Journal of Mechanical Science and Technology 25 (4) (2011) 885~893 www.springerlink.com/content/1738-494x

DOI 10.1007/s12206-011-0142-4

Qualitative and quantitative analysis of spray characteristics of diesel and biodiesel blend on common-rail injection system† Ainul Ghurri1,2,*, Kim Jae-duk2, Song Kyu-Keun3, Jung Jae-Youn4 and Kim Hyung Gon5 1

Department of Mechanical Engineering, Udayana University, Bali-Indonesia Graduate School, Precision Mechanical Eng., Chonbuk National University, Jeonju, 561-756, Korea 4 RCIT, Chonbuk National University, Jeonju, 561-756, Korea 5 CAMS Tech Co. Ltd., 726 Palbok-dong 2ga, Duckjin-gu, Jeonju, 561-844, Korea

2

(Manuscript Received June 22, 2010; Revised October 23, 2010; Accepted December 14, 2010) ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

Abstract An experimental study was performed on spray characteristics of spray diesel (D100) and biodiesel blend (BD65) injected into an atmospheric chamber. A qualitative analysis of spray images was conducted through exploiting the image processing with common image processing software. The results showed that the posterization of the images offered more detailed qualitative information on the spray compared to the more commonly-used threshold method. The posterized images showed the existence of layers in the spray with its transition at different grey levels. At lower injection, the spray tip penetration of BD65 was slightly lower than D100, whereas at high injection pressure, spray tip penetration of BD65 was higher than D100. Although BD65 had lower maximum velocity, the higher density of biodiesel may have resulted in greater momentum that enabled BD65 to have longer spray tip penetration at higher injection pressure. At higher injection pressure, the spray angle of BD65 tended to be less than that of D100. Keywords: Biodiesel; Common-rail injection; Image processing; Spray characteristics ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

1. Introduction Biodiesel has obtained much attention as an alternative fuel to reduce greenhouse and combustion emissions. There has been a wide-spread study on biodiesel-fuelled engine performance and combustion emission. The differences in physical properties between diesel and biodiesel fuel led to the study of biodiesel’s spray characteristics. Grimaldi and Postrioti [1]—among the first to study this topic—compared the characteristics of spray penetration length and cone angle of the fuel spray generated by a common-rail injection system in an atmospheric chamber. They reported that, generally, biodiesel (rapeseed oil) was characterised by greater penetration length and lower spray angles compared to diesel fuel. Kim et al. tested biodiesel (soybean oil) and DME under different ambient pressure [2]. Their results for the spray injected into the atmospheric chamber condition were comparable in terms of spray tip penetration to that of Grimaldi and Postrioti [1]. Meanwhile, the spray cone angle tended to have a constant value, which differed from the †

This paper was recommended for publication in revised form by Associate Editor Kwang-Hyun Bang * Corresponding author. Tel.: +82 63 270 4267 E-mail address: [email protected] © KSME & Springer 2011

fluctuated result of Grimaldi and Postrioti [1]. Yamane et al.’s paper showed a temporal change of non-evaporated spray for rapeseed and corn oil biodiesels at the early time of fuel injection (first 0.2 ms) that were shorter than the diesel fuel [3]. Zhao et.al. investigated the effect of injection duration and back pressure on spray penetration length and spray cone angle and reported that both spray penetration length and angle were greater than those of diesel fuel [4]. Lee et al. found that biodiesel had similar spray tip penetration length and higher SMD compared to conventional diesel fuel [5]. Su et al. studied experimental and numerical analysis of spray-atomization characteristics of biodiesel fuel in various ambient temperature conditions and concluded that the spray tip penetration had almost the same pattern regardless of the variation in fuel properties caused by change in the fuel temperature [6]. Yuan et al. [7], in their experimental study of the microscopic and macroscopic characteristics of inedible oil biodiesel, found results similar to those of Grimaldi and Postrioti [1]. Spray image processing is a part of spray analysis procedures that is time-consuming and requires extra care but results in less detailed discussion, especially when produced by the ‘manual’ way that uses image processing software. Delacourt et al. developed software that enables automatic processing of the spray images to obtain macroscopic spray

886

A. Ghurri et al. / Journal of Mechanical Science and Technology 25 (4) (2011) 885~893

characteristic without intervention of the experimenter [8]. Shao et al. also developed software to measure macroscopic spray characteristics and compared the results to those obtained by the manual method [9]. Larsson was among the few authors that mentioned the use of a threshold 10 out of 255 in his spray images to suppress the noise without omitting the essential parts of the image [10]. Morgan et al. [11] applied a threshold to pick out the spray outline from its background. They applied the same threshold value for a series of images in the same batch. The threshold value was chosen by directionally comparing to the spray images. Since the manual method is still used by many researchers, the exploitation of this method is still needed to find the most convenient way to process spray images. This article reports both the examination results of spray image processing using a common image processing software and the results from investigating the spray characteristics of diesel and biodiesel blends injected into an atmospheric chamber.

Table 1. Fuel properties. Fuel properties Density (kg/m3) Viscosity (mm2/s) Flash point (oC) Pour point (oC) Cetane number

Diesel (D100) 850 3.25 68 -35 ÷ -15 54.6

Biodiesel 884 4 139 0 54 ÷ 56

Table 2. Experimental condition. Injection system Injector needle Hole diameter Spray cone angle Injection duration Injection pressure Ambient gas Ambient temperature High speed camera Frame rate Resolution

Common-rail 5 side holes 0.2 mm 156o 1000 μs 30, 60, 90 MPa Atmospheric air 283 K FASTCAM Ultima40k 18000 fps 256 x 64

2. Experimental procedures and image processing 2.1 Experimental procedures Fig. 1 displays the experimental set up consisting of a fuel tank, feed pump, fuel filter, high pressure pump driven by DC motor, common rail injection system, and an electronic high speed camera to capture the spray images. The fuel properties and experimental conditions are described in Table 1 and 2. Fuel temperature in the fuel tank was maintained constantly. Fuel was fed by a feed pump to pass a fuel filter and drawn by a DC motor-driven high pressure pump. The 3.7 kW DC motor could be adjusted up to 3000 rpm in speed. The high pressure fuel was then delivered to the injector via the common rail. One part of the fuel was injected into a quiescent atmospheric chamber at room temperature; the other part controlled the injection nozzle of the injector and then flowed back to the tank. Pressure sensors measured the fuel pressure in the rail. This signal was compared to a desired value adjusted in a controller drive (TEMS, TDA 1100H). The injection could be controlled internally or externally by an injector drive (TEMS, TDA 3200H). The injected fuel quantity was determined by the duration of injection, fuel pressure in the rail, and the flow area of the injector. Spray images from bottom view were captured by a high speed camera (FASTCAM Ultima40k) with a metal-halide lamp as a light source. A multi-channel digital delay/pulse generator and an image grabber connected to a computer were used to synchronize the fuel injection and camera shutter signal. A five-hole nozzle injector having 0.2 mm diameter and a 156o spray angle was used in this experiment. The spray images were taken at a shutter speed of 18000 FPS and used black as the background screen. The present work performed the experiments at injection pressure 30, 60, and 90 MPa, respectively, and 1000 μs of injection duration for each injection pressure.

Fig. 1. The experimental apparatus.

2.2 Image processing Image processing involves automating and integrating a wide range of processes and representations used for vision perception [12]. The outcomes are useful for various purposes. Generally, image processing has three steps: data acquisition, processing, and interpretation [12]. The image acquisition is described in Section 2.1. Image processing was then conducted in order to distinguish and segment the spray from its background. The boundary between the spray and its background can be interpreted based on the level of segmentation process. Threshold is the most commonly-used segmentation method. This is time-consuming due to the need to compare the threshold result to the original spray image until obtaining the optimum result; and subjectivity is unavoidable in this process. The threshold level on a spray image is not always transferrable to other images at the same sequence of injection. The present work performed a rather different approach in the segmentation process. Two types of segmentation methods—threshold and posterization—were applied at different

A. Ghurri et al. / Journal of Mechanical Science and Technology 25 (4) (2011) 885~893

887

(a) Posterize 12 (a) D100

(b) Posterize 20 (b) BD65 Fig. 2. Spray image series (60 MPa; 1000 μs).

(c) Threshold 18 out of 225

Fig. 3. Geometrical characteristics of spray. (d) Original image

levels to a series of spray images. Thresholding is an image adjustment that converts all shades of gray to pure black or to pure white; any shades of gray brighter than the threshold value will become white, and any shades darker than the threshold value will become black [13]. Instead of checking each image in the detailed way explained above, the present work selected a certain level of threshold after comparing to a single spray image and applied the level to all images in the same spray series. Posterization breaks up a smooth transition of colors into visible steps of solid color. This is often called stair-stepping, or banding, when referring to a gradient [13]. The qualitative and quantitative results in terms of spray tip penetration and spray angle were then compared and analyzed.

3. Results and discussion 3.1 Qualitative analysis Fig. 2 shows an example of the spray image series of D100 and BD65 fuels at injection pressure 60 MPa and an injection duration of 1000 μs during the first 1.4 ms after start of injection (ASOI). Each spray image was then processed qualitatively and quantitatively. The spray tip penetration and spray

Fig. 4. Thresholded, posterized, and original images.

angle, commonly mentioned as macroscopic or geometrical characteristics, are defined as in Fig. 3. Spray penetration length is determined by finding the furthest spray pixel from the nozzle. The near angle refers to the Grimaldi and Postrioti definition [1], i.e. the angle that includes the spray structure from the nozzle up to 1/3 of the penetration. The linear lines used to measure the near angle were the tangents to the contour that existed until the tip of spray. The far angle is the angle between the tangents to the spray envelope in the region between 1/3 to 2/3 of penetration. As mentioned in the previous section, in order to find the boundary between the spray and surrounding air, the spray images are processed by threshold and posterization segmentation methods. Fig. 4 shows the result of thresholding and posterization at different levels compared to the original image. Note that, in those experiments, the spray was injected into an atmospheric chamber with a black screen background. After posterization, the spray images were inverted to get a clearer boundary between the spray and its surrounding noise Compared to the thresholded images, the posterized ones

888

A. Ghurri et al. / Journal of Mechanical Science and Technology 25 (4) (2011) 885~893

showed the existence of layers in the fuel spray with transitions at different grey levels. This can be interpreted as a sequence of fuel sprayed during injection. Higher levels of posterization showed finer in the furthest layer than in lower level. Consequently, posterization results in difference in measurement. At this point, the posterized image must be carefully interpreted. Posterization might show the fuel layer that starts to evaporate or decay into the surroundings. The angle measurement of posterized spray images was more cumbersome than of the thresholded images. The waviness and unstable outermost layer complicated the selection of the spray angle. Particularly at high levels of posterization, the outermost spray layer that broke down and dissipated into the surrounding air was detected. As illustrated, at the level of posterization 12 the images had 11 spray layers, while at level 20 they had 19 layers. Finer transition images occur at higher posterization levels. We found more detailed qualitative information compared to thresholded images. For example, at posterization level 12, both D100 and BD65 resulted in 11 layers. The grey transition inside the images clearly showed that BD65 had a more compact shape. Further interpretation shows that the spray angle of BD65 tends to be lower than that of D100 fuel. At the outermost edges of the spray, BD65 had a larger scale of unstableness compared to D100. Fig. 5 shows different levels of posterization with different spray tip penetration length. More attention is required to interpret this image. Even though the spray images at higher levels of posterization are more similar to the original image; it might not produce better measurement result. Fig. 6 illustrates the measurement result on spray tip penetration with different levels of posterization and thresholding diesel fuel at injection pressure 60 MPa. The variation in measurement results were occurred up to 0.16 cm at 1 ms ASOI, and developed up to 0.5 cm at 1.4 ms ASOI. From this, it was predicted that the higher level of posterization resulted in longer spray tip penetration. Measurement results at posterization level 12 were very close to those obtained from thresholding. However, the tresholding results of the spray images at early injection time were very coarse and it was very difficult to determine the boundary between the spray and the surrounding air. This process required more time and care. When compared to the calculation results based on Hiroyasu equation [14], the measurement at posterization level 12 showed the closest result. Similar results were shown at other tested injection pressures and for BD65. In terms of spray angle, level of posterization made a difference in the measurement results in the range of 2o-4o. The higher level of posterization might not have resulted in higher far spray angle though it had more layers, as shown in Fig. 4 and 5, due to the outermost layer decaying into the surroundings. After examining the measurement results as shown in Fig. 6 and comparing them to the calculations based on Hiroyasu equations, we decided to use the measurement at level posterization 12 in further analysis.

Fig. 5. Spray tip penetration at different level of posterization.

Fig. 6. Spray tip penetration according to level of segmentation method.

Fig. 7. Spray tip penetration D100 and BD65 at various tested injection pressure.

3.2 Macroscopic spray characteristic analysis 3.2.1 Spray tip penetration Fig. 7 displays the comparison of spray tip penetration of both tested fuels i.e. D100 and BD65 at injection pressure 30, 60, and 90 MPa, respectively. From the figure, it can be seen that the spray tip penetration for the three tested fuels are almost same at all injection pressures. In spite of that, there are some interesting observation results, which are discussed below. The first image was captured at 0.6, 0.2 and 0.06 ms after start of injection (ASOI) at 30, 60 and 90 MPa, respectively. In all cases the first images were too coarse to measure, but clear enough to distinguish the sprays. All first images at different injection pressures showed that the spray tip penetration of D100 as slightly higher than the BD65. By cross-checking

A. Ghurri et al. / Journal of Mechanical Science and Technology 25 (4) (2011) 885~893

with the calculated break-up time based on Eq. (1), which shows slightly higher break-up times of BD65 compared to D100, it is concluded that the BD65 drop appeared in the test chamber with a relative delay to D100. This could be attributed to the higher density and viscosity of BD65. The higher viscosity prevented the spray from breaking up. The spray penetration length results as presented in Fig. 7 highlight the influence of the injection pressure. At an injection pressure of 30 Mpa, the spray tip penetration of both tested fuels were almost same, though D100 tended to have slightly higher results than BD65. This might be due to the fuel jet’s lack of energy to spray BD65 as well as D100. At injection pressure 60 MPa, the spray tip penetration of both tested fuels were almost the same every time. Furthermore, at 90 Mpa, all the spray tip penetration of BD65 were slightly higher than those of D100 except the first appearance in the test chamber. It showed more obviously the role of the biodiesel viscosity. At higher injection pressure the fuel spray had a higher kinetic energy to penetrate the surrounding air, and together with the higher viscosity of biodiesel, resulted in longer fuel penetration. With higher viscosity, the droplet size will be larger, which results in greater momentum that makes it easier to penetrate the surrounding resistance. Fig. 8 displays the measurement result on spray tip penetration of both tested fuels at injection pressure 60 Mpa compared to Hiroyasu’s empirical formula expressed in Eqs. (1)(3), and Sazhin equation [15] as formulated in Eqs. (4)-(5). Hiroyasu’s equation overestimated the spray penetration length with tendency to intersect at around 1.2 ms after start of injection to the experimental data, while the Sazhin equation predicted more closely to the experimental data. At injection pressure 90 MPa, both predicted results showed good agreement with the experimental results as shown in Fig. 9. However the intersection between the predicted result by Hiroyasu equation and the experimental data still existed at around 0.8 ms after start of injection. After the intersection, the Hiroyasu equation underestimated the spray penetration length. tb = 28.65ρ L d o ( ρ a ΔPinj ) −0.5 ⎡ ΔP ⎤ S = 2.95 ⎢ inj ⎥ ⎣ ρa ⎦

( d ot )

0.5

U in d ot (1 − α d )

U in = Cd

ρ a0.25 tan θ

2ΔPinj

ρL

(b) BD65 Fig. 8. Spray tip penetration length at injection pressure 60 compared to predicted results.

(a) D100

0.25

for t > tb

(2)

0.5

0.25

(a) D100

(1)

⎡ 2ΔPinj ⎤ S = 0.39 ⎢ ⎥ t for 0 < t < tb ⎣ ρL ⎦

S=

889

; ρ a =

ρa ρL

(3)

(b) BD65 Fig. 9. Spray tip penetration length at injection pressure 90 compared to predicted results.

(4)

(5)

3.2.2 Spray angle The comparison of far spray angle of both tested fuels at injection pressure 60 and 90 MPa are shown in Fig. 10. The far spray angle generally decreased along injection timing with small rise and fall irregularity at the final stage of injection. Both tested fuels had a relatively similar far spray angle at

890

A. Ghurri et al. / Journal of Mechanical Science and Technology 25 (4) (2011) 885~893

(a) Injection pressure 60 Mpa

(a) Injection pressure 60 Mpa

(b) Injection pressure 90 MPa (b) Injection pressure 90 Mpa

Fig. 12. Near spray angle according to injection pressure.

Fig. 10. Far spray angle according to injection pressure.

(a) D100

(b) BD65 Fig. 11. Far spray angle according to fuel type.

Both tested fuels had a relatively similar far spray angle at injection pressure 60 MPa, while at injection pressure 90 MPa the far spray angle of BD65 at the final stage of injection was lower than that of D100. This might be attributable to biodiesel content in the fuel and the kinetic energy available within the injected fuel as well as the interaction with aerodynamic resistance of surrounding air at far distances from the nozzle. The comparison between fuel types showed that, generally, far spray angle was lower at higher injection pressure, as shown in Fig. 11. The experimental result of near spray angle at injection pressure 60 and 90 MPa, are shown in Fig. 12. The near angle is measured to observe the spray evolution at a near distance from the nozzle. The behaviour of near spray angle is relatively similar compared to that of far spray angle. The near spray angle decreased with injection time. At injection pressure 60 MPa, the near spray angle of both tested fuel were almost same, while at 90 MPa the BD65 showed lower near spray angle compared to D100. Furthermore, at 90 MPa the near spray angle showed a slope change at the end stage of injection with a tendency to remain constant. The difference of near spray angles between D100 and BD65 again showed the effect of biodiesel content in the fuel. With higher viscosity, BD65 has stronger resistance to the surrounding air when injected and resulted in a denser shape of spray. Fig. 13 displays the near spray angle of each tested fuel at injection pressure 60 and 90 MPa. Unlike the far spray angle, both fuels indicated clearly that the near spray angle of BD65

A. Ghurri et al. / Journal of Mechanical Science and Technology 25 (4) (2011) 885~893

891

decreased gradually. These fluctuations occurred due to the fluctuation pressure inside the high pressure pipeline. The spray tip velocity of BD65 was similar to D100, with the maximum velocity of D100 at each injection pressure slightly higher than that of BD65. Although BD65 had lower maximum velocity, the higher density of biodiesel might have resulted in higher momentum, allowing BD65 to have longer spray tip penetration at higher injection pressure.

4. Conclusions (a) D100

(b) BD65 Fig. 13. Near spray angle according to fuel type.

Fig. 14. Spray tip velocity (D100, 60 and 90 MPa).

was lower than that of D100 during all injection times. It was observed that there were changes in spray development at different distances from the nozzle. When compared to other reports, the wide spans of spray angles shown in these results are similar to those of Grimaldi and Postrioti [1]. As a comparison, the spray cone angle of biodiesel tested by Kim et al. [2] tended to have constant value as shown at the end stage of injection pressure 90 MPa in this experiment. 3.2.3 Spray tip velocity The spray tip velocity of D100 according to the injection pressure was presented in Fig. 14. Spray tip velocity increased at the early stages of injection, and after reaching the maximum value, the velocity unstably

In this work, an experimental study was performed on characteristics of spray diesel and biodiesel blend injected into an atmospheric chamber by common-rail injection system. Exploitation of the manual method to measure spray penetration and angle by using common image processing software was put before the main analysis. By examining the results, it can be concluded that: Posterization of the spray images offered more detailed qualitative information for the spray compared to the more commonly-used threshold method. The posterized images showed the existence of layers in the fuel spray with transitions at different grey levels. However, it still requires care to select the level of posterization to get the optimum result compared to the original image. The first images at different injection pressures showed that the spray tip penetration of D100 was slightly higher than of BD65. At low injection pressure (30 MPa), the spray tip penetrations were almost the same, with that of D100 tending to vary slightly higher than that of BD65. At injection pressure 60 MPa, the spray tip penetrations of both tested fuels were almost the same all the time. At 90 Mpa, all the spray tip penetrations of BD65 were slightly higher than those of D100. Two empirical equations applied in this study to calculate spray tip penetration provide good agreement with the experimental data, especially for injection pressure 90 MPa. Both the tested fuels had a relatively similar far spray angle at injection pressures 60 MPa, while at injection pressure 90 MPa the far spray angles of BD65 at the final stage of injection were lower than those of D100. The comparison between fuel types showed that, generally, far spray angle was lower at higher injection pressure. At injection pressure 60 MPa, the near spray angle of both tested fuels were almost the same, while at 90 MPa BD65 showed lower near spray angle compared to D100 with a slope change at the final stage of injection with a tendency to remain constant. Spray tip velocity increased at the early stage of injection, and after reaching the maximum value the velocity decreased gradually. The spray tip velocity of BD65 was similar to D100, with the maximum velocity of D100 at each injection pressure slightly higher than that of BD65. Although BD65 had lower maximum velocity, the higher density of biodiesel might result in greater momentum, enabling BD65 to have longer spray tip penetration at higher injection pressure.

892

A. Ghurri et al. / Journal of Mechanical Science and Technology 25 (4) (2011) 885~893

Acknowledgment The first author thanks the Ministry of National Education of the Republic of Indonesia for supporting his doctoral study at Chonbuk National University, Korea.

Nomenclature-----------------------------------------------------------------------do S t tb Uin Cd ΔPinj ρL ρa αd θ

: : : : : : : : : : :

Nozzle diameter Spray tip penetration Time after start of injection Break-up time Initial velocity of the spray Discharge coefficient of the nozzle Injection pressure Fuel density Ambient air density Volume fraction of droplets in the spray Half angle of the spray cone

References [1] C. Grimaldi and L. Postrioti, Experimental Comparison between Conventional and Bio-derived Fuels Sprays from a Common Rail Injection System, SAE Technical paper (2000) 2000-01-1252. [2] H. J. Kim, S. H. Park and C. S. Lee, A study on the macroscopic spray behavior and atomization characteristics of biodiesel and dimethyl ether under increased ambient pressure, Fuel processing technology, 91 (2009) 354-363. [3] Y. Koji, U. Atsushi and S. Yuzuru, Influence of physical and chemical properties of biodiesel fuel on injection, combustion and exhaust emission characteristics in a DI-CI engine, The Fifth International Symposium on Diagnostics and Modeling of Combustion in Internal Combustion Engines (2001) 402-409. [4] X. W. Zhao, X. K. Han, C. He and J. W. Tan, Experimental study on spray characteristics of biodiesel oil, C. Intern Combustion Engine Eng., 1 (2008) 16-19. [5] C. S. Lee, S. W. Park and S. I. Kwon, An experimental study on atomization and combustion characteristics of biodieselblended fuels, Energy Fuels, 19 (2005) 2201-8. [6] S. H. Park, H. J. Kim, K. S. Hyun and S. L. Chang, Experimental and numerical analysis of spray-atomization characteristics of biodiesel fuel in various fuel and ambient temperatures conditions, International journal of heat and fluid flow, 30 (2009) 960-970. [7] Y. G., J. D., C. L., F. D., Z. L., Z. W. and L. L., Experimental study of the spray characteristics of biodiesel based on inedible oil, Biotechnology Advance, 27 (2009) 616-624. [8] E. Delacourt, B. Desmet and B. Besson, Characterization of very high pressure diesel sprays using digital imaging techniques, Fuel, 84 (2005) 859-867.

[9] J. Shao, Y. Yan, G. Greeves and S. Smith, Quantitative characterization of diesel sprays using digital imaging techniques, Meas. Sci. Technol., 14 (2003) 1110-1116. [10] A. Larsson, Optical studies in a DI diesel engine, SAE Technical paper (1999) 1999-01-3650. [11] R. Morgan, J. Wray, D.A. Kennaird, C. Crua and M.R. Heikal, The influence of injector parameter on the formation and break-up of diesel spray, SAE Technical Paper Series (2001) 2001-01-0529. [12] K. P. Sudheer and R. K. Panda, Digital image processing for determining drop size from irrigation spray nozzles, Agricultural Water Management, 45 (2000) 159-167. [13] B. Willmore, Definition Doctor, Digital Mastery (2006) 6-7. [14] H. Hiroyasu and M. Arai, Structures of fuel sprays in diesel engine (1990), SAE Paper 900475. [15] S.S. Sazhin, G. Feng and M.R. Heikal, A model for fuel spray penetration, Fuel, 80 (2001) 2171-2180. [16] H. K. Suh and C. S. Lee, Experimental and analytical study on the spray characteristics of DME and diesel fuels within a common-rail injection system in a diesel engine, Fuel, 87 (2008) 925-932. [17] S. L. Chang and W. P. Sung, An experimental and numerical study on fuel atomization characteristics of high-pressure diesel injection sprays, Fuel, 81 (2002) 2417–2423. [18] J. R., H. K. and K. L., A study on the spray structure and evaporation characteristic of common rail type high pressure injector in homogeneous charge compression ignition engine, Fuel, 84 (2005) 2341–2350.

Ainul Ghurri received the B.S. degree in mechanical engineering from Brawijaya University, Indonesia in 1995 and master degree in mechanical engineering from Indonesia University, Indonesia in 1998, respectively. He is a lecturer at mechanical engineering department of Udayana University, Indonesia. He is currently a Ph.D. student in Precision Mechanical Engineering Department of Chonbuk National University, Jeonju, South Korea. His research interests are in the area of fuel spray characteristics, biodiesel fuel, and its engine performance and emissions characteristics. Kim Jae Duk received the B.S. degree in Precision Mechanical Engineering from Chonbuk National University in 2009, and directly continued his study in M.S degrees. His research interests include fuel spray injection characteristics, and diesel engine performance and emission characteristics.

A. Ghurri et al. / Journal of Mechanical Science and Technology 25 (4) (2011) 885~893

Song Kyu Keun received the B.S. and M.S. degree in mechanical engineering from Chonbuk National University, South Korea and Ph.D. degree in mechanical engineering from Hokkaido University, Japan. He is currently a professor at Precision Mechanical Engineering Department, Chonbuk National University. His research interests include combustion and exhaust emission characteristics of diesel engine, spray characteristics of biodiesel in common rails injection system. Jung Jae Youn received the B.S. and M.S. degree in mechanical engineering from Chonbuk National University, South Korea and Ph.D. degree from Tokyo Institute of Technology, Japan. He is currently a professor at Precision Mechanical Engineering Department, Chonbuk National University. His major research field includes tribology, hydraulic pump and motor design, and its performance improvement.

893

Hyung Gon Kim received the B.S. and M.S. degrees in Precision mechanical Engineering from Chonbuk national University, Korea in 1995 and 1997; the Ph.D. degree from Kagoshima University, Japan in 2006. He currently works as a CEO in CAMSTech Co., Ltd, Korea. His current research interests are heat and fluid engineering, atomization system, and agricultural machinery.