The Effects of Pilot Injection on Combustion in Dimethyl-ether (DME) Direct Injection Compression Ignition Engine

Transcription

1 SAE TECHNICAL PAPER SERIES The Effects of Pilot Injection on Combustion in Dimethyl-ether () Direct Injection Compression Ignition Engine H. Yoon, K. Yeom, C. Bae Korea Advanced Institute of Science and Technology Capri, Naples Italy September 16-2, 27

2 The Effects of Pilot Injection on Combustion in Dimethyl-ether () Direct Injection Compression Ignition Engine Copyright 27 SAE International Hyeonsook Yoon, Kitae Yeom and Choongsik Bae Korea Advanced Institute of Science and Technology ABSTRACT Dimethyl-ether combustion with pilot injection was investigated in a single cylinder direct injection diesel engine equipped with a common-rail injection system. Combustion characteristics and emissions were tested with dimethyl-ether and compared with diesel fuel. The main injection timing was fixed to have the best timings for maximum power output. The total injected fuel mass corresponded to a low heating value of 45 joules per cycle at 8 rpm. The fuel quantity and the injection timing of the pilot injection were varied from 8 to 2% of the total injected mass and from 5 to 1 crank angle degrees before the main injection timing, respectively. Ignition delay decreased with pilot injection. The effects of pilot injection were less significant with combustion than with diesel. Pilot injection caused the main combustion to increase in intensity resulting in decreased emissions of hydrocarbons, carbon monoxide and particulate matter. The NOx emission with diesel combustion was increased because of the activation of main combustion by pilot injection. On the contrary, in combustion, the NOx emission was decreased below that of single injection when the pilot ratio was more than. INTRODUCTION Different kinds of alternative fuels for engines have been introduced to reduce toxic emissions. Dimethyl-ether () is considered one of the most promising alternative fuels for compression-ignition engines. is a gaseous fuel at standard temperature and pressure conditions, with markedly reduced emissions of particulate matter (PM) compared with standard diesel. Other reasons that account for the negligible emission of black smoke are the presence of oxygen and absence of carbon-to-carbon bonds in the fuel molecules. Small amounts of black smoke are oxidized during the expansion stroke based on the oxygen content. The properties of and diesel fuel are shown in Table 1 [1]. Much research on spray characteristics, combustion processes and development of engines with has been reported [2-11]. vaporizes more Table 1 Properties of and fuel [1] Property (unit/condition) Chemical Unit fuel CH 3 -O- Structure CH 3 - Molar mass g/mol Carbon content mass % Hydrogen content mass % Oxygen content mass % 34.8 Carbon-to hydrogen ratio Critical temperature K 4 78 Critical pressure MPa * Critical density kg/m Liquid density kg/m Relative gas density (air=1) Cetane number > Auto-ignition temperature K Stoichiometric air/fuel mass ratio Boiling point to K at 1 atm 643 Enthalpy of vaporization kj/kg Lower heating value MJ/kg Gaseous specific heat capacity kj/kgk Ignition limits Modulus of elasticity Kinematic viscosity of liquid Surface tension (at 298K) Vapour pressure (at 298K) Vol % in air N/m E+8 3.4/18.6.6/ E+8 cst <.1 3 N/m kpa 5 << 1

3 easily than diesel at the same injection pressure due to superior evaporation performance [2]. Lower nitric oxide (NOx) can be achieved with combustion than with diesel because of higher gaseous specific heat capacity and slightly adiabatic flame temperature than diesel [5]. An engine for a light duty truck was recently developed which reduced NOx emissions by 4%, maintained engine power output and provided the same or better specific fuel consumption [6]. Higher torque and power could be achieved with, and brake thermal efficiencies were higher than diesel with the same NOx level at high engine speed [7]. Total hydrocarbon (HC) exhaust emissions are less than or equal to diesel. The oxygen in the molecular structure and good mixing characteristics make combustion more complete. However, toxic oxygen compound emissions, such as formaldehyde, acetaldehyde and formic acid, were higher [8]. Carbon monoxide (CO) emissions depend on engine systems and operating conditions. Rapid evaporation and good mixing characteristics occur if the mixture is lean, and they cause the less rich region and less CO emissions. Multiple injection for diesel engines was introduced in order to reduce exhaust emissions and improve combustion characteristics. The multiple injection strategy for combustion may improve combustion and exhaust emission characteristics. Pilot injection is an effective way to reduce combustion noise and emissions. Pilot combustion increases incylinder pressure and temperature, which hastens the auto-ignition of the injected fuel. The shortened ignition delay before main combustion and increased in-cylinder pressure and temperature change the configuration of heat release, influencing engine performance and exhaust emission, especially NOx emission. The effects of pilot injection on NOx emission are determined according to engine operating conditions. NOx emission was reduced by injecting the proper amount of fuel just before the main injection [12], though it could vary considerably with the timing and quantity of pilot injection [13, 14]. The effects of pilot injection on combustion were investigated in a single cylinder direct injection compression ignition (DICI) engine with a common-rail injection system. The effects of pilot injection timing and quantity were tested with a fixed amount of total injected fuel mass at a fixed main injection timing. The combustion and exhaust emissions of diesel and were investigated and compared. EXPERIMENTS EXPERIMENTAL SETUP The test engine was a single-cylinder DICI engine equipped with a common-rail injection system. The engine specifications are described in Table 2. The schematic diagram of the experimental setup for is shown in Fig. 1. The speed and load of the research engine were controlled by a DC dynamometer (37 kw). and diesel were supplied at a pressure of 65 MPa with a common-rail injection system. Experiments were completed with two separate fuels. Each fuel was supplied with individual fuel injection system. was pressurized by a pneumatic pump (Haksel Inc.; 4B-) and supplied to a common-rail. was supplied through a commercial common-rail injection system (Bosch) dedicated for diesel fuel. The test injector was a 7-hole solenoid type commercial diesel injector. The fuel injection parameters, injection pressure, injected quantity and injection timing were controlled by an injector driving Figure 1 Schematic diagram of the experimental setup

4 Table 2 Engine specifications Research engine type Valves per cylinder Bore Stroke Connecting rod length Displacement per cylinder Single-cylinder DICI engine 4 (2 intake and 2 exhaust) 83 mm 92 mm 1 mm 498 cm 3 Compression ratio 14.8 : 1 Injector type / nozzle style Solenoid / mini sac Number of holes 7 Table 3 Operating conditions Engine speed Injection pressure Injection timing 8 rpm 65 MPa injection Pilot injection injection Pilot injection Total injected fuel quantity 22 mm mm 3 TDC 5 CAD BTDC ~ 1 CAD BTDC 1 CAD BTDC CAD BTDC ~ 2 CAD BTDC Corresponding to LHV of 45 J/cycle system (Zenobalti Co.; IDU 5B). A precise rotary encoder (Koyo; 3 pulses/revolution) mounted on a camshaft was used to control fuel injection timing at a resolution of 1 crank angle degree (CAD). Engine out emission was measured by an exhaust gas analyzer (HORIBA; MEXA D); composed of a nondispersive infrared absorption (NDIR) analyzer for the carbon monoxide (CO) and carbon dioxide (CO 2 ) measurements, chemiluminescence detector (CLD) analyzer for the NOx measurements and a flame ionization detector (FID) analyzer for the HC measurements. PM emissions were analyzed by a smoke-meter (AVL; 4S). A filter smoke number (FSN) of 1, relating to the smoke-meter output, means that the PM concentration is mg/m 3 at the sampling point. FSN can be convertible to soot concentration (ρ) in mg/m 3 from the following correlation. ρ = ( 1/.45) * α * FSN * exp( β * FSN) This correlation is applicable when the FSN is less than 8. The values of the constants are α=5.32 and β=.31 []. OPERATING CONDITIONS The engine operating conditions are given in Table 3. The speed of the engine was fixed at 8 rpm. Total injected fuel quantity was held constant, corresponding to a low heating value (LHV) of 45 joule per cycle at 8 rpm; 22 mm 3 for and 11.5 mm 3 for diesel. injection timing was fixed at top dead center (TDC) for and 1 CAD before TDC for diesel, which achieved the maximum indicated mean effective pressure (IMEP) with single injection. Fuel quantity of the pilot injection varied from 8% to 2% of the total injected fuel quantity, and pilot injection timing was varied from 5 to 1 CAD before the main injection. The fuel quantity ratio of the pilot injection was represented as pilot ratio. Pilot ratio was defined as the amount of pilot injected fuel versus total injected fuel as shown in equation (2). (1) Fuel quantity of the pilot injection Pilot ratio = (2) Total injected fuel quantity MEASUREMENTS In-cylinder pressure data were recoded with a resolution of 1 CAD by a piezoelectric transducer (KISTLER; 52A). Pressure data and engine out emissions were collected by a data acquisition system (Iotech Ltd.; WaveBook512) at 2 khz. Data from 5 engine cycles were averaged and used for the calculation of the heat release rate (HRR). The HRR was calculated by means of the traditional single-zone first law equation [16]. Heat transfer to the cylinder walls was estimated using the Woschni model [17]. RESULTS AND DISCUSSIONS MAIN INJECTION TIMING To determine the main injection timing, the engine was operated under single injection conditions and the injection timing was swept from 2 CAD before top dead center (BTDC) to 5 CAD after top dead center (ATDC). The maximum IMEP was achieved when the injection timing was 1 CAD BTDC during diesel combustion. Engine combustion became unstable for diesel fuel injected after 2 CAD BTDC. IMEP was maximized when the injection timing was TDC for. The HC and CO emissions increased considerably when the injection timing was retarded after 2 CAD ATDC, until misfire occurred at 5 CAD ATDC of injection timing. According to the results, the main injection timing was fixed for

5 maximum IMEP for each fuel, at 1 CAD BTDC for diesel and TDC for. EFFECTS OF PILOT RATIO ON COMBUSTION The effects of pilot injection on combustion were investigated in a single cylinder DICI engine, which were compared with standard diesel fuel injection. The HRR, in-cylinder pressure and exhaust emissions were obtained and then analyzed to understand the effect of pilot injection on and diesel combustion. injection timing was fixed to TDC for and 1 CAD BTDC for diesel. The pilot injection timing was varied from 1 CAD to 5 CAD before main injection timing. The pilot ratio was varied from 8% to 2% with fixed total injected fuel quantity corresponding to the LHV of 45 joules per cycle at 8 rpm, 22 mm 3 for and 11.5 HRR of [J/deg] Pilot 2% Pilot Crank angle degree HRR of diesel [J/deg] Figure 2 Heat release rate at different pilot ratios for and diesel [ injection timing: TDC for and 1 CAD BTDC for diesel, Pilot injection timing: 1 CAD before main injection, Total injection quantity: 22 mm 3 for and 11.5 mm 3 for diesel] In-cylinder P of [bar] 75 Pilot 2% Pilot Crank angle degree In-cylinder P of diesel [bar] mm 3 for diesel. Figures 2 and 3 illustrate HRR and in-cylinder pressure of and diesel combustion under various pilot ratios with a fixed pilot injection timing of 1 CAD before main injection. The ignition delay of the main combustion was shortened with pilot injection compared to single injection with increased in-cylinder pressure and temperature by pilot combustion. The peak HRR of pilot combustion increased at high pilot ratios as a result of increased heat supply from pilot combustion (Fig. 2). At the same time, the decreased fuel quantity during main injection caused a reduced peak HRR during main combustion. HRR of [J/deg] pilot at 1 CAD before main pilot at 25 CAD before main pilot at 5 CAD before main injection timing Crank angle degree HRR of diesel [J/deg] Figure 4 Heat release rate at different pilot injection timings for and diesel [ injection timing: TDC for and 1 CAD BTDC for diesel, Pilot ratio: 2%, Total injection quantity: 22 mm 3 for and 11.5 mm 3 for diesel] In-cylinder P of [bar] pilot at 1 CAD before main pilot at 25 CAD before main pilot at 5 CAD before main injection timing Crank angle degree In-cylinder P of diesel [bar] Figure 3 In-cylinder pressure at different pilot ratios for and diesel [ injection timing: TDC for and 1 CAD BTDC for diesel, Pilot injection timing: 1 CAD before main injection, Total injection quantity: 22 mm 3 for and 11.5 mm 3 for diesel] Figure 5 In-cylinder pressure at different pilot injection timings for and diesel [ injection timing: TDC for and 1 CAD BTDC for diesel, Pilot ratio: 2%, Total injection quantity: 22 mm 3 for and 11.5 mm 3 for diesel]

6 EFFECTS OF PILOT INJECTION TIMING ON AND DIESEL COMBUSTION Figures 4 and 5 show HRR and in-cylinder pressure at different pilot injection timings with a fixed pilot ratio of 2%. The peak HRR of pilot combustion was increased during advanced pilot injection timing. The in-cylinder pressure and temperature were lower than auto-ignition conditions at advanced pilot injection timings. This caused a longer ignition delay, which is the time for fuel vaporization and mixing. As a result, the advanced pilot injection may cause the formation of a more homogeneous mixture. Fuel quantity during pilot injection was small, making the mixture too lean to burn in some regions of the combustion chamber [13]. This lean and homogeneous mixture caused a lower peak HRR, longer burn duration of pilot combustion, and a lower in-cylinder pressure and temperature at the start of main combustion; these all caused the longer ignition delay of main combustion. The peak HRR of main combustion was increased when the pilot injection timing was advanced (Fig. 4). Remaining fuel-air mixture from pilot combustion may be increased with advanced pilot injection timing. Less fuel from pilot injection was burned because of the lower incylinder pressure and temperature at advanced injection timing. Residual fuel from pilot combustion was burned during main combustion, increasing the peak HRR of main combustion. However, some part of the fuel remained under over-lean conditions, and did not participate in main combustion. Consequently, this affected HC and CO emissions. SPECULATION ABOUT THE EFFECT OF PILOT RATIO AND INJECTION TIMING ON AND DIESEL COMBUSTION In this section, previous explanations for the effect of pilot injection on and diesel combustion are summarized and then further discussed. Figures 6 and 7 show the peak HRR of and diesel combustion under various pilot ratios and pilot injection timings. Peak HRR of pilot combustion was increased with high pilot ratios and retarded pilot injection timings. The peak HRR of main combustion decreased at higher pilot ratios and advanced pilot injection timings. and diesel showed similar trends with pilot injection timing and ratios; however, the decrease in main peak HRR by the advanced pilot injection timing was more significant in diesel combustion. The difference in diesel combustion between the main peak HRR for early pilot and for late pilot was more than twice that of (Fig. 6). has lower auto-ignition temperature, higher evaporation rate, and a higher cetane number than diesel. These permit the pilot injected fuel to be ignited and burned with a small quantity of fuel and advanced injection timing. However, for the diesel pilot, large portions of the pilot injection may impinge and remain on the in-cylinder surfaces, and may not vaporize sufficiently. Therefore, the peak HRR of main combustion varied with the pilot injection timing and ratio. In other words, the pilot injection burned out independently before main combustion; on the contrary, diesel pilot remained vaporized till the start of main combustion. Therefore, the amount of remaining mixture after pilot combustion was much smaller than with the diesel pilot. According to this, the peak heat release of diesel increased sufficiently as the pilot injection timing advanced. Although and diesel showed similar trends with variation in pilot conditions, the specific value of heat release, emerging time and duration of pilot and main combustion showed differences, owing to the already-mentioned differences in physical properties, volatility and auto-ignition characteristics. The primary difference between and diesel is volatility. Considering that the has high volatility and evaporates more quickly, has a greater possibility of generating an over-lean mixture Peak HRR of main [J/deg] % Peak HRR of diesel main [J/deg] Figure 6 Peak heat release rate of main combustion under various pilot injection timings and pilot ratios [Injection quantity: 22 mm 3 for and 11.5 mm 3 for diesel] Peak HRR of pilot [J/deg] % Peak HRR of diesel pilot [J/deg] Figure 7 Peak heat release rate of pilot combustion under various pilot injection timings and pilot ratios [Injection quantity: 22 mm 3 for and 11.5 mm 3 for diesel]

7 compared with diesel. This causes less variation in heat release and in-cylinder pressure at different pilot conditions than diesel combustion. Furthermore, this can cause higher HC emissions and less NOx emissions from combustion compared with diesel combustion. In the power output, the IMEP of combustion decreased with pilot injection; in contrast, diesel pilotinjection made the IMEP increase. The pilot combustion created negative work in both cases. Nevertheless, the negative work was bigger in combustion. The large proportion of the pilot injection burned out before the main combustion. Moreover, the effect of pilot combustion on vaporization of main injected fuel was insignificant than diesel fuel, because the has good evaporation characteristic. With diesel, the portion of negative work by the pilot combustion was small and the increased in-cylinder temperature at main injection timing made the main injected fuel evaporate more; the activation of the main combustion is more effective than the negative work of the pilot injection. As a result, the power output was bigger for diesel combustion than. THE EFFECTS OF PILOT INJECTION ON ENGINE- OUT EMISSIONS OF AND DIESEL COMBUSTION To further understand the effect of pilot injection on and diesel combustion, the engine-out emissions were measured and then analyzed. Figure 8 shows the HC emission of and diesel combustion at different pilot injection conditions. HC emission was reduced in both fuels by % for and 8% for diesel, respectively, when the pilot injection timing was retarded. When the pilot injection timing was retarded, a higher portion of pilot injected fuel was burned in pilot combustion and the HC % HC Figure 8 HC emission of and diesel combustion at different pilot injection timings and pilot ratios [Injection quantity: 22 mm 3 for and 11.5 mm 3 for diesel] NOx % NOx Figure 1 NOx emission of and diesel combustion at different pilot injection timings and pilot ratios [Injection quantity: 22 mm 3 for and 11.5 mm 3 for diesel] CO % CO Soot Concentration [mg/m 3 ] % Soot Concentration [mg/m 3 ] Figure 9 CO emission of and diesel combustion at different pilot injection timings and pilot ratios [Injection quantity: 22 mm 3 for and 11.5 mm 3 for diesel] Figure 11 PM emission of and diesel combustion at different pilot injection timings and pilot ratios [Injection quantity: 22 mm 3 for and 11.5 mm 3 for diesel]

8 remaining lean mixture was decreased. This caused reduced HC emission at retarded pilot injection timings. The combustion of the advanced pilot injection showed higher HC emissions due to the incomplete burn-out of vapor phase fuel from crevices and bulk-gas. In the case of combustion, the HC emission was higher and the reduction in HC with retarded pilot injection timing was increased compared with diesel. This is because of the higher portion of over-lean mixture, because of the higher fuel volatility of and the combustion characteristics similar to low temperature combustion of the premixed fuel. Figure 9 shows the CO emission of and diesel combustion at different pilot injection conditions. CO emissions were decreased with retarded pilot injection timings by up to 7% for and 86% for diesel. This could be explained in the same way as HC emissions. However, diesel combustion emitted higher CO compared with, owing to the wider rich mixture region, since diesel has less fuel volatility compared with. Figures 1 and 11 illustrate the NOx emission of and diesel combustion at different pilot injection conditions. In combustion, the NOx emission increased compared with single injection only when the pilot ratio was 8%, and decreased as the pilot ratio increased. In contrast, the NOx emission was doubled with pilot injection in diesel combustion. The NOx emission is strongly influenced by local temperature and equivalence ratios. Pilot combustion did not activate the fuel vaporization process in combustion, since has sufficient evaporation performance. NOx was decreased due to a decrease in HRR from main combustion. combustion for diesel was activated with pilot combustion by the improvement in fuel evaporation, thus the in-cylinder pressure and temperature increased. This caused the increase of NOx emission in diesel combustion. The soot concentration, the quantity of the PM emission, was nearly zero for the combustion which was less than.66 mg/m 3. In diesel combustion, the PM and NOx showed the trade-off relation; NOx increased and PM decreased as the combustion was activated by pilot injection. SUMMARY AND CONCLUSIONS Pilot injection is an effective way to control combustion noise and emissions. The influences of pilot injection on combustion characteristics and exhaust emissions were investigated and compared with diesel. Pilot injection timing and quantity were varied with fixed main injection timing and the total injected fuel quantity. The main injection timing was fixed at TDC for, and at 1 CAD BTDC for diesel, which achieved the maximum IMEP for each case. Pilot injection timing was varied from 1 CAD to 5 CAD before main injection timing, and the pilot ratio was varied from 8% to 2%. The following are the findings of combustion characteristics and engine-out emission of affected by pilot injection, compared with diesel combustion. 1. The pilot combustion increased the in-cylinder pressure and temperature under main injection timing; as a result, the start of main combustion was advanced with pilot injection. 2. The peak HRR of main combustion decreased with the pilot injection. As the pilot injection timing was advanced, the peak HRR of main combustion was increased; the peak HRR of pilot combustion was decreased. The variation in diesel combustion was more significant than that of due to diesel s inferior evaporation rate compared with. 3. The emissions of HC, CO and PM decreased with retarded pilot injection. However, these increased with advanced pilot injection timing due to the low temperature combustion of the premixed pilot fuel. 4. The NOx emission in diesel combustion was increased because of the activation of main combustion by pilot injection. On the contrary, in combustion, the NOx emission was decreased below that of single injection when the pilot ratio was more that 12 %. 5. PM was not emitted in combustion without a tradeoff with NOx since has good evaporation performance and good auto-ignition characteristics. Consequently, combustion could achieve reduction in HC and CO without an increase in NOx emissions. The effects of pilot injection were less significant in combustion than in diesel. The pilot injection with diesel combustion helped evaporate the main injected fuel which made the main combustion more intense. Therefore, the pilot injection resulted in decreased emissions of HC, CO and PM; but doubled NOx emissions in diesel combustion. In combustion, the pilot injected fuel had its combustion phase independently, and the fuel injected during the main injection timing had sufficient evaporation performance even without the help of pilot combustion. Although pilot combustion activated the main combustion with increased in-cylinder pressure and temperature during main injection timing, the decreased main HRR with the decreased main fuel quantity caused NOx emissions to decrease. ACKNOWLEDGMENTS This work is part of CERC (Combustion Engineering Research Center) and the project "Development of Partial Zero Emission Technology for Future Vehicle" funded by the Ministry of Commerce, Industry and Energy. The authors are grateful for its financial support.

9 REFERENCES 1. Arcoumanis, C., Bae, C., Crookes, R., and Kinoshita, E., The Potential of di-methyl ether () as an Alternative Fuel for Compression-Ignition engines : a review, Fuel, in press. 2. Yu, J., and Bae, C., Dimethyl Ether () Spray Characteristics in a Common-rail Fuel Injection System, Journal of Automobile Engineering, Proc. IMechE, Part D, Vol. 217, No. 12, pp , Curran, H. J., Pitz, W. J., Westbrook, C. K., Dagaut, P., Boettner, J-C., and Cathonnet, M., A Wide Range Modeling Study of Dimethyl Ether Oxidation, International Journal of Chemical Kinetics, Vol., No. 3, pp , Yamada, H., Sakanashi, H., Choi, N., and Tezaki, A., Simplified Oxidation Mechanism of Applicable for Compression Ignition, SAE Paper , Cipolat, D., Combustion Aspects of a compression Ignition Engines Fueled on, 7th International Conference on Energy for a Clean Environment, Lisbon, pp.1-13, Sato, Y., Nozaki, S., and Noda, T., The Performance of a Engine for Light Duty Truck Using a Jerk Type In-Line Injection System, SAE Paper , Oguma, M., Goto, S., and Watanabe, T., Engine Performance and Emission Characteristics of Engine With Inline Injection Pump Developed for, SAE Paper , Oguma, M., Shiotani, H., Goto, S., and Suzuki, S., Measurement of Trace Levels of Harmful Substances Emitted from a DI Engine, SAE Paper , Goto, S., Oguma, M., and Suzuki, S., Research and Development of a Medium Duty Truck, SAE Paper , Tsuchiya, and T., Sato, Y., Development of Engine for Heavy-duty Truck, SAE Paper , Oguma, M., and Goto, S., Evaluation of Medium Duty Truck Performance Field Test Results and PM Characteristics-, SAE Paper , Tow, T. C., Pierpont, D. A., and Reitz, R. D., Reducing Particulate and NOx Emissions by Using Multiple Injections in a Heavy Duty D.I. Engine, SAE Paper 94897, Tennison, P. J., and Reitz, R., An Experimental Investigation of the Effects of Common-rail Injection System Parameters on Emissions and Performance in a High-Speed Direct-Injection Engine, Journal of Engineering for Gas Turbines and Power, Vol. 123, No. 1, pp , Tanaka, T., Ando, A., and Ishizaka, K., Study on pilot injection of DI diesel engine using common-rail injection system, JSAE Review, Vol. 23, No. 3, pp.297-2, 22.. AVL Model 4S Smoke Meter Product Literature, Updated October, Heywood, JB., Internal Combustion Engine Fundamentals, McGraw-Hill, New York, Woschni, G., A Universally Applicable Equation for the Instantaneous Heat transfer Coefficient in the Internal Combustion Engine, SAE Paper 67931, CONTACT Choongsik Bae Dept. Mechanical engineering Korea Advanced Institute of Science and Technology (KAIST) 373-1, Kusong-dong, Yusong-gu, Taejon 5-71, Republic of Korea Tel.: Fax: csbae@kaist.ac.kr ABBREVIATIONS PM NOx HC DICI CAD NDIR CLD FID FSN LHV TDC IMEP HRR BTDC ATDC GREEK dimethyl-ether particulate matter nitric oxide hydrocarbon direct injection compression ignition crank angle degree non-dispersive infrared absorption chemiluminescence detector flame ionization detector filter smoke number low heating value top dead center indicated mean effective pressure heat release rate before top dead center after top dead center ρ soot concentration, mg/m 3 α constant β constan