Experimental Analysis of the effect of very early pilot injection on pollutant formation for a PCCI Diesel engine A. Vanegas, N. Peters Institut für Technische Verbrennung RWTH Aachen University, Aachen, Germany In the present work, the influence of a very early pilot injection on pollutant formation was investigated in a Common-Rail DI Diesel engine. The engine was operated at conventional part-load conditions at 2000 rpm, an EGR variation was done and the injected fuel mass was 15 mm^3/cycle. The Nozzle type were conical and flow optimized geometries (ks nozzle) with hole diameters of 0.141 mm, length of hole of 1mm and the Spray Cone Angles were 148° and 120° Introduction a Common-Rail DI Diesel engine was carried out In Common-Rail DI Diesel Engines, a low on a production-type GM FIAT 1.9 l CDTI ECOTEC combustion temperature process is considered as Diesel engine. The 4-cylinder engine utilizes a one of the most important possibilities to achieve Common-Rail fuel injection system, variable ge- very small emissions and optimum performance. ometry turbocharger (VGT), an exhaust gas recir- To reduce NOx and Soot strongly, it is necessary culation system, and an intake throttle valve. The to achieve a homogenization of the mixture in or- engine has four valves per cylinder, centrally lo- der to avoid the higher local temperatures which cated injectors, and a re-entrant type combustion are responsible for the NOx formation [1]. Through chamber. All relevant engine data are given in the homogenization it is also possible to obtain a Table 1. The mounting of the engine on the test stoichiometric air-fuel ratio in order to significantly bench is shown in Fig. 1. The production of this reduce the Soot emissions. One way to achieve engine is certified to meet EURO IV emission this homogeneous condition is to start injection standards. very early together with the use of higher EGR rates. The direct effect of these conditions cause a DI, 4-cylinder, Engine Type longer ignition delay (this is the time between start charged, 4-stroke of the injection and auto-ignition during physical Bore [mm] 82.0 and chemical sub processes such as fuel atomiza- Stroke [mm] 90.4 tion, evaporation, fuel air mixing and chemical pre- 3 Displacement [cm ] 1900 reactions take place) so that the mixture formation has more time to achieve a homogeneous state. Compression Ratio 18.3 However there are some problems that must be Combustion Chamber Re-entrant type solved before this concept can be use completely. Max. Power [kW (PS)] 110 (150) @4000rpm The first problem consists on the higher production -1 of HC and CO emissions, due to spray is impinging Max. Torque (Nm / min ) 320 / 2000-2750 onto the wall surface. The second problem con- Injection System Bosch Common Rail sists on the position of the combustion phasing, Max. Rail Pressure [bar] 1600 bar which take place before the top death center and this situation influences negatively the engine effi- Nozzle hole diameter [mm] 0.141 ciency. Therefore is necessary to develop many Injector Nozzle 7 holes strategies in order to solve mainly these two prob- 3 Hydraulic flow rate [cm lems. [2]. In Previous Works was found optimal (30s) at 100 bar] 440 spray cone angle and piston bowl geometry, now this work try to find an optimal injection strategy. Table 1. Engine specifications and injection sys tem specifications Experimental Setup and Measurement Tech- niques The experimental investigation of the effect of a very early pilot injection on pollutant formation in along with an EGR valve. An intake throttle valve supports high flow rates of exhaust gas recircula- tion. According to conditioning modules for engine testing, the following applications were used: - Intake Air Conditioning - Fuel Conditioning and Measurement - Engine Oil Conditioning and Engine Coolant Conditioning. The air-flow rate was measured using a laminar flow element. AVL 733, a dynamic fuel meter, was used for fuel metering. An air-conditioning system determined and maintained the preset temperature and pressure of the intake air. Pressure data was collected for one cylinder. Measured emission data included smoke number, NOX, HC, and CO. AVL 415, a variable sampling smoke meter, provided exhaust smoke levels. An engine dynamometer (AC/DC type) was used to measure engine torque. The measuring equipment used in the experiments Figure 1. Engine test bench is summarized in Table 3. The engine is equipped with a second-generation Advance Options Ana- Bosch Common-Rail injection system that was CO, CO2 lysatormodul Magnos 16 used for all experiments reported in this study. ECO Physics CLD 700 NOX Regular Diesel fuel was used in the experiments. A EL ht Advance Options FID- summary of the most important properties of the HC Analysatormodul Multi- used Diesel fuel is shown in Table 2. FID 14 (THC C3) Variable Sampling Particulate matter Smoke Meter AVL 415 Fuel AVL 733 Cetane number 53.6 Density at 288K 834.2 (kg/m3) Table 3: Emission and fuel consumption measur- Viscosity at 313K ing equipment 2.892 (mm2/s) Distillation T 50% (K) 539.7 A Bosch-type flow bench [3, 4] was used to meas- ure the fuel injection rate. Fuel is injected into a Distillation T 95% (K) 619.0 tube of constant diameter and known length, filled Sulphur (mg/kg- with liquid fuel. Then, a pressure wave propagates 9.0 fuel) through the tube and provides a signal which is Lower Heating detected by a Kistler piezo electric pressure trans- 43.163 Value (MJ/kg-fuel) ducer (type 70061B) mounted close to the nozzle exit. This signal correlates to the instantaneous Table 2. Fuel properties injection flow-rate. The Piezo signal, fuel delivery pipe pressure, injector energizing time, injector A Common-Rail fuel system allows for vari- current, and voltage are measured with a fast data able pressures (up to 1600 bar), timing, and num- acquisition system (sample = 30 kHz, IMTEC com- bers of injections. Second-generation Bosch injec- pany). All measurements are carried out over 2000 tor systems allow for up to five injections (for ex- single injection events, where the injected fuel is ample, two pilot, one main, and two post injections) collected and weighted. The measured pressure per cycle. VGTs have flexible vanes, which move wave signal is calibrated by the time-integral of the and let more air into the engine, depending on the signal which is equal to the averaged injected fuel load. This technology increases both performance mass per injection. From a comparison of the in- and fuel economy. Turbo lag is reduced, as the jection rate and the current signal, the injection turbo impeller inertia is compensated using VGT delay and duration can be derived. spray cone angle 148 and 120 are showed in the figure 3 and 4. Description of experiments The investigations were carried out for a conven- tional medium-load point at 2000 rpm, a fuel amount of 15 mm^3/cycle, an external EGR rate variation and 700 bar rail pressure. In order to get a reference point in the first experi- ment the fuel mass of 15 mm^3/cycle was injected via a single injection. Secondly, 1/15 of the total fuel mass was pre-injected at a distance (dSOI) of the 70 CA deg respect to the SOI main injection or 80 deg. BTDC, in this case the fuel mass of the main injection was 14 mm^3/cycle and Thirdly, a pilot injection was done 60 deg. BTDC also with a mass fuel of 1 mm^3/cycle and a Main injection with 14 mm^3/cycle. See figure 2 Figure 3. Soot and NOx emissions for SCA 148°. At a SOI pilot injection by 60 deg. BDTC the Soot and NOx emissions were reduced in comparison to the single injection event, for the second case with SOI 80 deg. BTDC the situation was similar. If the Soot emissions of this experiment with the SCA 148 are compared to the SCA 120 results is possi- ble to see that the Soot and NOx emissions in- crease due at the higher temperatures resultants of a better quality of the combustion. In the figure 4 is showed that the Soot emissions are reduced by using a pilot injection at 80 deg. BTDC in compari- son to the single event. These results could be mean that through the pilot injection the mixture Figure 2: Duration between the starts of two injec- formation is more homogeneous with respect to tions the single injection. All experiments were carried out first with the spray cone angle 148° and the same experiments also were done with the spray cone angle 120°. See Figure 3. Figure 3. Spray Cone Angle Variation Figure 4. Soot and NOx emissions for SCA 120°. Results and Discussion For the pilot injection events the HC and CO emis- sions are increased. These results were expected This work was only experimental but numerical because at very early injection times like 60 or 80 support is expected. A very early pilot injection has deg. BTDC the pressure, temperature and density the intention to prepare good conditions for the in the combustion chamber are very lower and combustion. The Soot and NOx emissions for the therefore the spray penetration is longer, that means, the fuel spray is impinging on the cylinder- Due a better quality of the combustion the HC and wall. In the figure 5 and 6 the HC and CO emis- CO are lower but unfortunately the NOx emission sions are showed for a SCA 148°. increases due at the high temperature. In relation- ship to the pilot events there are not advantages with respect to the HC and CO emissions. Figure 5. HC and NOx emissions for SCA 148° Figure 8. CO and NOx emissions for SCA 120° For the nozzle with the SCA 148° the IMEP is for the single injection the largest. That means, the fuel consumption is higher for the cases with pilot events. This is a consequence that the spray of the pilot injections is impinging on the cylinder walls. In the figure 9 is showed the IMEP versus NOx for the SCA 148. Figure 6. CO and NOx emissions for SCA 148° However the HC and CO emissions for the SCA 120 are decreased with respect to the SCA 148°. See figure 7 and 8. Figure 9. IMEP and NOx emissions for SCA 148° For the nozzle with SCA 120 the IMEP for the sin- gle injection event is similar to the experiment with the SCA 148. However, the pilot injection at 60 deg. BTDC showed a similar value of IMEP in compari- son to the single injection. Figure 7. HC and NOx emissions for SCA 120° Summary and Conclusions An optimal injection strategy is necessary in order to reduce the NOx, Soot, HC and CO emissions and to move the combustion phasing after the top death center. Therefore is necessary to do a lot of experiments using different start of injection times for the main- and pilot injections in order to find the optimal. Use of a very early pilot injection reduces mainly the Soot and NOx emissions but increase the HC and CO emissions. A drastic reduction of HC and CO is possible by using of one smaller Spray Cone Angle. References [1] Merker, G., Schwarz, C., Stiesch, Otto, F. “Verbrennungsmotoren“ , 2004. [2] H. Won, A. Vanegas, “Experimental Study of HC Emissions Using Narrow Spray Cone An- gles and Different Surrogate Fuels in Low Temperature Diesel Combustion Systems FISITA 2008 World Automotive Congress, September 2008 Munich Germany. [3] Bosch, W., ”Der Einspritzgesetzindikator – ein neues Messgerät zur direkten Bestimmung des Einspritzgesetzes von Einzeleinspritzun- gen”, MTZ-Motortechnische Zeitschrift, 25(7), 1964 [4] Smith, W. J., and Timoney, D. J., “Fuel Injec tion Rate Analysis - A New Diagnostic Tool for Combustion Research”, Paper No. SAE 92224, 1992.