- Email: [email protected]
Spray and atomization of diesel fuel and its alternatives from a single-hole injector using a common rail fuel injection system
Fuel 103 (2013) 850–861
Contents lists available at SciVerse ScienceDirect
Fuel journal homepage: www.elsevier.com/locate/fuel
Spray and atomization of diesel fuel and its alternatives from a single-hole injector using a common rail fuel injection system Pin-Chia Chen a, Wei-Cheng Wang a, William L. Roberts a,b, Tiegang Fang a,⇑ a b
Department of Mechanical and Aerospace Engineering, North Carolina State University, 911 Oval Drive-Campus, Box 7910, Raleigh, NC, 27695, United States Department of Mechanical Engineering, Clean Combustion Research Center, King Abdullah University of Science and Technology, 4220 Al Kindi (West), Thuwal, Saudi Arabia
h i g h l i g h t s " Spray and atomization of diesel fuel and its renewable alternatives. " Increasing injection pressure reduces droplet size with narrow droplet size distributions. " Newly developed renewable diesel results in small droplet size. " Biodiesel results in large droplet than Jet A fuel, No. 2 diesel, and renewable diesel fuel.
a r t i c l e
i n f o
Article history: Received 26 April 2012 Received in revised form 31 July 2012 Accepted 9 August 2012 Available online 24 August 2012 Keywords: Biofuel Diesel Jet fuel Common rail High-pressure injection
a b s t r a c t Fuel spray and atomization characteristics play an important role in the performance of internal combustion engines. As the reserves of petroleum fuel are expected to be depleted within a few decades, finding alternative fuels that are economically viable and sustainable to replace the petroleum fuel has attracted much research attention. In this work, the spray and atomization characteristics were investigated for commercial No. 2 diesel fuel, biodiesel (FAME) derived from waste cooking oil (B100), 20% biodiesel blended diesel fuel (B20), renewable diesel fuel produced in house, and civil aircraft jet fuel (Jet-A). Droplet diameters and particle size distributions were measured by a laser diffraction particle analyzing system and the spray tip penetrations and cone angles were acquired using a high speed imaging technique. All experiments were conducted by employing a common-rail high-pressure fuel injection system with a single-hole nozzle under room temperature and pressure. The experimental results showed that biodiesel and jet fuel had different features compared with diesel. Longer spray tip penetration and larger droplet diameters were observed for B100. The smaller droplet size of the Jet-A were believed to be caused by its relatively lower viscosity and surface tension. B20 showed similar characteristics to diesel but with slightly larger droplet sizes and shorter tip penetration. Renewable diesel fuel showed closer droplet size and spray penetration to Jet-A with both smaller than diesel. As a result, optimizing the trade-off between spray volume and droplet size for different fuels remains a great challenge. However, high-pressure injection helps to optimize the trade-off of spray volume and droplet sizes. Furthermore, it was observed that the smallest droplets were within a region near the injector nozzle tip and grew larger along the axial and radial direction. The variation of droplet diameters became smaller with increasing injection pressure. Ó 2012 Elsevier Ltd. All rights reserved.
1. Introduction Fossil fuels such as petroleum, natural gas and coal have been meeting most industrial and commercial demands for relatively low cost, high energy density, transportable fuels for decades. However, being the main source of transportation fuels, petroleum is estimated to be running out within 50 years due to the limited storage under the stratum [1]. In addition, burning fossil fuels gen⇑ Corresponding author. Tel.: +1 919 515 5230. E-mail address: [email protected] (T. Fang). 0016-2361/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.fuel.2012.08.013
erates pollutant emissions and greenhouse gases (primarily concerned with carbon dioxide emissions). According to a US Environmental Protection Agency (EPA) report, more than 90% of greenhouse gas emissions come from the combustion of fossil fuels [2]. Pollutants from internal combustion (IC) engines include CO, unburned fuel, SO2 (sulfur oxide), NOx (nitrogen oxides), volatile organic compounds (VOCs) and particulate matter (PM). For compression–ignition (CI) engines (diesel engines), PM is a major concern. In order to retard the depletion of fossil fuels and reduce the global warming effect of CO2 emission, discovering alternative, renewable, cost effective, and reliable fuel sources to replace fossil
P.-C. Chen et al. / Fuel 103 (2013) 850–861
fuels had become an essential issue for scientists. On the other hand, optimizing combustion process in IC engines also contributes to the reduction of both fossil fuel consumption and pollutant emissions by increasing efficiency. As highly efficient IC engines, diesel engines offer superior thermal efficiency and low fuel consumption compared with spark ignited engines. They are widely used in both light duty vehicles and heavy duty trucks. The spray characteristics of diesel fuel (and its alternatives) fuel plays a critical role in the engine performance and it is essential to optimize these characteristics. Studies of the fuel spray behavior can be classified into two categories: macroscopic and microscopic. Macroscopic parameters such as spray tip penetration and cone angle can be measured by direct visualization methods. Charge-coupled device (CCD) and intensified charge-couple device (ICCD) cameras are commonly used in research labs for taking spray images with a light source. Microscopic parameters such as droplet velocity, droplet size and size distribution can be measured through particle image velocimetry (PIV), phase Doppler particle analyzer (PDPA) or laser diffraction particle analyzer (LDPA) systems [3]. 1.1. Macroscopic spray properties The breakup length of diesel spray injecting into a high-pressure gas was measured by Yule and Filipovic [4]. Their results showed the breakup length of diesel sprays were consistently about 35% of the spray penetration length. In addition, the breakup length slightly increased for a larger nozzle diameter but it decreased with increasing injection pressure. Ryu et al. compared spray experimental and simulation results [5]. They found that the spray penetration was proportional to the 1=4 power of the injection pressure, which is similar to the findings of Hiroyasu and Arai [6]. Their fuel evaporation and air fuel mixture results indicated that the influence of the ambient density on spray impingement was stronger than the effects of temperature conditions when the fuel was injected early during the compression stroke. Diesel spray formation from a common-rail multi-holed injector was studied in a high-pressure, high-temperature cell by Klein-Douwel et al. [7]. Their results showed that the growth of spray penetration was proportional to tb where t is the time after injection, and b = 0.57 ± 0.02. The spray cone angle was not constant, but varied with time during an injection, mainly as a result of spray shape changes. An investigation of the effect of injection pressure on the macroscopic spray characteristics was performed by Delacourt et al. [8]. The experimental results showed that the spray tip penetration increased with an increase in injection pressure. The spray angle was slightly influenced by injection pressure and remained nearly constant during the entire injection process. Delacourt et al. also evaluated the spray volume by combining the information of spray tip penetration and spray angle, and observed that spray volume increased with injection pressure but seemed to reach a limit. Suh et al. [9] studied the effects of injection duration and ambient pressure on diesel spray characteristics. Their results illustrated that spray tip penetration increased with injection pressure and duration time. They also indicated that spray tip penetration decreased as ambient pressure increases, while spray cone angle was enhanced by increase in ambient pressure. Alam et al. [10] investigated spray and combustion in an optical turbocharged direct-injection (DI) diesel engine. As a result, the start of injection (SOI) and start of combustion (SOC) of neat biodiesel were the earliest compared with diesel and blended fuels. These were mainly due to the higher density and bulk modulus of biodiesel compared to diesel fuel. Spray characteristics of diesel and biodiesels under ultra-high injection pressure (up to 300 MPa) were studied by Wang et al. [11]. This study showed that biodiesel
851
had longer spray tip penetration than diesel, but the difference became smaller at higher ambient density (>30 kg/m3). The spray cone angles of biodiesel were smaller than that of diesel, generating smaller spray area and volume. Gao et al. found that although biodiesels were produced from different inedible oils, their viscosities and densities, and therefore their macroscopic spray characteristics, were similar [12]. Spray tip penetration and spray velocity increased with increase in the blending ratios of biodiesel, but the spray cone angle decreased. This indicated a reduction in spray atomization quality while increasing biodiesel blending ratios. Fang et al. investigated the spray and combustion of diesel and biodiesel blends in an optical diesel engine using a commonrail fuel injection system [13]. Longer spray penetration was observed for blends with more biodiesel content and lower flame luminosity and higher NOx emissions were found for biodiesel fuels. However, applying advanced multiple-injection strategies could reduce the emissions of biodiesel fuels under clean diesel combustion modes [14]. In addition, due to the ‘‘one fuel forward’’ policy adopted by the US military, application of jet fuels in diesel engines is important for ground vehicles [15]. Spray tip penetration of JP-8 were measured in an optical combustion vessel and compared with diesel fuel by Pickett and Hoogterp [16]. Results showed that the liquid-phase penetration of JP-8 was shorter than that of diesel due to the lower boiling point of JP-8. However, the vapor penetration rate of JP-8 was not much different from that of diesel. Lee et al. [17] used the Mie-scattering method to analyze the macroscopic spray characteristics of JP-8 and diesel in an optically-accessible single-cylinder heavy duty diesel engine equipped with a high-pressure common-rail injection system. It was found that JP-8 had a shorter spray tip penetration and wider spray angle than diesel fuel mainly due to the faster vaporization characteristics of JP-8. 1.2. Microscopic spray properties An experimental investigation of diesel and coal-water slurry Sauter mean diameters (SMDs) near the spray tip region was conducted by Kihm and Terracina [18]. Their results showed that average SMDs of both fuels increased with increasing ambient pressure and distance from the nozzle but decreased significantly with increasing injection pressure. No significant effects of nozzle diameters on SMD were found with nozzle diameters between 0.2 and 0.4 mm. Suh et al. [19] studied the effects of injector driver type on spray and atomization characteristics using a PDPA system. They concluded that piezo-driven injection system had a smaller droplet size than solenoid-driven injection system due to the more rapid opening of injector nozzle and piezo-driven injector generated higher spray velocity at the beginning of injection. Labs and Parker [20] characterized the interior properties of diesel spray near the injector nozzle. Their results indicated that SMD increased with radial distance and axial distance from the nozzle tip. Thus, the smallest droplets were on axis and close to the injector tip. An analytical study of atomization characteristics of seven different kinds of biodiesels were carried out by Ejim et al. [21]. Results from statistical analysis showed that pure coconut biodiesel had similar atomization characteristics to No. 2 diesel, because of its similar properties, i.e., density, surface tension, and viscosity. Other biodiesels, such as palm, soybean, cottonseed, peanut and canola, had slightly larger droplet sizes compared with diesel. Allen et al. mathematically analyzed and compared the atomization characteristics of 15 biodiesel fuels by predicting their viscosities and surface tensions directly from their fatty acid compositions [22]. According to their comparison, all kinds of biodiesels have similar atomization characteristics, except for biodiesel derived from coconut and rapeseed oils. Gao et al. conducted experiments of the spray characteristics of inedible oils (such as jatropha oil,
852
P.-C. Chen et al. / Fuel 103 (2013) 850–861
palm oil and used cooking oil) blended fuel and compared with #0 diesel [12]. SMD of blended fuels was greater than diesel, and the spray was more concentrated, due to the higher viscosity and surface tension of biodiesel, compared with conventional diesel. However, at lower blending ratios, the spray properties of biodiesel were similar to diesel. Choi et al. investigated the droplet properties under different blending ratios of biodiesel (palm oil) [23]. They found that the peak of the droplet size volume frequency distribution increased with increasing injection pressure but the peak decreased while increasing the blending ratio. Furthermore, the cumulative volume distribution moved toward the large droplet direction when increasing the blending ratio. As the biodiesel blending ratio increased, the SMD, mass median diameter (MMD), and span factor (which is a dimensionless parameter indicative of the uniformity of the drop size distribution) increased. The deviation between spray size distribution and mean diameter became larger while increasing blending ratio, i.e., the spray size was distributed in a wider range than that of common diesel. Mao et al. [24] compared the SMD of diesel, Jet-A and JP-8 under different injection pressures with a swirl atomizer. The results indicated the SMDs of Jet-A and JP-8 are both smaller than that of diesel fuel. The smaller SMDs of jet fuels were mainly due to their relatively lower viscosity than diesel fuel. The droplet size change among different fuels when exposed to a hot atmosphere (800 K) at 1 bar ambient pressure was demonstrated by Prommersberger et al. [25]. Their result showed that the droplet sizes of both Jet A and JP-4 decreased faster than that of diesel due to their lower boiling points. In this study, we compared the differences of commercial No. 2 diesel fuel (D) and several alternative fuels by investigating their spray and atomization characteristics. The alternative fuels we studied are biodiesel derived from waste cooking oil (B100), 20% of biodiesel blend (B20), renewable diesel fuel (R) synthesized at NC State University using a thermo-catalytic approach yielding a mixture of normal and iso-alkanes [26,27] and jet-A fuel (J) purchased from a local civil airport. The picture of the fuels used in this work is shown in Fig. 1 and different colors are observed for different fuels. Table 1 lists the physical properties of the five fuels used. Biodiesel has the largest values of density, viscosity and surface tension. Renewable diesel fuel has the lowest density but it is close to Jet-A, which has a low viscosity and surface tension. Renewable diesel fuel has similar viscosity and surface tension to diesel fuel. The experiments were conducted with various pressures, injection durations, and locations of spray for microscopic spray properties. Macroscopic spray properties were investigated with different injection pressures and fuels. 2. Experimental methods 2.1. Alternative fuels for diesel The conventional biodiesel chosen in this study is derived from waste cooking oil and produced by Piedmont Biofuels located in
Table 1 Physical properties of fuels. Fuel
Abbreviation
Density (g/ml) (40 °C)
Kinematic viscosity (mm2/s) (40 °C)
Surface tension (m N/m)
Diesel Biodiesel 20% biodiesel blend Renewable diesel fuel Jet A
D B100 B20 R J
0.83 0.87 0.84 0.76 0.79
2.65 4.60 3.04 2.48 1.15
30.3 34.7 27.7 30.3 28.7
Pittsboro, North Carolina. Producing biodiesel based on waste cooking oil has better environment benefits and less land usage compared with other feed stocks. These advantages are the main reasons of studying waste cooking oil biodiesel in this work. The renewable diesel fuel (biofuel) used in this study was made in the Department of Mechanical and Aerospace Engineering, North Carolina State University [26,27]. This biofuel is composed of a mixture of n- and iso-alkanes derived from canola oil using a multi-step process including hydrolysis, deoxygenation, and isomerization, yielding a fuel whose relatively lower viscosity and density make it be closer to those properties of diesel fuel than biodiesel. In addition, this new biofuel is a pure hydrocarbon fuel with no oxygen content, which is significantly different from conventional biodiesel. Jet fuel is commonly used in ground support vehicles at airports, in place of diesel, to reduce the number of different fuels for logistic reasons. It is also a basic requirement for military vehicles to run jet fuels in diesel engines under the ‘‘one fuel forward’’ policy [15]. Jet fuel has higher flash point and lower freezing point than diesel for safe usage in aircraft. The jet fuel used for our work is Jet A purchased from a local airport. It has the lowest viscosity but most pungent smell among the fuels in this study due to the higher aromatic content. 2.2. Common-rail high-pressure injection system The fuel injection system is illustrated in Fig. 2. A first generation common-rail fuel injection system was built to control and maintain the fuel pressure at a given level (up to 1350 bar). The common rail has a pressure sensor providing a measure of the pressure inside the rail. A specially machined single holed injector nozzle with a diameter of 140 lm was used in this work. The fuel pressure in the rail was controlled by a DAQ (NI 195509D-01L, National Instruments) with a PID control program built in LabView 2010. The DAQ acquired the pressure signal from the rail pressure sensor and compared the data with desired pressure set by the operator. The PID controller adjusted the opening and closing of the pressure regulator solenoid valve to maintain the rail pressure at a given level. The activation of fuel injection was controlled by a pulse/delay generator, by which we were able to control the width of injection pulse (injection duration) and output a single shot signal to the injector driver, which directly controlled the opening and closing of the injector valve. The chosen injection pressures were 300, 500, 800, and 1000 bar. The fuels were stored in room temperature at 25 °C and also injected into air at room temperature and pressure. 2.3. Droplet size measurement system
Fig. 1. Picture of fuels used in this study (from left to right: Jet-A, renewable biofuel, biodiesel, 20% biodiesel blend diesel, and diesel fuel, respectively).
The droplet size of fuel spray was measured by a laser diffraction particle analyzer (LDPA) system (SprayTecÒ, MAL 1009475, Malvern Instruments Inc.). The data acquisition rate was set at 10 kHz and a 300 mm lens was used for droplet size range of
853
P.-C. Chen et al. / Fuel 103 (2013) 850–861
Fig. 2. Schematic of the common rail fuel injection system setup.
Fig. 3. Setup of the high speed imaging system (left) and the definitions of spray tip penetration and cone angle (right).
0.1–900 lm. The fuel injector and the LDPA system were synchronously triggered by the delay/pulse generator. The measurement duration was 5 ms after the trigger, which covered the whole process of spray formation and atomization. The detector range was selected as 15–26 of 36 detectors (larger number of detector catches signal with wider scattering angle which is caused by larger droplet) to eliminate false signals due to beam steering effects and to avoid erroneous results. The LDPA instrument was located at different positions downstream of the spray to study the droplet size distribution inside the spray plume. The droplet sizes shown in this study are the average values of 10 experimental runs. The error bars of the droplet size data are the standard deviations calculated based on the 10 runs of each case.
Table 2 Experiment parameters. Equipment
Parameter
Particle size measurement system Data acquisition rate Lens Detect droplet size range Detector range Spray image acquisition system Lens Frame rate Resolution Exposure time
SprayTec, MAL 1009475 10 kHz 300 mm 0.1–900 lm 15–26 Phantom 4.3 NikonÒ 50-mm f/1.8 7312 fps 608 128 7 ls
2.4. Spray image acquisition system A high-speed digital video camera (Phantom 4.3, Vision Research Inc.) was used to capture the images of the fuel spray. The frame rate was set at 7312 frames per second at a resolution of 608 128 with the exposure time of 7 lm in this study. A 1000 W light source was used to volumetrically illuminate the spray plume. A picture of the setup is shown in Fig. 3 (left). The high-speed camera is also synchronized with the injector using the same trigger pulse. The spray images were acquired up to 15 ms, which covered the entire spray appearance time. The spray tip penetration and spray cone angle data were calculated based on the spray images as illustrated in Fig. 3 (right). Based on the spray
tip penetration and the cone angle, spray volume was calculated using the following equation [8]:
V ¼ ðp=3ÞS3 ½tan2 ðh=2Þ
1 þ 2 tanðh=2Þ ½1 þ tanðh=2Þ3
ð3:1Þ
where S is the spray tip penetration and h is the spray cone angle. The spray tip penetration and cone angle data in this study were the averaged values derived from 5 sets of spray images for each case. The experimental parameters are listed in Table 2 for both droplet size measurements and high-speed spray imaging.
854
P.-C. Chen et al. / Fuel 103 (2013) 850–861
Fig. 4. SMD variation under different parameters: (a) injection pressure (axial distance: 8 cm, radius distance: 0 cm, injection duration: 1.5 ms); (b) spray axial distance from injector nozzle tip (injection pressure: 800 bar, injection duration: 1.5 ms, radial distance: 0 cm); and (c) radial distance from the spray axis (axial distance: 8 cm, injection pressure: 500 bar, injection duration: 1.5 ms).
Fig. 5. Derived diameter parameters of Dv10, Dv50, and Dv90 for different fuels with injection pressure: (a) Dv10, (b) Dv50, and (c) Dv90 (axial distance: 8 cm, radius distance: 0 cm, injection duration: 1.5 ms).
P.-C. Chen et al. / Fuel 103 (2013) 850–861
855
Fig. 6. Comparison of particle size distribution for different fuels under different injection pressures: (a) 300 bar; (b) 500 bar; (c) 800 bar; and (d) 1000 bar.
3. Results and discussion 3.1. Sauter mean diameter (SMD) Fig. 4a–c depict the SMD results for different conditions. Fig. 4a shows the SMDs of the different fuels under 300, 500, 800 and 1000 bar injection pressures, all with the injection duration of 1.5 ms. For the figures shown in this study, the symbol J stands for Jet-A and R stands for renewable diesel fuel. Conventional biodiesel and 20% biodiesel blend are represented by B100 and B20,
and D is for diesel. It is clear from Fig. 4a that SMDs decrease with an increase in the injection pressure for all the fuels. The SMD difference among the fuels is mainly due to the differences in their viscosity and surface tension (Table 1). A higher viscosity leads to a lower fuel jet velocity, leading to larger droplet size. A lower surface tension makes the spray easier to break up into small droplets. According to Fig. 4a, biodiesel has much bigger droplets than diesel due to its highest viscosity and surface tension among the fuels. The viscosity of B20 is higher than diesel but its surface tension is lower. These combined fuel properties affect the SMD of B20
856
P.-C. Chen et al. / Fuel 103 (2013) 850–861
by showing a close value to diesel at lower pressures (<500 bar). Jet-A has the smallest droplet size due to its lowest viscosity and surface tension as well. Renewable diesel fuel has a similar surface tension and viscosity to diesel; however, it has smaller droplet size than diesel and is closer to Jet-A, particularly under high injection pressure. This is a very interesting result indicating unique atomization characteristics of the newly developed biofuel. According to the results in Fig. 4a and fuel properties in Table 1, the current results also show that the effect of viscosity is more significant on the SMD than that of surface tension due to the fact that viscosity variations among fuels are higher than that of surface tension. Locating the laser beams of the LDPA system at different positions can help increase the level of understanding of the droplet
sizes distribution within the spray plume. Fig. 4b and c demonstrate the SMD changes at different spray axial and radial distances. The SMDs increase with increasing distance between the measuring point and the injector nozzle tip due to momentum loss along the penetration development. SMDs get larger while moving away from the nozzle tip. Fig. 4c shows the SMDs at different radial distances from the axis of the spray. The smallest SMDs are found along the axis of sprays and increase about 1 lm at the spray peripheries. This phenomenon is believed to be due to the spray pressure dropping to ambient pressure on the peripheries. At a radial distance of 1 cm, it is already outside of the spray periphery, and the sprays become loose and dilute. Therefore, their droplet sizes decrease again. These results are consistent with findings in Ref. [20].
Fig. 7. Spray tip penetration for different fuels under different injection pressures: (a) 300 bar; (b) 500 bar; (c) 800 bar; and (d) 1000 bar.
P.-C. Chen et al. / Fuel 103 (2013) 850–861
3.2. Derived parameters The derived parameters for droplet size include Dv10, Dv50, and Dv90 as shown in Fig. 5a–c. Dv50, also known as volume median diameter (VMD), is a value indicates that half the spray volume was contained in droplets that were smaller than this value while the other half were contained in droplets that were larger than this value. Similarly, Dv10 and Dv90 values indicate that 10% and 90%, respectively, of the spray volume was contained in droplets that were smaller than these values. Fig. 5a shows the Dv10 of different fuels with injection pressures. The difference between diesel and B20 for Dv10s is larger than that of SMD. Furthermore, Dv10 of Jet-A decreases more slowly than renewable diesel fuel while increasing injection pressure, thus it becomes larger than renewable diesel fuel at higher injection pressures (>800 bar). This phenomenon indicates that the effect of injection pressure on
857
droplet size distribution of Jet-A is less pronounced than that of renewable diesel fuel. Fig. 5b shows the Dv50 of different fuels under different injection pressures and the data are a little higher compared with the SMDs, which indicates that the SMD of spray is slightly smaller than the droplet size at median volume. As illustrated in Fig. 5c, large droplets increase with an increase in viscosity. Dv90s of B20 are close to diesel fuel because the composition of B20 is similar to diesel. Interestingly, the standard deviations of Dv90s were smaller than those of Dv10s and Dv50s indicating the diameter variation is relatively smaller for larger droplets. 3.3. Droplet size distribution Droplet size distribution of alternative fuels compared with conventional diesel under different injection pressures are illustrated in Fig. 6a–d. At 300 bar, biodiesel, B20 and renewable diesel
Fig. 8. Spray cone angles for different fuels under different injection pressures: (a) 300 bar; (b) 500 bar; (c) 800 bar; and (d) 1000 bar.
858
P.-C. Chen et al. / Fuel 103 (2013) 850–861
Fig. 9. Spray volume for different fuels under different injection pressures: (a) 300 bar; (b) 500 bar; (c) 800 bar; and (d) 1000 bar.
fuel shows similar distribution, slightly narrower than diesel, but the peak of volume frequency of these fuels and diesel are all about 16 lm. Jet-A shows a distinct distribution in smaller diameter regions and its peak is located at about 13 lm for 300 bar injection pressure. At 500 bar, the droplet size distribution of biodiesel and B20 are still narrower than diesel but that of renewable diesel fuel becomes closer to diesel fuel, while Jet-A remains in the smallest region. All distributions of different fuels at 500 bar injection pressure show a trend of moving toward smaller diameters compared with 300 bar. For 800 bar, all the fuels have similar droplet size distributions within a smaller range (from 1 lm to 30 lm) with peaks around 10 lm. This is the reason 800 bar injection pressure was
chosen as the standard condition for the experiments that follow. The variation of droplet sizes of different fuels is the smallest under 800 bar injection pressure; therefore, the effect of changing other experimental conditions on droplet size can be clearly distinguished. Comparing with 800 bar, under the highest injection pressure used in this study (1000 bar), the peaks of renewable diesel fuel, diesel and Jet-A further go down into smaller diameter regions while this change is not obvious for biodiesel and B20. In summary, the droplet sizes of all the fuels are distributed in a relatively smaller region with increasing injection pressure. From these results, it is interesting to note that although B20 has a similar composition and SMD of diesel fuel, it shows similar
P.-C. Chen et al. / Fuel 103 (2013) 850–861
frequencies of larger droplet diameter distribution to biodiesel under lower injection pressures (i.e., 300 and 500 bar). 3.4. Spray tip penetration Spray tip penetrations of all the fuels at different injection pressures are shown in Fig. 7a–d. Tip penetrations of the five fuels increase with time after being injected from the injector and a higher injection pressure promotes this increase. This trend can be easily observed by looking at diesel penetration under 300 bar in Fig. 7a. At 1.5 ms after injection, the spray tip penetration is 10 cm compared with 14 cm under 1000 bar in Fig. 7d. Biodiesel shows significantly longer penetration than other fuels at low injection pressure (300 bar) but the difference becomes small at higher injection pressures. This may be attributed to the fact that the relatively higher density and bulk modulus of biodiesel increase the velocity of the fluid exiting the nozzle compared with the other fuels, thus increasing injection pressure has less effect on biodiesel to enhance the growth of spray tip penetration. Another interesting phenomenon is that B20 has shorter penetrations than diesel under all injection pressures. This might be due to the higher viscosity of B20 raising the friction between fuel and the injector nozzle surface. However, for pure biodiesel, B100, this friction effect is
Injection Pressure: 300 bar (cm) J R B100 B20 D
859
overwhelmed by its large density and bulk modulus, which make biodiesel have the longest penetration. Renewable diesel fuel and Jet-A have shorter spray tip penetrations than diesel and biodiesel due to their smaller densities. Furthermore, it can be seen from the figures that the spray tip penetration of the renewable diesel fuel is slightly shorter than Jet-A. This phenomenon may be caused by the effect of the significantly smaller viscosity of Jet-A (1.15 mm2/s) compared with renewable diesel fuel which has a viscosity at about 2.48 mm2/s. A lower viscosity reduces the friction between fuel and the surface of injector nozzle hole and yields a higher fluid velocity exiting from the injector nozzle. 3.5. Spray cone angle Fig. 8a–d illustrate the spray cone angle results under different injection pressures with time after the start of injection. As the spray penetrates, the droplets on the boundaries become smaller and diffuse easily, generating a decreasing trend of spray cone angle. The spray cone angles decrease rapidly after injection to about 10° for each injection pressure, however, at different rates. As seen from these figures, the spray cone angles of the five fuels converge to a constant value (steady state of the spray) at around 0.8 ms after start
Injection Pressure: 500 bar R B100 B20 D (cm) J
Fig. 10. Comparison of spray development images with time after injection trigger signal at 300 bar (left) and 500 bar (right) injection pressure at different times 0f 0.7 ms, 1.0 ms, and 1.5 ms (from top to bottom) after the injection trigger.
860
P.-C. Chen et al. / Fuel 103 (2013) 850–861
Injection Pressure: 800 bar (cm) J R B100 B20 D
Injection Pressure: 1000 bar (cm) J R B100 B20 D
Fig. 11. Comparison of spray development images with time after injection trigger signal at 800 bar (left) and 1000 bar (right) injection pressure at different times 0f 0.7 ms, 1.0 ms, and 1.5 ms (from top to bottom) after the injection trigger.
of injection at 300 bar injection pressure, and steady state is achieved sooner after injection as the injection pressure increases, i.e., 0.7 ms for 500 and 800 bar. Under 1000 bar injection pressure, the spray cone angles converge at 0.6 ms after injection. These results show that the distinctions between the different fuels after achieving steady state decrease with increasing injection pressure. More obvious distinctions between the different fuels can be seen before their convergence. The spray of biodiesel starts with the smallest cone angle, resulting from its highest viscosity. From the above observations, it is found that increasing the biodiesel blend ratio reduces the cone angle of fuel spray due to the increase of viscosity. As for renewable diesel fuel and Jet-A, the effect of density is more significant than the effect of viscosity and makes the cone angle of Jet-A smaller than the renewable diesel fuel under all injection pressures. 3.6. Spray volume Fig. 9a–d shows the calculated spray volumes of different fuels used in this study with time after the start of injection. The spray volumes develop as exponential curves under low injection pressure (300 bar) and become proportional when increasing the pressure. This is considered as the effect of higher injection pressure delivering the fuel faster than low pressure and causing the sprays
to reach steady state sooner. Larger spray volumes can be found at higher injection pressure by comparing Fig. 9a and d. For example, spray volume of diesel increases from 9 cm3 to 19 cm3 at 1.5 ms after the start of injection trigger as the injection pressure is increased from 300 to 1000 bar. The distribution in spray volume for the different fuels follows the same trend as spray tip penetration, which is biodiesel, diesel, B20, Jet-A, and then renewable diesel fuel, from large to small, respectively. According to this result, it is seen that spray volume is mainly determined by the spray penetration length, since the variation of spray cone angles among different fuels is not significant. Compared with diesel fuel, Jet-A and renewable diesel fuel have smaller spray volumes due to their shorter spray penetration. The results of B20 are interesting compared with diesel and B100. As a blended fuel, a common expectation is that B20 would show spray and atomization characteristics between diesel and B100. Our results show that B20 has a spray volume less than diesel fuel; the same trend is also observed in the spray tip penetration. 3.7. Spray image The spray images acquired with the high-speed camera capture the transient shape of sprays at different timings. These images are
P.-C. Chen et al. / Fuel 103 (2013) 850–861
sorted in Figs. 10 and 11 by three time periods (0.7, 1.0, and 1.5 ms) after the injection trigger signal. It can be seen from the figures that biodiesel has the largest penetration length, spray projected area, and thus the largest spray volume. This phenomenon is most obvious in the images at 1.0 ms after injection trigger. At this time point, the spray has already achieved steady state and develops at a constant velocity. This trend can be observed by comparing the volume differences at 0.7 ms and 1.5 ms with 1.0 ms. The penetration slows down and the perturbation area increases at 1.5 ms due to interaction with ambient air. An interesting characteristic of these fuel sprays is that the perturbation regions of the spray occurs earlier at higher injection pressures by comparing the images under different pressures at 0.7 ms after injection. This effect is due to the higher spray speeds under high injection pressures making the sprays penetration and interaction with air faster and stronger. In addition, the spray images show a shorter spray penetration of B20 compared with diesel fuel, which is consistent with our previous discussion of spray tip penetration and spray volume. 4. Conclusions In this work, the spray and atomization characteristics were investigated for diesel fuel and its four alternatives, including biodiesel derived from waste cooking oil, blend of biodiesel and diesel, a non-oxygenated hydrocarbon biofuel, and Jet-A fuel. Both microscopic and macroscopic parameters were measured. The observations and findings can be summarized as follows: 1. Biodiesel and its blend lead to larger droplets compared with diesel. However, Jet-A and renewable diesel have smaller droplet sizes due to their relatively lower viscosity and surface tension compared with diesel fuel. Increasing injection pressure is effective at reducing the droplet size. For example, the droplet size decreased 50% from 300 bar to 1000 bar injection pressure for all the fuels studied; 2. The current results further confirm that droplets at the spray periphery have larger diameters than those in the center of the spray due to the faster pressure drop at the spray periphery. Droplet diameters become larger as the distance from the nozzle tip increases due to pressure drop along the spray axial distance; 3. Under higher injection pressures, small droplets distribute in a narrower region and have less variation of their diameters; 4. Results from the measurement of spray tip penetration show a combined effect of density, bulk modulus and viscosity. B100 has the longest penetration of the fuels investigated. Longer penetration of fuel spray is observed under higher injection pressure. For example, the spray penetrations at 1000 bar injection pressure are about 40% longer than those at 300 bar, and this increase is more pronounced for Jet-A and renewable diesel; 5. Effect of injection pressure on spray cone angle is not evident compared with that on droplet size and spray tip penetration. Spray cone angle converges to its steady state very rapidly with increasing injection pressure; 6. Both microscopic and macroscopic spray properties influence the atomization quality of a certain fuel. There is a trade-off between smaller droplet size and larger spray volume as they both are determined by the properties of the fuel. The renewable diesel stands out among the biofuels by showing smaller droplet sizes (SMD, Dv10, Dv50, and Dv 90) than diesel. Using jet fuel in diesel engines might yield better atomization due to its small droplet size and similar spray volume to diesel under different injection pressures.
861
Acknowledgements This material is based, in part, upon work supported by the Faculty Research and Professional Development (FRPD) Fund from the North Carolina State University, the Natural Science Foundation under Grant No. CBET-0854174, and the National Science Foundation EFRI Program under Grant EFRI-093772. The authors also thank Dr. Brian Farkas and Dr. Allen Foegeding of the Department of Food Science at the North Carolina State University for their help to measure the fuel properties. References [1] BP Amoco Statistical Review of World Energy, June 2007. [2] Inventory of US greenhouse gas emissions and sinks: 1990–2009. US Environmental Protection Agency; 2011. [3] He C, Ge Y, Tan J, Han X. Spray properties of alternative fuels: a comparative analysis of biodiesel and diesel. Int J Energy Res 2008;32:1329–38. [4] Yule AJ, Filipovic I. On the break-up times and lengths of diesel sprays. Int J Heat Fluid Flow 1992;13:197–206. [5] Ryu J, Kim H, Lee K. A study on the spray structure and evaporation characteristic of common rail type high pressure injector in homogeneous charge compression ignition engine. Fuel 2005;84:2341–50. [6] Hiroyasu H, Arai M. Structure of fuel sprays in diesel engines. SAE paper 900475; 1990. [7] Klein-Douwei RJH, Frijters PJM, Somers LMT. Macroscopic diesel fuel spray shadowgraphy using high speed digital imaging in a high pressure cell. Fuel 2007;86:1994–2007. [8] Delacourt E, Desmet B, Besson B. Characterization of very high pressure diesel sprays using digital imaging techniques. Fuel 2005;84:857–67. [9] Suh HK, Park SW, Lee CS. Effect of piezo-driven injection system on the macroscopic and microscopic atomization characteristics of diesel fuel spray. Fuel 2007;86:2833–45. [10] Alam M, Song J, Zello V. Spray and combustion visualization of a directinjection diesel engine operated with oxygenated fuel blends. Int J Engine Res 2006;7:503–21. [11] Wang X, Huang Z, Kuti OA. Experimental and analytical study on biodiesel and diesel spray characteristics under ultra-high injection pressure. Int J Heat Fluid Flow 2010;31:659–66. [12] Gao Y, Deng J, Li C. Experimental study of the spray characteristics of biodiesel on inedible oil. Biotechnol Adv 2009;27:616–24. [13] Fang T, Lin YC, Foong TM, Lee CF. Biodiesel combustion in an optical HSDI diesel engine under low load premixed combustion conditions. Fuel 2009;88(11):2154–62. [14] Fang T, Lee CF. Biodiesel effects on the combustion processes in an HSDI diesel engine employing advanced multiple injection strategies. Proceedings of the Combustion Institute 2009;32(2):2785–92. [15] DOD 4140.25-M, DOD Directive 4140.25; 2004. [16] Pickett LM, Hoogterp L. Fundamental spray and combustion measurements of JP-8 at diesel conditions. SAE Int J Commer Veh 2009;1(1):108–18. [17] Lee CS, Park SW, Kwon SI. An experimental study on the atomization and combustion characteristics of biodiesel-blended fuels. Energy Fuels 2005;19:2201–8. [18] Kihm KD, Terracina DP. Synchronized droplet size measurements for coalwater slurry sprays generated from a high-pressure diesel injection system. Department of Mechanical Engineering, Texas A&M University; June 1993. [19] Suh HK, Roh HG, Lee CS. Spray and combustion characteristics of biodiesel/ diesel blended fuel in a direct injection common-rail diesel engine. J Eng Gas Turbines Power 2008;130(3):032807. [20] Labs J, Parker T. Two-dimensional droplet size and volume fraction distributions from the near-injector region of high-pressure diesel sprays. Atomization Spray 2006;16:843–55. [21] Ejim CE, Fleck BA, Amirfazli A. Analytical study for atomization of biodiesels and their blends in a typical injector: surface tension and viscosity effects. Fuel 2007;86:1534–44. [22] Allen CAW, Watts KC. Comparative analysis of the atomization characteristics of fifteen biodiesel fuel types. Transactions ASAE 2000;43(2):207–11. [23] Choi S, Oh YT. The study on spray behavior and atomization characteristics of biodiesel. Adv Mater Dev Perform (AMDP2011) Int J Mod Phys: Conf Ser 2012;6:419–24. [24] Mao CP, Short J, Buelow P. Innovative injection and mixing systems for diesel fuel reforming. SECA 6th annual, workshop; 2005. [25] Prommersberger K, Maier G, Wittig S. Validation and application of a droplet evaporation model for real aviation fuel. RTO MP-14, RTO AVT symposium on ‘‘gas turbine engine combustion, emissions and alternative fuels’’, Lisbon, Portugal; October 12–16, 1998. [26] Roberts WL, Lamp HH, Stikeleather LF, Turner TL. Process for conversion of biomass to fuel. US Patent#: US7816570B2, 2010. [27] Wang WC, Thapaliya N, Campos A, Stikeleather LF, Roberts WL. Hydrocarbon fuels from vegetable oils via hydrolysis and thermo-catalytic decarboxylation. Fuel 2012;95:622–9.
Contents lists available at SciVerse ScienceDirect
Fuel journal homepage: www.elsevier.com/locate/fuel
Spray and atomization of diesel fuel and its alternatives from a single-hole injector using a common rail fuel injection system Pin-Chia Chen a, Wei-Cheng Wang a, William L. Roberts a,b, Tiegang Fang a,⇑ a b
Department of Mechanical and Aerospace Engineering, North Carolina State University, 911 Oval Drive-Campus, Box 7910, Raleigh, NC, 27695, United States Department of Mechanical Engineering, Clean Combustion Research Center, King Abdullah University of Science and Technology, 4220 Al Kindi (West), Thuwal, Saudi Arabia
h i g h l i g h t s " Spray and atomization of diesel fuel and its renewable alternatives. " Increasing injection pressure reduces droplet size with narrow droplet size distributions. " Newly developed renewable diesel results in small droplet size. " Biodiesel results in large droplet than Jet A fuel, No. 2 diesel, and renewable diesel fuel.
a r t i c l e
i n f o
Article history: Received 26 April 2012 Received in revised form 31 July 2012 Accepted 9 August 2012 Available online 24 August 2012 Keywords: Biofuel Diesel Jet fuel Common rail High-pressure injection
a b s t r a c t Fuel spray and atomization characteristics play an important role in the performance of internal combustion engines. As the reserves of petroleum fuel are expected to be depleted within a few decades, finding alternative fuels that are economically viable and sustainable to replace the petroleum fuel has attracted much research attention. In this work, the spray and atomization characteristics were investigated for commercial No. 2 diesel fuel, biodiesel (FAME) derived from waste cooking oil (B100), 20% biodiesel blended diesel fuel (B20), renewable diesel fuel produced in house, and civil aircraft jet fuel (Jet-A). Droplet diameters and particle size distributions were measured by a laser diffraction particle analyzing system and the spray tip penetrations and cone angles were acquired using a high speed imaging technique. All experiments were conducted by employing a common-rail high-pressure fuel injection system with a single-hole nozzle under room temperature and pressure. The experimental results showed that biodiesel and jet fuel had different features compared with diesel. Longer spray tip penetration and larger droplet diameters were observed for B100. The smaller droplet size of the Jet-A were believed to be caused by its relatively lower viscosity and surface tension. B20 showed similar characteristics to diesel but with slightly larger droplet sizes and shorter tip penetration. Renewable diesel fuel showed closer droplet size and spray penetration to Jet-A with both smaller than diesel. As a result, optimizing the trade-off between spray volume and droplet size for different fuels remains a great challenge. However, high-pressure injection helps to optimize the trade-off of spray volume and droplet sizes. Furthermore, it was observed that the smallest droplets were within a region near the injector nozzle tip and grew larger along the axial and radial direction. The variation of droplet diameters became smaller with increasing injection pressure. Ó 2012 Elsevier Ltd. All rights reserved.
1. Introduction Fossil fuels such as petroleum, natural gas and coal have been meeting most industrial and commercial demands for relatively low cost, high energy density, transportable fuels for decades. However, being the main source of transportation fuels, petroleum is estimated to be running out within 50 years due to the limited storage under the stratum [1]. In addition, burning fossil fuels gen⇑ Corresponding author. Tel.: +1 919 515 5230. E-mail address: [email protected] (T. Fang). 0016-2361/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.fuel.2012.08.013
erates pollutant emissions and greenhouse gases (primarily concerned with carbon dioxide emissions). According to a US Environmental Protection Agency (EPA) report, more than 90% of greenhouse gas emissions come from the combustion of fossil fuels [2]. Pollutants from internal combustion (IC) engines include CO, unburned fuel, SO2 (sulfur oxide), NOx (nitrogen oxides), volatile organic compounds (VOCs) and particulate matter (PM). For compression–ignition (CI) engines (diesel engines), PM is a major concern. In order to retard the depletion of fossil fuels and reduce the global warming effect of CO2 emission, discovering alternative, renewable, cost effective, and reliable fuel sources to replace fossil
P.-C. Chen et al. / Fuel 103 (2013) 850–861
fuels had become an essential issue for scientists. On the other hand, optimizing combustion process in IC engines also contributes to the reduction of both fossil fuel consumption and pollutant emissions by increasing efficiency. As highly efficient IC engines, diesel engines offer superior thermal efficiency and low fuel consumption compared with spark ignited engines. They are widely used in both light duty vehicles and heavy duty trucks. The spray characteristics of diesel fuel (and its alternatives) fuel plays a critical role in the engine performance and it is essential to optimize these characteristics. Studies of the fuel spray behavior can be classified into two categories: macroscopic and microscopic. Macroscopic parameters such as spray tip penetration and cone angle can be measured by direct visualization methods. Charge-coupled device (CCD) and intensified charge-couple device (ICCD) cameras are commonly used in research labs for taking spray images with a light source. Microscopic parameters such as droplet velocity, droplet size and size distribution can be measured through particle image velocimetry (PIV), phase Doppler particle analyzer (PDPA) or laser diffraction particle analyzer (LDPA) systems [3]. 1.1. Macroscopic spray properties The breakup length of diesel spray injecting into a high-pressure gas was measured by Yule and Filipovic [4]. Their results showed the breakup length of diesel sprays were consistently about 35% of the spray penetration length. In addition, the breakup length slightly increased for a larger nozzle diameter but it decreased with increasing injection pressure. Ryu et al. compared spray experimental and simulation results [5]. They found that the spray penetration was proportional to the 1=4 power of the injection pressure, which is similar to the findings of Hiroyasu and Arai [6]. Their fuel evaporation and air fuel mixture results indicated that the influence of the ambient density on spray impingement was stronger than the effects of temperature conditions when the fuel was injected early during the compression stroke. Diesel spray formation from a common-rail multi-holed injector was studied in a high-pressure, high-temperature cell by Klein-Douwel et al. [7]. Their results showed that the growth of spray penetration was proportional to tb where t is the time after injection, and b = 0.57 ± 0.02. The spray cone angle was not constant, but varied with time during an injection, mainly as a result of spray shape changes. An investigation of the effect of injection pressure on the macroscopic spray characteristics was performed by Delacourt et al. [8]. The experimental results showed that the spray tip penetration increased with an increase in injection pressure. The spray angle was slightly influenced by injection pressure and remained nearly constant during the entire injection process. Delacourt et al. also evaluated the spray volume by combining the information of spray tip penetration and spray angle, and observed that spray volume increased with injection pressure but seemed to reach a limit. Suh et al. [9] studied the effects of injection duration and ambient pressure on diesel spray characteristics. Their results illustrated that spray tip penetration increased with injection pressure and duration time. They also indicated that spray tip penetration decreased as ambient pressure increases, while spray cone angle was enhanced by increase in ambient pressure. Alam et al. [10] investigated spray and combustion in an optical turbocharged direct-injection (DI) diesel engine. As a result, the start of injection (SOI) and start of combustion (SOC) of neat biodiesel were the earliest compared with diesel and blended fuels. These were mainly due to the higher density and bulk modulus of biodiesel compared to diesel fuel. Spray characteristics of diesel and biodiesels under ultra-high injection pressure (up to 300 MPa) were studied by Wang et al. [11]. This study showed that biodiesel
851
had longer spray tip penetration than diesel, but the difference became smaller at higher ambient density (>30 kg/m3). The spray cone angles of biodiesel were smaller than that of diesel, generating smaller spray area and volume. Gao et al. found that although biodiesels were produced from different inedible oils, their viscosities and densities, and therefore their macroscopic spray characteristics, were similar [12]. Spray tip penetration and spray velocity increased with increase in the blending ratios of biodiesel, but the spray cone angle decreased. This indicated a reduction in spray atomization quality while increasing biodiesel blending ratios. Fang et al. investigated the spray and combustion of diesel and biodiesel blends in an optical diesel engine using a commonrail fuel injection system [13]. Longer spray penetration was observed for blends with more biodiesel content and lower flame luminosity and higher NOx emissions were found for biodiesel fuels. However, applying advanced multiple-injection strategies could reduce the emissions of biodiesel fuels under clean diesel combustion modes [14]. In addition, due to the ‘‘one fuel forward’’ policy adopted by the US military, application of jet fuels in diesel engines is important for ground vehicles [15]. Spray tip penetration of JP-8 were measured in an optical combustion vessel and compared with diesel fuel by Pickett and Hoogterp [16]. Results showed that the liquid-phase penetration of JP-8 was shorter than that of diesel due to the lower boiling point of JP-8. However, the vapor penetration rate of JP-8 was not much different from that of diesel. Lee et al. [17] used the Mie-scattering method to analyze the macroscopic spray characteristics of JP-8 and diesel in an optically-accessible single-cylinder heavy duty diesel engine equipped with a high-pressure common-rail injection system. It was found that JP-8 had a shorter spray tip penetration and wider spray angle than diesel fuel mainly due to the faster vaporization characteristics of JP-8. 1.2. Microscopic spray properties An experimental investigation of diesel and coal-water slurry Sauter mean diameters (SMDs) near the spray tip region was conducted by Kihm and Terracina [18]. Their results showed that average SMDs of both fuels increased with increasing ambient pressure and distance from the nozzle but decreased significantly with increasing injection pressure. No significant effects of nozzle diameters on SMD were found with nozzle diameters between 0.2 and 0.4 mm. Suh et al. [19] studied the effects of injector driver type on spray and atomization characteristics using a PDPA system. They concluded that piezo-driven injection system had a smaller droplet size than solenoid-driven injection system due to the more rapid opening of injector nozzle and piezo-driven injector generated higher spray velocity at the beginning of injection. Labs and Parker [20] characterized the interior properties of diesel spray near the injector nozzle. Their results indicated that SMD increased with radial distance and axial distance from the nozzle tip. Thus, the smallest droplets were on axis and close to the injector tip. An analytical study of atomization characteristics of seven different kinds of biodiesels were carried out by Ejim et al. [21]. Results from statistical analysis showed that pure coconut biodiesel had similar atomization characteristics to No. 2 diesel, because of its similar properties, i.e., density, surface tension, and viscosity. Other biodiesels, such as palm, soybean, cottonseed, peanut and canola, had slightly larger droplet sizes compared with diesel. Allen et al. mathematically analyzed and compared the atomization characteristics of 15 biodiesel fuels by predicting their viscosities and surface tensions directly from their fatty acid compositions [22]. According to their comparison, all kinds of biodiesels have similar atomization characteristics, except for biodiesel derived from coconut and rapeseed oils. Gao et al. conducted experiments of the spray characteristics of inedible oils (such as jatropha oil,
852
P.-C. Chen et al. / Fuel 103 (2013) 850–861
palm oil and used cooking oil) blended fuel and compared with #0 diesel [12]. SMD of blended fuels was greater than diesel, and the spray was more concentrated, due to the higher viscosity and surface tension of biodiesel, compared with conventional diesel. However, at lower blending ratios, the spray properties of biodiesel were similar to diesel. Choi et al. investigated the droplet properties under different blending ratios of biodiesel (palm oil) [23]. They found that the peak of the droplet size volume frequency distribution increased with increasing injection pressure but the peak decreased while increasing the blending ratio. Furthermore, the cumulative volume distribution moved toward the large droplet direction when increasing the blending ratio. As the biodiesel blending ratio increased, the SMD, mass median diameter (MMD), and span factor (which is a dimensionless parameter indicative of the uniformity of the drop size distribution) increased. The deviation between spray size distribution and mean diameter became larger while increasing blending ratio, i.e., the spray size was distributed in a wider range than that of common diesel. Mao et al. [24] compared the SMD of diesel, Jet-A and JP-8 under different injection pressures with a swirl atomizer. The results indicated the SMDs of Jet-A and JP-8 are both smaller than that of diesel fuel. The smaller SMDs of jet fuels were mainly due to their relatively lower viscosity than diesel fuel. The droplet size change among different fuels when exposed to a hot atmosphere (800 K) at 1 bar ambient pressure was demonstrated by Prommersberger et al. [25]. Their result showed that the droplet sizes of both Jet A and JP-4 decreased faster than that of diesel due to their lower boiling points. In this study, we compared the differences of commercial No. 2 diesel fuel (D) and several alternative fuels by investigating their spray and atomization characteristics. The alternative fuels we studied are biodiesel derived from waste cooking oil (B100), 20% of biodiesel blend (B20), renewable diesel fuel (R) synthesized at NC State University using a thermo-catalytic approach yielding a mixture of normal and iso-alkanes [26,27] and jet-A fuel (J) purchased from a local civil airport. The picture of the fuels used in this work is shown in Fig. 1 and different colors are observed for different fuels. Table 1 lists the physical properties of the five fuels used. Biodiesel has the largest values of density, viscosity and surface tension. Renewable diesel fuel has the lowest density but it is close to Jet-A, which has a low viscosity and surface tension. Renewable diesel fuel has similar viscosity and surface tension to diesel fuel. The experiments were conducted with various pressures, injection durations, and locations of spray for microscopic spray properties. Macroscopic spray properties were investigated with different injection pressures and fuels. 2. Experimental methods 2.1. Alternative fuels for diesel The conventional biodiesel chosen in this study is derived from waste cooking oil and produced by Piedmont Biofuels located in
Table 1 Physical properties of fuels. Fuel
Abbreviation
Density (g/ml) (40 °C)
Kinematic viscosity (mm2/s) (40 °C)
Surface tension (m N/m)
Diesel Biodiesel 20% biodiesel blend Renewable diesel fuel Jet A
D B100 B20 R J
0.83 0.87 0.84 0.76 0.79
2.65 4.60 3.04 2.48 1.15
30.3 34.7 27.7 30.3 28.7
Pittsboro, North Carolina. Producing biodiesel based on waste cooking oil has better environment benefits and less land usage compared with other feed stocks. These advantages are the main reasons of studying waste cooking oil biodiesel in this work. The renewable diesel fuel (biofuel) used in this study was made in the Department of Mechanical and Aerospace Engineering, North Carolina State University [26,27]. This biofuel is composed of a mixture of n- and iso-alkanes derived from canola oil using a multi-step process including hydrolysis, deoxygenation, and isomerization, yielding a fuel whose relatively lower viscosity and density make it be closer to those properties of diesel fuel than biodiesel. In addition, this new biofuel is a pure hydrocarbon fuel with no oxygen content, which is significantly different from conventional biodiesel. Jet fuel is commonly used in ground support vehicles at airports, in place of diesel, to reduce the number of different fuels for logistic reasons. It is also a basic requirement for military vehicles to run jet fuels in diesel engines under the ‘‘one fuel forward’’ policy [15]. Jet fuel has higher flash point and lower freezing point than diesel for safe usage in aircraft. The jet fuel used for our work is Jet A purchased from a local airport. It has the lowest viscosity but most pungent smell among the fuels in this study due to the higher aromatic content. 2.2. Common-rail high-pressure injection system The fuel injection system is illustrated in Fig. 2. A first generation common-rail fuel injection system was built to control and maintain the fuel pressure at a given level (up to 1350 bar). The common rail has a pressure sensor providing a measure of the pressure inside the rail. A specially machined single holed injector nozzle with a diameter of 140 lm was used in this work. The fuel pressure in the rail was controlled by a DAQ (NI 195509D-01L, National Instruments) with a PID control program built in LabView 2010. The DAQ acquired the pressure signal from the rail pressure sensor and compared the data with desired pressure set by the operator. The PID controller adjusted the opening and closing of the pressure regulator solenoid valve to maintain the rail pressure at a given level. The activation of fuel injection was controlled by a pulse/delay generator, by which we were able to control the width of injection pulse (injection duration) and output a single shot signal to the injector driver, which directly controlled the opening and closing of the injector valve. The chosen injection pressures were 300, 500, 800, and 1000 bar. The fuels were stored in room temperature at 25 °C and also injected into air at room temperature and pressure. 2.3. Droplet size measurement system
Fig. 1. Picture of fuels used in this study (from left to right: Jet-A, renewable biofuel, biodiesel, 20% biodiesel blend diesel, and diesel fuel, respectively).
The droplet size of fuel spray was measured by a laser diffraction particle analyzer (LDPA) system (SprayTecÒ, MAL 1009475, Malvern Instruments Inc.). The data acquisition rate was set at 10 kHz and a 300 mm lens was used for droplet size range of
853
P.-C. Chen et al. / Fuel 103 (2013) 850–861
Fig. 2. Schematic of the common rail fuel injection system setup.
Fig. 3. Setup of the high speed imaging system (left) and the definitions of spray tip penetration and cone angle (right).
0.1–900 lm. The fuel injector and the LDPA system were synchronously triggered by the delay/pulse generator. The measurement duration was 5 ms after the trigger, which covered the whole process of spray formation and atomization. The detector range was selected as 15–26 of 36 detectors (larger number of detector catches signal with wider scattering angle which is caused by larger droplet) to eliminate false signals due to beam steering effects and to avoid erroneous results. The LDPA instrument was located at different positions downstream of the spray to study the droplet size distribution inside the spray plume. The droplet sizes shown in this study are the average values of 10 experimental runs. The error bars of the droplet size data are the standard deviations calculated based on the 10 runs of each case.
Table 2 Experiment parameters. Equipment
Parameter
Particle size measurement system Data acquisition rate Lens Detect droplet size range Detector range Spray image acquisition system Lens Frame rate Resolution Exposure time
SprayTec, MAL 1009475 10 kHz 300 mm 0.1–900 lm 15–26 Phantom 4.3 NikonÒ 50-mm f/1.8 7312 fps 608 128 7 ls
2.4. Spray image acquisition system A high-speed digital video camera (Phantom 4.3, Vision Research Inc.) was used to capture the images of the fuel spray. The frame rate was set at 7312 frames per second at a resolution of 608 128 with the exposure time of 7 lm in this study. A 1000 W light source was used to volumetrically illuminate the spray plume. A picture of the setup is shown in Fig. 3 (left). The high-speed camera is also synchronized with the injector using the same trigger pulse. The spray images were acquired up to 15 ms, which covered the entire spray appearance time. The spray tip penetration and spray cone angle data were calculated based on the spray images as illustrated in Fig. 3 (right). Based on the spray
tip penetration and the cone angle, spray volume was calculated using the following equation [8]:
V ¼ ðp=3ÞS3 ½tan2 ðh=2Þ
1 þ 2 tanðh=2Þ ½1 þ tanðh=2Þ3
ð3:1Þ
where S is the spray tip penetration and h is the spray cone angle. The spray tip penetration and cone angle data in this study were the averaged values derived from 5 sets of spray images for each case. The experimental parameters are listed in Table 2 for both droplet size measurements and high-speed spray imaging.
854
P.-C. Chen et al. / Fuel 103 (2013) 850–861
Fig. 4. SMD variation under different parameters: (a) injection pressure (axial distance: 8 cm, radius distance: 0 cm, injection duration: 1.5 ms); (b) spray axial distance from injector nozzle tip (injection pressure: 800 bar, injection duration: 1.5 ms, radial distance: 0 cm); and (c) radial distance from the spray axis (axial distance: 8 cm, injection pressure: 500 bar, injection duration: 1.5 ms).
Fig. 5. Derived diameter parameters of Dv10, Dv50, and Dv90 for different fuels with injection pressure: (a) Dv10, (b) Dv50, and (c) Dv90 (axial distance: 8 cm, radius distance: 0 cm, injection duration: 1.5 ms).
P.-C. Chen et al. / Fuel 103 (2013) 850–861
855
Fig. 6. Comparison of particle size distribution for different fuels under different injection pressures: (a) 300 bar; (b) 500 bar; (c) 800 bar; and (d) 1000 bar.
3. Results and discussion 3.1. Sauter mean diameter (SMD) Fig. 4a–c depict the SMD results for different conditions. Fig. 4a shows the SMDs of the different fuels under 300, 500, 800 and 1000 bar injection pressures, all with the injection duration of 1.5 ms. For the figures shown in this study, the symbol J stands for Jet-A and R stands for renewable diesel fuel. Conventional biodiesel and 20% biodiesel blend are represented by B100 and B20,
and D is for diesel. It is clear from Fig. 4a that SMDs decrease with an increase in the injection pressure for all the fuels. The SMD difference among the fuels is mainly due to the differences in their viscosity and surface tension (Table 1). A higher viscosity leads to a lower fuel jet velocity, leading to larger droplet size. A lower surface tension makes the spray easier to break up into small droplets. According to Fig. 4a, biodiesel has much bigger droplets than diesel due to its highest viscosity and surface tension among the fuels. The viscosity of B20 is higher than diesel but its surface tension is lower. These combined fuel properties affect the SMD of B20
856
P.-C. Chen et al. / Fuel 103 (2013) 850–861
by showing a close value to diesel at lower pressures (<500 bar). Jet-A has the smallest droplet size due to its lowest viscosity and surface tension as well. Renewable diesel fuel has a similar surface tension and viscosity to diesel; however, it has smaller droplet size than diesel and is closer to Jet-A, particularly under high injection pressure. This is a very interesting result indicating unique atomization characteristics of the newly developed biofuel. According to the results in Fig. 4a and fuel properties in Table 1, the current results also show that the effect of viscosity is more significant on the SMD than that of surface tension due to the fact that viscosity variations among fuels are higher than that of surface tension. Locating the laser beams of the LDPA system at different positions can help increase the level of understanding of the droplet
sizes distribution within the spray plume. Fig. 4b and c demonstrate the SMD changes at different spray axial and radial distances. The SMDs increase with increasing distance between the measuring point and the injector nozzle tip due to momentum loss along the penetration development. SMDs get larger while moving away from the nozzle tip. Fig. 4c shows the SMDs at different radial distances from the axis of the spray. The smallest SMDs are found along the axis of sprays and increase about 1 lm at the spray peripheries. This phenomenon is believed to be due to the spray pressure dropping to ambient pressure on the peripheries. At a radial distance of 1 cm, it is already outside of the spray periphery, and the sprays become loose and dilute. Therefore, their droplet sizes decrease again. These results are consistent with findings in Ref. [20].
Fig. 7. Spray tip penetration for different fuels under different injection pressures: (a) 300 bar; (b) 500 bar; (c) 800 bar; and (d) 1000 bar.
P.-C. Chen et al. / Fuel 103 (2013) 850–861
3.2. Derived parameters The derived parameters for droplet size include Dv10, Dv50, and Dv90 as shown in Fig. 5a–c. Dv50, also known as volume median diameter (VMD), is a value indicates that half the spray volume was contained in droplets that were smaller than this value while the other half were contained in droplets that were larger than this value. Similarly, Dv10 and Dv90 values indicate that 10% and 90%, respectively, of the spray volume was contained in droplets that were smaller than these values. Fig. 5a shows the Dv10 of different fuels with injection pressures. The difference between diesel and B20 for Dv10s is larger than that of SMD. Furthermore, Dv10 of Jet-A decreases more slowly than renewable diesel fuel while increasing injection pressure, thus it becomes larger than renewable diesel fuel at higher injection pressures (>800 bar). This phenomenon indicates that the effect of injection pressure on
857
droplet size distribution of Jet-A is less pronounced than that of renewable diesel fuel. Fig. 5b shows the Dv50 of different fuels under different injection pressures and the data are a little higher compared with the SMDs, which indicates that the SMD of spray is slightly smaller than the droplet size at median volume. As illustrated in Fig. 5c, large droplets increase with an increase in viscosity. Dv90s of B20 are close to diesel fuel because the composition of B20 is similar to diesel. Interestingly, the standard deviations of Dv90s were smaller than those of Dv10s and Dv50s indicating the diameter variation is relatively smaller for larger droplets. 3.3. Droplet size distribution Droplet size distribution of alternative fuels compared with conventional diesel under different injection pressures are illustrated in Fig. 6a–d. At 300 bar, biodiesel, B20 and renewable diesel
Fig. 8. Spray cone angles for different fuels under different injection pressures: (a) 300 bar; (b) 500 bar; (c) 800 bar; and (d) 1000 bar.
858
P.-C. Chen et al. / Fuel 103 (2013) 850–861
Fig. 9. Spray volume for different fuels under different injection pressures: (a) 300 bar; (b) 500 bar; (c) 800 bar; and (d) 1000 bar.
fuel shows similar distribution, slightly narrower than diesel, but the peak of volume frequency of these fuels and diesel are all about 16 lm. Jet-A shows a distinct distribution in smaller diameter regions and its peak is located at about 13 lm for 300 bar injection pressure. At 500 bar, the droplet size distribution of biodiesel and B20 are still narrower than diesel but that of renewable diesel fuel becomes closer to diesel fuel, while Jet-A remains in the smallest region. All distributions of different fuels at 500 bar injection pressure show a trend of moving toward smaller diameters compared with 300 bar. For 800 bar, all the fuels have similar droplet size distributions within a smaller range (from 1 lm to 30 lm) with peaks around 10 lm. This is the reason 800 bar injection pressure was
chosen as the standard condition for the experiments that follow. The variation of droplet sizes of different fuels is the smallest under 800 bar injection pressure; therefore, the effect of changing other experimental conditions on droplet size can be clearly distinguished. Comparing with 800 bar, under the highest injection pressure used in this study (1000 bar), the peaks of renewable diesel fuel, diesel and Jet-A further go down into smaller diameter regions while this change is not obvious for biodiesel and B20. In summary, the droplet sizes of all the fuels are distributed in a relatively smaller region with increasing injection pressure. From these results, it is interesting to note that although B20 has a similar composition and SMD of diesel fuel, it shows similar
P.-C. Chen et al. / Fuel 103 (2013) 850–861
frequencies of larger droplet diameter distribution to biodiesel under lower injection pressures (i.e., 300 and 500 bar). 3.4. Spray tip penetration Spray tip penetrations of all the fuels at different injection pressures are shown in Fig. 7a–d. Tip penetrations of the five fuels increase with time after being injected from the injector and a higher injection pressure promotes this increase. This trend can be easily observed by looking at diesel penetration under 300 bar in Fig. 7a. At 1.5 ms after injection, the spray tip penetration is 10 cm compared with 14 cm under 1000 bar in Fig. 7d. Biodiesel shows significantly longer penetration than other fuels at low injection pressure (300 bar) but the difference becomes small at higher injection pressures. This may be attributed to the fact that the relatively higher density and bulk modulus of biodiesel increase the velocity of the fluid exiting the nozzle compared with the other fuels, thus increasing injection pressure has less effect on biodiesel to enhance the growth of spray tip penetration. Another interesting phenomenon is that B20 has shorter penetrations than diesel under all injection pressures. This might be due to the higher viscosity of B20 raising the friction between fuel and the injector nozzle surface. However, for pure biodiesel, B100, this friction effect is
Injection Pressure: 300 bar (cm) J R B100 B20 D
859
overwhelmed by its large density and bulk modulus, which make biodiesel have the longest penetration. Renewable diesel fuel and Jet-A have shorter spray tip penetrations than diesel and biodiesel due to their smaller densities. Furthermore, it can be seen from the figures that the spray tip penetration of the renewable diesel fuel is slightly shorter than Jet-A. This phenomenon may be caused by the effect of the significantly smaller viscosity of Jet-A (1.15 mm2/s) compared with renewable diesel fuel which has a viscosity at about 2.48 mm2/s. A lower viscosity reduces the friction between fuel and the surface of injector nozzle hole and yields a higher fluid velocity exiting from the injector nozzle. 3.5. Spray cone angle Fig. 8a–d illustrate the spray cone angle results under different injection pressures with time after the start of injection. As the spray penetrates, the droplets on the boundaries become smaller and diffuse easily, generating a decreasing trend of spray cone angle. The spray cone angles decrease rapidly after injection to about 10° for each injection pressure, however, at different rates. As seen from these figures, the spray cone angles of the five fuels converge to a constant value (steady state of the spray) at around 0.8 ms after start
Injection Pressure: 500 bar R B100 B20 D (cm) J
Fig. 10. Comparison of spray development images with time after injection trigger signal at 300 bar (left) and 500 bar (right) injection pressure at different times 0f 0.7 ms, 1.0 ms, and 1.5 ms (from top to bottom) after the injection trigger.
860
P.-C. Chen et al. / Fuel 103 (2013) 850–861
Injection Pressure: 800 bar (cm) J R B100 B20 D
Injection Pressure: 1000 bar (cm) J R B100 B20 D
Fig. 11. Comparison of spray development images with time after injection trigger signal at 800 bar (left) and 1000 bar (right) injection pressure at different times 0f 0.7 ms, 1.0 ms, and 1.5 ms (from top to bottom) after the injection trigger.
of injection at 300 bar injection pressure, and steady state is achieved sooner after injection as the injection pressure increases, i.e., 0.7 ms for 500 and 800 bar. Under 1000 bar injection pressure, the spray cone angles converge at 0.6 ms after injection. These results show that the distinctions between the different fuels after achieving steady state decrease with increasing injection pressure. More obvious distinctions between the different fuels can be seen before their convergence. The spray of biodiesel starts with the smallest cone angle, resulting from its highest viscosity. From the above observations, it is found that increasing the biodiesel blend ratio reduces the cone angle of fuel spray due to the increase of viscosity. As for renewable diesel fuel and Jet-A, the effect of density is more significant than the effect of viscosity and makes the cone angle of Jet-A smaller than the renewable diesel fuel under all injection pressures. 3.6. Spray volume Fig. 9a–d shows the calculated spray volumes of different fuels used in this study with time after the start of injection. The spray volumes develop as exponential curves under low injection pressure (300 bar) and become proportional when increasing the pressure. This is considered as the effect of higher injection pressure delivering the fuel faster than low pressure and causing the sprays
to reach steady state sooner. Larger spray volumes can be found at higher injection pressure by comparing Fig. 9a and d. For example, spray volume of diesel increases from 9 cm3 to 19 cm3 at 1.5 ms after the start of injection trigger as the injection pressure is increased from 300 to 1000 bar. The distribution in spray volume for the different fuels follows the same trend as spray tip penetration, which is biodiesel, diesel, B20, Jet-A, and then renewable diesel fuel, from large to small, respectively. According to this result, it is seen that spray volume is mainly determined by the spray penetration length, since the variation of spray cone angles among different fuels is not significant. Compared with diesel fuel, Jet-A and renewable diesel fuel have smaller spray volumes due to their shorter spray penetration. The results of B20 are interesting compared with diesel and B100. As a blended fuel, a common expectation is that B20 would show spray and atomization characteristics between diesel and B100. Our results show that B20 has a spray volume less than diesel fuel; the same trend is also observed in the spray tip penetration. 3.7. Spray image The spray images acquired with the high-speed camera capture the transient shape of sprays at different timings. These images are
P.-C. Chen et al. / Fuel 103 (2013) 850–861
sorted in Figs. 10 and 11 by three time periods (0.7, 1.0, and 1.5 ms) after the injection trigger signal. It can be seen from the figures that biodiesel has the largest penetration length, spray projected area, and thus the largest spray volume. This phenomenon is most obvious in the images at 1.0 ms after injection trigger. At this time point, the spray has already achieved steady state and develops at a constant velocity. This trend can be observed by comparing the volume differences at 0.7 ms and 1.5 ms with 1.0 ms. The penetration slows down and the perturbation area increases at 1.5 ms due to interaction with ambient air. An interesting characteristic of these fuel sprays is that the perturbation regions of the spray occurs earlier at higher injection pressures by comparing the images under different pressures at 0.7 ms after injection. This effect is due to the higher spray speeds under high injection pressures making the sprays penetration and interaction with air faster and stronger. In addition, the spray images show a shorter spray penetration of B20 compared with diesel fuel, which is consistent with our previous discussion of spray tip penetration and spray volume. 4. Conclusions In this work, the spray and atomization characteristics were investigated for diesel fuel and its four alternatives, including biodiesel derived from waste cooking oil, blend of biodiesel and diesel, a non-oxygenated hydrocarbon biofuel, and Jet-A fuel. Both microscopic and macroscopic parameters were measured. The observations and findings can be summarized as follows: 1. Biodiesel and its blend lead to larger droplets compared with diesel. However, Jet-A and renewable diesel have smaller droplet sizes due to their relatively lower viscosity and surface tension compared with diesel fuel. Increasing injection pressure is effective at reducing the droplet size. For example, the droplet size decreased 50% from 300 bar to 1000 bar injection pressure for all the fuels studied; 2. The current results further confirm that droplets at the spray periphery have larger diameters than those in the center of the spray due to the faster pressure drop at the spray periphery. Droplet diameters become larger as the distance from the nozzle tip increases due to pressure drop along the spray axial distance; 3. Under higher injection pressures, small droplets distribute in a narrower region and have less variation of their diameters; 4. Results from the measurement of spray tip penetration show a combined effect of density, bulk modulus and viscosity. B100 has the longest penetration of the fuels investigated. Longer penetration of fuel spray is observed under higher injection pressure. For example, the spray penetrations at 1000 bar injection pressure are about 40% longer than those at 300 bar, and this increase is more pronounced for Jet-A and renewable diesel; 5. Effect of injection pressure on spray cone angle is not evident compared with that on droplet size and spray tip penetration. Spray cone angle converges to its steady state very rapidly with increasing injection pressure; 6. Both microscopic and macroscopic spray properties influence the atomization quality of a certain fuel. There is a trade-off between smaller droplet size and larger spray volume as they both are determined by the properties of the fuel. The renewable diesel stands out among the biofuels by showing smaller droplet sizes (SMD, Dv10, Dv50, and Dv 90) than diesel. Using jet fuel in diesel engines might yield better atomization due to its small droplet size and similar spray volume to diesel under different injection pressures.
861
Acknowledgements This material is based, in part, upon work supported by the Faculty Research and Professional Development (FRPD) Fund from the North Carolina State University, the Natural Science Foundation under Grant No. CBET-0854174, and the National Science Foundation EFRI Program under Grant EFRI-093772. The authors also thank Dr. Brian Farkas and Dr. Allen Foegeding of the Department of Food Science at the North Carolina State University for their help to measure the fuel properties. References [1] BP Amoco Statistical Review of World Energy, June 2007. [2] Inventory of US greenhouse gas emissions and sinks: 1990–2009. US Environmental Protection Agency; 2011. [3] He C, Ge Y, Tan J, Han X. Spray properties of alternative fuels: a comparative analysis of biodiesel and diesel. Int J Energy Res 2008;32:1329–38. [4] Yule AJ, Filipovic I. On the break-up times and lengths of diesel sprays. Int J Heat Fluid Flow 1992;13:197–206. [5] Ryu J, Kim H, Lee K. A study on the spray structure and evaporation characteristic of common rail type high pressure injector in homogeneous charge compression ignition engine. Fuel 2005;84:2341–50. [6] Hiroyasu H, Arai M. Structure of fuel sprays in diesel engines. SAE paper 900475; 1990. [7] Klein-Douwei RJH, Frijters PJM, Somers LMT. Macroscopic diesel fuel spray shadowgraphy using high speed digital imaging in a high pressure cell. Fuel 2007;86:1994–2007. [8] Delacourt E, Desmet B, Besson B. Characterization of very high pressure diesel sprays using digital imaging techniques. Fuel 2005;84:857–67. [9] Suh HK, Park SW, Lee CS. Effect of piezo-driven injection system on the macroscopic and microscopic atomization characteristics of diesel fuel spray. Fuel 2007;86:2833–45. [10] Alam M, Song J, Zello V. Spray and combustion visualization of a directinjection diesel engine operated with oxygenated fuel blends. Int J Engine Res 2006;7:503–21. [11] Wang X, Huang Z, Kuti OA. Experimental and analytical study on biodiesel and diesel spray characteristics under ultra-high injection pressure. Int J Heat Fluid Flow 2010;31:659–66. [12] Gao Y, Deng J, Li C. Experimental study of the spray characteristics of biodiesel on inedible oil. Biotechnol Adv 2009;27:616–24. [13] Fang T, Lin YC, Foong TM, Lee CF. Biodiesel combustion in an optical HSDI diesel engine under low load premixed combustion conditions. Fuel 2009;88(11):2154–62. [14] Fang T, Lee CF. Biodiesel effects on the combustion processes in an HSDI diesel engine employing advanced multiple injection strategies. Proceedings of the Combustion Institute 2009;32(2):2785–92. [15] DOD 4140.25-M, DOD Directive 4140.25; 2004. [16] Pickett LM, Hoogterp L. Fundamental spray and combustion measurements of JP-8 at diesel conditions. SAE Int J Commer Veh 2009;1(1):108–18. [17] Lee CS, Park SW, Kwon SI. An experimental study on the atomization and combustion characteristics of biodiesel-blended fuels. Energy Fuels 2005;19:2201–8. [18] Kihm KD, Terracina DP. Synchronized droplet size measurements for coalwater slurry sprays generated from a high-pressure diesel injection system. Department of Mechanical Engineering, Texas A&M University; June 1993. [19] Suh HK, Roh HG, Lee CS. Spray and combustion characteristics of biodiesel/ diesel blended fuel in a direct injection common-rail diesel engine. J Eng Gas Turbines Power 2008;130(3):032807. [20] Labs J, Parker T. Two-dimensional droplet size and volume fraction distributions from the near-injector region of high-pressure diesel sprays. Atomization Spray 2006;16:843–55. [21] Ejim CE, Fleck BA, Amirfazli A. Analytical study for atomization of biodiesels and their blends in a typical injector: surface tension and viscosity effects. Fuel 2007;86:1534–44. [22] Allen CAW, Watts KC. Comparative analysis of the atomization characteristics of fifteen biodiesel fuel types. Transactions ASAE 2000;43(2):207–11. [23] Choi S, Oh YT. The study on spray behavior and atomization characteristics of biodiesel. Adv Mater Dev Perform (AMDP2011) Int J Mod Phys: Conf Ser 2012;6:419–24. [24] Mao CP, Short J, Buelow P. Innovative injection and mixing systems for diesel fuel reforming. SECA 6th annual, workshop; 2005. [25] Prommersberger K, Maier G, Wittig S. Validation and application of a droplet evaporation model for real aviation fuel. RTO MP-14, RTO AVT symposium on ‘‘gas turbine engine combustion, emissions and alternative fuels’’, Lisbon, Portugal; October 12–16, 1998. [26] Roberts WL, Lamp HH, Stikeleather LF, Turner TL. Process for conversion of biomass to fuel. US Patent#: US7816570B2, 2010. [27] Wang WC, Thapaliya N, Campos A, Stikeleather LF, Roberts WL. Hydrocarbon fuels from vegetable oils via hydrolysis and thermo-catalytic decarboxylation. Fuel 2012;95:622–9.