HCCI operation of an optically accessible diesel engine fuelled with RME fuel Ezio Mancaruso, Bianca M. Vaglieco Istituto Motori CNR – Napoli (Italy) Introduction In order to overcome the pollutant emission li- mitations the use of homogeneous charge com- ICCD pression ignition (HCCI) mode must be considered in direct injection diesel engine using biodiesel fuel. HCCI mode reduces PM and NOx emissions without penalize the performances [1]. This occurs because the combustion develops with low tem- Spectrograph perature and burns a premixed air/fuel mixture. Filter Biodiesel is a renewable fuel that can be produced ICCD from a variety of vegetable oils including rapeseed oil, soybean oil, sunflower oil and palm oil [2]. The goal of this paper is the evaluation of the HCCI CCD combustion in an optically accessible diesel engine realized with four small early injections using RME Fig. 1: Experimental setup. fuel. In particular, the ICCD camera was used and Experimental apparatus coupled with narrow filters corresponding to λ=310 The optically accessible engine used during ex- nm for OH, λ=330 nm for HCO* and λ=431nm for periments was a single cylinder, direct injection, CH, respectively. The second CCD was sensible in four-stroke, diesel engine, with EURO IV multi the visible. Moreover, two colour pyrometry method valves production head. The engine used a classic was applied by using an “ad hoc” filter in order to extended piston with piston crown window (46 mm evaluate soot temperature and concentration [4]. diameter) and a 45° UV-visible mirror located in- Synchronization of engine with ICCD and CCD side the elongated piston. The engine was camera was obtained by the unit delay connected equipped with Common Rail injection system and a to the engine shaft encoder. The effect of RME fuel fully flexible Electronic Control Unit (ECU) that with respect to reference fuel (REF) at two engine controlled the number of injections up to 5 per speeds, 1500 and 2000 rpm, respectively, and cycle. To analyse the injection signals, a Hall-effect varying the injection pressure was analysed. sensor was applied to the line of the solenoid cur- rent and a piezoelectric pressure transducer was Results and discussions located in the injection line between the rail and In figure 2 typical histories of cylinder pressure, the injector. To acquire the cylinder pressure in rate of heat release and drive current of HCCI at motored and fired condition, a piezoelectric pres- 2000 rpm for REF and RME fuels, respectively, are sure transducer was set in the glow plug seat of reported. In particular, in order to realize a well the engine head. Imaging measurements from mixed air/fuel charge in the cylinder four early in- ultraviolet (UV) to visible were performed by means jections were performed during the compression of the optical set up shown in figure 1. stroke. The same fuel amount was injected varying Two different CCD cameras, with a different the injection pressure, even if higher quantities for capability, were used. The first (ICCD) had high the engine operation with RME fuel because of its sensibility both in the UV and visible range. The lower energy content. It can be noted that the rates intensifier-gate duration of 41 µs was used in order of heat release have typical shape of HCCI com- to have a good accuracy in the timing of the onset bustion with two well resolvable peaks not corre- of the combustion process. Previous investigations lated with the injection strategy investigated. The showed the presence of characteristics radicals first is characteristic of low temperature reactions during the low temperature and premixed combus- that occur in the chamber at autoignition; and the tion development [3]. second is due to the development of high tempera- ture reactions [3]. 60 REF Combustion UV images were recorded by means a filter 400 500 pressure (bandwidth 10nm) at the characteristic emission 50 600 wavelength of OH (310 nm). In figure 3, the OH Rate Of Heat Release [kJ/kg °] 700 40 images and the relative maximum intensity at the lower left corner are reported. Cylinder pressure [bar] 150 30 100 Drive injector current [A] 20 50 10 ROHR 30 0 0 20 10 Current 0 -70 -60 -50 -40 -30 -20 -10 0 10 20 30 40 Crank angle [°] 60 RME Combustion 400 pressure 50 500 600 Rate Of Heat Release [kJ/kg °] 700 40 Cylinder pressure [bar] 150 30 100 Drive injector current [A] 20 50 10 ROHR 30 0 0 20 10 Current 0 -70 -60 -50 -40 -30 -20 -10 0 10 20 30 40 Crank angle [°] Fig. 2: Combustion pressure, rate of heat release and drive current at 2000 rpm for REF (up) and RME (down) fuels at several injection pressures. To examine the temporal and spatial evolution of HCCI combustion digital imaging measurements from UV to visible were performed. In figure 3 the images of visible combustion and the soot KL fac- tor are reported. They are related to RME fuel, the engine speed of 2000 rpm and an injection pres- sure of 500 bar. In figure 3, the first column shows the flame evolution, it can be noted the presence of luminous spots randomly distributed in the bowl. Increasing the crank angle, the luminous spots move in the cylinder, due to the air swirl motion, and increase their density in the whole chamber. These bright spots are distributed not only in the bowl but also in the space between the engine head and the top of the piston [3]. By means of the two colour pyrometry method, the soot KL factor proportional to the soot mass concentration were calculated and reported in figure 3. The soot mass concentration in the cylinder is very low during the whole combustion process. In particular, the max- Fig. 3: HCCI combustion images, KL factor propor- imum intensity is detected at 6° before top dead tional to soot mass concentration, and OH intensity for center (BTDC). RME fuel at 2000 rpm. Previous analysis of extinction and chemilumi- nescence measurements showed that the pre- They show very low emission because only mixed combustion process is widely dominated by 15% of incident light passed through the filter. The the presence of OH radical [3]. For this reason, the OH radical is detected close to the bowl wall at the start of combustion, then it fills the whole chamber, This is due to the high unburned hydrocarbon con- and it shows maximum intensity at 4° BTDC, three centrations of the homogeneous combustion. crank angle degrees later than the heat release In conclusion, the use of oxygenated fuel helps peak. to keep low the production and emission of HC with consequently positive effect on the emission 4.E+07 soot REF of particulate matter. 1.E+08 4.E+07 soot RME Integral soot mass concentration [a.u.] OH REF 1.E+08 3.E+07 OH integral emission [a.u.] OH RME 1.E+08 3.E+07 2.E+07 8.E+07 Acknowledgments 6.E+07 The authors wish to thank Mr. Carlo Rossi and 2.E+07 Bruno Sgammato for maintaining the experimental 1.E+07 4.E+07 apparatus and for their precious help. 5.E+06 2.E+07 0.E+00 0.E+00 400 500 600 700 Fig. 4: Integral OH radical and soot mass concentra- References tion measured in the combustion chamber at 2000 rpm [1] Assanis, D. N., Najt, P. M., Dec J. E., Eng J. A., for REF and RME fuels at different injection pressure. Asmus T. N., Zhao F., “Homogeneous Charge Com- pression Ignition (HCCI) Engines”. Key research and In figure 4 the integral OH measured in the Development Issues. SAE International 2003 [2] Sustainable biofuels: prospects and challenges-The combustion chamber at 2000 rpm and for several Royal Society, 2008. injection pressures are reported. The integral con- [3] Merola, S.S., Vaglieco, B. M., Mancaruso, E., “Ex- centrations were computed for both REF and RME tinction and Chemiluminescence Measurements in fuels. It can be noted that increasing the injection CR DI Diesel Engine Operating in HCCI mode”, SAE pressure the OH intensity decreases for both fuels Paper n° 2007-01-0192, 2007. and the lowest intensity is detected for RME at 700 [4] Mancaruso, E., Merola, S.S., Vaglieco, B.M. “Study bar. Previous paper showed that OH radical was of the Multi-Injection Combustion Process in a responsible of the soot oxidation in the chamber Transparent DI Common Rail Diesel Engine by [5]. For this reason the integral soot concentrations means of Optical Techniques”. Int. Journal of Engine Research, vol.9, n.6, pp. 483-498, 2008. are reported in same figure. The higher injection [5] Heywood, J. B., “Internal Combustion Engine Fun- pressure produces better atomization of the fuel, damentals” - Mc Graw-Hill, NewYork, 1988. thus lower soot production. Moreover, the oxygen content of the RME fuel contributes strongly to the reduction of soot. Finally, the lowest OH concen- tration is detected at the lowest soot concentration, due to the higher in-cylinder soot reduction. 8 7.E+07 REF Integral soot mass concentratiuon [a.u.] RME 7 6.E+07 REF cc RME cc 6 5.E+07 5 PM [mg/m3] 4.E+07 4 3.E+07 3 2.E+07 2 1 1.E+07 0 0.E+00 400 500 600 700 Fig. 5: Comparison of in-cylinder soot mass concen- tration and PM exhaust emission at 2000 rpm for REF and RME fuels at different injection pressure. Figure 5 compares the integral value of soot concentration curves with the PM measured at the exhaust pipe of the optical engine by means of an opacimeter. The data in Figure 5 indicate similar trends in soot emission even if the difference be- tween the two fuels is less evident at the exhaust.