained combustion products within the recirculation zone creates a continuous ignition source for the main flow and resulting in a stabilized flame. The region of uniform intensity distribution below and downstream of the jet represents pure combustion products suggesting high combustion efficiency.
Movie 1 Schlieren movie of the ignition process at MR = 0.2,
f = 0.86, fj/fm = 1, and Uo = 3 m/s
Movie 4 shows a stabilized flame at MR = 0.591 and equivalence ratio of f = 1.1. The image shows that at a high momentum ratio the flame spread extends completely to the top combustor wall, promoting a high combustion efficiency. As shown, it is clear to identify the recirculation zone from the region of uniform intensity distribution below the jet representing nominally pure combustion products within the recirculation zone.
Figure 5 shows the mean flow streamline pattern and streamwise velocity distribution for a high blowing case having a momentum ratio of 0.467 and mass flow ratio of 5.1%. It is clear from the streamline pattern that the jet penetrates into the main flow; however the higher momentum of the main flow causes the jet to deflect in the crossflow direction. The velocity difference between the accelerated flow and the reverse flow causes an intense shear between the mean and the reverse flow. The shear produces intense and large-scale turbulence that is expected to enhance combustion. Under reacting conditions the accelerated mean flow caused by the injected jet has increased by a factor of 1.6 times the average velocity of that under nonreacting condition. The difference between the accelerated flow and the reverse flow is on the order of 7 times the average velocity of the channel flow increasing the shear between the accelerated flow and the reverse flow compared to the nonreacting flow field. The two shear regions identified in Fig. 5 are clearly shown in the mean velocity field, where the first region is present between the accelerated flow above the recirculation zone and the reverse flow within the recirculation. The second shear region is shown in the downstream end of the combustor with a velocity difference between the accelerated combustion products and the reactants of nearly 1.6 times the average velocity.
The intense shear between the main and reverse flows will produce high turbulence levels that will play a role in flame stabilization as well as creating volumetrically-efficient turbulent flames. Figure 6 shows the normalized turbulence level distributions for a high momentum ratio. The figure shows that the jet creates a highly turbulent flow that remains near the injection wall that eventually decays with increasing downstream distance due to expected high dissipation levels in the presence of reduced mean shear. Turbulence levels increased in reacting flow conditions. Three peak turbulent regions are clearly shown in the figure.
The integral length scales for the flame holders are shown in Fig. 11. The integral length scale was computed through a cross-stream integration of the streamwise velocity spatial correlation function. The large integral length scale will produce highly wrinkled flames which will enhance burning rates. Under nonreacting conditions, Fig. 11a shows at a high MR the integral length scale peak is nominally similar to the V-gutter length scale although the recirculation zone for the V-gutter was slightly larger than the fluidic flame holder. It is clear that the integral length scale for each MR follows a similar trend near the early section of the jet trajectory. Figure 9b shows the integral length scales under reacting conditions. Smaller length scales were produced under reacting conditions. A noticeable increase in the length scale for the fluidic flame holder is shown at x/H > 2.6H which is due to the presence of the second shear region produced by the hot combustion products escaping the combustor. At this region, the fluidic flame holder length scales peak at higher levels than the V-gutter.
An approach to asses the relative heat release rate is achieved by integrating the streamwise mean velocity along the combustor height resulting in the average bulk flow velocity . The average velocity is directly proportional to the integral of the dilatation across the combustor height assessing relative heat release rate (due to volumetric expansion). Figure 12 shows the change in the average streamwise velocity across the combustor length for the fluidic flame holder at two different momentum ratios and the V-gutter. The fluidic flame holder shows a larger volumetric expansion due to heat release than the V-gutter. The momentum ratio shows minimal effect on the heat release rate of the fluidic flame holder which is expected since a larger geometrical flame holder will tend to enhance blowout limits with a relatively low efficiency increase.
The fluidic flame holder stabilized the flame using a confined transverse slot jet in a channel flow without the thrust penalties associated with typical bluff body flame holders. The fluidic flame holder recirculation zone size is controlled dynamically using the MR to optimize for stability and flame spread. The approach has potential to utilize a shorter combustor for the same recirculation height as typical bluff body flame holders, hence reducing weight and improving efficiency. High turbulent burning rates were produced by the fluidic flame holder assessing a more efficient combustion process.