Comparison of Measured PFI Spray Characterizations of E85 and N-heptane Fuels for a Flex-Fuel Vehicle

Transcription

1 ILASS Americas, 21st Annual Conference on Liquid Atomization and Spray Systems, Orlando, Florida, May Comparison of Measured PFI Spray Characterizations of E85 and N-heptane Fuels for a Flex-Fuel Vehicle A. H. Gandhi*, M. A. Meinhart Departments of Powertrain Research and Vehicle Evaluation and Verification Ford Motor Company Dearborn, Mi USA Abstract The continuous effort to develop a more efficient automotive internal combustion engine, while simultaneously reducing exhaust emissions, has led to a focus on improving nearly every process relating to combustion. One such area concentrates on analyzing the delivery of atomized fuel from the fuel injector. Experimental work was conducted at the Powertrain and Fuel Subsystems Laboratory of Ford Motor Company to determine the effect the two limiting cases for fuel in a flex-fuel vehicle would have on spray characterization. This work included spray pattern visualization, high-speed imaging and phase-doppler interferometry. With n-heptane selected as a representative substitute for gasoline, the results show that the use of n-heptane results in a larger spray angle and a smaller mean drop size compared to E85. *Corresponding author: 2400 Village Road Dearborn, MI Phone/ Fax agandhi1@ford.com

2 INTRODUCTION The continuous effort to develop automotive Internal Combustion (IC) engines that have low emissions and that can operate on a renewable fuel source has led to a renewed interest in alternative fuels. Ethanol is considered one of the more promising alternative fuels for spark-ignited IC engines because it is renewable, and can be blended with gasoline in any proportion. Also, it may be injected in the same manner as gasoline, and can be transported using the existing fuel delivery infrastructure. Other factors that favor ethanol include: the Renewable Fuels Standard of the Energy Policy Act of 2005 (mandating that 4.7 billion gallons of renewable fuel be added to the gasoline supply in 2007 and escalating to 7.5 billion gallons in 2012) [1]; enables a reduction in greenhouse gases; can be made from biomass; is considered renewable; enables a reduction in the U.S. dependency on foreign crude oil; and provides support for the agricultural infrastructure. For a number of these reasons, many automotive companies are currently designing and marketing flexible-fuel vehicles that can operate on any blend of gasoline and ethanol, ranging from 100% gasoline up to 15% gasoline with 85% ethanol (E85). However, E85 has a lower energy density than gasoline which requires a significantly larger volume of the fuel blend to be injected for each cycle. It is important for automotive manufacturers to optimize these vehicles for the use of gasoline and ethanol blends, thereby enabling an increased usage of renewable energy sources [2 4]. One important research area concentrates on analyzing the delivery of fuel from the injector. With the range of fluid blends employed in flex-fuel vehicles, it is vital that the limiting fuel spray characteristics be known to ensure appropriate injector targeting. Proper fuel targeting affects not only cold start emissions and transient fuel calibration requirements, but is also an important step in attaining optimal combustion efficiency and reduced unburned HC emissions. This optimization requires that the fuel spray characteristics be known and understood during the design and development process. EXPERIMENTAL SETUP Three different experimental techniques were utilized for this spray characterization study at Ford Motor Company's Powertrain and Fuel Subsystems Laboratory (PFSL). These were spray pattern visualization, high-speed imaging and phase-doppler interferometry. Each of the three techniques will be briefly outlined in the following sub-sections. Spray pattern visualization: It is important to qualitatively visualize the wetted footprint of the fuel spray so that the distribution of liquid fuel can be observed and documented. With regards to the engine, the positioning of the spray pattern is known as "targeting". In this technique the fuel injector is mounted a fixed distance above special calligraphy paper and is commanded to inject one time. The resultant fuel spray leaves a wetted footprint on this paper. This footprint is then imaged by means of a digital camera. An overview of the test configuration is shown in Figure 1, and an example resultant spray pattern is provided in Figure 2. Figure 1. Test Fixture for Spray Pattern Visualization Figure 2. Sample Result from Spray Pattern Visualization High-speed imaging: The fuel spray from an automotive fuel injector may be visualized using a high-speed camera combined with a light source. The resultant image may be used to characterize the spray angle, spray-tip penetration and the symmetry of the fuel spray. Although imaging is not the conventional means to determine the spray angle of a Port Fuel Injection (PFI) fuel injector, a highresolution patternator was not available for this study. The light source may be continuous or pulsed and illuminate the entire fuel spray from the front, back, or side. For these experiments, a digital camera was used,

3 with front lighting provided by a copper-vapor laser. A schematic of the test configuration is shown in Figure 3, along with a sample resultant image. The details for calculating the spray angle and spray-tip penetration length are subjective, and in this application a commercial software package titled SprayMaster from LaVision is used for image processing. Scaling, background subtraction, and image thresholding are some of the capabilities that this software provides. The algorithm for calculating the spray angle first detects the boundary of the spray for each horizontal pixel line and then interpolates a linear fit for both the left and right boundaries. This is based upon the inclusion of a user-defined percentage of the spray (95% for this study). The intersection of these lines is defined by the injector position, and the resultant angle formed is the spray angle. The spray-tip penetration is calculated as the axial distance along the injector axis from the injector tip to the leading edge of the spray. This algorithm employs the area that is defined by the left and right boundaries created by the spray angle, and determines a penetration radius of a circle having a center at the injector tip. The radius of the circle is determined by counting all vertical pixels inside the spray angle that have intensities above the minimum threshold, and then including a user-defined percentage (95% for this study) for determining the tip penetration. within the fuel spray. A TSI Phase-Doppler Particle Analyzer was used for these characterization experiments. The mean drop velocity was measured in the direction along the injector axis. The drop size and velocity data taken were initially time-windowed, with the measurement synchronized to the injection event. A test schematic for obtaining the drop size and velocity distributions is shown in Figure 4, and sample output data are shown in the two panels in Figure 5. TSI Incorporated Schematic of Phase Doppler Optics Beamsplitting Optics Laser Fringes Move Transmitting Optics FireWire PDM FSA Photodetectors Signal Processor Copyright 2004 TSI Incorporated A B C Fiberoptic θ Receiving Angle Receiving Optics Figure 4. Phase-Doppler Interferometry Schematic Velocity 1 Histogram Velocity Count Ch LASER Velocity Ch. 1 (m/sec) C A M E R A Figure 3. Schematic of the High-speed Imaging Test and a Representative Spray Image Diameter Count Diameter Histogram Phase-Doppler interferometry: In addition to the values for the spray pattern, spray angle and spray-tip penetration, it is important to also characterize the distributions of drop size and velocity Diameter (um) Figure 5. Sample Velocity and Diameter Histogram Data from Phase-Doppler Interferometry

4 Fuel injectors characterized: PFI fuel injectors of three different designs were evaluated in this study, with all the injectors being from the same manufacturer. The beta and alpha angles of the fuel spray are defined in Figure 6. Table 1 provides the specifications for each of the fuel injectors that were tested. The information from Table 1 and Figure 6 was provided by the injector manufacturer using n-heptane in conjunction with a high-resolution mechanical patternator. Therefore, the manufacturer's specifications will be referred to as "cone angle", whereas the results from imaging will be referred to as "spray angle" in accordance with the SAE J2715 Recommend Practice for spray measurement and characterization [5]. Figure 6. Spray Characterization Nomenclature FI # Number of holes Static flow 270 kpa (g/min) Alpha angle (º) Beta angle (º) Table 1. Fuel Injector Specifications High-speed imaging: High-speed imaging for determining the spray angle and spray-tip penetration was performed at PFSL. Data were taken with a Redlake HG-100K camera and an Oxford Lasers LS 10:10 copper-vapor laser for total volume illumination. The testing conditions are provided in Table 2. When the injector's electrical connector was oriented towards the camera, two distinct spray plumes are evident, hence the designation 2-spray for that image view. If the same measurement technique employed by the injector manufacturer were to be used, the calculated 2-spray spray angle should match the alpha + beta values provided. When the electrical connector is rotated 90, only 1 spray plume is evident, hence the designation of 1-spray for that image view. For this case also, if the same measurement technique as the manufacturer were to be used, the calculated 1-spray spray angle should match the beta value provided by the injector manufacturer. Each of the three fuel injectors tested were on-axis designs, and thus had 0 spray offsets from the injector axis. Each test condition in Table 2 was repeated five times with the resultant average values reported. Figures 7, 8 and 9 show the fuel spray images using n-heptane and E85 as the test fluids. Figures 10 through 13 show the spray angle and spray-tip penetration plotted versus time after Start of Fuel (SOF): this is the time when fuel is first visible in the high-speed image sequence, which accounts for the non-zero spray angle and the small spray-tip penetration at zero time. Injector #179 is also referred to as the 2.0L injector. Condition 1 2 Fuel type n-heptane E85 Fuel pressure (kpa) Back pressure (atm) Pulse width (ms) Frame rate (fps) Resolution 384 x x 720 Table 2. High-speed Imaging Test Conditions It is important to note that all E85 testing was performed using a 5.0 ms injector pulse width, whereas n-heptane testing employed 3.6 ms. The lower heating value of E85 requires a longer injection pulse width to deliver the same amount of energy to the engine. The same saturated-switch injector driver was used for all testing. RESULTS AND DISCUSSION The high-speed imaging, spray pattern visualization and phase-doppler interferometry results will be discussed in the following sub-sections. Figure 7. Spray Images for Injector #105 with N-heptane (left) and E85 (right) at 1 ms after SOF

5 Figure 8. Spray Images for Injector #103 with N-heptane (left) and E85 (right) at 1 ms after SOF Figure 9. Spray Images for Injector #179 with N-heptane (left) and E85 (right) at 1 ms after SOF Penetration [mm] L 4-hole_nhep_1 spray Figure 12. Spray-tip Penetration vs. Time after SOF for 2-spray 9 12-hole_nhep_1 spray 12-hole_E85_1 spray 4-hole_nhep_1 spray 4-hole_E85_1 spray 2.0L 4-hole_E85_1 spray 1 12-hole_nhep_2 spray hole_E85_2 spray 4-hole_nhep_2 spray 7 Spray Angle [º] hole_E85_2 spray 2.0L 4-hole_nhep_2 spray 2.0L 4-hole_E85_2 spray 4-hole 2-spray specification: 27º +- 8º 12-hole 2-spray specification: 32.5º +- 8º Penetration [mm] hole_nhep_2 spray 12-hole_E85_2 spray 4-hole_nhep_2 spray 1 4-hole_E85_2 spray 2.0L 4-hole_E85_2 spray Figure 10. Spray Angle vs. Time after SOF for 2-spray Spray Angle [º] 9 12-hole_nhep_1 spray 8 12-hole_E85_1 spray 4-hole_nhep_1 spray 7 4-hole_E85_1 spray 6 2.0L 4-hole_nhep_1 spray 2.0L 4-hole_E85_1 spray 5 4-hole 1-spray specification: 7º +- 4º 12-hole 1-spray specification: 10.5º +- 4º 3 1 Figure 11. Spray Angle vs. Time after SOF for 1-spray 2.0L 4-hole_nhep_2 spray Figure 13. Spray-tip Penetration vs. Time after SOF for 1-spray Figures 10 and 11 show that having n-heptane as the fuel generated a larger spray angle than using E85, regardless of the injector design. A comparison of the experimental results for the spray angle to the manufacturer's specifications (with n-heptane as the fuel type) is provided in Table 3. The discrepancy in data may be primarily attributed to the different measurement techniques used in obtaining the data; high-resolution patternation versus imaging. The largest difference in spray angle was exhibited by injector #179. Regardless of injector design or orientation (1-spray vs. 2-spray), it was found that changing from n-heptane to E85 had little effect on spray-tip penetration. This is shown in Figures 12 and 13.

6 FI # Number of holes spray injector specification (º) spray test results For n-heptane (º) spray test results for E85 (º) spray injector specification (º) 2-spray test results For n-heptane (º) spray test results for E85 (º) Table 3. Comparison of the Measured Spray Angle with the Specification Cone Angle Spray Pattern Visualization: Spray pattern visualization for qualitatively determining the spatial distribution of the fuel spray was also performed at PFSL. The testing conditions for all three injectors are given in Table 4. Figures 14, 15, and 16 show the results with n-heptane and E85 as the test fluid. The injector was positioned 50 mm above the paper. The electrical connector is oriented downward. Condition 1 2 Injector height (mm) Fuel type n-heptane E85 Fuel pressure (kpa) Back pressure (atm) Pulse width (ms) Table 4. Spray Pattern Visualization Test Conditions Figure 14. Spray Pattern for #105 with N-heptane, E85 Figure 16. Spray Pattern for #179 with N-heptane, E85 The results from spray pattern visualization are illustrated in Figures 14 through 16 and confirm that the change from n-heptane to E85 did not affect the spray pattern. The slightly larger footprint for E85 is attributed to the ~40% increase of injected fuel volume. Phase-Doppler interferometry: Phase-Doppler interferometry was performed at PFSL to determine the distributions of drop size and velocity within the fuel spray. The drop velocity distribution was included because it was a parameter requested by combustion modelers. The testing conditions for all three fuel injectors are provided in Table 5. The data were taken 50 mm below the injector tip while traversing the x-axis as shown in Figure 17. Figures 18 and 19 illustrate the variations in Sauter Mean Diameter (SMD) and mean drop velocity versus the PDI spatial location within the fuel spray. For these results all of the time-windowed data were timeintegrated. The results include n-heptane and E85 as the two test fluids. SMD represents the diameter of a drop having the same ratio of volume to surface area as the collection of all drops in the measured distribution. Condition 1 2 Injector height (mm) Grid spacing (mm) Fuel type n-heptane E85 Fuel pressure (kpa) Back pressure (atm) Pulse width (ms) Table 5. PDI Test Conditions -X axis X=0 +X axis Figure 15. Spray Pattern for #103 with N-heptane, E85 Figure 17. PDI Sampling Locations along the X-Axis

7 Sauter Mean Diameter [µm] Mean Drop Velocity [m/s] X coordinate Figure 18. SMD vs. Spatial Location in the Spray #103_nhep #103_E85 #105_nhep 4 #105_E85 #179_nhep #179_E X coordinate Figure 19. Mean Drop Velocity vs. Spatial Location in the Spray #103_nhep #103_E85 #105_nhep #105_E85 #179_nhep #179_E85 The results from Figure 18 confirm that the use of E85 resulted in larger drop sizes than n-heptane (0.2 37% depending upon spatial location) regardless of the injector type. Also, the largest drop sizes were typically present near the geometric center of each plume. The largest value of SMD for the positive and negative X- axis was spaced closer for injector #105 and #179 than for injector #103. This agrees with the injector manufacturer's specifications, as injector #103 had the largest alpha and beta angle of all the injectors. Typically the largest drop sizes are present in the area of maximum mass flux, which is generally near in the geometric center of each plume for this type of injector. The results from Figure 19 illustrate that the use of E85 as a fuel typically generated larger mean drop velocities than n-heptane (0.4 56% depending upon spatial location). This result was primarily due to the larger pulse width for E85. The largest mean drop velocities were typically present near the geometric center of each plume. The largest value of mean drop velocity for the positive and negative X-axis was spaced closer to the injector axis for injector #105 and #179 than injector #103. Once again, this agrees with the injector manufacturer's specifications as injector #103 has the largest alpha and beta angle of all the injectors. Typically the largest velocities are present in the area of maximum mass flux, which is generally near the geometric center of each plume for this type of injector. RECOMMENDATIONS AND CONCLUSIONS Experiments were conducted to determine the effect that the two limiting cases for the fuel in a flexfuel vehicle would have on the spray characterization, specifically comparing the sprays resulting from the use of n-heptane and E85. Three different PFI injector designs were evaluated using high-speed imaging, spray pattern visualization and phase-doppler interferometry. The following conclusions and recommendations are drawn from this experimental work. 1. Using n-heptane generated a larger spray angle than E85; for example, injector #105 had an 11% larger spray angle for the 2-spray view. 2. The spray-tip penetration was determined to be essentially the same for both n-heptane and E The pattern of the wetted spray footprint was similar for both n-heptane and E The use of E85 yielded larger drop sizes than n- heptane; for example, injector #105 had a % increase depending upon spatial location. ACKNOWLEDGEMENTS The authors would like to especially thank Stefano Guida for maintaining the equipment used in this study. REFERENCES 1. U.S. Energy Policy Act of 2005, U.S. Government Printing Office, Washington D.C., Dai, W., Cheemalamarri, S., Curtis, E., et al, "Engine Cycle Simulation of Ethanol and Gasoline Blends," SAE paper , Dumont, R.J., Cunningham, L.J., Oliver, M.K., "Controlling Induction System Deposits in Flexible Fuel Vehicles Operating on E85," SAE paper , West, B.H., Lopez, A.J., Theiss, T.J., "Fuel Economy and Emissions of the Ethanol-Optimized Saab 9-5 Biopower," SAE paper , "Gasoline Fuel Injector Spray Measurements and Characterizations, SAE Standards Document J2715, (2007), SAE International, Warrendale, Pennsylvania.