AIAA 2009-5415 45th AIAA/ASME/SAE/ASEE Joint Propulsion Conference & Exhibit 2 - 5 August 2009, Denver, Colorado AIAA 2009-5415 Scramjet Mixing Control Using Fin-Guided Fuel Injection C. Aguilera*, B. Pang†, A. Ghosh†, A. Winkelmann‡, and K.H. Yu§ Department of Aerospace Engineering, University of Maryland, College Park, MD 20742 A specially designed fin was used to assist fuel jet penetration into a scramjet combustor flow-field. The fin served two specific purposes – first, to enhance mixing between fuel and air, and second, to minimize the stagnation pressure loss associated with the fuel injection process. Non-reacting mixing experiments were conducted in a Mach 2.1 blowdown air facility with transverse injection of choked helium flow simulating the fuel injection process. The penetration of helium jet and the mixing with supersonic airflow were characterized using spark schlieren and laser-sheet visualization techniques. Strength of injection-caused shocks was determined using a series of wall pressure measurements. The present results were consistent with the findings from our earlier study, which showed substantial increase in the fuel-jet-penetration height while at the same time reducing the stagnation pressure loss. In comparison with the baseline injector, the fin-guided injector caused at least a twofold increase in the fuel-jet-penetration height and up to 45% reduction in the shock-caused static pressure rise. The present results suggest that a properly designed fin could be used to increase both combustion efficiency from mixing enhancement and propulsion efficiency from reduced pressure loss. IXING enhancement is a problem in scramjet mainly due to extremely short flow residence time, which is on the order of msec. Also, supersonic mixing layers spread at a much slower rate in comparison to incompressible mixing layers subjected to a similar level of shear stress. This reduced rate of growth, which is widely known as compressibility effect1-3, makes mixing in scramjet an even more challenging problem. For typical Mach numbers encountered in scramjet internal flow field, one may expect up to a five-fold decrease in shear layer spreading rate due to the compressibility effect.4. There have been many previous studies5-7 for mixing enhancement in supersonic shear flows, especially for propulsion and power applications. Although these studies have shown that a variety of techniques can be used to enhance mixing over a wide range of flow length scales that may be particularly useful for reacting flows, the associated cost of mixing on the system performance has not been rigorously considered. Since mixing enhancement must result in an increase of flow entropy, there will be some reduction in the stagnation pressure. Then the amount of decrease in stagnation pressure can 100 Stagnation Pressure Loss (%) M I. Introduction 10 1 M=2 (weak shock) M=3 (weak shock) M=2 (strong shock) M=3 (strong shock) 0.1 0.01 0 5 10 15 Deflection Angle (deg) Figure 1: Stagnation pressure losses associated with planar deflection * Graduate Research Assistant, Aerospace Engineering Dept., University of Maryland, AIAA Student Member. Postdoctoral Scientist, Aerospace Engineering Dept., University of Maryland, AIAA Member. ‡ Associate Professor, Aerospace Engineering Dept., University of Maryland. § Associate Professor, Aerospace Engineering Dept., University of Maryland, AIAA Associate Fellow. 1 American Institute of Aeronautics and Astronautics † Copyright © 2009 by Aguilera et al. Published by the American Institute of Aeronautics and Astronautics, Inc., with permission. 20 Figure 2: Experimental setup for fuel injection experiment. be used to quantitatively assess the cost of mixing. In this study, we consider a tradeoff between mixing enhancement and stagnation pressure loss by making measurements on these quantities. The mixing enhancement technique we have been developing utilizes a sharp fin not only for guiding fuel jet penetration into the air flow but also for triggering a weak oblique shock just upstream of the fuel injection location.8-10 The fin is designed to minimize the pressure losses associated with jet-induced shocks. For instance, when a transverse fuel jet is introduced into a supersonic cross flow, a blunt shock is created upstream which reduces the stagnation pressure downstream. This loss in stagnation pressure is mostly a wasted cost as the shock helps very little in terms of enhancing turbulent mixing at scales important for combustion. Typical transverse fuel injection causes deflection in supersonic airflow by setting up an oblique shock. Potential amount of stagnation pressure loss due to oblique shock is calculated for a two-dimensional flow and the results are shown in Fig. 1 as a function of shock strength and flow deflection angle. As shown in Fig. 1, the amount of stagnation pressure loss due to oblique shock can vary widely between a small fraction of a percent and tens of percent. Our ultimate objective is to optimize the tradeoff between combustion enhancement and stagnation pressure loss. For now, we quantify the amounts of mixing and pressure loss from our controlled experiments. II. Experimental Set-up Fuel injection and mixing experiments were conducted using a blow-down supersonic tunnel driven with a vacuum tank from the downstream end. This tunnel uses suction from the vacuum tank to establish Mach 2.1 airflow in the test section for approximately 15 seconds. The flow path dimensions and the fin shape are shown in Fig. 2. Helium, which was used as surrogate fuel, was injected transversely into the supersonic airflow through a 12.7-mm diameter choked orifice, flush mounted on the lower wall. Wall pressure was measured using a series of pressure ports that lined the combustor upper wall every 12.7 mm. The fin employed in this study was a triangular blade with a rectangular base of 12.7 mm wide and 127 mm long mm placed just upstream of the orifice injector. The height of the triangular cross-section varied from 0 at the leading edge to 38.1 mm at the trailing edge. The fin was placed such that the injector orifice was just downstream of the fin trailing edge allowing fuel injection into the fin’s wake. The resulting displacement in the tunnel crosssectional area was equivalent to a constant-angle ramp, corresponding to a simple 8.5° ramp for the same footprint area as the fin base, or a 0.7° ramp if the entire span of the wind tunnel was considered. A Scanivalve DSA-3217 Digital Sensor Array was used to make static pressure measurements along the top wall of the combustor. This Array consists of 16 temperature compensated piezoresistive pressure sensors with a 2 American Institute of Aeronautics and Astronautics pneumatic calibration valve. The 16 sensors, or channels, all have a range of 0-1.5 MPa. Their associated error is ±0.2% of scale for pressures less than 0.1 MPa, ±0.12% of scale for pressures between 0.1 and 0.14 MPa, and ±0.05% of scale for pressures above 0.14 MPa. The measured pressures are sent via a TCP/IP connection to a desktop computer and into a LabView virtual control panel. This virtual interface allows for monitoring of all 16 channels and the DSA’s settings as well as writing of the data to a text file to be read by post-processing software. For flow visualization, a conventional schlieren technique was used to obtain side profiles while laser sheet visualization was used to obtain cross-sectional view at selected downstream locations. Fuel injection experiments were conducted with and without the guide fin in place. The configuration without the fin was used as the baseline for performance comparison purpose. Table 1 shows the summary of various fuel injection conditions that were tested. Table 1: Flow conditions for fuel injection experiment. m! air m! fuel (kg/s) (g/s) FuelAir Ratio 1 1.56 2.7 2 1.56 3 4 case Stagnation Pressure (psi) p0,air p0,fuel 0.0017 14.7 75 3.5 0.0022 14.7 95 1.56 4.2 0.0027 14.7 115 1.56 5.0 0.0032 14.7 135 III. Results and Discussion For each of the injection conditions shown in Table 1, the flow mixing behavior is compared for the two injector configurations, one with the fin (Fin-Guided) and one without (Baseline). The results are presented in Figs. 3-6, each containing the schlieren images comparison as well as the corresponding wall pressure distributions. The surrogated fuel, helium, was injected at x/D=0. With increasing injection pressure or fuel flow rate, the fuel jet penetrated deeper into the free-stream supersonic airflow. The images show clearly that fuel penetration depth was substantially improved with the use of the fin. As the fuel penetration into the free stream was guided by the fin wake, the fuel was pushed to the top of the fin relatively unobstructed. The fuel penetrating beyond the fin height appeared to follow a similar trajectory as in the baseline case. Static pressure distributions along the upper wall for the corresponding cases were also compared. In these plots, the x-axis is normalized with the injector orifice diameter, while the y-axis is normalized with the initial stagnation pressure, p0,air. Also shown here are the corresponding wall pressure distributions prior to the fuel injection, that are shown as dotted lines. There was a slight increase in the starting pressure for the configuration with the fin as the fin acting as a shallow ramp induced a weak oblique shock. With fuel injection, the wall pressure was observed to increase from the pre-injection values due to jet-induced shocks. The amount of pressure increase was thus related to the strength of shocks caused by fuel injection. Comparison of these pressure plots suggests that the injection shock strength was substantially lower in the fin-guided injection case than that in the baseline case. Also, as the fuel flow rate was increased from 2.7 g/s in Fig. 3 to 5.0 g/s in Fig. 6, the shock strength in each case increased as expected. For another type of comparison, the amount of static pressure rise associated with fuel injection was normalized using the pressure distribution before the injection. The injection shocks caused static pressure rise depending on the shock strength. The results are shown in Fig. 7, which suggest that the fin effectively reduced the strength of the shock that the flow encountered. The pressure rise associated with fin-guided fuel injection was significantly smaller. In Fig. 8, the amount of reduction in wall pressure rise associated with shocks is plotted for various injection pressures. This reduction reached nearly 33% to 47% depending on the injection pressure. 3 American Institute of Aeronautics and Astronautics 0.14 Baseline w/o injec 0.14 Fin-Assisted w/o injec 0.12 p/p 0 0 p/p Fin-Assisted w/o injec 0.12 0.10 0.08 0.10 0.08 0.06 0.06 0 5 10 15 x/D Figure 3: Schlieren images of fuel injection at 75 psi and the corresponding pressure distribution on the upper wall 0 0.14 Baseline w/o injec 0.12 5 10 15 x/D Figure 5: Schlieren images of fuel injection at 115 psi and the corresponding pressure distribution on the upper wall 0.14 Fin-Assisted w/o injec Baseline Fin-Assisted w/o injec w/o injec 0.12 p/p 0 0 p/p Baseline w/o injec 0.10 0.10 0.08 0.08 0.06 0.06 0 5 10 15 x/D Figure 4: Schlieren images of fuel injection at 95 psi and the corresponding pressure distribution on the upper wall 0 5 10 15 x/D Figure 6: Schlieren images of fuel injection at 135 psi and the corresponding pressure distribution on the upper wall 4 American Institute of Aeronautics and Astronautics Baseline, 75psi Baseline, 95psi Baseline, 115psi Baseline, 135psi 1.5 Normalized Pressure 50 Reduction in Pressure Rise (%) Fin, 75psi Fin, 95psi Fin, 115psi Fin, 135psi 1.4 1.3 1.2 1.1 1.0 45 40 35 30 60 0.9 0 5 x/D 10 Figure 7. Static pressure rise associated with fuel injection. The pressure results were normalized with static pressure before the injection. 15 80 100 120 Injection Pressure (psi) 140 Figure 8. Amount of reduction in shock-induced static pressure rise associated with fin-guided fuel injection Cross-sectional views of fuel-air mixing were obtained at several axial locations using planar Mie-scattering technique. These images are presented in Figs. 9-11. Figures 9 and 10 represent instantaneous images of the flow fields at various flow conditions and distances downstream from the injector, corresponding to the baseline case and the fin-guided case respectively. The air flow was illuminated by fog condensation from moist air, while the surrogated fuel, helium, remained dry and thus appeared as dark spot. As the density of the fog particles in the airflow were affected by pressure and velocity, the post-shock region appeared brighter than the pre-shock region and so did the boundary layer where the velocity was slower. The size of the viewing area being presented is 4 injector diameters wide over a full height of the combustor. It can be inferred from Fig. 9 that the fuel jet penetration in the baseline case was relatively shallow and the fuel jet dispersed rather slowly. Comparing the corresponding images in Fig. 10, it can be vividly shown that the fuel penetration depth substantially increased with the fin-guided fuel injection. Averaged images are presented in Fig. 11. Between 80 and 100 instantaneous images were averaged for each case, spanning over 11-14 seconds of run time. Also, remarkably, much of the fuel in the fin-guided injection case was delivered to the core of the supersonic airflow rather than staying in the boundary layer as in the baseline case. This difference in mixing characteristics could have an additional implication if combustion efficiency and wall heat transfer characteristics were considered in the reacting flow. IV. Summary and Conclusions An experimental study on fin-guided fuel injection was conducted to characterize the enhanced mixing and the reduced pressure loss. Non-reacting fuel injection experiments were conducted in a simulated scramjet flow path. The results clearly showed that the fin-guided injection approach could not only enhance the mixing between the supersonic air stream and the injected fuel flow, but it could also reduce the stagnation pressure loss by lowering the strength of the jet-induced shocks. From the experimental results, it is clear that the location of fuel delivery into the supersonic airflow can be controlled by the fin height. In the current configuration, the width of the fin base was matched to the size of the injector orifice, which resulted in most of the fuel being delivered to the top of the fin. This resulted in more than doubling the fuel penetration depth in comparison with the baseline transverse injection case. It should also be possible to control the vertical distribution of fuel by adjusting the fin base width. 5 American Institute of Aeronautics and Astronautics The pressure distribution results showed that the amount of stagnation pressure loss, as inferred by the shockinduced pressure rise, was effectively controlled by the fin initiating the weak oblique shock rather than the fuel jet. In the current configuration, the maximum saving was achieved with the lowest fuel flowrate case which yielded more than 45% reduction in shock-caused static pressure rise. Extrapolating from such results, it would be possible to further reduce the shock-induced pressure loss by optimizing the fin design in the future. The ultimate optimization test will require combustion experiments with actual thrust measurements. x/D = 0 x/D = 2 x/D = 4 x/D = 6 x/D = 8 x/D = 10 (a) p0,fuel /p0,air =5.1 (b) p0,fuel /p0,air =6.4 (c) p0,fuel /p0,air =7.8 (d) p0,fuel /p0,air =9.2 Figure 9. Instantaneous images of baseline case fuel injection visualized with a laser sheet at various axial distances downstream. (a) fuel flow rate = 2.7 g/sec (b) fuel flow rate = 3.5 g/sec (c) fuel flow rate = 4.2 g/sec (d) fuel flow rate = 5.0 g/sec 6 American Institute of Aeronautics and Astronautics x/D = 12 x/D = 0 x/D = 2 x/D = 4 x/D = 6 x/D = 8 x/D = 10 (a) p0,fuel /p0,air =5.1 (b) p0,fuel /p0,air =6.4 (c) p0,fuel /p0,air =7.8 (d) p0,fuel /p0,air =9.2 Figure 10. Instantaneous images of fin-guided fuel injection visualized with a laser sheet at various axial distances downstream. (a) fuel flow rate = 2.7 g/sec (b) fuel flow rate = 3.5 g/sec (c) fuel flow rate = 4.2 g/sec (d) fuel flow rate = 5.0 g/sec 7 American Institute of Aeronautics and Astronautics x/D = 12 x/D = 0 x/D = 2 x/D = 4 x/D = 6 x/D = 8 x/D = 10 x/D = 12 (a) p0,fuel /p0,air =5.1 (b) p0,fuel /p0,air =5.1 (c) p0,fuel /p0,air =9.2 (d) p0,fuel /p0,air =9.2 Figure 11. Time-averaged images comparing the fin-guided fuel injection and the baseline fuel injection under various conditions. (a) baseline case with fuel flow rate = 2.7 g/sec (b) fin-guided case with fuel flow rate = 3.5 g/sec (c) baseline case with fuel flow rate = 4.2 g/sec (d) fin-guided case with fuel flow rate = 5.0 g/sec 8 American Institute of Aeronautics and Astronautics Acknowledgment This study was sponsored by the Office of Naval Research with Dr. Gabriel D. Roy as the scientific officer. References 1 Bogdanoff, D.W., "Compressibility effects in turbulent shear layers," AIAA J., Vol. 21(6), 926-927, 1983. Chinzei, N., Masuya, G., Komuro, T., Murakami, A., and Kudou, K., "Spreading of two-stream supersonic turbulent mixing layers," Phys. Fluids, Vol. 29(5), 1345-1347, 1986. 3 Papamoschou, D., and Roshko, A., "The compressible turbulent shear layer: an experimental study," J. Fluid Mech., Vol. 197, 453-77, 1988. 4 Gutmark, E.J., Schadow, K.C., and Yu, K.H., “Mixing Enhancement in Supersonic Free Shear Flows,” Annual Review of Fluid Mechanics, Vol. 27, 1995, pp. 375-417. 5 Yu, K.H., Schadow, K.C., Kraeutle, K.J., and Gutmark, E.J., "Supersonic Flow Mixing and Combustion Using Ramp Nozzle," Journal of Propulsion & Power, Vol. 11 (6), pp. 1147-1153, (1995) 6 Yu, K.H., and Schadow, K.C., "Role of Large Coherent Structures in Turbulent Compressible Mixing," Experimental Thermal and Fluid Science, Vol. 14, pp. 75-84, (1997) 7 Yu, K.H., Wilson, K.J., and Schadow, K.C., “Effect of Flame-Holding Cavities on Supersonic Combustion Performance,” Journal of Propulsion & Power, Vol. 17 (6), pp. 1287-1295, (2001) 8 Balar, R., Young, G., Pang, B., Gupta, A., Yu, K.H., and Kothari, A. "Comparison of Parallel and Normal Fuel Injection in a Supersonic Combustor," 42nd AIAA/ASME/SAE/ASEE Joint Propulsion Conference, Sacramento, CA, AIAA-2006-4442, July (2006) 9 Balar, R., Gupta, A.K., Yu, K.H., and Kothari, A.P. "Pylon-Aided Fuel Injection into Supersonic Flow,” 45th AIAA Aerospace Sciences Meeting, Reno, NV, AIAA-2007-834, January 8-11, 2007 10 Yu, K.H., Balar, R., Pang, B., Winkelmann, A., and Gupta, A., “Characterization of Fin-Assisted Supersonic Mixing Control,” International Symposium on Recent Advances in Combustion and Noise Control for Propulsion, Kauai, HW, December (2008) 2 9 American Institute of Aeronautics and Astronautics